John A. Hird (Editor) - Policy Analysis in The United States-Policy Press (2018)

INTERNATIONAL LIBRARY OF POLICY ANALYSIS
SERIES EDITORS:
IRIS GEVA-MAY & MICHAEL HOWLETT
POLICY ANALYSIS IN
the United States
Edited by John Hird

POLICY ANALYSIS
IN THE UNITED STATES
International Library of
Policy Analysis
Series editors: Iris Geva-May and Michael Howlett,
Simon Fraser University, Canada
This major new series brings together for the first time a detailed
examination of the theory and practice of policy analysis systems
at different levels of government and by non-governmental actors
in a specific country. It therefore provides a key addition to
research and teaching in comparative policy analysis and policy
studies more generally.
Each volume includes a history of the country’s policy analysis which

offers a broad comparative overview with other countries as well as the
country in question. In doing so, the books in the series provide the
data and empirical case studies essential for instruction and for further
research in the area. They also include expert analysis of different
approaches to policy analysis and an assessment of their evolution and
operation.
Early volumes in the series will cover the following countries:

Australia • Brazil • China • Czech Republic • France • Germany •
India • Israel • Netherlands • New Zealand • Norway •
Russia • South Africa • Taiwan • UK • USA
and will build into an essential library of key reference works. The series
will be of interest to academics and students in public policy, public
administration and management, comparative politics and government,
public organisations and individual policy areas.
It will also interest people working in the countries in question
and internationally.
In association with the ICPA-Forum and Journal of Comparative Policy Analysis.

See more at http://goo.gl/raJUX
POLICY ANALYSIS IN THE
UNITED STATES
Edited by John A. Hird
International Library of Policy Analysis, Vol 12

First published in Great Britain in 2018 by
Policy Press North America office:
University of Bristol Policy Press
1-9 Old Park Hill c/o The University of Chicago Press
Bristol BS2 8BB 1427 East 60th Street
UK Chicago, IL 60637, USA
+44 (0)117 954 5940 t: +1 773 702 7700
pp-info@bristol.ac.uk f: +1 773 702 9756
www.policypress.co.uk sales@press.uchicago.edu
www.press.uchicago.edu
© Policy Press 2018
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library.
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested.
ISBN 978-1-4473-3382-1 hardcover
ISBN 978-1-4473-4600-5 paperback
ISBN 978-1-4473-4601-2 ePub
ISBN 978-1-4473-4602-9 mobi
ISBN 978-1-4473-3383-8 ePdf
The right of John A. Hird to be identified as editor of this work has been asserted by him in
accordance with the Copyright, Designs and Patents Act 1988.
All rights reserved: no part of this publication may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording,
or otherwise without the prior permission of Policy Press.
The statements and opinions contained within this publication are solely those of the editor
and contributors and not of the University of Bristol or Policy Press. The University of Bristol
and Policy Press disclaim responsibility for any injury to persons or property resulting from any
material published in this publication.
Policy Press works to counter discrimination on grounds of gender, race, disability, age and
sexuality.
Cover design by Qube Design Associates, Bristol
Front cover: image kindly supplied by www.istock.com
Printed and bound in Great Britain by CPI Group (UK) Ltd,
Croydon, CR0 4YY
Policy Press uses environmentally responsible print partners
Contents
List of tables and figures vii

Notes on contributors viii
Editors’ introduction to the series xiv
Introduction by John A. Hird 1
Part One: History, styles, and methods of policy analysis in 7

the United States
one Policy analysis in the United States 9

David L.Weimer
two The evolution of the policy analysis profession in the United States 31
Beryl A. Radin
three The argumentative turn in public policy inquiry: deliberative 55
policy analysis for usable advice
Frank Fischer
four Reflections on 50 years of policy advice in the United States 73
Laurence E. Lynn, Jr.
Part Two: Policy analysis by governments 91
five The practice and promise of policy analysis and program evaluation 93
to improve decision making within the U.S. federal government
Rebecca A. Maynard
six Policy analysis in the states 113
Gary VanLandingham
seven Policy analysis and evidence-based decision making at the local level 131
Karen Mossberger, David Swindell, Nicholet Deschine Parkhurst,
and Kuang-Ting Tai
eight Committees and legislatures 153
Philip Joyce
Part Three: Policy analysis outside of government 171
nine Policy advisory committees: an operational view 173

Michael Holland and Julia Lane
ten Public opinion and public policy in the United States 183
Saundra K. Schneider and William G. Jacoby
eleven Political parties and policy analysis 205
Zachary Albert and Raymond J. La Raja
v
Policy analysis in the United States
twelve Policy analysis by corporations and trade associations 223

Erik Godwin, Kenneth Godwin, and Scott Ainsworth
thirteen Policy analysis and the nonprofit sector 245
Steven Rathgeb Smith
fourteen The media 265
Annelise Russell and Maxwell McCombs
fifteen Think tanks and policy analysis 281
Andrew Rich
Part Four: Policy analysis education and impact internationally 295
sixteen Public policy education in the United States 297

Michael O’Hare
seventeen The status of the profession: the role of PhD and masters 319
programs in public policy education
Nadia Rubaii
eighteen The influence of policy analysis in the United States on the 339
international experience
B. Guy Peters
Index 353
vi
List of tables and figures
Tables
5.1 Selected evidence clearinghouses 2017 102
5.2 Sample snapshots of evidence on the effectiveness of education 103
programs and practices reviewed by the What Works Clearinghouse
5.3 Illustrative evidence-guided funding streams 105
6.1 Relative strengths and weaknesses of state-level policy analysis 121
organizations
7.1 Local government sources for research on new policies and 141
practices (%)
8.1 Congressional staffing, 1979–2009 154
8.2 Selected items from GAO high risk list 161
12.1 Participation levels and likelihood of success of business and citizen 228
organizations
17.1 Top 10 masters in public affairs according to USNWR, 2016 330
Figures
7.1 Frequency of policy analysis tool usage (% sometimes or often) 138
7.2 Resource utilization for decisions (% to a fair or large extent) 139
7.3 Influence on outcome of policy/management decision 140
(% to a fair or large extent)
7.4 Extent of performance data usage by purpose (% used moderately 142
or extensively)
vii
Notes on contributors
Scott H. Ainsworth is Professor and Head of Political Science in the School of

Public and International Affairs at the University of Georgia. Professor Ainsworth’s
research focuses on rational choice models of politics, policy making, and the
intersection of public institutions (Congress, courts, agencies, and so on) and
private interests. He is the co-author of Lobbying and Policymaking: The Public
Pursuit of Private Interests and Abortion Politics in Congress: Strategic Incrementalism
and Policy Change and the author of Analyzing Interest Groups: Group Influence on
People and Policies.
Zachary Albert is a PhD student in political science at the University of

Massachusetts Amherst. His research examines political parties and polarization,
campaign financing, and the development of public policy in an increasingly
partisan political landscape, often using network analysis.
Frank Fischer is Professor Emeritus of Politics and Global Affairs at Rutgers

University and is currently research scholar at Humboldt University in Berlin,
Germany. Professor Fischer’s research specialties are environmental policy and
public policy; U.S. foreign policy; public policy analysis; politics of climate change;
and scientific expertise and democratic governance. He is co-editor of Critical
Policy Studies journal published by Routledge and editor of Handbook of Public
Policy Series for Edward Elgar. Professor Fischer received a number of awards,
including the Policy Studies Organization’s Harold Lasswell award for scholarship
in the field of public policy.
Erik Godwin is a lecturer at Brown University’s Taubman Center for Public

Policy and American Institutions. He is the co-author of Lobbying and Policymaking:
The Public Pursuit of Private Interests. Godwin is also the Director of the Office of
Regulatory Reform at the state of Rhode Island’s Office of Management and
Budget and instructor at the U.S. Office of Personnel Management.
Kenneth Godwin is Professor Emeritus of Political Science and Public

Administration at the University of North Carolina Charlotte. Previously, he
was named the Marshall Rauch Distinguished Professor of Political Science at
the University of North Carolina Charlotte and he has served as the Rockefeller
Environmental Fellow at Resources for the Future. Professor Godwin is the author
or co-author of seven books concerning public policy issues and interest groups,
including Lobbying and Policymaking: The Public Pursuit of Private Interests. From
2000 to 2006, he served as the co-editor of Political Research Quarterly.
John A. Hird is Professor of Political Science and Public Policy, and Dean of
the College of Social and Behavioral Sciences, at the University of Massachusetts
viii
Amherst, where he was the founding director of its Center (now School) of
Public Policy. His research focuses on the use of expertise in policy making, and
he has authored or co-authored several books, including Superfund: The Political
Economy of Environmental Risk and Power, Knowledge, and Politics: Policy Analysis
in the States. He is currently engaged in an NSF-funded project on the use of
science and other forms of evidence in regulatory policy making.
Michael Holland is the Executive Director of the Center for Urban Science and
Progress at New York University. He has a background in research policy and the
oversight of federal research programs at the White House Offices of Management
& Budget and Science & Technology Policy, the US House of Representatives
Science Committee, and the US Department of Energy.
William G. Jacoby is Professor in the Department of Political Science at Michigan

State University, Editor of the American Journal of Political Science, and 2017
President of the Southern Political Science Association. Professor Jacoby is also
a Faculty Research Associate at the Inter-university Consortium for Political and
Social Research (ICPSR) where he is an instructor in, and the former Director
of, the ICPSR Summer Program in Quantitative Methods of Social Research.
His areas of interest include mass political behavior (public opinion and voting
behavior) and quantitative methodology (measurement theory, scaling methods,
and statistical graphics).
Philip Joyce is Senior Associate Dean and a Professor of Public Policy in the
University of Maryland’s School of Public Policy. Professor Joyce’s teaching and
research interests include public budgeting, performance measurement, and
intergovernmental relations. He is the author of The Congressional Budget Office:
Honest Numbers, Power, and Policymaking and co-author of Government Performance:
Why Management Matters and Public Budgeting Systems (9th edition). Professor
Joyce is a Fellow of the National Academy of Public Administration. He is the
recipient of several national awards, including the Aaron Wildavsky Award for
lifetime scholarship in public budgeting and finance.
Julia Lane is Professor in the Wagner School of Public Policy at New York
University. She is also a Provostial Fellow in Innovation Analytics and a Professor
in the Center for Urban Science and Policy. She also co-founded the Institute
for Research on Innovation and Science (IRIS) at the University of Michigan.
Professor Lane has published over 70 articles in leading journals, and authored or
edited 10 books, including her most recent Big Data and Social Science Research:
Theory and Practical Approaches, Privacy, Big Data, and the Public Good: Frameworks
for Engagement, and The Handbook of Science of Science Policy (co-edited with Kaye
Husbands-Fealing, Jack Marburger, Stephanie Shipp, and Bill Valdez, Stanford
University Press, 2011).
ix
Raymond J. La Raja is Professor of Political Science at the University of

Massachusetts Amherst. His research interests include political parties, interest
groups, elections, campaign finance, political participation, American state
politics, public policy, and political reform. Professor La Raja is the co-founder
and former co-editor of The Forum: A Journal of Applied Research in Contemporary
Politics and he is a member of the Academic Advisory Board of the Campaign
Finance Institute. He has co-authored or authored several books, including
Campaign Finance and Political Polarization: When Purists Prevail and Small Change:
Money, Political Parties and Campaign Finance Reform.
Laurence E. Lynn, Jr. is the Sydney Stein, Jr. Professor of Public Management
Emeritus at the University of Chicago. Professor Lynn’s most recent books are
Public Management: Old and New, Madison’s Managers: Public Administration and
the Constitution (with Anthony M. Bertelli), and a textbook, Public Management:
Thinking and Acting in Three Dimensions (with Carolyn J. Hill). For lifetime
contributions to public administration research and practice, he was selected as a
John Gaus lecturer by the American Political Science Association and is a recipient
of the Dwight Waldo and Paul Van Riper awards by the American Society for
Public Administration, the recipient of the inaugural H. George Frederickson
award by the Public Management Research Association, and the Charles H.
Levine Memorial Lecturer at American University.”
Rebecca A. Maynard is University Trustee Chair Professor of Education and

Social Policy at the University of Pennsylvania and former Commissioner for
the National Center for Education Evaluation and Regional Assistance at the
Institute of Education Sciences. Over her career, she has directed many large scale
social experiments to test policies and programs to improve health, education,
welfare, and workforce outcomes. Her current research focuses on methods for
integrating program evaluation and improvement science and on improving the
quality and utility of research syntheses.
Maxwell McCombs is the Jesse H. Jones Centennial Chair in Communication

Emeritus in the School of Journalism at the University of Texas at Austin. Professor
McCombs is internationally recognized for his research on the agenda-setting
role of mass communication, the influence of the media on the focus of public
attention. Since the original Chapel Hill study with his colleague Donald Shaw
coined the term ‘agenda setting’ in 1968, more than 400 studies of agenda setting
have been conducted worldwide. In 2011 Professors McCombs and Shaw were
awarded the Helen Dinerman Award of the World Association for Public Opinion
Research for their continuing work in this area.
Karen Mossberger is Professor of the School of Public Affairs at Arizona State

University. Her research interests include local governance, urban policy, digital
inequality, evaluation of broadband programs, and e-government. Her most recent
x
books are Digital Cities: The Internet and the Geography of Opportunity (with C.
Tolbert and W. Franko), as well as the Oxford Handbook of Urban Politics (with S.
Clarke and P. John), which includes contributions from around the globe.
Michael O’Hare is Professor of Public Policy in the Goldman School of

Public Policy at the University of California Berekley. He has been editor of
the Curriculum and Case Notes section of the Journal of Policy Analysis and
Management, is currently an associate editor of the Journal of Public Affairs Education,
and has published frequently on quality assurance and best practices in professional
teaching. Professor O’Hare’s research history has included periods of attention to
biofuels and global warming policy, environmental policy generally, including the
‘NIMBY problem’ and facility siting, arts and cultural policy, public management,
and higher education pedagogy.
Nicholet Deschine Parkhurst is a PhD student in public administration and

policy in the School of Public Affairs at Arizona State University. Parkhurt’s
research interests include American Indians, substance abuse prevention, health
disparities, tribal governance, tribal consultation, American Indian public policy,
and tribal telecom.
B. Guy Peters is the Maurice Falk Professor of American Government at

the University of Pittsburgh. He has written extensively in the areas of public
administration and public policy, both for the United States and comparatively.
Professor Peter’s most recent books include Contemporary Approaches to Public
Policy, Handbook of Public Administration in Latin America and Governance and
Comparative Politics. Professor Peters has received the Lifetime Achievement
Award from NISPACEE and the Fred Riggs Award for Lifetime Achievement
in Public Administration.
Beryl A. Radin is a member of the faculty at the Georgetown Public Policy

Institute of Georgetown University. She is Editor of the Georgetown University
Press book series Public Management and Change. Her government service
included two years as a Special Advisor to the Assistant Secretary for Management
and Budget of the US Department of Health and Human Services and other
agencies. Much of her work has focused on policy analysis, intergovernmental
relationships, and federal management change. Her most recent books are the
second edition of her book on policy analysis, Beyond Machiavelli: Policy Analysis
Reaches Midlife and Federal Management Reform In a World of Contradictions. She
has published nine other books and a wide range of articles.
Andrew Rich is the Executive Secretary of the Harry S. Truman Scholarship

Foundation. As Executive Secretary, he directs the independent federal agency
that provides merit-based Truman Scholarships to college students who plan to
attend graduate school in preparation for careers in public service. Dr. Rich is the
xi
author of Think Tanks, Public Policy, and the Politics of Expertise as well as a number
of wide-ranging articles about think tanks, interest groups, foundations, individual
donors, and the role of experts and ideas in the American policy process. Before
joining the Truman Scholarship Foundation, Dr. Rich was President and CEO
of the Roosevelt Institute and Associate Professor and Chairman of the Political
Science Department at the City College of New York (CCNY).
Nadia Rubaii is Associate Professor of Public Administration at Binghamton

University and Co-Director of the Institute for Genocide and Mass Atrocity
Prevention. Dr. Rubaii has extensive consulting experience with local
governments, nonprofit organizations, and graduate public affairs programs
around the world, with a particular emphasis on Latin America. Dr. Rubaii’s
research focuses on how local government managers can more effectively respond
to increasing diversity among employees and the population, how public affairs
programs can most effectively prepare the next generation of public service
professionals, and the opportunities and barriers to international accreditation
in public affairs.
Annelise Russell is a PhD student in Government at the University of Texas

at Austin. Her research interests include public policy within U.S. institutions,
specifically Congress and the media, with an emphasis on how new media
platforms serve elite interests. She currently serves as director of the undergraduate
program for the Policy Agendas Project, which collects and organizes data to
track changes in the national policy agenda.
Saundra K. Schneider is Professor in the Department of Political Science at

Michigan State University and the Director of the ICPSR Summer Program in
Quantitative Methods of Social Research at the University of Michigan. She
is also the 2020 President of the Southern Political Science Association. Her
current research examines the policy priorities of the American states, the role
of administrative influences in social welfare program developments, the linkages
between public opinion and public policy, and the effectiveness of governmental
responses to natural disasters. Professor Schneider is the author of Dealing with
Disaster: Public Management in Crisis Situations.
Steven Rathgeb Smith is the Executive Director of the American Political Science
Association. Previously, he was the Louis A. Bantle Chair in Public Administration
at the Maxwell School of Citizenship and Public Affairs at Syracuse University.
He also taught for many years at the University of Washington where he was the
Nancy Bell Evans Professor of Public Affairs at the Evans School of Public Affairs
and director of the Nancy Bell Evans Center for Nonprofits & Philanthropy.
Dr. Smith has authored and edited several books, including Nonprofits for Hire:
The Welfare State in the Age of Contracting, Governance and Regulation in the Third
xii
Sector: International Perspectives, and Nonprofits and Advocacy: Engaging Community

and Government in an Era of Retrenchment.
David Swindell is the Director of the Center for Urban Innovation and Associate
Professor in the School of Public Affairs at Arizona State University. His work
focuses primarily on community and economic development, especially public
financing of sports facilities, the contribution of sports facilities to the economic
development of urban space, collaborative arrangements with public, private,
and nonprofit organizations for service delivery, and citizen satisfaction and
performance measurement standards for public management and decision making.
Professor Swindell’s latest book, American Cities and the Politics of Party Conventions
(with co-authors Eric Heberlig and Suzanne Leland)will be published in 2017.
Kuang-Ting Tai is a Graduate Research Assistant at the School of Public Affairs

and Administration, Rutgers University-Newark.
Gary VanLandingham is Professor and Reubin O’D. Askew Senior Practitioner

in Residence at the Reubin O’D. Askew School of Public Administration and
Policy at Florida State University. His areas of specialization include evidence-
based policy making, program evaluation, policy analysis, performance
measurement, and Florida government. Prior to joining the Askew School, Dr.
VanLandingham was founding Director of the Pew-MacArthur Results First
Initiative, a national initiative which supports evidence-based policy making in
states and local governments, and as Director of the Florida Legislature’s Office
of Program Policy Analysis and Government Accountability.
David L. Weimer is the Edwin E. Witte Professor of Political Economy at the

La Follette School of Public Affairs at the University of Wisconsin-Madison.
His research focuses broadly on policy craft and institutional design. He has
done policy-relevant research in the areas of energy security, natural resource
policy, education, criminal justice, and research methods. Professor Weimer is the
author of Medical Governance: Values, Expertise, and Interests in Organ Transplantation
and Behavioral Economics for Cost-Benefit Analysis: Benefit Validity when Sovereign
Consumers Seem to Make Mistakes. He is co-author of Organizational Report Cards,
Policy Analysis: Concepts and Practice (6th edition) and Cost-Benefit Analysis: Concepts
and Practice (4th edition).
xiii
Editors’ introduction to the series
Professor Iris Geva-May and Professor Michael Howlett, ILPA series editors
Policy analysis is a relatively new area of social scientific inquiry, owing its origins
to developments in the US in the early 1960s. Its main rationale is systematic,
evidence-based, transparent, efficient, and implementable policymaking. This
component of policymaking is deemed key in democratic structures allowing
for accountable public policies. From the US, policy analysis has spread to other
countries, notably in Europe in the 1980s and 1990s and in Asia in the 1990s and
2000s. It has taken, respectively one to two more decades for programmes of public
policy to be established in these regions preparing cadres for policy analysis as a
profession. However, this movement has been accompanied by variations in the
kinds of analysis undertaken as US-inspired analytical and evaluative techniques
have been adapted to local traditions and circumstances, and new techniques
shaped in these settings.
In the late 1990s this led to the development of the field of comparative policy
analysis, pioneered by Iris Geva-May, who initiated and founded the Journal
of Comparative Policy Analysis, and whose mission has been advanced with
the support of editorial board members such as Laurence E. Lynn Jr., first co-
editor, Peter deLeon, Duncan McRae, David Weimer, Beryl Radin, Frans van
Nispen, Yukio Adachi, Claudia Scott, Allan Maslove and others in the US and
elsewhere. While current studies have underlined differences and similarities in
national approaches to policy analysis, the different national regimes which have
developed over the past two to three decades have not been thoroughly explored
and systematically evaluated in their entirety, examining both sub-national and
non-executive governmental organisations as well as the non-governmental sector;
nor have these prior studies allowed for either a longitudinal or a latitudinal
comparison of similar policy analysis perceptions, applications, and themes across
countries and time periods.
The International Library for Policy Analysis (ILPA) series fills this gap in the
literature and empirics of the subject. It features edited volumes created by experts
in each country, which inventory and analyse their respective policy analysis
systems. To a certain extent the series replicates the template of Policy Analysis
in Canada edited by Dobuzinskis, Howlett and Laycock (Toronto: University of
Toronto Press, 2007).
Each ILPA volume surveys the state of the art of policy analysis in governmental
and non-governmental organisations in each country using the common template
derived from the Canadian collection in order to provide for each volume in the
series comparability in terms of coverage and approach.
Each volume addresses questions such as: What do policy analysts do? What
techniques and approaches do they use? What is their influence on policymaking
in that country? Is there a policy analysis deficit? What norms and values guide
xiv
the work done by policy analysts working in different institutional settings?

Contributors focus on the sociology of policy analysis, demonstrating how analysts
working in different organisations tend to have different interests and to utilise
different techniques. The central theme of each volume includes historical works
on the origins of policy analysis in the jurisdiction concerned, and then proceeds
to investigate the nature and types, and quality, of policy analysis conducted by
governments (including different levels and orders of government). It then moves
on to examine the nature and kinds of policy analytical work and practices found
in non-governmental actors such as think tanks, interest groups, business, labour,
media, political parties, non-profits and others.
Each volume in the series aims to compare and analyse the significance of the
different styles and approaches found in each country and organisation studied,
and to understand the impact these differences have on the policy process.
Together, the volumes included in the ILPA series serve to provide the basic
data and empirical case studies required for an international dialogue in the area
of policy analysis, and an eye-opener on the nuances of policy analysis applications
and implications in national and international jurisdictions. Each volume in the
series is leading edge and has the promise to dominate its field and the textbook
market for policy analysis in the country concerned, as well as being of broad
comparative interest to markets in other countries.
The ILPA is published in association with the International Comparative Policy
Analysis Forum, and the Journal of Comparative Policy Analysis, whose mission is
to advance international comparative policy analytic studies. The editors of each
volume are leading members of this network and are the best-known scholars
in each respective country, as are the authors contributing to each volume in
their particular domain. The book series as a whole provides learning insights for
instruction and for further research in the area and constitutes a major addition
to research and pedagogy in the field of comparative policy analysis and policy
studies in general.
We welcome to the ILPA series Volume 12, Policy Analysis in the United States,
edited by John A. Hird, and thank the editor and the authors for their outstanding
contribution to this important encyclopedic database.
Iris Geva-May
Professor of Policy Studies, Baruch College at the City University of
New York, Professor Emerita Simon Fraser University; Founding President
and Editor-in-chief, International Comparative PolicyForum and Journal of
Comparative Policy Analysis
Michael Howlett
Burnaby Mountain Professor, Department of Political Science, Simon Fraser
University, and Yong Pung How Chair Professor, Lee Kuan Yew School of
Public Policy, National University of Singapore
xv
Introduction
John A. Hird
Policy analysis traverses and aims to bridge the space between truth and power.
Indeed, policy analysts are trained to be in a position to ‘speak truth to power.’
With its political origins in the United States during the progressive era, the
operational origins of policy analysis occurred in the 1950s when both truth and
power were more clearly defined and when those in power presumably were
interested in truth. Policy analysis emerged from systems analysis, which was
applied successfully to clearly delineated problems with unambiguous objective
functions.1 The success of systems analysis and related methods gave rise to an
early technocratic orientation to policy analysis, extrapolating from relatively
straightforward problems to the belief that complex systems, such as the U.S.
economy, could be fine-tuned through appropriate government intervention
guided by analysis. Good policy outcomes would follow if only ‘the best and
the brightest’ wielded the right tools. The country did not go to war, including
the ‘war on poverty,’ without thinking it would win, and a belief in analysis
bolstered that view.
This volume engages leading U.S. public policy scholars describing changes
in policy analysis in the United States over time and in various sectors as well as
projecting its future. Representing the U.S. portion of an international series on
policy analysis, the first four chapters provide an overview of policy analysis and
the profession in the United States in historical, philosophical and, in one instance,
personal terms. The next four chapters focus on policy analysis in government,
including federal, state, and local, as well as in the U.S. Congress. The following
seven chapters outline policy analysis outside of government, involving public
opinion, corporations and trade associations, the media, political parties, and
so on. The final three chapters focus on public policy education, the role of
graduate programs in public policy, and the influence of the United States on the
international experience with policy analysis. As the concluding chapter by Guy
Peters notes, the formal recognition of policy analysis as a field and profession
began in the United States, though it has improved through international exposure
and is now arguably more influential in other parts of the world.
The Vietnam war, ‘stagflation,’ the war on poverty, and Watergate diminished
confidence in government and served to remind citizens and policymakers that
expertise was insufficient to solve complex public problems. This in turn shifted
the orientation of policy analysis away from purely technocratic solutions,
engendering greater humility in the application of analysis to addressing public
problems, not necessarily solving them. The presumption at the time was that
both analysts and decision makers needed each other, and because political
leaders frequently cooperated across party lines, there was greater potential for
persuasion by analysis. The chapters by David Weimer and Beryl Radin trace the
1
lineage of policy analysis and the role of policy analysts from its origins through
contemporary use. Laurence Lynn’s chapter is a personal reflection on the
development and implementation of policy analysis in the United States over the
last 50 years, while Frank Fischer’s chapter outlines an alternative, ‘argumentative,’
role of policy analysis in policymaking. The chapters by Rebecca Maynard, Gary
VanLandingham, Karen Mossberger et al., and Philip Joyce outline the use of
policy analysis in federal and sub-national governments, while Michael Holland
and Julia Lane cover its use in policy advisory committees. This era also saw rapid
growth in schools of public policy, established to train policy analysts for agencies
throughout federal and sub-national governments.2 The chapters by Michael
O’Hare and Nadia Rubaii cover public policy analysis education.
Many became disillusioned that policy analysis in practice failed to live up to its
promise. Policy analysis often came to be viewed by policymakers as late, off-target,
incomplete, and therefore of questionable value.3 However, the presumption
was that policy analysis—if it could overcome these hurdles—would be utilized
because decision makers were thought to value analysis in weighing various policy
options.4 This argument rests on the assumption that enough voters are paying
sufficient attention to link politicians’ positions and votes with policy outcomes,
and will therefore hold them accountable. Now, however, we face the prospect
that no matter how well policy analysis performs its functions, policymakers may
simply disregard analysis without political consequence.
Contingent truth, limited trust in elites, heightened political partisanship, wealth
concentration, and technological changes have altered the role of policy analysis.
Rapid and free exchange of information was once thought to contribute to an
educated citizenry, yet the sheer volume of and ability to target information
(and misinformation and disinformation) at the individual level produces an
informationally polarized citizenry with seemingly orthogonal information
conduits. Coupled with diminished trust in large institutions—the media,
government, political parties—the curatorial role that such institutions once
played in filtering fiction from fact and offering shared societal understanding
and perspectives has faded. Significant partisanship in national policymaking
diminishes the role of analysis as well. This volume’s chapters by Saundra
Schneider and William Jacoby, Zachary Albert and Raymond La Raja, Annelise
Russell and Maxwell McCombs, Steven Rathgeb Smith, Erik Godwin et al.,
and others underscore these points. The combination of extreme partisanship in
national politics and an under-informed public suggests that the future of policy
analysis could be tenuous. If decision makers no longer need, or believe they
need, analytical insights to make decisions, and if the public does not hold them
accountable for policy outcomes, policy analysis will have limited impact. Policy
analysis needs both truth and power to succeed.
Public policy is a massive and complex weave of activities, from various
sectoral interventions—private, public, voluntary—to engagement at the local,
sub-national, national, and international levels in virtually all areas of human
activity. As with any complex system, change in one area engenders changes
2
Introduction
in another. With current (as of this writing, mid-2017) U.S. national policy
holding a decidedly oppositional stance to the use of facts and evidence, not to
mention science, in reaching policy decisions, responses have been many, from
individuals to governments. For example, scientists held hundreds of marches
across the nation on Earth Day in 2017, and reports indicate that more scientists
are becoming engaged in public policy and even running for elected office. Court
challenges to federal regulations still require confirmation that decisions are not
arbitrary and capricious, which in many cases requires establishing fidelity to
scientific research findings. More fundamentally, the locus of policy activity in
some areas, such as climate change, has shifted within the United States from
the federal government to cities and states, and internationally to other nations
taking leadership roles. Other responses are likely, including the possibility that
partisanship has reached its apex.
Because of the complexity of policy issues coupled with limited understanding
of the impacts of various interventions, policymakers will always make decisions
under conditions of uncertainty and ambiguity. Science is far better at illuminating
problems than determining solutions; climate change and obesity are but two
recent examples. Policy analysis—analyzing the disparate impacts of various
means to address public problems—is exceedingly difficult and, because it is
not scientific, has not been a research focus in the academy. Instead, as several
chapters in this volume show, various forms of policy analysis emerge not mainly
from government agencies or the scientific community but from lobbyists, trade
associations, ‘think tanks,’ corporations, and other institutions that are rarely
disinterested. Several chapters also underscore inequities in the production of
policy analysis. The chapter by Erik Godwin et al. highlights how business interests
dominate the regulatory comment process, fueled by deep pockets and analytical
expertise unmatched by citizen groups. Zachary Albert and Raymond La Raja
note how increasing partisanship throughout a candidate’s career leads to political
actors receiving information and analysis from divergent ideological sources. As
well, increasing concentration of wealth means that well-heeled donors on the
Left and Right can create ‘think tanks,’ publishing houses, and other outlets for
specific ideological causes. This creates the potential for policy analysis to tip
toward moneyed interests to the detriment of the broader public interest and can
potentially exacerbate widening inequalities. Andrew Rich notes the influence
of money on think tanks, once the home of frequently disinterested analysis and
now far more ideological and politically conservative than in the past.
Policy analysis is as important as ever in the face of ‘alternative facts’ and frequent
repudiation of evidence and science by political leaders. Disinterested analysis of
policy proposals is critical to effective governance; U.S. Congressional Budget
Office (CBO) budget scores still affect Congressional deliberations, even when
the results are not congenial to the party in power. The need to distinguish fact
from fiction and provide policymakers with dispassionate analysis is all the more
important with the cacophony of voices hawking policy ideas. As several chapters
note, recipients of policy analysis are not only elected officials, but also political
3
parties, nonprofits, and other institutions in the policy sphere. Yet paradoxically
today’s political environment, where ‘truth’ is frequently unmoored from facts and
analysis is disparaged, also challenges the future of disinterested policy analysis.
Policymakers and advocates have always cherry-picked analysis compatible with
their understanding of how to change the world, but the outright rejection of
evidence, analysis, and science should worry all proponents of democracy. The
willful rejection of facts by some national political leaders as well as the rejection
of broad scientific consensus foreclose the demand for policy analysis and serious
political debate.
Policy analysis has over the past 60 years moved from the expectation that truth
would help define if not dictate policy outcomes to the point where some in
power openly disregard evidence, analysis, and science. Whether policy analysis is
ignored, becomes another tool to advance particular interests, or retains its original
commitment to serve the broad public interest will depend on what policymakers
seek from analysis. Speaking truth only matters if those in power are listening.
Notes
1 Stokey and Zeckhauser (1978) outline many policy analytical methods taught as part of policy
analysis training. Earlier volumes include Hitch (1955), Eckstein (1958), McKean (1958), Hitch
and McKean (1965), Schlesinger (1968), Schultze (1969), and Rivlin (1971).
2 Earlier policy school graduates worked mostly in the public and nonprofit sectors. Subsequently,
and to the consternation of many concerned that the ethic of public service was being lost,
graduates increasingly worked in the private sector as well. See the chapter by Rubaii in this
volume.
3 Despite initial expectations, policy analysis rarely if ever determined policy outcomes.
Policymakers on the Left and Right have long ignored analysis when it was inconvenient,
such as President Johnson ignoring cost estimates of Medicare or Presidents Bush and Trump
repudiating the implications and even veracity of climate science. Indeed, attempts to demonstrate
the direct utilization of traditional policy analysis in driving policy decisions have found few
examples (Shulock 1999; Hird 2017).
4 Policymakers are presumed to demand policy analysis because they seek to understand the
anticipated outcomes of proposed policies since these policies will affect their constituents
(Lupia and McCubbins 1994). It does not depend on policy makers’ concern for the public
interest, though of course sometimes that is the case. Policymakers have an electoral self-interest
in understanding the impacts of proposed policies.
References
Eckstein, O. (1958) Water Resource Development: The Economics of Project Evaluation,
Cambridge, MA: Harvard University Press.
Hird, J. A. (2017) ‘‘How Effective Is Policy Analysis?’’ in L. S. Friedman (ed)
The Effectiveness of Public Policy Analysis, Berkeley, CA: University of California
Press, pp 44–84.
Hitch, C. (1955) ‘‘An Appreciation of Systems Analysis’,’ Journal of the Operations
Research Society of America 3(4): 466–81.
Hitch, C. J. and McKean, R. N. (1965) The Economics of Defense in the Nuclear
Age, New York: Atheneum.
4
Introduction
Lupia, A. and McCubbins, M. D. (1994) ‘‘Who Controls? Information and

the Structure of Legislative Decision Making’,’ Legislative Studies Quarterly 19:
361–84.
McKean, R. N. (1958) Efficiency in Government Through Systems Analysis, New
York, NY: J. Wiley.
Rivlin, A. (1971) Systematic Thinking for Social Action, Washington, DC: Brookings
Institution.
Schlesinger, J. R. (1968) Defense Planning and Budgeting, Santa Monica, CA:
RAND.
Schultze, C. (1969) The Politics and Economics of Public Spending, Washington, DC:
Brookings Institution.
Shulock, N. (1999) ‘‘The Paradox of Policy Analysis: If It Is Not Used, Why Do
We Produce So Much of It?’’ Journal of Policy Analysis and Management 18(2):
226–44.
Stokey, E. and Zeckhauser R. (1978) A Primer for Policy Analysis. New York:
W.W. Norton & Company.
5
Part One
History, styles, and methods
of policy analysis in the
United States
7
ONE

David L. Weimer
A narrow definition of policy analysis as a professional activity emphasizes client

orientation and public perspective: ‘policy analysis is client-oriented advice relevant
to public decisions and informed by social values’ (Weimer and Vining 2011,
p 24). A full assessment of the role of policy analysis in the United States requires
a somewhat broader definition that encompasses policy research by relaxing the
requirement that the analysis be for a specific individual or institutional client.
So, for purposes of this review, I define policy analysis as professionally provided
advice relevant to public decisions and informed by social values. In addition to
government employees who occupy positions specifically labeled as policy analyst,
I include other public officials, such as public managers, who devote some of their
professional time to policy analysis, as well as persons in the for- and nonprofit
sectors whose professional duties involve providing policy-relevant advice. Under
this definition many people in the United States work as policy analysts. What
factors have shaped the emergence of policy analysis as a common professional
activity in public life?1
In this introductory chapter, I identify what I believe to be the four primary
sources of demand for policy analysis in the United States: first, evidence-based
support for reform (or advocacy); second, specialized expertise to inform legislative
and executive decisions; third, a response to the increasing scope and complexity
of government involvement in society; and fourth, a mechanism for politicians
to discipline themselves and others from pursuing politically attractive policies
that do not promote the public interest. For each of these sources of demand I
briefly discuss its historical roots and sketch its implications for contemporary
practice. I also note the incentives these demands have created for higher education
to produce people with relevant policy-analytic skills.
Demand arising from the desire for reform (advocacy)

The United States has a plethora of nonprofit organizations with missions to bring
evidence to bear in specific policy areas or in support of various governmental
‘reforms.’ As one person’s desired reform may be another person’s feared folly,
perhaps advocacy would be a better term. Nonetheless, in retrospect, I believe
that reform is the appropriate term for the first such organizations, the Municipal
Research Bureaus, the first American ‘think tanks.’ Although reform remains
a generally appropriate description of the goals of contemporary think thanks,
9
increasing ideological differentiation sometimes brings competing reforms into

conflict.
By the beginning of the 20th Century, the Progressive movement had scored a
number of electoral victories in major U.S. cities. However, these victories were
often reversed in subsequent elections by local political machines and consequently
did not result in lasting improvements in terms of reduced corruption in, and
increased efficiency of, municipal governments. An alternative approach, proposed
by progressive reformers and funded originally by wealthy philanthropists, was
the creation of organizations to gather data about municipal services and propose
changes to operations that would make them more efficient and less prone to
corruption.
The first such organization, financed by the likes of Andrew Carnegie and John
D. Rockefeller, was founded in 1906 in New York (Schacter 1995). It became
the model for similar organizations in other cities (Gill 1944; Schacter 2002).
Within 10 years, the model had spread to 21 major cities under names such as
the Citizens’ Bureau in Milwaukee, the Civic Federation in Chicago, and the
Survey Commission in Atlantic City (Gill 1944). A primary tool of the bureaus
was the systematic collection of data about the operation and organization of
city governments through ‘surveys,’ which have been described as the empirical
basis for the emerging field of public administration (Bertelli and Lynn 2006).
These surveys were extensive. For example, the newly created Rochester
Bureau of Municipal Research, supported by George Eastman and other local
philanthropists, commissioned the New York Bureau of Municipal Research to
do a survey of the City of Rochester government. The resulting report of 546
pages was packed with facts about municipal operations and recommendations
to support its stated goals: ‘1. To get things done for the community through
cooperation with persons who are in office, by increasing efficiency and eliminating
waste. 2. To serve as an independent non-partisan agency for keeping citizens
informed about the city’s business’ (New York Bureau of Municipal Research 1915).
The bureaus were certainly advocates for efficiency. However, they claimed
that they were not trying to influence broad policy, but rather foreshadowing
contemporary rhetoric about evidence-based policy making, that they were
providing neutral facts and advice aimed at improving efficiency in the
administration of policy. Nonetheless, advice about how to reorganize and staff
municipal governments clearly meets the definition of policy analysis, both
narrowly focused on operational issues and broadly focused on institutional design.
Further, ‘… while the bureau movement is based upon the fact-finding
approach, bureaus must not be regarded as mere thinking machines. As the
director of the Detroit agency has said: “The bureaus think, dream, consult,
stimulate, educate, irritate and somehow change governmental outlook and
public conscience of the city if they are good”’ (Gill 1944, p 37).
In addition to the spread of private municipal bureaus, a number of cities
created their own units, usually called efficiency bureaus, to do research on
city operations (Lee 2008). One of the earliest examples was a research agency
10
established in Milwaukee by its socialist administration in 1910 (Gill 1944). The

efficiency bureaus conducted studies similar to the surveys done by municipal
bureaus and many evolved into the municipal offices that currently advise city
officials on budgets and other fiscal matters.
Expanding the perspective beyond specific cities, a number of think tanks,
now commonly defined as public policy research, analysis, and engagement
organizations (McGann 2005), were created to provide research relevant to
national policy. In 1916, the Brookings Institution began providing fact-based
research relevant to national policy issues. Subsequently, a number of other think
tanks were created, including the National Bureau of Economic Research in
1920, the American Enterprise Institute for Public Policy in 1938, Resources
for the Future in 1952, and the Urban Institute in 1968. The growth of think
tanks in the United States has accelerated over the last 20 years with the total
now exceeding 1,830 with almost 400 located in Washington, D.C. (McGann
2016, p 31).
Although many of the early think tanks were created at least in part to advocate
for some type of reform, until the 1980s most of the prominent think tanks
could be described as either universities without students or contract research
organizations, with the former providing more policy research and the latter more
policy analysis (in the narrow definition). However, since then a third type, the
advocacy tank, has proliferated.
Advocacy tanks combine a strong policy, partisan or ideological

bent with aggressive salesmanship and an effort to influence current
policy debates. Advocacy tanks synthesize and put a distinctive ‘spin’
on existing research rather than carrying out original research. What
may be lacking in scholarship is made up for in their accessibility to
policymakers (Weaver 1989, p 567).
Advocacy tends to be along identifiable ideological lines (McGann 2005,

pp 11–13): conservative (for example Heritage Foundation, American Enterprise
Institute, Manhattan Institute), libertarian (for example Cato Institute, Reason
Foundation), center-Right (for example Center for Strategic and International
Studies), centrist (for example Resources for the Future, National Bureau for
Economic Research, Public Policy Institute of California), center-Left (for
example Urban Institute, Brookings Institution), Left (for example Center on
Budget and Policy Priorities, Center for American Progress).
The growth and increased public and political prominence of think tanks have
both positive and negative implications for policy analysis. On the positive side
of the ledger, they provide venues for policy research and analysis more directly
relevant to public policy issues than is often the case with academic policy research,
which is usually driven by disciplinary rewards for generality rather than relevance
to public policy decisions. Their sheer number also creates many opportunities
for policy analysts to work outside of the constraints imposed by government
11
employment. Their incentives to please donors and gain favorable media coverage
encourage them to invest in translating research and analysis into information
directly relevant to policy decisions and disseminating it widely. On the negative
side of the ledger, ideological orientations not adequately disciplined by widely
accepted research standards may undercut the credibility of analysis generally
and lead to ‘dueling studies’ rather than a search for the best available answers to
policy-relevant questions. Concern about the reaction of potential donors may
also affect think tank agendas. It has even been suggested that the dominance of
think tanks at the ‘crossroads of the academic, political, economic, and media
spheres’ has ‘undermined the value of independently produced knowledge ... by
institutionalizing a mode of intellectual practice that relegates its producers to
the margins of public and political life’ (Medvetz 2012, p 7). However one tallies
the ledger, think tanks certainly play an important role in policy analysis in the
contemporary United States.
Demand arising from the need for expertise

From its very beginning, the executive branch of the federal government sought
advice from committees of citizens—President Washington convened the Whiskey
Rebellion Commission, among others, to supplement the meager capacity of
the fledgling executive branch for gathering and applying information to policy
problems (Smith 1992, p 14). As early as 1842, Congress attempted to limit
expenditures on the numerous advisory committees that had been created, but
use of advisory committees continued to grow in step with the expanding role
played by the federal government in the economy (Croley and Funk 1997). By
the time the current framework requiring that advisory committee membership
be balanced, their functioning transparent, and their durations limited was put in
place by the Federal Advisory Committee Act (P.L. 92-463) in 1972, the federal
government had almost 1,500 advisory committees. The number had fallen to
approximately 800 by 1978 (Petracca 1986, p 85) but had grown to over 900 by
2008 (U.S. GAO (Government Accountability Office) 2008, p 1) and currently
exceeds 1,000 (U.S. GSA [Government Services Administration] 2015).
A driving force for the use of advisory committees was the difficulty of
maintaining specialized expertise within the bureaucracy, initially because of
its limited capacity overall and later because civil service rules made hiring and
firing more difficult and therefore less amenable to adding personnel with needed
expertise quickly. The need for such expertise was especially urgent during
wartime. In 1863, during the Civil War, Congress created the National Academy
of Science at the request of President Lincoln to advise federal departments
on scientific matters. The increasing demand for scientific expertise during
World War I led to the creation of the National Research Council to provide
administrative infrastructure for the National Academy of Science in 1916.
Increasingly, the National Research Council, which now serves the National
Academy of Science, the National Academy of Engineering, and the Institute of
12
Medicine, undertakes projects requested by Congress as well as by administrative

agencies (Morgan and Peha 2003). These projects often involve the production
of policy analyses that assess social problems and compare possible alternative
policies for addressing them.
World War II created demand not just for expertise drawn directly from the
natural sciences, but also for expertise on how to use scarce resources more
efficiently both across and within organizations. This expertise, drawing on
mathematics, statistics, and the new game theory, eventually became known
as operations research. Motivated by the analytical successes of the British
‘Blackett’s circus’ during the Battle of Britain, operations research units were set
up throughout the U.S. military (Fortun and Schweber 1993). They worked on
topics ranging from intelligence questions, such as how to use serial numbers on
captured equipment to estimate enemy rates of production (Champkin 2013),
to operational issues, such as the size of convoys that would maximize the ratio
of U-boat losses to merchant ship losses (Hitch 1953).
The U.S. military maintained capacity in operations research following the
War. It also became the primary client for newly established contract research
organizations such as the RAND Corporation. Analysts in the Pentagon and
the RAND Corporation were resources for the implementation of Program
Planning and Budgeting Systems (PPBS) in the Department of Defense, which
focused attention on programs rather than categories of inputs. For example,
rather than focusing budgeting on categories of expenditures, such as personnel
and equipment aggregated across programs, PPBS sought to organize budgets
by purposes, such as air defense or logistical support. It was initiated in 1961 by
Secretary Robert McNamara, who had been an operations researcher during
the War and later President of the Ford Motor Company. The apparent success
of PPBS in the Defense Department led President Johnson in 1965 to ask all
agency heads to implement it. It had at best modest success, largely because
the circumstances that yielded benefits in the Defense Department, including
analytical capacity, relevant information, and clear goals, generally were not present
in most other federal agencies at the time (Wildavsky 1969). Nonetheless, failed
efforts to implement PPBS and its successors, such as Zero-Based Budgeting
during the Carter Administration, required the addition of technically trained
personnel, some of whom outlasted these top-down initiatives and contributed
to the development of analytical capacity within agencies in the longer run. (See
also the chapter by Lynn in this volume.)
Several factors encouraged state and local governments to adopt PPBS (Harty
1971). The Ford Foundation funded the George Washington University State-
Local Finances Project, which worked with five states, five counties, and five
cities to assess the feasibility and desirability of adopting PPBS. This effort was
subsequently supported by the Department of Housing and Urban Development
and encouraged through incentives in federal legislation. As was the case at the
federal level, these efforts generally did not implement the full PPBS model
and appear to have produced modest improvements in budgeting processes but
13
relatively little increase in general capacity in policy analysis (Harty 1971, Exhibit
1).
Another sort of expertise began to be demanded by the federal government
in the late 1960s as those implementing President Johnson’s Great Society
searched for ways to address poverty. Large-scale social experiments, with
random assignment of participants into treatment and control groups, offered
the possibility of assessing whether programs could produce desired outcomes.
Beginning with the New Jersey Income Maintenance Experiment in 1968, federal
agencies have funded contract research organizations to conduct over 240 major
social experiments (Greenberg and Schroder 2004). These experiments require
expertise in research design, implementation, and statistical analysis, within both
the contracted organizations and the bureaus that conceive, select, fund, and
monitor them. Beyond the specific information these experiments created, they
also contributed to the general capacity of the federal government to do and
support policy research.
The New Deal greatly expanded the role of the federal government in the
economy, both through programs implemented by existing federal bureaus, such
as the Department of Agriculture, and newly created agencies, and through
increased economic regulation by the Securities and Exchange Commission
(SEC), the organization that replaced the Bureau of Corporations, the Federal
Trade Commission (FTC), and the Federal Communications Commission (FCC).
In the 1970s, the federal role in health and safety, so-called social regulation,
was expanded through the creation of the Occupational Safety and Health
Administration (OSHA), the Consumer Product Safety Commission (CPSC),
and the Environmental Protection Agency (EPA). The federal website (www.
regulations.gov), created to increase the transparency of rulemaking, now carries
notices, proposed rules, final rules, and other materials related to rulemaking for
approximately 300 federal agencies.
The expanded federal role required personnel to fill in the details of laws
passed by Congress and executive orders issued by the president. The expansion
of federal rulemaking required agencies to add personnel with substantive and
legal expertise to design and justify rules. These regulators often served as policy
analysts, diagnosing problems with existing rules, proposing alternatives, and
assessing the alternatives in terms of legislative (and agency) goals. Although the
Administrative Procedure Act of 1946 (P.L.79-404) regularized the process of
notifying potentially interested parties through the general requirement for the
publication of proposed rules in the Federal Register and accepting the comments
of these stakeholders before the finalization of rules, much of the analysis of
proposed rules, especially those proposed by agencies rather than independent
commissions, was and is typically done in camera, or, when open, influenced by
informal and idiosyncratic public participation (West 2009; Yackee 2012). This
lack of transparency tends to obscure the analysis that underlies public proposals.
Nonetheless, most rules are public policies that require at least minimal analysis
if only in descriptive comparison with current policy.
14
The American states also regulate. Indeed, they created agencies to regulate
the professions, corporations, the insurance industry, public utilities, and some
natural resources during a period when the federal role was limited largely to
the regulation of railroads by the Interstate Commerce Commission. Although
federal regulation has preempted the states in areas such as transportation and
finance, states continue to regulate but they vary considerably in terms of the
professional resources they make available for the task (Teske 2004). Often their
work is supported by national associations, such as the National Association of
Insurance Commissioners, that provide policy research relevant to the work of
the state regulators.
In addition to augmenting in-house expertise through the use of advisory
committees, many federal and some state agencies extend their analytical
capabilities through contracts with consulting firms. The contracts allow agencies
to procure the services of variously skilled personnel on a temporary basis. At one
extreme, the contracted personnel may work independently to deliver a specific
product, such as a policy analysis or a database. At the other extreme, contracted
personnel may perform ongoing tasks as if they were part of the agency staff. In
either case, the writing and monitoring of effective contracts requires some in-
house expertise. Indeed, an irony of such privatization of analysis is that, for it to
be effective, agencies may have to develop advanced expertise not only in contract
management, but also in the substance of the issue (Vining and Weimer 1990).
In addition to the need to secure specialized expertise quickly, occasional agency
downsizing and routinely strict limits on numbers of employees have also driven
contracting for analysis. In periods of retrenchment, agencies often find it easier,
or at least faster, to contract for expertise than to retrain employees, especially
when confronted with new issues or policy responsibilities.
The increasing use of such contracting, variously referred to as ‘third-party
government, the hollow state, the shadow bureaucracy, the blended public
workforce, articulated chains of third parties, public programs and their partners,
steering rather than rowing,’ raises general concerns about accountability within
U.S. public administration (Dubnick and Frederickson 2010, p i143). Specifically
with respect to policy analysis, it means that the content of many analyses
nominally done by government agencies may actually have been produced by
private entities. It also means that any counts of policy analysts within the federal
and state bureaucracies would likely grossly underestimate the number of people
actually doing policy analysis on their behalf.
In a few cases, Congress has delegated analytical responsibility to nongovernmental
actors. These delegations tend to be in policy areas, such as health, characterized
by rapid technological or scientific advances and substantial complexity, which
places a premium on the tacit knowledge of those actually working in the area.
For example, Congress gave the Advisory Committee on Immunization Practices
authority to specify the vaccines covered by the Vaccines for Children Program
through which the Centers for Disease Control and Prevention make bulk
15
purchases of vaccines for distribution to state and local public health agencies
(Weimer 2010a).
The difficulty civil service systems face in maintaining specialized, cutting-
edge expertise within the bureaucracy provides a rationale for delegation. For
policies from which there will be identifiable losers, such as children who have
negative side-effects from vaccination or the allocation of very scarce transplant
organs, legislators may seek to insulate themselves from demands by delegating
policy design to groups of professionals and other stakeholders rather than to a
bureau that constituents would expect their representatives to influence on their
behalf (Weimer 2006). A prominent example of such stakeholder rulemaking
is the Organ Procurement and Transplantation Network (OPTN), which has
responsibility for developing the content of rules with literally life-and-death
consequences for the procurement and allocation of cadaveric organs. The rules
must be issued by the Department of Health and Human Services to have the force
of law. However, no substantive rules have ever been issued by the department.
Instead, the rules gain force because federal funding requires transplant centers
to be members of the network in good standing to receive Medicare funding,
which in turn requires them to adhere to the rules. Participants in the OPTN
benefit from a near-universal longitudinal database on all entrants to the transplant
waiting lists as well as on their own tacit knowledge about transplantation to
analyze rules continually and to change them incrementally (Weimer 2010b).
The increasing demands for specialized expertise have resulted in many
individuals and organizations outside of government participating in policy
analysis through advisory committees, contracts, and delegated responsibilities.
This participation has been both a substitute for and a complement to analytical
capacity and production within governments. At least at the federal level, on net
it almost certainly has contributed to an increase in the quantity and technical
sophistication of policy analysis done on a routine basis by government agencies
over the last 50 years.
Demand arising from the need to manage complexity

The expanding role of government in the economy created a demand not only
for specialized expertise, but also for analysts to support management of the
concurrent complexity of public policy. Both legislators and executives required
information to construct and execute increasingly complicated budgets. New
programs often gave considerable discretion to public administrators, who required
them to consider alternatives for implementation. As dramatically evidenced
recently by the Patient Protection and Affordable Care Act 2010 (P.L. 111-148),
complexity continues to arise from the shared governance required when the
federal government attempts to induce change in policies under the direct purview
of the states (Haeder and Weimer 2013).
16
General analytical support for executives and legislatures
The modern era of efforts to create general analytical capacity within the federal
government followed the revelation of limited administrative capacity during
World War I. In 1921 Congress created twin agencies, the Bureau of the Budget
(now the Office of Management and Budget), and the General Accounting Office
(now the Government Accountability Office) to provide neutral and competent
support for improving federal fiscal policy (Mosher 1984).
The Bureau of the Budget originally focused exclusively on reigning in
agency budgets, a sometimes politically difficult task when agencies had strong
congressional constituencies, but later developed roles as ‘referee’ and ‘whipping
boy.’ The former involved resolving jurisdictional disputes among agencies; the
latter involved cutting earmarked funds of questionable value that were inserted
in the budget to please particular constituencies—members of Congress could
get credit for the insertions and place blame on the Bureau of the Budget for the
removals (Wolf 1999). In 1970 the Bureau of the Budget was reconstituted as
the Office of Management and Budget. An increase in the number of political
appointees undercut its reputation for neutrality but did result in it playing a
greater analytical role in policy development for successive administrations. It
currently employs about 470 people.
The General Accounting Office (GAO) initially served as an auditor of the
financial transactions of agencies. The immense volume of such transactions
during World War II led it to refocus on agency procedures and practices,
providing Congress with information about agency operations that would
otherwise be opaque, typically through publicly available policy-analytic reports
that included recommendations. Congress reinforced the policy analysis function
of the GAO with the Legislative Reorganization Act of 1970 (P.L. 91-510), which
calls for the GAO to conduct program evaluations and cost-benefit analyses. The
office currently employs almost 3,000 people, many of whom do policy analysis.
In a typical year, it makes almost 1,500 specific recommendations to improve
government operations—it reports that 79% of the recommendations it made in
fiscal year 2011 were implemented by the end of fiscal year 2015 (U.S. GAO 2016,
p 31). Although the GAO certainly has been influential in many policy areas, its
quantitative claims of efficacy may be too optimistic because recommendations
often come from the agencies whose programs are being evaluated.
The Legislative Reorganization Act also gave more independence and greater
responsibility for providing policy analysis in support of congressional committees
to the Legislative Research Service of the Library of Congress and renamed it
the Congressional Research Service (Kravitz 1990). The Congressional Research
Service (CRS), which employs about 600 people, conducts analyses for both
congressional committees and individual members of Congress. Its relatively low
profile but high level of service to members of Congress has contributed to its
stability, which in turn enables it to serve as a repository of ‘institutional memory.’
In fiscal year 2015, it reported that its staff responded to ‘62,000 individual
17
requests; hosted over 7,400 congressional participants at seminars, briefings and

trainings; provided over 3,600 new or refreshed products; and summarized over
8,000 pieces of legislation’ (U.S. CRS (Congressional Research Service) 2016,
p iii). Unlike the GAO, the CRS generally does not make recommendations
and only selectively makes its reports public.
Battles with the Nixon Administration over the impoundment of congressionally
appropriated funds led Congress to strengthen its capacity for analyzing budgets.
It created the Congressional Budget Office (CBO) through the Congressional
Budget and Impoundment Control Act of 1974 (P.L. 94-344). The CBO is
organized into eight divisions. Of these, by far the largest is the Budget Analysis
Division, which analyzes the president’s budget and scores the fiscal implications
of proposed legislation, among other tasks. Of the approximately 235 employees,
over half hold positions with the formal title of analyst, with most focusing on
budgetary and fiscal impacts of policies.
A fourth organization, the Office of Technology Assessment (OTA), provided
Congress with analyses of issues with major technological or scientific components.
Created in 1972 (P.L. 92-484), it was eliminated in 1995 along with reductions
made to the staffs of all of the congressional support agencies after Republicans
gained a majority in the House of Representatives after many years in the minority
and wished to signal their commitment to deficit reduction. One factor in the
demise of the OTA may have been that Congress could easily obtain substitute
analyses from universities, think tanks, and professional societies; another factor
may have been that its analyses, often done by analysts without deep connections
to those working in substantive areas, sometimes challenged the preferred policies
of powerful interests, undercutting the support from representatives sympathetic
to these interests who did not rely on the OTA for the sorts of ongoing support
provided by the other three congressional support organizations (Bimber 1996).
It appears that the GAO, the CRS, and the CBO have succeeded in, and
established reputations for, providing analyses that embody neutral competence,
which Kaufman (1956, p 1060) defined as the ‘ability to do the work of
government expertly, and to do it according to explicit, objective standards
rather than to personal or party or other obligations and loyalties.’ Unlike the
administrative agencies that bear responsibility for advancing the president’s policy
agenda, these congressional agencies are designed to serve both the majority and
minority parties in each chamber. As the control of the chambers can change
as often as every two years, neutrality, and the perception of neutrality, reduces
the chances that changes of partisan control will threaten the existence of the
agencies. The perception of neutrality is furthered by stable professional staffs and
by the appointment of nonpartisan heads with long terms (Weimer 2005). For
example, the Director of the CBO is appointed for a term of four years and the
Comptroller General, who heads the GAO, is appointed for a term of 15 years.
Although the Director of the CRS does not have a fixed term, appointment is
made by the Librarian of Congress in consultation with the Joint Committee
on the Library rather than directly by congressional leadership (Brudnick 2013).
18
These organizations contribute to the policy analysis that ‘… clearly does flow
through congressional communication networks’ (Whiteman 1995, p 181).
The states have created agencies to support legislative work analogous to those
serving the federal government. In some cases the states actually led the way. The
first professional, nonpartisan organization to provide drafting and research support
to a state legislature was the Wisconsin Legislative Reference Bureau, which was
created in 1901. Only a few states followed during the next 40 years, but in 1941
California created the Legislative Analyst’s Office, which has substantial analytical
capacity and served as a model for the creation of similar offices in other states
as well as for the CBO. In addition to the sorts of services provided by the CBO
and the CRS, it also provides institutional memory relevant to policy debates, a
function that has increased in importance with the adoption of term limits for
legislators (Radin 2013). Based on information provided by the Council of State
Governments and the National Conference of State Legislatures, Hird (2005)
identified 105 nonpartisan agencies providing analytical and research support to
state legislatures. Of 80 agencies responding to questions about their activities,
81% reported policy analysis as a major activity.
A particularly influential organization supporting policy making by a state
legislature is the nonpartisan Washington State Institute for Public Policy (WSIPP).
Beginning with its efforts to identify and assess alternatives to prison expansion,
WSIPP has conducted numerous meta-analyses of evaluations of various types
of programs and uses the estimated effect sizes to conduct sophisticated and
influential cost-benefit analyses of alternative portfolios of programs for the
Washington State legislature (Vining and Weimer 2009). Over the last four years,
23 states have adapted for their own use the WSIPP tool through the Results First
Initiative,2 a joint effort of the MacArthur Foundation and the Pew Charitable
Trusts. The initiative helps states organize stakeholders as a constituency for the
use of cost-benefit analysis and provides technical assistance for implementing
the WSIPP tool with state-specific data.
Some larger cities also have created nonpartisan organizations that provide
analysis across the range of substantive issues. The most prominent example is the
Independent Budget Office of the City of New York. Created in 1989 during a
major revision of the city charter, it produces issue briefs and other policy analyses
in addition to carrying out its responsibilities for fiscal forecasting and providing
assessments of the mayor’s preliminary and final budgets. Although a general
assessment of the prevalence of such agencies at the local government level is
not available, it does appear that the related activity of performance measurement
is pervasive in local government departments (Melkers and Willoughby 2005).
Analytical support within agencies
All the major federal agencies have offices that provide policy analysis in support
of the agencies’ missions. Prior to World War II, agency heads generally relied
on their legal staffs for agency-wide policy analysis; subsequently, they began
19
creating offices that specialized in providing policy analysis. For example, during
the 1930s the Secretary of the Interior relied on the Office of the Solicitor for
policy analysis; it was not until 1947 that an office with departmental-wide
responsibility for policy analysis was created and eventually became the Office of
Policy Analysis in 1974 (Nelson 1989).
The coordinating role at the department level is especially important. For
instance, consider the Office of the Assistant Secretary for Planning and Evaluation
in the large and complex Department of Health and Human Services. Analysts
play a role in policy development, including the construction of alternatives and
the assessment of their consequences. In addition to also playing roles in oversight
and research, other obviously analytical roles, and ‘firefighting’—the immediate
response to requests for information relevant to urgent issues—guidance for
new employees identifies the role of ‘desk officer,’ which involves coordinating
policy relevant to particular programmatic areas (Assistant Secretary for Policy
and Evaluation n.d.). As a long-time member of the staff reflected: ‘Coordination
with other staff analysts in the budget office, the legislative office, and other staff
offices is critical. Assistant secretaries don’t like to be a lone voice of dissent on
a policy issue’ (Greenberg 2003). In many circumstances public managers must
play the role of policy analyst in choosing among alternatives for implementing
policies to be effective in pursuing the missions of their agencies; in the central
analytical offices of major departments, it appears that analysts must play the role of
manager in coordinating with other analysts and program personnel to be effective.
Policy analysis staffs carrying out functions similar to those of analysts in federal
agencies can also be found in most large state agencies. The extensive use of for-
and nonprofit organizations to deliver social services means that policy-relevant
information and even analytical capacity are often distributed within policy
networks consisting of diverse organizations sharing some common goal (Milward
and Provan 2000). One can speculate that the role of ‘desk officer’ for analysts
in state agencies has grown in importance along with the increase in provision
of social services by other organizations. One should also recognize that some
organizations within the network may be primary sources for, or contributors
to, policy analysis.
Policy networks may also play a role in shaping the process of policy diffusion
across multiple jurisdictions by providing analysis that can be used by advocates
during consideration of adoption and by administrators during implementation.
For example, policy-relevant information produced by local, state, and national
organizations has played an important role in promoting the diffusion of drug
courts, which seek to break the link between substance abuse and criminal
behavior through greater reliance on treatment rather than punishment. A private
organization, the National Association of Drug Court Professionals, championed
drug courts through information sharing, which has led to their adoption in over
2,000 jurisdictions since the feasibility of the concept was first demonstrated in
1989 (Hale 2011).
20
Politicians wishing to tie their own (or their colleagues’) hands
The imposition of procedural rules on decision processes by politicians has been

viewed normatively as a way of ensuring fairness (Stewart 1975) and positively
as a way of controlling bureaus through the forced generation of information
and the enfranchisement of specific constituencies (McCubbins et al.. 1987). In
the case of requirements for the routine production of policy analysis, politicians
can also be thought of as trying to counter future incentives to support policies
favored by influential stakeholders but not necessarily in the public interest. They
may be willing to tie their own hands and, perhaps more importantly from their
perspective, the hands of their future colleagues in this way because of uncertainty
about what policies will be considered in the future. Whereas the political cost
of forgoing a constituent benefit in the present is likely to be high, the perceived
political cost of adopting a rule to make such constituent deference in the future
more difficult may be low behind a ‘veil of ignorance’ about what future demands
will actually materialize (Weimer 1992). In other words, politicians who favor
analysis in general may oppose the conduct of ad hoc analyses in specific cases when
they anticipate the possibility that the analyses will not support their preferred
policies. Nonetheless, they may accept the risk that analyses in the future will
sometimes not support their preferred policies because the analyses will also help
them prevent policies that they oppose.
Procedural rules requiring analysis in specified circumstances have been adopted
by both Congress and the president. The environmental impact statement process
is a prominent example of a congressionally imposed rule requiring analysis. It
fits most obviously within the conventional view of procedural requirements as
the empowerment of constituencies, but can also be viewed as Congress setting
the status quo of required analysis that would have to be explicitly overruled in
future legislation. A more mundane example is the routine requirement of cost
estimates for pending legislation.
Under the National Environmental Policy Act (PL 91-190) adopted in 1970,
federal agencies must prepare environmental impact statements before taking
any action that would significantly affect the human environment. Although
environmental impact statements typically require substantial analytical
resources, they have been found to have a number of common deficiencies that
undercut their value as policy analyses, including failure to compare alternatives
and consideration of only a narrow scope of impacts (Gregory et al. 1992).
Nonetheless, the preparation of environmental impact statements involves analysts
in agencies and consulting firms as well as in the interest groups that choose to
mount challenges to them during administrative processes or in court.
Congress sought to tie its own hands in the consideration of water resource
projects, which historically were one of the most common ingredients in ‘pork
barrel’ legislation. Early in the last century Congress began requiring predictions
of costs and benefits for proposed projects, and by mid-century it required the
Army Corps of Engineers to conduct cost-benefit analyses (Hammond 1966).
21
The imposition of the hurdle of having to show positive net benefits made it
somewhat easier to weed out the worst projects. The process became sufficiently
important that the Bureau of the Budget compiled guidelines on the application
of cost-benefit analysis to water and related land projects (Bureau of the Budget
1952).
Cost-benefit analysis has since become an important part of agency rulemaking.
It evolved out of a series of executive actions intended to force agencies to consider
the likely consequences of their proposed rules (Copeland 2011). President Nixon
created Quality of Life reviews of rules dealing with the environment, consumer
and occupational safety, and public health that were overseen by the Office of
Management and Budget (Shultz 1971); President Ford required agencies to
consider the impacts of rules on inflation (Executive Order 11821); and President
Carter required agencies to conduct cost-effectiveness analyses of proposed rules
relative to plausible alternatives (Executive Order 12044).
The application of cost-benefit analysis to all major agency rules was initiated by
President Reagan in 1981 (Executive Order 12291). A proposed rule involving
more than $100 million in annual costs, benefits, or transfers requires an assessment
of its costs and benefits through a Regulatory Impact Analysis (RIA), which is
subject to review by the Office of Management and Budget. RIAs, which often
are hundreds or more pages in length, are summarized in proposed rules published
in the Federal Register.
Subsequent presidents have maintained the requirement for cost-benefit analysis
while extending the scope and analytical requirement of the RIA. President
Clinton expanded the rules subject to analysis by creating the category of
economically significant rules to include rules that potentially affect in a material
way a sector of the economy, productivity, competition, employment, the
environment, public health or safety, or state and local governments (Executive
Order 12866). President G. W. Bush’s Office of Management and Budget (OMB)
and Council of Economic Advisors formalized OMB guidance for conducting
cost-benefit analyses, including the requirement that a reasonable number of
alternatives to the proposed rule be considered (OMB (Office of Management
and Budget) 2003). President Obama expanded the scope of cost-benefit analysis
of rules by directing agencies to conduct retrospective assessments to determine if
continuation of rules in place was economically justified (Executive Order 13563)
and to submit their plans for reviews to the OMB (Executive Order 13610).
As executive orders apply only to administrative agencies, independent
regulatory commissions (those agencies that adopt rules by majority voting of
commissioners appointed for fixed terms) such as the FCC and the SEC are not
required to conduct RIAs. Congress sought to extend the RIA requirement to
the independent regulatory commissions through a provision of the Contract with
America Advancement Act (PL 104-121) by requiring that they file summaries
of the analyses with the Comptroller General. It appears, however, that these
summaries are pro forma and largely uninformative about how cost-benefit
analyses are conducted (Sherwin 2006). President Obama urged the independent
22
regulatory commissions to conduct the analyses mandated for other agencies

(Executive Order 13579).
Congress has imposed other analytical requirements on rulemaking. The
Regulatory Flexibility Act (P.L. 96-254) seeks to force agencies to anticipate
the effects of rules on small entities such as businesses with few employees and
local governments serving small populations. The Paperwork Reduction Act (P.L
104-13) requires agencies to justify the collection of data from the public. The
Unfunded Mandates Reform Act (P.L. 104-4) nominally requires analysis of rules
that involve costly mandates on the private sector, state or local governments,
or Native American tribes. However, because its conditions overlap with RIA
requirements, it generally does not require substantial additional analysis in support
of proposed rules.
Compliance with procedural requirements for analysis forces the development
of analytical capacity within government. To the extent that the analyses
become influential, either through informing policy choice or justifying it,
various interests outside of government have an incentive to invest at least in
the capability for critiquing these analyses. In this way, analysis contributes to
democratic participation, perhaps more so than to immediate problem solving
(Shulock 1999).
Supply response to policy analysis demands

Prior to the 1970s those entering careers as policy analysts did so with training
from many different fields of study, including public administration, law,
substantively oriented professional programs such as urban planning and public
health, operations research and statistics, business, and social science disciplines.
Especially within public administration and the substantively oriented professions,
specializations in policy analysis developed to provide training more oriented
toward problem solving.
Beginning in the late 1960s, graduate programs designed to provide preparation
specifically for careers in policy analysis began to be established at prominent U.S.
universities. These programs, which were created as alternatives to traditional
public administration programs that focused primarily on ‘the institutional and
operational aspects of government’ (Engelbert 1977, pp 228–9), emphasized
interdisciplinary approaches to policy design and implementation. The movement
toward stand-alone programs in public policy benefitted from Ford Foundation
grants awarded in 1973 to programs located in seven major universities and
the RAND Corporation. The foundation decided ‘to bypass traditional public
administration schools and look to new public policy programs—already
underway,’ all of which displayed a conviction that ‘a blend of social sciences can
be fashioned to educate people to analyze and manage complex problems’ (Ford
Foundation 1976, p 5).
The difference between public administration and policy analysis programs
diminished substantially over time. The increasing inclusion of policy analysis
23
training within public administration schools resulted in its recognition by the

National Association of Schools of Public Affairs and Administration as a major
subject area in 1974 (Engelbert 1977). At the same time, two factors pushed
the new policy analysis schools to be more like public administration programs.
First, the faculty of the schools recognized the importance of implementation
in determining the extent to which policies actually produce their intended
outcomes. Organizational processes and leadership, which are traditional topics
within public administration, are critical to implementation—administrators
must often fill in details of policy designs as they put them in place, determine
options for change and their likely consequences, and take the initiative to change
existing policies when they do not appear to be yielding desirable outcomes.
Second, the university incentives that many programs faced led them to introduce
public management concentrations to expand their student markets beyond the
relatively small numbers of students both interested in policy analysis careers and
quantitatively prepared to develop technical skills demanded in many policy
analysis positions.
The new schools of public policy, along with several major research organizations,
founded the Association for Public Policy Analysis and Management in 1979 to
promote the practice and teaching of policy analysis and public management.
Currently among its institutional members are over 70 training programs based
in U.S. universities. It sponsors the Journal of Policy Analysis and Management and
holds annual research conferences that attract both academics and practitioners.
A number of other organizations, including the Society for Risk Analysis and the
Society for Benefit-Cost Analysis, also seek to engage both those who conduct
and those who teach specific types of policy analysis.
Conclusion
The four demands for policy analysis will not abate. Indeed, demands arising
from the need for expertise and the management of public sector complexity can
reasonably be expected to continue to increase. Further, the analytical capacity
arising from these demands has been embedded in both governmental and non-
governmental organizations. Therefore, the prevalence of policy analysis in the
United States is unlikely to decrease and may very well continue to increase. Yet,
in an environment of crass public discourse and heightened partisan polarization,
one can question whether policy analysis actually contributes to public policies
that improve society. Policy analysis abounds, but to what avail?
Governments routinely face choices about how to use scarce resources. Often
these choices do not involve great political controversy. We may disagree over
whether refuse should be collected by public employees working for a government
agency or by private employees working for a private contractor. We may even
disagree about what weights should be placed on the appropriateness, reliability,
and cost-effectiveness of services. Nonetheless, we would likely value analysis
that helped us understand the tradeoffs in things we value across the various
24
alternatives. In some cases, despite our differences in party or ideology, we may

want even very important decisions to be informed by neutrally competent
analysts—my guess is that most of us want decisions about how to minimize
the harm caused by infectious diseases to be based on epidemiological expertise
rather than purely ideological predisposition.
Although there seems to be a greater willingness in today’s polarized
environment to challenge official statistics and predictions that previously were
taken at face value, there also continues to be a willingness to take the products
of some analyses as the basis for common assumptions in budgeting and other
policy making. Projections and cost estimates by legislative support agencies
like the CBO may be accepted because of the credibility or consistency of their
source, or simply because without some common assumptions policy making
would just be too chaotic.
Analyses that must be done during the federal rulemaking process, the primary
source of the detailed substantive content of public policy in the United States,
almost certainly are not the primary basis for the choice of rules. Nonetheless,
the analyses probably improve the social value of adopted rules overall by forcing
consideration of alternatives and by helping to weed out those with large negative
net benefits. The capacity developed by agencies to comply with analytical
requirements may also have spillover effects to other agency tasks, such as the
identification of problems deserving attention.
Many politicians have embraced the rhetoric of the value of evidence-based
policy making. Of course, good policy analysis is evidence-based whenever
relevant evidence exists or can be assembled within the constraints under which
the analysis must be produced. Fortunately, the large policy research enterprise
in the United States generates considerable evidence relevant to policy through
experimental and quasi-experimental program evaluations, natural experiments,
and creative observational studies. The digital revolution has already reduced the
costs of finding evidence, and ‘big data’ may yet provide new ways of creating
it. When relevant evidence does exist, the rhetoric of evidence-based policy
making may provide a bridge over ideological differences. This may be happening
today with respect to criminal justice policy. Both conservatives and liberals
share the goal of reducing crime. However, conservatives have tended to favor
the punishment goals of isolation, deterrence, and retribution, while liberals
have tended to favor rehabilitation. However, increasing incarceration costs have
made conservatives in many states open to consideration of alternatives to long
prison terms; continuing recidivism has made liberals more concerned about
how increasingly scarce program funds are used. Identifying evidence-based
interventions that appear to reduce both correctional costs and recidivism may
be policies that can gain support across the ideological spectrum.
Technological change has dramatically changed the availability of data and
information since policy analysis first arose as a recognized profession. Early
practitioners and students spent time physically searching for information in
libraries, archives, and visits to organizations. The telephone was often the
25
most valuable tool for gathering information needed in short-term analyses.

Computation was costly, often involving decks of punch cards submitted to
mainframe computers. Those trained and practicing today have ready access to
amazing amounts of data and information through the Internet and World Wide
Web. They also have personal computers with great varieties of accessible software
that put sophisticated statistical and computational capabilities at their fingertips.
These technological changes have reshaped the world of the policy analyst from
one in which overcoming the scarcity of data and information was routinely
a challenge to one in which sifting through the great abundance of data and
information to find that which is valid, relevant, and useful is more often the
challenge. The onset of ‘big data’ makes the reshaping of the analytical world
even more dramatic, opening up new possibilities for analysis but also risking false
inferences—an inconsistent statistical estimator remains inconsistent no matter
how large the sample size. The ease of computation offers great possibilities for
doing more sophisticated statistical estimation and prediction, assessing more
complex relationships through techniques such as computable general equilibrium
and agent-based models, and taking better account of uncertainties in predictions
through Monte Carlo simulations. However, the availability of data and the ease
of computation without solid grounding in the relevant underlying theory risks
misuse, or what I have heard Aidan Vining colorfully refer to as the problem of
‘chimps with uzis.’
Just as technological change offers opportunities for, and poses challenges to,
rationalist policy analysis done primarily by professionals (Weimer and Vining
2011), it does so as well for participatory policy analysis in its various forms
(Durning 1993). The cost of participation generally discourages all but the most
interested individuals from participating in public decisions and contributes to
relatively more influence for stakeholders with more resources. Most obviously,
the Internet facilitates transparency and reduces the cost of access to existing
administrative processes, such as rulemaking. More dramatically, social media
opens up the possibility for new fora in which individuals can potentially engage
in the analytical process at very low cost. Advocates for participatory policy
analysis thus have a new set of tools available. Can they develop protocols that
elicit meaningful participation in policy analysis despite manipulation by organized
interests and disruption from Internet trolls? Will such participation lead to policy
choices that are better for the broader society?
In conclusion, the logics of the historical demands for policy analysis in the
United States will likely continue. Finding ways to make it more useful should
continue to be a major concern for public affairs scholars.
Notes
1 Parts of this chapter originally appeared in David L. Weimer (2015) ‘La Evolución del Análisis de
Políticas en Estados Unidos: Cuatro Fuentes de Demanda’, Foro Inernacional 40(2): 540–75. I thank
Angela Evans, John Hird, Laura Flammad, Reynaldo Ortega, and the faculty of the Centro de
Estudios Internacionales, El Colegio de México, for helpful comments.
2 http://www.pewtrusts.org/en/projects/pew-macarthur-results-first-initiative.
26
References
Assistant Secretary for Policy and Evaluation, n.d., All about ASPE: A Guide for
ASPE Staff.
Bertelli, A. and Lynn, L. Jr. (2006) Madison’s Managers: Public Administration and
the Constitution, Baltimore, MD: Johns Hopkins University Press.
Bimber, B. (1996) The Politics of Expertise in Congress: The Rise and Fall of the
Office of Technology Assessment, Albany, NY: State University of New York Press.
Brudnick, I. (2013) Legislative Branch Agency Appointments: History, Processes, and
Recent Proposals, CRS Report for Congress, August 6.
Bureau of the Budget (1952) Circular No. A-47, December 31.
Champkin, J. (2013) ‘The German Tank Problem’, Significance 10(5): 28
Copeland, C. (2011) Cost-Benefit and Other Analysis Requirements in the Rulemaking
Process, Congressional Research Service, August 30.
Croley, S. and Funk, W. (1997) ‘The Federal Advisory Committee Act and Good
Government’, Yale Journal on Regulation 14(2): 452–557.
Dubnick, M. and Frederickson, H. G. (2010) ‘Accountable Agents: Federal
Performance Measurement and Third-Party Government’, Journal of Public
Administration Research and Theory 20(1): i143–59.
Durning, D. (1993) ‘Participatory Policy Analysis in a Social Service Agency: A
Case Study’, Journal of Policy Analysis and Management 12(2): 297–322.
Engelbert, Ernest A. (1977) University Education for Public Policy Analysis,
Public Administration Review 37(3): 220–36.
Ford Foundation (1976) Ford Foundation Program in Public Policy and Social
Organization, New York: Ford Foundation.
Fortun, M. and Schweber, S.S. (1993) ‘Scientists and the Legacy of World War II:
The Case of Operations Research (OR)’, Social Studies of Science 23(4): 595–642.
Gill, N. (1944) Municipal Research Bureaus: A Study of the Nation’s Citizen-Supported
Agencies, Washington, D.C.: American Council on Public Affairs.
Greenberg, D. and Schroder, M. (2004) The Digest of Social Experiments,
Washington, D.C.: Urban Institute Press.
Greenberg, G. (2003) ‘Policy Advice at the Department of Health and Human
Services, Then and Now’, Journal of Policy Analysis and Management 22(2): 304–7.
Gregory, R., Keeney, R. and von Winterfeldt, D. (1992) ‘Adapting the
Environmental Impact Statement Process to Inform Decisionmakers’, Journal
of Policy Analysis and Management 11(1): 58–75.
Haeder, S. and Weimer, D. (2013) ‘You Can’t Make Me Do It! State
Implementation of Insurance Exchanges under the Affordable Care Act’, Public
Administration Review 73(1): S34–47.
Hale, K. (2011) How Information Matters: Networks and Public Policy Innovation,
Washington, D.C.: Georgetown University Press.
Hammond, R. (1966) ‘Convention and Limitation in Benefit-Cost Analysis’,
Natural Resources Journal 6(2): 195–222.
Harty, H. (1971) ‘Status of PPBS in Local and State Governments in the United
States’, Policy Sciences 2(2): 177–89.
27
Hird, J. (2005) Policy Analysis in the States: Power, Knowledge, and Politics,
Hitch, C. (1953) ‘Sub-Optimization in Operations Problems’, Journal of the
Operations Research Society of America 1(3): 87–99.
Kaufman, H. (1956) ‘Emerging Conflicts in the Doctrines of Public
Administration’, American Political Science Review, 50(4): 1057–73.
Kravitz, W. (1990) ‘The Advent of the Modern Congress: The Legislative
Reorganization Act of 1970’, Legislative Studies Quarterly 15(3): 375–99.
Lee, M. (2008) Bureaus of Efficiency: Reforming Local Government in the Progressive
Era, Milwaukee, WI: Marquette University.
McCubbins, M., Noll, R., and Weingast, B. (1987) ‘Administrative Procedures
as Instruments of Political Control’, Journal of Law, Economics and Organization
3(2): 243–77. Philadelphia, PA:
McGann, James G. (2016) 2015 Global Go To Think Tank Index Report. Think
Tanks and Civil Society Program, Philadelphia, PA: University of Pennsylvania,
February 9.
Medvetz, T. (2012) Think Tanks in America, Chicago, IL: University of Chicago
Press.
Melkers, J. and Willoughby, K. (2005) ‘Models of Performance-Measurement
Use in Local Governments: Understanding Budgeting, Communication, and
Lasting Effects’, Public Administration Review 65(2): 180–90.
Nelson, R. (1989) ‘The Office of Policy Analysis in the Department of the
Interior’, Journal of Policy Analysis and Management 8(3): 395–410.
Milward, H. and Provan, K. (2000) ‘Governing the Hollow State’, Journal of Public
Administration Research and Theory 10(2): 359–80.
Morgan, M. and Peha, J. (eds) (2003) Science and Technology Advice for Congress,
Washington, D.C.: Resources for the Future.
Mosher, F. (1984) A Tale of Two Agencies: A Comparative Analysis of the General
Accounting Office and the Office of Management and Budget, Baton Rouge, LA:
Louisiana State University Press.
New York Bureau of Municipal Research (1915) ‘Government of the City
of Rochester, N.Y.: General Survey, Critical Appraisal, and Constructive
Suggestions’, Prepared for the Rochester Bureau of Municipal Research. Albany,
NY: J.B. Lyon Company (Printers).
Office of Management and Budget (2003) OMB Circular A-4, ‘Regulatory
Analysis’, September 17.
Petracca, M. (1986) ‘Federal Advisory Committees, Interest Groups, and the
Administrative State’, Congress & The Presidency 13(1): 83–114.
Radin, B. (2013) Beyond Machiavelli: Policy Analysis Reaches Midlife (2nd edn),
Schacter, H. (2002) ‘Philadelphia’s Progressive-Era Bureau of Municipal
Research’, Administrative Theory & Praxis 24(3): 555–70.
28
Sherwin, E. (2006) ‘The Cost-Benefit Analysis of Financial Regulation: Lessons

from the SEC’s Stalled Mutual Fund Reform Effort’, Stanford Journal of Law,
Business, and Finance 12 (Fall): 1–60.
Shulock, N. (1999) ‘The Paradox of Policy Analysis: If It Is Not Used, Why Do We
Produce so Much of It?’ Journal of Policy Analysis and Management 18(2): 226–44.
Shultz, G. (1971) ‘Memorandum for the Heads of Departments and Agencies’,
Office of Management and Budget, October 5.
Smith, B. L. R. (1992) The Advisors: Scientists in the Policy Process, Washington,
D.C.: Brookings Institution Press.
Stewart, R. (1975) ‘The Reformation of Administrative Law’, Harvard Law Review
88(8): 1669–813.
Teske, P. (2004) Regulation in the States, Washington, D.C.: Brookings Institution
Press.
U.S. Congressional Research Service (2016) Annual Report of the Congressional
Research Service of the Library of Congress for Fiscal Year 2015, January.
U.S. Government Accountability Office (2016) Summary of GAO’s Performance
and Accountability Report: Fiscal Year 2015, GAO-16-3SP.
U.S. Government Services Administration (2015) Federal Advisory Committee Act
(FACA) Management Overview (http://www.gsa.gov/portal/content/104514)
Vining, A. and Weimer, D. (2009) ‘Assessing the Costs and Benefits of Social
Policies’, in D. Weimer and A. Vining (eds), Investing in the Disadvantaged:
Assessing the Benefits and Costs of Social Programs, Washington, D.C.: Georgetown
University Press 1–16.
Weaver, K. R. (1989) ‘The Changing World of Think Tanks’, PS: Political Science
& Politics 22(3): 563–78.
Weimer, D. (2010a) Medical Governance: Values, Expertise, and Interests in Organ
Transplantation, Washington, D.C.: Georgetown University Press.
Weimer, D. (2010b) ‘Stakeholder Governance of Organ Transplantation: A
Desirable Model for Inducing Evidence-Based Medicine?’, Regulation &
Governance 4(3): 281–302.
Weimer, D. and Vining, A. (2011) Policy Analysis: Concepts and Practice (5th edn),
Upper Saddle River, NJ: Pearson.
West, W. (2009) ‘Inside the Black Box: The Development of Proposed Rules
and the Limits of Procedural Controls’, Administration and Society 41(5): 576–99.
Whiteman, D. (1995) Communication in Congress: Members, Staff, and the Search
for Information, Lawrence, KS: University of Kansas Press.
Wildavsky, A. (1969) ‘Rescuing Policy Analysis from PPBS’, Public Administration
Review 29(2): 189–202.
Wolf, P. (1999) ‘Neutral and Responsive Competence: The Bureau of the Budget,
1939–1948 Revisited’, Administration and Society 31(1): 142–67.
Yackee, S. (2012) ‘The Politics of Ex Parte Lobbying: Pre-proposal Agenda
Building and Blocking during Agency Rulemaking’, Journal of Public
Administration Research and Theory 22(2): 373–93.
29
TWO
The evolution of the policy analysis

profession in the United States
Beryl A. Radin
What does it mean to be a policy analyst? Despite the growth of the field in the
United States over the past several decades, this is not a profession that the general
public understands. Indeed, as time has gone by since the profession developed,
there sometimes appears to be less agreement about the activity than there was
in its earliest days. It is obvious that policy analysis has not gained a place in
the world of professions equal to that of law, medicine, or engineering. Since
many of the original policy analysts worked for the government, it was difficult
for those outside of the field to differentiate between public administrators and
policy analysts.1
There are many reasons for this situation, but at least one of them is the inability
of this field to define itself in the context of a changing economic, political, and
social environment in the United States and, at the same time, to find a way to
describe and appreciate the contributions of the fluid profession. For the answer
to the question ‘What is a policy analyst?’ is different today than it was in the
1960s, when the profession first defined itself.
In order to understand these changes, one must view these developments as an
example of the intellectual history of the development of a profession that—in
its practice—must directly confront the broader changes in American society.
Policy analysis, perhaps more than the traditional professions (for example, law,
medicine, or engineering) directly confront the changing political values of the
day. The America of the 1960s was characterized by a strong belief in progress,
abundance, and the possibilities of public sector change. The economic, political,
and social fabric of the society supported a reliance on expertise and experts.
Soon afterward, however, the climate of opinion within the society had shifted.
Americans were less confident about the future, focused on the realities of scarcity,
and pulled back from public sector responsibilities. And as many policy issues
confront ideological and value differences, there has been an attempt to avoid
them through reliance on technocratic solutions.
Yet, at the same time, many in the society are much less willing to defer to
so-called experts than they were in the past. According to one experienced
practitioner, by the 1990s the terms ‘fear, paranoia, apprehension, and denial’
described the life of a policy analyst (quoted in Beam 1996). But at the same time
that this shift has occurred, ironically the number and range of jobs in the policy
analysis field have increased markedly during that period. But the 21st century has
brought less clarity about what a policy analyst would or could do. In addition,
31
while new definitions of the field have emerged, there are still examples of past
approaches to policy analysis in both the practice and training of policy analysts.
There are those who bemoan the modifications in the profession and continue
to point to a golden day when decision makers listened to experts and seemed to
take the work of the policy analysts seriously (Williams 1998). Yet there are others
who believe that the dimensions of the field in the 1990s were more appropriate
to the institutions, values, and processes of American democracy than were the
more elite practices of the world of the 1960s. These modifications, it could be
argued, have produced a profession that is classically American as it has become
more pluralistic and open, whereas the earlier version of the field had more in
common with societies that had traditions of elitism and secrecy. Yet as the 21st
century developed, it sometimes appears that pluralism and openness have led to
increased conflict and inability to deal with multiple perspectives.
Policy analysis in three eras

One can depict the realities of the policy analysis profession as occurring during
three periods: 1960 to 1989, 1990 to 2009, and the 2010s. Each of these
eras approached many of the same questions that confront the policy analysis
practitioner today: the conflict between the imperatives of analysis and the
world of politics; the analytic tools that have been used, created, or discarded
over those 50 years; the relationship between decision makers and analysts as the
field has multiplied and spread; and the assumptions about the availability and
appropriateness of information that can be used in the analytic task. Each also
reflects the shifts in society’s attitudes toward public action, the availability of
resources to meet public needs, and the dimensions of policy making.
Using three points of time to frame this contrast does provide a way to emphasize
the shifts that have occurred in the profession. This treatment thus is not a history
but a contrast between three generations. But all of these generations continue
to be active today. Thus, the newer generations can expect the approaches
of the past to continue in some form today. During these three eras, policy
analysts operate with quite different standards of success, reflecting the essentially
normative character of the profession. For some, success comes when decision
makers seek and accept their advice. For others, the reward is involvement—even
tangentially—in the decision process. Some measure success as the ability to do
interesting work and apply a range of analytical methodologies. Still others define
their success in terms of their ability to advance substantive policy positions or at
least to influence the views of others involved in the decision-making process.
Some analysts believe that they must include all possible options and strive for a
comprehensive approach to policy problems. Others, by contrast, begin with a
set of values (or even a specific agenda) and define their roles as advocacy analysts.
Although policy analysts initially focused on the stage of the policy process
when alternative approaches to issues were being formulated, it is clear that the
scope of the role has widened. Policy analysts are now involved in bringing policy
32
The evolution of the policy analysis profession in the United States
issues to the policy agenda, working with the decision makers (particularly at
the legislative level) in the actual adoption of policy, ascertaining the details of
implementation, and engaging in the evaluation process.
Era one: 1960–89

Although policy analysis as a distinct field and emerging profession is relatively
recent, the practice of providing policy advice to decision makers is an ancient
art. Our tendencies to focus on the uniqueness of the contemporary American
environment have often evoked approaches that ignore the lessons of history. A
field that seeks to create its own special identity works hard to differentiate itself
from the past.
Creating a profession
The early stages of policy analysis grew out of the experience of World War
II. In the post-World War II period, social scientists began to play a role in the
decision-making process. The imperatives of war had stimulated new analytic
techniques—among them systems analysis and operations research—that sought
to apply principles of rationality to strategic decision making. The precepts
developed in the natural sciences found expression in the social sciences in two
modes: positivism (using the concept of laboratory experimentation to seek to
differentiate the true from the false)2 and normative economic reasoning (using
the concept of a market made up of individual rational actors as the basis for
prescription). Although still in an embryonic form, the computer technology of
that period did allow individuals to manipulate what were then considered large
data sets in ways that had been unthinkable in the past. In a sense, the techniques
that were available in the late 1950s cried out for new forms of use.
Yehezkel Dror, one of the earliest advocates for the creation of policy analysis as
a new profession, described the early phases of this search for new expressions as
‘an invasion of public decision-making by economics’ (1971, p 226). Further, he
wrote, ‘Going far beyond the domain of economic policy making, the economic
approach to decision making views every decision as an allocation of resources
between alternatives, that is, as an economic problem’ (Dror 1971, p 226).
DoD: organizing a policy analysis office
The office that was established in the Department of Defense to carry out
McNamara’s analytical agenda became the model for future analytic activity
throughout the federal government. As the office developed, its goal of providing
systematic, rational, and science-based counsel to decision makers included
what has become the classic policy analysis litany: problem identification,
development of options or alternatives, delineation of objectives and criteria,
33
evaluation of impacts of these options, estimate of future effects, and—of course—

recommendations for action.
In many ways, the establishment of this office represented a top-level strategy
to avoid what were viewed as the pitfalls of traditional bureaucratic behavior.
Rather than move through a complex chain of command, the analysts in this
office—regardless of their rank—had direct access to the top officials in the
department. Their loyalty was to the secretary of the department, the individual
at the top of the organization who sought control and efficiencies in running
the huge department.
This office was viewed as an autonomous unit, made up of individuals with
well-hewn skills who were not likely to have detailed knowledge of the substance
of the policy assignments given them. Their specializations were in the techniques
of analysis, not in the substantive details of their application (see the chapters
by Lynn and Weimer in this volume). There was an assumption that the policy
analysts who were assembled would constitute a small, elite corps, made up of
individuals who would be expected to spend only a few years in the federal
government. Their frame of reference and support system were atypical for a
federal employee. Sometimes called the ‘Whiz Kids,’ this staff was highly visible:
both its PPBS activity and its general expertise came to the attention of President
Lyndon Johnson. In October 1965, the Bureau of the Budget issued a directive to
all federal departments and agencies, calling on them to establish central analytic
offices that would apply the PPBS approach to all their budget submissions.
Even during these days, however, there were those who were skeptical about
the fit between the PPBS demands and the developing field of policy analysis
(Wildavsky 1969, pp 189–202).
Creating a new profession
The policy analysis institutions of the early 1960s shared many of the problems
and characteristics of those who practiced the ancient art of providing advice
to decision makers, but though the functions contained similarities, something
new had occurred. What had been a relatively informal set of relationships and
behaviors in the past had moved into a formalized, institutionalized world.
Although the process began with responsibility for the PPBS process, it became
obvious very quickly that the policy analysis task had moved beyond a single
analytic technique. Much still depended on the personal relationships between
the analyst/adviser and the decision maker/ruler, but these organizations took
their place in public, open, and legally constituted organizations.
This field was conceptualized as integral to the formulation stage of the policy
process—the stage of the process where analysts would explore alternative
approaches to ‘solve’ a policy problem that had gained the attention of decision
makers and had reached the policy agenda. Both decision makers and analysts saw
this early stage as one that was separable from other aspects of policy making. It
did not focus on the imperatives of adopting the preferred alternative (particularly
34
in the legislative branch) or on the details of implementing an enacted policy

inside an administrative agency. Instead, it focused on the collection of as
much information and data as was available to help decision makers address the
substantive aspects of the problem at hand.
Less than two years after the diffusion of PPBS throughout the federal
government, Yehezkel Dror sounded a clarion call that defined a new profession
in what has become a classic article in the Public Administration Review. Published
in September 1967, Dror’s ‘Policy Analysts: A New Professional Role in
Government’3 sought to differentiate policy analysis from systems analysis and
to define characteristics of policy analysts as government staff officers. Dror
emphasized the confidential relationship between the decision makers and the
analyst. He envisioned a relatively small staff, with a minimum of between 10
and 15 persons but a preference for a staff of between 25 and 30.
By the early 1970s, the field began to assume both visibility and self-definition.
Through support from private foundations, between 1967 and 1970, graduate
programs in public policy were begun at Harvard, the University of California
at Berkeley, Carnegie-Mellon, the RAND Graduate Institute, the University of
Michigan, the University of Pennsylvania, the University of Minnesota, and the
University of Texas at Austin (Heineman et al. 1990; Wildavsky 1997). Almost
all of these programs were at the master’s degree level, focused on training
professionals to enter the policy analysis field (see chapters by O’Hare and Rubaii
in this volume).
There were sufficient examples of the new breed within federal agencies for
Arnold Meltsner to begin a study of their experience in the early 1970s. His
work Policy Analysts in the Bureaucracy (1976) suggested that policy analysts used
two different types of skills—technical and political skills. The individual who
emphasized technical skills has a ‘desk and shelves … cluttered with papers. His
books will be the same as those of academic social scientists’ (Meltsner 1976, pp
19–20). The individual who highlighted technical skills ‘is convinced that he
is objective, a scientist of sorts’ (Meltsner 1976, p 23). That individual sees the
policy analyst as someone who can do high-quality, usually quantitative policy-
oriented research that satisfies him and his peers.
By contrast, the individual who emphasizes political skills is, in many ways,
more like a traditional bureaucrat. But that individual wants to be able to work
on the major problems facing the agency. ‘They want to be where the action is’
(Meltsner 1976, p 32). Meltsner’s work indicated that there were a number of
attributes that contributed to the variety of behaviors; some had to do with the
analyst’s own skill level and personal comfort with various approaches. Others
had to do with client expectations.
The profession after a decade
By the mid 1970s, the policy analysis field did seem to be moving toward
a professional identity. It had its own journals, a range of university-based
35
professional schools, and individual practitioners (largely in government agencies)

who identified themselves as policy analysts. Yet many issues remained unresolved
that would surface as problems for the future. It appeared that policy analysis
functions legitimately included planning, evaluation, and research but the
relationship between policy analysis and budgeting was not clear. Some agencies
had followed the PPBS path and directly linked the analytical activity to fiscal
decision making. Others separated the two functions.
Many policy analysis shops consciously separated their responsibilities from
those of the operating divisions in the department or agency, following Dror’s
advice. Even those analysts who exerted political skills did not focus on the
implementation stage of the policy process. If they accentuated politics at all,
they highlighted the political realities of the agenda-setting stage of the process,
looking to the agenda of their political masters and to the political intrigue
involved in achieving passage of a proposal. It was a rare analyst who believed
that effectiveness was related to the ability to raise implementation issues in the
analysis, requiring interaction with the operating units.
Some policy analysts found that they were assuming two quite different roles
in their activity. Not only were they the neutral analysts, laying out all possible
options, but they were also advocates for their preferred positions, which they
believed best represented the ‘national interest.’ Unlike advisers in other societies,
American policy analysts often described their role as including concern about
the best policy option that was in the ‘public interest.’ Despite the difficulty of
defining a single ‘public interest’ in the diverse American society, this language
was often used by analysts. This created a tension that was not easy to manage
(see Nelson 1989, p 401).
Finally, as the field took shape, the definition of success in the practice of policy
analysis was not always obvious or agreed on. Was success the ability to convince
the client decision maker to adopt your recommendation? Was an analyst successful
when he or she helped the client understand the complexity and dimensions
of a policy choice situation? Or was the analyst successful when the work that
was performed was of a quality that was publishable and well received by peers
within the profession? These problems continued to define the profession in the
coming years.
Era two: 1990–2003

The world that faced the early policy analysts in the federal government from
1960 to 1989 was very different from that confronting the policy analyst in the
1990s. The years that had elapsed since policy analysis activity began in the
1960s were characterized by dramatic economic, social, and political changes
in American society. The environment of possibility and optimism of the early
1960s was replaced by a mood of scarcity and skepticism. Indeed, by the end of
the 1960s, the experience of the Vietnam War and demands for equality within
36
the society forced many to acknowledge that there were large cleavages among
groups within the nation.
When it became clear that policy analysis had become a part of the decision-
making process and language, multiple actors within the policy environment
began to use its forms and sometimes its substance. As a result, policy analysis has
become a field with many voices, approaches, and interests. As it has matured,
the profession has taken on the structure and culture of American democracy,
replacing the quite limited practice found only in the top reaches of government
bureaucracies, which characterized the early stages.
The impact of these macro-level changes was not always obvious to policy
analysis practitioners in the government. Perhaps the most direct effect of the
modifications was in the range of possibilities for policy change that were before
them. As Alice Rivlin has noted, the crusading spirit attached to the early days
of the field was closely linked to the large number of new programs that were
being created (Rivlin 1998). Soon, however, the agenda for many analysts focused
mainly on changes in existing policies, not on the creation of new policies and
programs.
Policy analysts did experience other shifts in their world. In the early 1960s,
beginning with the PPBS activity in the Department of Defense, it was assumed
that policy analysis units would be established at the top of organizations, looking
to the top executives and senior line staffers as the clients for analysis.4 The focus
of the activity was upward, and the separate nature of the policy analysis unit
minimized concern about the organizational setting in which the analysis took
place.
By the mid 1970s policy analysis activities had dramatically increased throughout
the structure of federal government agencies. Most of the units began in the top
reaches of departments or agencies but as time went on, policy analysis offices
appeared throughout federal agency structures, attached to specific program units
as well as middle-level structures. As it became clear that the terms of policy
discourse would be conducted (at least superficially) in analytic terms, those who
had responsibility for subunits within departments or agencies were not willing to
allow policy shops attached to the top of the organization to maintain a monopoly
on analytic activity. By the mid 1980s, any respectable program unit had its own
policy staff—individuals who were not overwhelmed by the techniques and
language of the staff of the original policy units and who could engage in debate
with them or even convince them of other ways of approaching issues. Both
budget and legislative development processes were the locations of competing
policy advice, often packaged in the language and form of policy analysis.
As a result, staffers appeared to become increasingly socialized to the policy
cultures and political structures with which they were dealing. Those staffers
who stayed in an office for significant periods of their careers became attached
to and identified with specific program areas and took on the characteristics of
policy area specialists, sometimes serving as the institutional memory within the
department (Rein and White 1977).
37
The institutionalization and proliferation of policy analysis through federal

departments (not simply at the top of the structure) also contributed to changes
in the behavior of those in the centralized policy analysis units. Those officials
became highly interactive in their dealings with analysts and officials in other
parts of the departments. Increasingly, policy development took on the quality
of debate or bargaining between policy analysts found in different parts of the
agency. In addition, policy analysis staff found they shared an interest in activities
performed by other offices; in many ways, staffers behaved more like career
bureaucrats than the in-and-out analysts envisioned in the early writings on the
field. Longevity and specialization combined to create organizational units made
up of staff members with commitments to specific policy areas (and sometimes
to particular approaches to those policies).
As a result of the spread of the function, staff implicitly redefined the concept of
‘client.’ Though there had always been some tension over the client definition—it
could be your boss, the secretary, the president, the society, or the profession—the
proliferation of the profession made this even more complicated. In part, this was
understandable as analysts developed predictable relationships with programs and
program officials over time. Not only were clients the individuals (or groups)
who occupied top-level positions within the organization, but clients for analysis
often became those involved in the institutional processes within the agency as
well as the maintenance of the organization itself (see Nelson 1989).
Analysts, not their clients, became the first-line conduit for policy bargaining.
The relationships among policy analysts became even more important by the
late 1970s, when the reality of limited resources meant that policy debate was
frequently focused on the modification of existing programs, not on the creation
of new policies. Analysts were likely to be working on issues where they already
had a track record, either in terms of substantive policy approaches or involving
relationships with others within the organization.
By the mid 1980s, policy analysts activities across the federal government took
on the coloration of the agencies and policy issues with which they worked. The
policy office in the Department of Health and Human Services had an approach
and organizational culture that was distinct from that in the Department of
Labor or in the Environmental Protection Agency (EPA). Agencies with highly
specialized and separate subunits invested more resources in the development
of policy analysis activities in those subunits than in the offices in centralized
locations. EPA’s responsibilities in water, air, and other environmental problems
were different enough to make such an approach effective. Some policy offices
were organized around substantive policy areas (for example, HHS had units in
health, social security, social services, and welfare) while others were created
around functional approaches to analysis.
38
Policy analysis beyond the bureaucracy
As the policy analysis function proliferated within the federal government, it

also spread outside the government to include a variety of actors engaged in the
policy process. The dynamic that spawned the proliferation of policy units within
the bureaucracy similarly stimulated the developments outside the organizations.
If the policy discourse were to be conducted in the language of analysis, then
one needed to have appropriate resources available to engage in that discussion.
One could find a discernible increase in interest in policy analysis in the U.S.
Congress by the 1970s (Malbin 1980). Members of Congress sought to develop
their own analytic capacity. Indeed, one observer noted that these amendments
‘constitute[d] the congressional declaration of analytical independence from the
executive branch’ (Robinson 1992, p 184) and, in some cases, actually exceeded
the capacity of the executive branch (Robinson 1989, p 2).
Despite the interest in policy analysis, the expectations of legislative institutions
were viewed as different from those of executive branch policy analysts. Carol H.
Weiss noted that at least some congressional staffers treat analysts ‘less as bearers
of truth and clarity than as just another interest group beating on the door’
(Weiss 1989, p 411). She has suggested that the structure of Congress does act
as a deterrent to the use of analysis—tangled committee jurisdictions, tenure of
committee members, shortage of time, the oral tradition, staff fragmentation, and,
of course, the dynamics of decision making in a legislative body. Yet analysis was
valued as staff on committees brought with them respect for analytic procedures
and members of Congress found that they must be in a position to evaluate the
work of others; they must be on top of the analytic work produced by agencies
and others and help their bosses avoid surprises.
Think tanks
Think tanks were seen to be influential in American society during this period.
(See chapters by Weimer and Rich in this volume). Think tanks do vary in the
degree to which they take positions. When they do take positions, they behave
much like interest groups with an advocacy posture. Some groups are identified
with a general policy direction, but the work that is produced by individuals
within the organization does not fall into a clear or predictable approach. Some
think tanks have actually become extensions of government agencies because
agencies now depend on them to do their essential work. In an era of government
staff reductions and contracting out, there has been an increasing use of outside
groups as surrogate federal officials, relying on procurement mechanisms such as
standing-task-order agreements to carry out these relationships (Metcalf 1998;
also see the chapter by Rich in this volume).
39
Policy analysis and interest groups
Because policy debate does often deal with analytical findings, it is not surprising
that many interest groups established their own in-house policy analysis
capabilities. To help carry out their well-known policy agendas, these groups
have found it useful to have staff who are comfortable with analytical techniques,
methodologies, and research findings. Some interest groups have made investments
in staff who are able to critique the work of others, particularly emphasizing areas
in which so-called objective analysis actually reflects differences in values or where
methodological determinations skew the recommendations or findings. Other
groups have actually conducted their own analyses, even encouraging original
data collection to support their positions or, at the least, to show that there is
an alternative to the ‘official’ analysis that has been performed (Primus 1989).
Ironically, although analysts in interest groups are thought to be far from the
image of the policy analyst without a policy agenda, sometimes they may be
valuable sources of information to others in society. But it is not always easy to
define the boundaries between advocacy and fairness or between objectivity and
fairness. The value or policy perspectives of interest group analysts are often much
clearer than are those of either government or think tanks. In the latter, values
may be embedded in the analytic techniques used or in the way the problems
are framed. Some groups stake their reputation on the quality of information
they produce and often become the source of information for decision makers
as well as members of the press. Their experience—focusing on the production
of rigorous, high-quality work that is organized around a commitment to low-
income citizens—suggests that it is possible to exhibit both commitment and rigor
The spread of the policy analyst role across government and nongovernmental
locations has created new opportunities that were unthought of in the 1960s.
Students who graduated from the policy schools found jobs in the field and
learned that they were, indeed, part of a new profession. In addition, during
this period policy shifts took place in many domestic policy areas that devolve
new authorities and responsibilities to players at the state and local level, which
stimulated a new demand for analytic activity outside of Washington (see the
chapter by Godwin, Godwin, and Ainsworth in this volume).
A response to complexity: institutionalizing the policy analysis function
As the years progressed, there were major changes in several of the attributes
that were associated with the early stages of the policy analysis field. As fiscal
scarcity and limited budgets became the reality in many policy areas, changes
occurred in the availability of funds for new or enlarged programs. Ideological
shifts demanded changes in the assumptions of policy analysts working inside
government. Experience with analysis that was both positive and negative
contributed to a growing sophistication both inside and outside government.
At the same time, staffing in a number of policy analysis offices was reduced.
40
One analysis suggested that the size of eight federal offices declined from 746
individuals in 1980 to 426 in 1988 (May 1998).
The bureaucratization and spread of the government policy analysis function
created a reality that was quite different from that envisioned by Dror and the
field’s early practitioners. The personal relationship between the top official in
the organization and a middle-level policy analyst became tenuous, if it existed
at all. An analyst is much more likely to provide advice to another analyst, who
in turn provides it to another analyst. Rarely did an analyst assume the role of
personal counselor to the top decision maker.
These changes suggest a broader range of clients than that described by Meltsner
(1976). Clients are not only decision makers but the client relationship may also
be defined by the institutional processes of governing (planning, budgeting,
regulation development, legislative drafting), maintenance of the organization,
and—perhaps as a result of strong attacks on the programs of some domestic
agencies—the analyst himself or herself. In addition, the work of the analyst
demands a broader range of skills than those described by Meltsner (see Radin
1992). The technical and political skills that he described continue to be required,
but so also were skills in organizational analysis, specific knowledge of the program
area, and evaluation techniques.
The growth of evaluation
The early policy analysis practitioners emphasized analytic functions that were
prospective (that is, they analyzed new program possibilities that were being
considered). The evaluations conducted during the early years of the profession
were devised as evaluations of experiments that created new ways to deliver or
define programs. But shifts in the policy environment diminished the number and
size of new programs that were created and instead concentrated on fine-tuning
existing programs. Thus it is not surprising that later analytic shops would also
be concerned about employing retrospective analytic techniques (examining the
effects of program interventions through various evaluation techniques).
By the 1970s, however, evaluation activities became more commonplace in the
policy analysis units although some agencies did create offices for evaluation that
were separate from the policy analysis shops. Evaluation activities were also found
in other settings; for example, several federal offices of inspector general created
units that did undertake evaluations. Though these evaluations did not adopt the
classic social science approach to evaluation, they emphasized the production
of information that would be of direct use to decision makers (Thompson and
Yessian 1992).
During this period, significant resources were available to think tanks, research
centers, and consulting firms to undertake federally financed evaluations. A series
of social experiments were conducted to test policy and program possibilities
before their enactment (Weiss 1998, p 13) but because the experiments had fairly
41
long time requirements, their results were usually not available in a timely fashion,
at least for the decision makers who had commissioned the studies.
By the 1980s, many evaluation studies were much more modest in scale and
usually sought to produce findings that would be timely for decision makers. A
series of evaluation projects in the education field were mandated by Congress,
seeking to provide information that would be useful to the reauthorization
process in Congress (the decision cycle that required Congress to renew, revise,
or eliminate programs with a finite life cycle). Evaluations were conducted to
examine different approaches to welfare reform. Though specifying a user of the
evaluation did not avoid all the problems common to the evaluation enterprise,
the information produced by several of these efforts did play a role in the decision-
making process (see Radin 1984).
Moving toward implementation
The policy analysts of the 1960s emphasized the early stages of the policy process—
what was viewed as the formulation stage. Some analysts focused on work that
would bring issues to the policy agenda; most, however, moved into action when
they were confronted with a policy problem and sought to formulate alternatives
that would advise the decision maker. By the 1980s, policy analysts were found
in later stages of the policy process.
Others began to pay more attention to issues related to implementation and to
raise those implementation concerns when policies were being devised. By the
mid 1970s, both academics and practitioners in the field realized that a significant
number of problems (or what some viewed as failures) associated with the Great
Society programs had occurred because program designers had not considered
the implementation of these efforts when they constructed new programs or
policies (see, for example, Nakamura and Smallwood 1980).
The academic response
Aaron Wildavsky’s essay ‘Principles for a Graduate School of Public Policy,’

written in 1976, discussed the academic dimensions of the policy analysis field.
As he described the process of organizing the Graduate School of Public Policy at
Berkeley, Wildavsky highlighted the importance of problem solving ‘rather than
fact grubbing’ in the creation process (Wildavsky 1976, p 408). As the schools
aged through the 1980s and into the 1990s, however, it was difficult to sustain
the spirit of experimentation within them. There was some tendency for the
faculty to revert to the academic disciplines from which they came.
Era three: 2003–present

The field of policy analysis that exists in the 21st century is quite different from
that found in the two earlier phases. The world of the 1960s that gave rise to this
42
field in the United States often seems unrelated to the world we experience today.
These shifts have occurred as a result of a range of developments – technological
changes, changes in the structure and processes of government both internally and
globally, new expectations about accountability and transparency, economic and
fiscal problems, and increased political and ideological conflict. It is interesting
that increasingly policy observers have been using the phrase ‘wicked problem’
to describe problems that are difficult or impossible to solve.5 But at the same
time that these developments require shifts in the way that the field operates,
they also continue many of the expectations from the past.
Until circa 2000, the literature on policy analysis was focused almost entirely
on the experience within the United States. The exception to this pattern was
found in a 1984 book authored by Brian Hogwood and Lewis Gunn, Policy
Analysis for the Real World. They noted:
The problem of generating suitable and readily available teaching

material for British courses has not yet been overcome. Although
our primary interest was in producing policy analysis materials for
British students, our experience has made us aware of the limitations
of much American literature for American students, particularly those
who have previously thought of policy analysis as merely American
politics rehashed, or as arcane mathematical techniques (Hogwood
and Gunn 1984, pp v–vi).
By the first decade of the 21st century, the concerns expressed by Hogwood
and Gun were being addressed by a range of scholars around the world. The
introduction of these issues produced a literature that indicated that the field
was rich and varied and, in particular, attentive to the context of the systems in
which policy analysis took place.
Evidence of a global perspective on policy analysis was found in the programs of
a variety of professional meetings (for example the International Political Science
Association created a new section on Comparative Policy Analysis) and in the
publication of a number of journals (for example Journal of Comparative Policy
Analysis), textbooks and edited volumes from around the world.
The developments in the 21st century suggest that there are many ways to
sort out topics relevant to the policy analysis field. Some topics concerned the
earlier eras—types of policy issues, the diverse relationships between analysts and
clients, the type of analysis required, its time frame, the stage of the policy process
at which it occurs, where in the system it occurs (for example, whether it takes
place inside government or outside government), the impact of the structure
of the government involved, the placement of analysis in central agencies vs.
program agencies, whether analysts and clients are career or political actors, the
appropriate skill set found in analysts, and the boundaries between policy analysis
and management.
43
For the first time it became clear that it was important to give attention to
the structure of the relevant government and the history of its operations. Thus,
variation in policy analysis approaches can be attributed to the structure of
government (for example, whether it is a centralized or federal system) or to the
historical demands of eliminating colonialism, achieving democracy, or responding
to the end of the Soviet Union. For many, the policy analysis field was evidence
of American exceptionalism.
But if anyone was pushed to come up with those comparisons, they are likely to
have emphasized the differences in the political structure between a parliamentary
system where the executive branch is viewed as a part of the legislative branch
and the shared power institutional design found in the United States between
legislative, executive, and judicial systems. Variations that are found between
parliamentary systems and the U.S. system have made it more difficult to generalize
about a single approach that undergirds the policy analysis function. Policy analysis
in the United States has moved far beyond government. It is found in all nodes
of the policy system; not only does it occur inside of government, but it is found
in interest groups, nongovernmental organizations, and state and local agencies.
The shifts that have taken place in the United States in the field have changed
the view of clients; clients are now not limited to top officials within the agencies.
The proliferation of policy analysts and policy analysis offices throughout the
nooks and crannies of an agency means that the offices found at the top reaches
of the departments or agencies no longer have a monopoly on the analysis
production. As a result, analysts attached to these offices now have to negotiate
with analysts in other parts of the department. In some cases, ideology—not
information—has become the currency of policy debate. In addition, clients
have moved from individual officials to collective bodies or even institutional
processes within the agency.
These developments have spawned an acknowledgement that analysis could take
place in all of the stages of the policy process. Analysis plays a role in bringing
policy issues to the policy agenda and is involved in the formulation of policy
details, in ascertaining the details of implementation, and in the evaluation process.
In some instances, analysts associated with particular policies or programs have
personally moved from an analytic role to an implementation role (being actually
involved in managing a program). The focus on implementation has led to a
blurring of boundaries between management and policy development; this has
also been supported by the contemporary interest in performance management,
which sometimes links evaluation activities to performance assessment.
Where are we today in the United States?

Level of formalization of the analytical process
The image of the policy analyst as a quasi-academic staffer whose product is a set
of written documents no longer fits many practicing US policy analysts, however
44
there are still those who conform to the original model of the profession. Yet
there are an increasing number of analysts who make their contribution to the
process through meetings and other forms of more informal interaction. In some
settings the stylized approach to analysis, ending with formal recommendations,
has been replaced by interactions in which the analyst is one of a number of
participants in policy discussions. In addition, there is a sense of urgency about
the need for decisions, a blurring of lines between managers and analysts, and,
as well, a blurring of differences between analysts and more traditional academic
researchers. Availability of information on the Internet allows broader access to
information than was possible in the past when information was largely controlled
by the analysts.
The proliferation of policy analysis both inside and outside of government has
supported an approach to information and data that moves far beyond a positivist
orientation. As Carol Weiss (1983) described it, it is difficult to think about
information without also acknowledging interests and ideology.
While in the past U.S. policy analysts seemed to have discovered ways to
minimize the impact of politicization on issues on their task, this had become
more difficult to do by the second decade of the 21st century. Often the demand
for analysis was effectively a request by decision makers for justifications for their
political agendas. As such, the clients for analysis were not interested in thinking
about alternatives to that position. This was similar to the way that lawyers used
evidence for their arguments.
These changes (and others) have contributed to the existence of a policy analysis
profession in the United States that has a number of attributes. Perhaps the most
dramatic set of changes dealing with globalization in the field in the United States
has taken place in the classroom. The boundaries between policy analysis and
management have become increasingly blurred and, at the same time, policies are
actually implemented by state and local governments and private organizations
which may not be willing to share information about their experiences. Perhaps
most importantly, they have contributed to a significant modification in the skills
that are viewed as essential for individuals hiring the field.
Networks as clients
Yet another development has occurred in the 21st century that has had a major
impact on the world of the policy analysts and the way they think about clients.
Views about decision-making processes have moved to quite a different approach.
The views of an individual client were often framed in the hierarchical decision-
making model. In that model, decisions followed the traditional hierarchical
structure and thus the assumed client would have authority and power to make a
decision. By the end of the 20th century, however, policy analysts often assumed
that decision making was the result of a bargaining process. Thus, the proliferation
of analysts and analytic organizations fits nicely into the bargaining relationships
45
between multiple players, most of whom were located somewhere within the
governmental structure.
By the first decade of the 21st century, another approach was added to the
decision-making repertoire: the use of networks. Although networks have
captured the interest of scholars in a variety of fields both in the United States
and abroad, it is not always clear how they operate as a formal decision-making
approach. Agranoff and McGuire describe a network as a structure that ‘includes
governmental and nongovernmental agencies that are connected through
involvement in a public policy-making and/or administrative structure through
which information and/or public goods and services may be planned, designed,
produced, and delivered’ (McGuire and Agranoff 2011, p 266). Their analysis
focuses on the operational limitations of networks caused by power imbalances,
overprocessing, and policy barriers. They note that networks represent both
formal and informal approaches and many of them do not possess formal authority
to act while others merely exchange information and identify problem-solving
approaches. When networks contain a mixture of actors with different resources,
it is not always clear how those without formal authority can operate within
those relationships.
These are issues that are embedded in situations where the network itself is the
policy analyst’s client. Since the network is not an entity with clear or simple goals,
how does the analyst determine the interests of the body when—by design—it
contains players drawn from multiple interests and settings? And many of those
interests represent substantive policy conflicts. These conflicts can emerge from
the combination of public sector and private sector players, representatives of
interest groups, multiple public agencies, and players from the various nodes of
the intergovernmental system. McGuire and Agranoff note that it is difficult to
conclude that networks are replacing governmental agencies or that networks
are controlling government. It is not always clear where the boundaries lie
between government agency and network. But they (and others) suggest that it
is important for researchers to look at the reality of who actually makes decisions
in and through networks.
While there is fairly widespread acknowledgement that decisions are often
made by networks (rather than individual or only government officials), it is not
clear how these developments affect the policy analysis process.
The economic environment
Policy analysts have always been concerned about questions related to the cost of
proposed action. The early interest in cost-benefit analysis not only pushed analysts
to think about the overall cost of an action but also gave them a framework to
think about who pays and who benefits from decisions. In that sense, the analytic
approach of cost-benefit analysis required analysts to think about the consequences
of their recommendations for real people and to acknowledge that there might
be differential costs and benefits to different categories of individuals.
46
During the early years of the profession analysts rarely based their decisions only
on cost calculations. Recommendations were not expected to be based only on
the lowest cost alternative but, rather, analysts tried to trade off multiple values
and to determine the ratio between costs and benefits. It certainly helped that
the environment of the early 1960s was largely one of growth and possibilities.
By the end of the 1960s, analysts were operating in a Vietnam War environment
where there was an acknowledgement that it was hard to have both guns and
butter. But it was still possible to ask questions that involved determinations of
the abstraction, ‘the public interest’. During the second phase of the profession,
beginning in the mid 1970s, the proliferation of policy analysis offices made
it obvious that there were multiple interests at play in most policy decisions.
Analysts worried about both costs and benefits but calculated them in terms
of their clients. It was the political process that eventually made the trade-offs
between the interests of multiple actors in terms of bargaining and negotiation.
As the 21st century began, 9/11 and two subsequent wars interacted with more
conservative views that advocated a limited role for government and sought to
lower tax rates. Increasingly, analysts inside of government saw themselves as
writing briefs for decision makers that justified their preferred positions. These
and other factors brought a fiscal condition which led to high debt, resistance
to increase taxes, and skepticism about the effectiveness of government action.
This created a climate in which cost was discussed without thinking about
benefits. It pushed analysts to think in short-term rather than long-term
frameworks. Debate was often posed in ideological terms and issues were captured
by highly politicized interests. Debates became increasingly based on budget
numbers not on real assessments of program effectiveness.
Changing skills and staff
The original depiction of policy analysts was conceptualized by Arnold Meltsner

in the 1970s based on his interviews with the first generation of policy analysts in
the federal government. His framework differentiated between types of clients,
emphasizing the dynamics that occurred as a result of the proliferation of policy
analysis inside of government agencies. However, this typology did not reflect
the growth of the policy analysis profession outside of government.
Today a large percentage of policy analyst jobs are found outside of government.
Many organizations are looking for staff who have substantive knowledge of
specific policy areas. Individuals come into policy analyst positions because of
their specialized knowledge of a policy sector. In addition, organizations seem to
be seeking written communication skills, collaborative team work, and computer-
related skills. They also focus on candidates’ oral communication skills as well as
evidence of their organizational abilities. A number of organizations are looking
for individuals who have the ability to deal with policymakers. If the organization
is involved in international work, it often seeks evidence of international
experience as well as language skills. If a policy analysis organization is identified
47
with particular ideologies or policy positions, it might require potential staffers

to show their commitment to their positions.
During the 21st century changes have occurred both in the world of practice as
well as in the academy. The modifications seem to be more visible and pronounced
in the academic world as the field reaches its mid life. While the policy analyst
practitioners represent diverse and varied roles and skills, the academic face of
the field seems to have swung back to more traditional academic perspectives.
As a result, there does appear to be a disjuncture between the world of practice
and the academy.
There are several patterns that seem to be evidence of this retreat to traditional
academic perspectives. First, the differentiation between policy research and
policy analysis has become much less clear (Vining and Weimer 2010). Policy
research approaches draw on broad social science research and do not focus on
the relationship between the policy analyst and the client he or she is advising.
In that sense, the work is less sensitive to the advising role of policy analysts.
Second, faculty members are increasingly recruited from traditional academic
fields and have been evaluated through the performance criteria of those fields.
Third, fewer faculty members than in earlier years either come to the academy
with experience as practitioners or are encouraged to spend time in a practitioner
role during sabbaticals or other forms of leave. Fourth, the academic work of
the faculty tends to retreat into technocratic postures and to avoid policy issues,
which require attention to framing questions and problems related to defining
appropriate strategies. Much of the work that is accomplished may be important
methodologically but does not confront the types of policy problems where there
are limited formal data sets.
Conclusion
At least three different generations can be identified in this depiction of the
development of the policy analysis field in the United States. It is not surprising
that this field is constantly developing and moving since it operates within a
highly turbulent environment. After more than 50 years, the policy analysis field
has taken its place in the contemporary world of decision making but it is also
clear that the original view of the policy analysis activity as a process of advising
the prince or ruler is very limited.
Policy analysis is now a field in the United States and around the globe with
multiple languages, values, and forms, and with multiple individuals and groups
as clients. Its conversations take place across the policy landscape. Though its
diversity is frustrating and dissonant to some, it also mirrors American society
itself. Instead of being a profession captured and controlled by an elite few, it is
now democratized, with an eclectic set of practices and people who represent
very different interests. As the field has become more complex, however, it is
increasingly difficult to establish a clear sense of professional norms and a definition
of success.
48
As the profession has changed over time, policy analysts have struggled to figure
out what criteria they should use to evaluate their own work. Analysts have fewer
opportunities to point to new programs and take pride in the role they played
in bringing them to life. More frequently than not, analysts circumscribe their
activity to small-scale issues and working at the margins of policy debates. Given
the range of actors and analysts involved in policy discussions, it is difficult to
point to one’s individual contribution to the process.
Analysts continue to confront the balancing act that has been a part of the
profession since its early days: the conflict between responsiveness to the client
and values of rigor and fairness. Although few analysts would wrap themselves
around abstract concepts of neutrality and objectivity, recognizing that values and
interests are implicit in most policy discussions, they do want to produce work
and advice that are as complete and balanced as possible. The norms established
for the academic expression of the field are not always sensitive to the real world
of decision making. Though this has been a challenge for analysts since the early
days of the profession, the boundary lines between analysis and advocacy have
made this task more difficult in recent years.
There seems to be a disconnect between the analyst’s perception of self-
worth (often drawn from Era one) and the real contribution that the individual
makes in the nooks and crannies of the policy process. Many analysts seem to
believe that formal studies or reports are the only types of products that are
noteworthy. They seem to need a language to describe what they do and to
convince themselves—as well as others—that they contribute to the process.
They are challenged to think about ways of developing standards for their work
that reflect their diverse experiences. What are standards against which work is
measured? Is it drawn from a peer review model? Should accountability revolve
around questions of utilization? How should one deal with the perception that
this is an elitist profession? Neither has there been a commitment to encourage
self-reflection on the part of seasoned professionals and to give them the tools to
think about their contributions.
There has been very little attention given to patterns of career development in
the field, the kinds of skills that are now appropriate for the profession, and the
expectations about packaging or marketing the results of the analytic process.
Students should be familiar with techniques such as organizing and running
focus groups; they have to figure out ways to gauge the intensity level around
certain issues, assessing not only classic political feasibility questions but also the
level of venom around some policy areas. They are required to be sensitive to the
demands of making presentations to busy clients and to acknowledge that many
clients participate in networks with multiple players and diverse perspectives.
49
Future changes in the policy environment
Policy analysts can expect to deal with several shifts in the policy environment.
These include the end of budget deficits, globalization, devolution, and political
conflict, all of which challenge the current framework that has been developed
in the early years of the profession.
Fiscal scarcity
For many years, analysts operated within the confines of budget deficits, the
parameters of policy possibilities being constrained by very limited resources.
In the short run, the decision rules that have been adopted for creating budgets
will continue to constrain activity at the national level and largely emphasize
short-term proposals.
Globalization
Increasingly policy issues have global dimensions, and analysts will be required
to think about the ways that policy decisions framed in the context of the U.S.
system will affect policies in other countries. Conversely, when thinking about
policy issues in the United States, analysts are challenged to think about the effects
of international perspectives on domestic policy.
Decentralization and devolution
Increased interest in federal systems and other forms of decentralization, the

growth of the European Union, and the end of the Soviet Union have all created
complex political and social systems that share some attributes with the United
States. These developments seem to have contributed to interest in analytic and
advising processes that had not been contemplated in traditional parliamentary
or controlled systems.
Over the past several decades, devolution of responsibilities to state and local
government and reliance on contracting in the private sector has shifted the points
of leverage for policy change away from Washington. Devolution of policies to
state and local levels has also created new opportunities and interest in policy
analysis by nongovernmental groups.
Political and ideological conflict
It appears that the current environment of ideological and partisan conflict is likely
to continue into the future. Given the intensity of the conflict in the contemporary
setting, it is likely that—at least in the short run—policy analysts will continue
to confront the pressures of ideological and partisan debates. The highly charged
50
political and partisan environment has created increased pressure to focus on the
views of the client rather than the lessons that emerge from balanced analysis.
A commitment to democratic values
There have been commentators throughout the evolution of the policy analysis
field who have expressed fears about the monopoly on analysis found in the early
years of practice. Analysts are now attached to a pluralistic and fragmented policy
world. Though this has opened up the profession, it has also made it more difficult
to define in rigorous terms the attributes that make up ‘good’ policy analysis or
to find effective ways of dealing with networks.
More people are now aware of the limitations of focusing only on efficiency
values as they establish criteria for assessing options. Technology developments
and a commitment to open information have eroded a tendency for analysts to
hold information close at hand. Efforts to democratize data have increased the
ability of various players to have information available to them. Concern about
transparency has become stronger and has had an impact on the work of policy
analysts.
Today the situation is very complex. We are a field with multiple languages,
values, and forms, and with multiple individuals and groups as clients. We confront
conflict between analysts who strive for objectivity and those who begin with
an advocacy posture. Often our work involves dueling between analysts. Our
conversations take place across a very broad policy landscape. We operate with
many different definitions of success. Some of our efforts meet these expectations
and others do not. But policy analysis in the United States today has taken its
place in the contemporary world of decision making.
Notes
1 This chapter draws on Beryl A. Radin (2013) Beyond Machiavelli: Policy Analysis Reaches Midlife
(2nd edn) Washington, DC: Georgetown University Press.
2 Heclo (1972, p 117) suggests that developments in political science did not always support an
enthusiastic move toward policy advising.
3 Reprinted in Dror (1971, pp 225–34).
4 Much of this discussion is drawn from Radin (1997).
5 The term was originally coined by Rittel and Weber (1973).
References
Beam, D. (1996) ‘If Public Ideas Are So Important Now, Why Are Policy Analysts
So Depressed?’ Journal of Policy Analysis and Management, 15(3): 430–37.
Dror, Y. (1971) Ventures in Policy Sciences, New York: Elsevier.
Heclo, H. (1972) ‘Modes and Moods of Policy Analysis’, British Journal of Political
Science, 2(1): 83–108.
Heineman, R., Bluhm, W., Peterson, S., and Kearny, E. (1990) The World of the
Policy Analyst: Rationality, Values, and Politics, Chatham, N.J.: Chatham House
Publishers.
51
Hogwood, B. and Gunn, L. (1984) Policy Analysis for the Real World, Oxford:
Oxford University Press.
Malbin, M. (1980) Unelected Representatives: Congressional Staff and the Future of
Representative Government, New York: Basic Books.
May, P. (1998) ‘Policy Analysis: Past, Present, and Future’, International Journal of
Public Administration 21(6–8): 1098.
McGuire, M. and Agranoff, R. (2011) ‘The Limitations of Public Management
Networks’, Public Administration 89(2): 266
Meltsner, A. (1976) Policy Analysts in the Bureaucracy, Berkeley, CA: University
of California Press.
Metcalf, C. (1998) ‘Presidential Address: Research Ownership, Communication
of Results, and Threats to Objectivity in Client-Driven Research’, Journal of
Policy Analysis and Management 17(2): 153–63.
Nakamura, R. and Smallwood, F. (1980) The Politics of Policy Implementation, New
York: St. Martin’s Press.
Nelson, R. (1989) ‘The Office of Policy Analysis in the Department of the
Interior’, Journal of Policy Analysis and Management 8(3): 395–410.
Primus, W. (1989) ‘Children in Poverty: A Committee Prepares for an Informed
Debate’, Journal of Policy Analysis and Management 8(1): 23–34.
Radin, B. A. (1984) ‘Evaluations on demand: Two Congressionally mandated
education evaluations’, in R. Gilbert (ed) Making and managing policy, New
York, NY: Marcel Dekker.
Radin, B. A. (1992) ‘Policy Analysis in the Office of the Assistant Secretary for
Planning and Evaluation in HEW/HHS: Institutionalization and the Second
Generation’, in C. Weiss (ed), Organizations for Policy Analysis: Helping Government
Think, Newbury Park, CA: Sage Publications, 144–60.
Radin, B. A. (1997) ‘Presidential Address: The Evolution of the Policy Analysis
Field: From Conversation to Conversations’, Journal of Policy Analysis and
Management, 16(2): 204–18.
Rein, M. and White, S. (1977) ‘Policy Research: Belief and Doubt’, Policy
Analysis 21(1): 239–71.
Rittel, H. and Weber, M. (1973) ‘Dilemmas in a General Theory of Planning’,
Policy Sciences 4: 155–69.
Rivlin, A. M. (1998) ‘Does Good Policy Analysis Matter?’ Remarks at the
Dedication of the Alice Mitchell Rivlin Conference Room, Office of the
Assistant Secretary for Planning and Evaluation, Department of Health and
Human Services, Washington, D.C., February 17.
Robinson, W. (1989) ‘Policy Analysis for Congress: Lengthening the Time
Horizon’, Journal of Policy Analysis and Management 8(1): 2.
Robinson, W. (1992) ‘The Congressional Research Service: Policy Consultant,
Think Tank, and Information Factory’, in C. Weiss (ed), Organizations for Policy
Analysis: Helping Government Think, Newbury Park, CA: Sage Publications, 184.
52
Thompson, P. and Yessian, M. (1992) ‘Policy Analysis in the Office of Inspector

General, U.S. Department of Health and Human Services’, in C. Weiss (ed),
Organizations for Policy Analysis: Helping Government Think, Newbury Park, CA:
Sage Publications, 161–77.
Vining, A. and Weimer, D. (2010) ‘Foundations of Public Administration: Policy
Analysis’, Public Administration Review.
Weiss, C. (1983) ‘Ideology, Interests and Information: The Basis of Policy
Positions’, in D. Callahan and B. Jennings (eds), Ethics, the Social Sciences, and
Policy Analysis, New York: Plenum Press.
Weiss, C. (1989) ‘Congressional Committees as Users of Analysis’, Journal of Policy
Analysis and Management 8(3): 411.
Weiss, C. (1998) Evaluation: Methods for Studying Programs and Policies, New Jersey,
NJ: Prentice Hall.
Wildavsky, A. (1969) ‘Rescuing Policy Analysis from PPBS’, Public Administration
Review 29: 189–202.
Wildavsky, A. (1976) ‘Principles for a Graduate School of Public Policy’, Journal
of Urban Analysis, in A. Wildavsky (1979) Speaking Truth to Power: The Art and
Craft of Policy Analysis, Boston, MA: Little, Brown & Co., 407–19.
Wildavsky, A. (1997) ‘Rescuing Policy Analysis from PPBS’, in Classics of Public
Administration, Fort Worth, TX: Harcourt Brace College Publishers, 275–88.
Williams, W. (1998) Honest Numbers and Democracy, Washington, D.C.:
Georgetown University Press.
53
THREE
The argumentative turn in public policy

inquiry: deliberative policy analysis for
usable advice
Frank Fischer
The argumentative turn in the development of the field in the United States
emerged from a practical problem. Early into the development of policy analysis it
became apparent that much of the work was either of questionable value or even
useless to policymakers. Toward the end of the 1970s scholars were becoming
increasingly aware of the significant disjunction between social research and the
practical world of policy decision making. An effort to explain this became a
pressing question for the newly emerging discipline. It was in response to this
disturbing problem that the argumentative turn had its origins.1
The discussion here opens with a discussion of the basic issues involved in
the effort to establish a discipline capable of offering effective advice for policy
decision makers. The focus then shifts to problems that appeared in the initial
stages of the field by the application of empirical social scientific methods to policy
problem solving. Two competing perspectives for confronting these difficulties are
outlined, one political in nature and the other dealing with epistemological and
methodological issues. The focus then turns to the methodological orientation as
delineated in the argumentative turn. Finally, after presenting a logic framework
for policy argumentation, the chapter closes with a discussion of the implications
of the approach for the relation between citizens and policy analysts.
Knowledge, politics, and policy: early developments

The role of political and policy advice are not new questions in politics and
governance. The significance of politically-relevant counsel, together with the
question of who should supply it, is found in the earliest treatises on political
wisdom and statecraft. It is a prominent theme in the philosophy of Plato, in
Machiavelli’s discussion of the role of the Prince, St. Simon and August Comte’s
theory of technocracy, the ‘Brain Trust’ of Franklin Roosevelt’s New Deal, the
writings of the U.S policy intellectuals during the Great Society of the 1960s,
and modern-day think tanks since Ronald Reagan and Margaret Thatcher, just
to name a few of the more prominent instances (Fischer 1990). Today it is a
prominent topic in the discourses of policy theory.
From the outset, a basic theme running through such writings has focused on
replacing or minimizing political debate with more rigorous or systematic modes
55
of thought. Within this tradition we recurrently find the idea that those with
knowledge should rule (Fischer 1990). By the time of St. Simon and Comte such
writings involved a call for a more ‘positivist’ epistemological form of knowledge,
emphasizing testable theory rather than the ‘negativism’ of earlier philosophers
based on critique and speculative reason. Whereas philosophical critique was
seen to provide no concrete foundation for social progress, the accumulation
of positivist knowledge provided the basis for the building of a harmonious,
prosperous society.
Policy analysis is in the first instance an American story with roots in the
Progressive Era of American politics around the turn into the 20th century. During
this period, there was an explicit call for technocratic expertise in politics. Herbert
Croly, a prominent Progressive thinker of the period, became a devotee of the
theories of Comte (Levy 1985). Drawing on Comte’s theory of technocracy, he
promoted the role of policy expertise as an approach to political reform designed
to deal with the pressing social and economic problems brought on by rapid urban
industrialization. Specifically, he and many leading Progressives called for the
adoption of the principles and practices of Taylorism and ‘Scientific Management’
to guide such reforms (Wiebe 1967).2 In this view, a ‘neutral’ scientific method
as opposed to political argumentation would show the way to the good society.
The influence of this ‘Progressive’ creed is attested to by the election of two
Progressive presidents in the United States, one a Republican—Theodore
Roosevelt—and the other a Democrat—Woodrow Wilson. These elections
and the politics surrounding them reflected the would-be ‘value-free’ nature of
scientific management and the ‘nonideological’ approach to good government
through expertise that it prescribed. Indeed, Wilson, himself a political scientist
(and often considered one of the founders of the American discipline of public
administration), called for scientifically explicating the efficient practices of
Prussian bureaucracy and applying them—independently of culture or context—
to American government. In the process, this new field of inquiry and its practices
would replace the traditional emphasis on legal-rational bureaucratic authority
with social scientific principles of organization and management (Wilson 1987;
Weber 1978).
Technocracy received renewed support in the 1920s from Thorstein Veblen
(1933) and his call to replace the capitalist price system with the decisions of
efficient engineers. The goal for Veblen was to eliminate waste and encourage
efficiency in the American economy through methods and practices of experts.
Pinning his hopes on the political possibilities posed by a newly emerging
professional class, he called on the ‘absentee owners’ of corporate America to
transfer their power to reform-minded technocrats and workers. Throughout
this early work the term ‘technocracy’ was seen as a positive new direction for
government. The message was straightforward: replace the talk of business owners
or politicians with the analysis of the experts. Some technocratic thought was
influenced by emerging socialist governments emphasizing planning, although it
56
The argumentative turn in public policy inquiry
was often difficult for this enthusiasm to be explicitly expressed, especially when
it related to the newly formed Soviet Union.
A major expansion of the Progressive ideas came with New Deal ‘Brain
Trust’ of the 1930s, also adding a new dimension of expert advice. Particularly
influential were the writings of John Maynard Keynes, holding out the possibility
of technically managing the economy (Graham 1976). This period saw a large-
scale influx of economists and social scientists in Washington, employed in policy
planning and related activities throughout the growing administrative state. It
also featured a change in the politics of the policy process, or what came to be
called the New Deal ‘Liberal Reform Strategy’ (Karl 1975). The political strategy
brought experts and politicians together, through so-called ‘Blue Ribbon’ advisory
commissions, to identify and analyze contemporary social and economic problems.
The reports of these commissions were then turned into central components of
the liberal reform agenda by party politicians, subsequently offered to voters in
electoral campaigns.
In the postwar period the emphasis on the role of empirical social science in
political advisement took full form in the development of the field of policy
analysis, from Harold Lasswell to Aaron Wildavsky. Although policy analysis
emerged in less grandiose terms than Lasswell had envisioned it, including his
recognition of the need to include more than empirical investigation, the field
and its practices that emerged in the 1960s were dominated by positivist-based
methods.
A worry here was the fact that the public seemed to be disturbingly uninformed
and often irrational in their thinking. A major manifestation of this was fascism
in Europe during World War II. Another was the emergence of various studies
in the United States that showed relatively high levels of ignorance on the part of
the public, raising important questions about the viability of democratic politics
in an increasingly complex society. It was a theme that resonated with Dewey’s
(1927) earlier apprehensions about the ability of the public to meaningfully engage
issues in a technological society. In fact, Lasswell emerged from the context of
Progressivism and was decisively influenced by American Pragmatism in the form
of Dewey’s approach to social problem solving. To address Dewey’s problem of the
public, Lasswell outlined what he referred to as the ‘policy sciences of democracy.’
It was an orientation designed as an approach for providing politicians, policy
decision makers, and the public with the relevant knowledge needed to render
informed decisions.
Toward this end, Lasswell (1951) advanced a grand multidisciplinary approach
that included qualitative disciplines like anthropology as well as economics and
political science. The approach recognized that policy analysis requires two
dimensions: It needs not only a contextual understanding of how policy processes
work, but also a body of causal knowledge related to the particular problems at
issue—poverty, environment, and so on. Taken together, this has proven to be
no small order.
57
Technocratic expertise and the rise of counter-expertise:

policy argumentation
By and large, policy analysis neglected Lasswell’s broader theoretical orientation.
Taking root in the United States, the actual practice started out with a much
narrower technical or technocratic perspective, influenced by World War II
operations research in combination with postwar developments in systems analysis.
The field, as we know it today, emerged with two US wars in the 1960s—the
War on Poverty and the Vietnam War—before traveling abroad. Both were driven
by political forces committed to ‘rational’ policy analysis.
The War on Poverty started out with a noble goal, namely to end poverty in
American society. This was based on a straightforward theory of policy and policy
analysis. It assumed a relatively direct connection between knowledge and decision
making; scientifically tested social scientific findings would speak directly to the
problems at hand. It assumed that facts and values could be neatly separated and
therefore policy research could be largely value-free. And it assumed that better
arguments based on such knowledge could diminish, if not eliminate, political
disagreement. When politicians and the public possess empirically demonstrated
facts and solutions, so went the contention, what was there left to argue about?
The approach also presumed, in theory, that we have adequate causal problem-
oriented knowledge to intervene effectively. That is, we know the reasons why
problems such as poverty exist and that such knowledge can be translated into
action programs (Farmbry 2014). But experience rather quickly showed this to
be a serious misjudgment. Much of the War on Poverty, based on research funded
earlier by the Ford Foundation, tended to deal more with the consequences of
poverty rather than attacking its underlying causes. And the Vietnam War, as
the Secretary of Defense Robert McNamara (1996) later admitted, was largely
based on a misunderstanding of Vietnamese culture and political motivations,
with disastrous results for both the Vietnamese and the United States.
In this early phase, there was a great flurry of activities devoted to policy
formulation, using causal findings about poverty from various disciplines, packaged
as a comprehensive program (Wood 1993). The simplicity of the policy thinking
was captured by the ‘Moon/Ghetto’ metaphor (Nelson 1977), which assumed a
straightforward equivalency between engineering and social engineering: If the
country could put a man on the moon, it could eliminate poverty in an ‘affluent’
society. At the same time, the Vietnam War was pursued technocratically by the
Pentagon, using elaborate systems-based planning methods and a reliance on
empirical measures, such as body counts and kill ratios.
The enthusiasm for such methods established policy analysis as both an attractive
academic specialty and an emerging job category (an interesting story unto itself).
There was a ‘policy analysis explosion’, as the editors of Society magazine titled
a symposium on the subject (Nagel 1979). Moreover, this was academically
contagious. Both policy theory and policy analysis started to travel across to
various European countries, even if only slowly at first. The Netherlands, a land
58
long known for its import/export industries, was one of the first to take up this
literature and bring this enterprise to Holland, thanks in significant part to Robert
Hoppe. And slowly other European countries began to follow suit, especially in
northern Europe. This gave rise to ongoing exchanges between European and
American policy scholars. Some Europeans even made their careers by more
or less representing U.S. theories in their own countries. But many also began
by making new contributions as well, especially in matters related to policy
implementation and evaluation.
Policy-analytic theory, however, encountered some unexpected and very
challenging questions along the way, particularly as they related to the promise
of problem solving. Early on the ‘utilization’ of knowledge proved to be more
difficult than conventional policy theory suggested. Not only did the United
States fail to solve the poverty problem, it flat out lost the Vietnam War, despite
the policy roles of the ‘best and brightest,’ as the phrase went (Halberstram 1972).
The concern was captured by the Secretary of Defense James Schlesinger
(1969, p 310), when he testified to Congress that ‘everyone is, in principle, in
favor of policy analysis but few are hopeful that its conclusions will be utilized
in real-world-policy making.’ And the Chairperson of the Council of Economic
Advisors, Charles Schultz (1968, p 89), concluded that ‘What we can do best
analytically, we find hardest to achieve politically.’
These concerns gave rise to a new specialization and journal in the field
concerned with the question of utilization. The journal Knowledge explored both
practical cases studies concerning the uses—successes and failures—of policy
findings and broader discussions of the relation of knowledge in society more
generally. Along the way, others such as Rittel and Webber (1973) spoke of ‘wicked
problems,’ in which not only the solutions are missing, but the definition of the
problem as well is unknown or uncertain. Today, scholars also refer to ‘messy
problems’ to capture the uncertainty and risks associated with complex problems
like climate change or global financial crisis (Ney 2009).
In search of usable advice: political and disciplinary responses

There were broadly speaking two competing responses to these setbacks, one
political and the other academic. On the political front, conservatives singled out
the methods and practices of policy analysis as ‘metaphysical madness,’ an argument
that picked up steam from the Nixon to the Reagan years. Send these ‘so-called
experts’ back to the ivory tower, these conservatives argued: bring back lawyers
and businessmen, viewed to be closer to the practical realities of main street.
However, these conservatives—both politicians and intellectuals—did not
end up rejecting policy analysis; rather they turned it to conservative purposes.
Instead of using policy analysis to create new social programs, they discovered
that they could also use it to cut or eliminate the same programs. Counseled by
conservative policy intellectuals, these administrations shifted first to ‘evaluation
research’ directed to program outcomes rather than policy formulation inputs.
59
Insofar as social programs often show mixed results, given the complexity of social
problems, they discovered that they could use rigorous, empirical evaluations to
provide arguments for cutting programs out of the budget.
It is not hard to challenge programs with either uncertain—or mixed—
outcomes or on methodological grounds. Questions related to how the data
were obtained, which definitions and assumptions were employed, the nature
of the sample groups, what kinds of statistical tests were employed, and the like.
Indeed, this gave rise to a kind of ‘politics of methodology.’
To further facilitate this orientation, the economists’ method of cost-benefit
analysis was elevated by conservatives to serve as the test for all programs. Given
the difficulties in monetizing social benefits as opposed to costs (most of which
are much easier to identify than benefits), many liberal programs—especially
social programs—have trouble surviving a cost-benefit test. A cost-benefit test
can thus function to filter out deliberation on social-democratic issues by adding
a business-friendly overlay on the decision process. This is particularly the case
for programs justified in the name of the ‘public interest’ or ‘common good.’
It is not that deliberative argumentation vanishes; the decision makers always
deliberate among themselves. It is rather the scope of the deliberation that is at
issue—in particular, who is excluded. In this regard, cost-benefit analysis and
evidence-based decision making tend to be used to justify political decisions
made behind closed doors.
For conservatives, usable policy knowledge involves a mix of program
performance data and comparisons of costs and benefits, presented as the essence
of rationality in decision making. This approach was further supported by ideas
about ‘evidence-based’ policy analysis, which emerged in Britain during the Blair
years, before spreading to other parts of the world. In short, empirical objectivity
became the privileged mode of thinking, as regularly reflected in the rhetoric of
politicians and policymakers. This orientation is not all that different from the
earlier approaches the conservatives were criticizing. But this was about partisan
politics, not usable knowledge per se.
The competing response to the problem of usable knowledge came mainly
from academic policy scholars, who recognized the need to confront the fact
that policy analysis—though widely commissioned as an approach to problem
solving—was not widely used. Research showed that some two thirds of policy
research was never used in one way or another by the decision makers who paid
for it. This led deLeon (1989) to ironically comment that policy analysis would
not be able to pass a cost-benefit analysis.
One of the first efforts to put this problem in theoretical perspective came from
Collingridge and Reeve (1989), two British researchers who referred in their
book, Science Speaks to Power, to the ‘unhappy marriage’ between social science
and policy. They developed a two-category model—one, an ‘over-critical model’
in which scientific contributions lead only to continuing technical debate and
thus fuel political conflict; the other an ‘under-critical model’ in which analysis
serves to legitimate predetermined political positions. These two models align
60
with how conservatives—as well as many politicians generally—have used analysis:

either to hamper programs they do not like with further expert debate (such as
in the case of climate change) or to justify their existing positions (for example,
fiscal austerity).
The problem led Lindblom and Cohen (1978) to ask: What is ‘usable
knowledge’? Insofar as giving up on policy analysis in a complex and uncertain
world seemed unwise, the challenge was to discover what sort of knowledge might
be used. The first consideration was the fairly obvious recognition that politicians,
in whichever country we find them, are wired differently than professional policy
analysts. While policy analysts seek to be problem solvers, this is only a secondary
concern of politicians. Uppermost in their minds is not whether their decisions
will solve a particular policy problem, but rather what impact their actions will
have on their ability to stay in office and/or to make further career gains. For
this reason, politicians choose personally trusted advisors to ensure that only ‘safe’
information reaches them. Despite the orientation of many policy-analytical
scholars, their advisors are often not chosen for their policy-analytic expertise,
but for their ability to protect their interests.
This tendency, of course, is not new, but we still tend to overlook its deeper
implications. More typically, policy analysts look for ways to break through to
decision makers—to bridge the gap—rather than to fully acknowledge and
confront the depth of the divide. Most mainstream practitioners—in the United
States, Europe, and elsewhere—still approach policy analysis as a form of rational—
or at least ‘semi-rational’—problem solving. Even if the process is not entirely
‘rational,’ so the argument goes, positivist analysts should strive to make it more
rational. But this has largely failed; we need to rethink our practices.
As a move in this direction, Lindblom and Cohen (1978) find the solution to
the problem in turning away from an exclusive emphasis on professional-analytic
inquiry to what they have called ‘ordinary knowledge.’ In this view, the idea would
be to improve our ‘ordinary’ policy knowledge rather than further pursue—in
vain—the effort to supply scientifically validated policy knowledge. This move
would, at the same time, break down the divide between analysts and decision
makers (practitioners who communicate and trade in ordinary knowledge). It
would, as such, help decision makers on their own terms, rather than only those
of academic social science (Fischer 2000).
Another approach that moved in a similar direction was advanced by Carol
Weiss (1977). In response to the failures to produce usable knowledge, she called
for a shift from the problem-solving focus to what she called an ‘enlightenment
function.’ In this view, policy analysis is seen to play a less technical, more
intellectual role. Rather than seeking technical solutions—which go wanting
anyway—the task is more appropriately understood as helping decision makers
think about the problems they face—that is, to improve their understanding with
the help of relevant findings and analytical perspectives (Fischer 2009). This role
might not be the direct one that policy analysts have sought, but it is not to be
underestimated. For one thing, policy concepts have clearly penetrated political
61
discourse; politicians speak of a ‘culture of poverty,’ a ‘broken window theory of

crime ‘clash of civilizations,’ and so on). Indeed, how to think about and frame
issues is what social scientists do well. Seen this way, social scientific policy inquiry
is scarcely a failure. It is on this perspective which the argumentative turn has
sought to build an alternative methodological orientation.
The argumentative turn

The concept of the ‘argumentative turn’ in policy analysis was first introduced
by Frank Fischer and John Forester in 1993. They set out a new orientation in
policy analysis that represented a shift away from the traditional empirical-analytic
approach to problem solving to include the study of language and argumentation
as essential dimensions of theory and analysis in policy making and planning. The
book was instrumental in stimulating a large body of work in policy research
both in the United States and Europe over the subsequent decades. Since the
publication of the book, the emphasis on argumentation has converged with
other developments in the social sciences focused on discourse, deliberation,
social constructivism, narration, and interpretative methods (See Fischer and
Gottweis 2012).
In pursuit of an alternative approach to policy inquiry, the argumentative
turn links postpositivist epistemology with social and political theory and the
search for a relevant methodology.3 At the outset, the approach emphasized
practical argumentation, policy judgment, frame analysis, narrative storytelling,
and rhetorical analysis, among other things (Gottweis 2006). From the early
1990s onward, argumentative policy analysis matured into a major strand in the
contemporary study of policy making and policy theory development. As one
leading policy theorist put it, this perspective became one of the competing
theoretical perspectives (Peters 2004).
Over these years the argumentative turn expanded to include work on discourse
analysis, deliberation, deliberative democracy, citizen juries, governance, expertise,
participatory inquiry, local and tacit knowledge, collaborative planning, the role
of the media, and interpretive analysis, among others. Although these research
emphases are scarcely synonymous, they all focus attention of communication
and argumentation, especially the processes of utilizing, mobilizing, and assessing
communicative practices in the interpretation and praxis of policy making
and analysis (Fischer 2003; Gottweis 2006). First and foremost, argumentative
policy inquiry challenges the view that public policy analysis can be a value-free
endeavor.4 Whereas neopositivist approaches generally embrace a technically
oriented rational approach to policy making—an attempt to provide unequivocal,
value-free answers to the major questions of policy making—the argumentative
approach rejects the idea that policy analysis can be a straightforward application
of scientific techniques.5 Instead of a narrow focus on empirical measurement of
inputs and outputs, it takes the policy argument as the starting point of analysis.
Without denying the importance of empirical analysis, the argumentative turn
62
seeks to understand the relationship between the empirical and the normative as
they are configured in the processes of policy argumentation. It thus concerns itself
with the validity of empirical and normative statements but moves beyond the
traditional empirical emphasis to examine the ways in which they are combined
and employed in the political process. This orientation is especially important
for an applied discipline such as policy analysis. Insofar as the field exists to serve
real-world decision makers, policy analysis needs to be relevant to those whom it
attempts to assist. The argumentative turn, in this regard, seeks to analyze policy to
inform the ordinary-language processes of policy argumentation, in particular as
reflected in the thought and deliberation of politicians, administrators, and citizens
(Lindblom and Cohen 1979). Rather than imposing scientific frameworks on the
processes of argumentation and decision making, theoretical perspectives generally
designed to inform specific academic disciplines, policy analysis thus takes the
practical argument as the unit of analysis. It rejects the ‘rational’ assumptions
underlying many approaches in policy inquiry and embraces an understanding
of human action as intermediated and embedded in symbolically rich social and
cultural contexts.
Recognizing that the policy process is constituted by and mediated through
communicative practices, the argumentative turn therefore attempts to understand
both the process of policy making and the analytical activities of policy inquiry
on their own terms. Instead of prescribing procedures based on abstract models,
the approach labors to understand and reconstruct what policy analysts do when
they do it, to understand how their findings and advice are communicated, and
how such advice is understood and employed by those who receive it. This
requires then close attention to the social construction of the normative—often
conflicting—policy frames of those who struggle over power and policy.
These concerns take on special significance in today’s increasingly turbulent
world. Contemporary policy problems facing governments are more uncertain,
complex, and often riskier than they were when many of the theories and methods
of policy analysis were first advanced. Often poorly defined, such problems have
been described as far ‘messier’ than their earlier counterparts—for example, climate
change, health, and transportation (Ney 2009). These are problems for which
clear-cut solutions are missing—especially technical solutions—despite concerted
attempts to identify them. In all of these areas, traditional approaches—often
technocratic—have proven inadequate or have failed. Indeed, for such messy
policy problems, science and scientific knowledge have often compounded
problem solving, becoming themselves sources of uncertainty and ambiguity. They
thus generate political conflict rather than helping to resolve it. In a disorderly
world that is in ‘generative flux,’ research methods that assume a stable reality
‘out there’ waiting to be discovered are of little help and prone to error and
misinterpretation (Law 2004, pp 6–7).
The remainder of the chapter explores a number of aspects of the approach,
particularly as they pertain to giving policy advice. While discussion of policy
expertise and advice-giving in politics is in no way new, the practice of expertise
63
and advice changed over the latter part of the 20th century (See the chapter by
Radin in this volume). The author seeks to show that the argumentative turn
can be usefully employed to interpret these changing patterns, and also to offer a
number of potentially useful prescriptions. Toward this end, the discussion begins
with a brief presentation of the long history of thought about political expertise
and advice, moving from political philosophy to social science. It then examines
the limits of the social scientific approach and presents the new argumentative
approach that has emerged as its challenger, especially as reflected in the field of
policy analysis. The advantages of a postempiricist argumentative approach to
policy inquiry are then presented and several practical contributions to policy
deliberation based on its tenets are outlined. The discussion closes with an
examination of the further postpositivist challenges posed by policy deliberation
and the argumentative approach.
The logic of policy argumentation

The turn to argumentation offers a useful alternative to the mainstream
technocratic approach to policy analysis and its problems. But an argumentative
approach is in itself not enough to qualify as critical policy analysis. The defining
characteristic of critical policy analysis is the task of assessing standard techno-
empirical policy findings against higher level norms and values. Such an approach,
in other words, requires that policy arguments be submitted to a higher-level
normative critique. To clarify this task, we can start by outlining a logic of policy
argumentation.
Following Majone (1989, p 63) we can understand a policy argument to involve
a complex mix of empirical findings and normative interpretations linked together
by an ‘argumentative thread.’ Whereas conventional policy analysis typically
focuses on the statistical analysis of the empirical elements, the goal of a critical
policy analysis is reflexive deliberation. To be reflexive means not only to focus
on the problems and decisions designed to deal with them, but also to examine
the normative assumption on which they are based. Toward this end, the aim
is to explore and establish the full range of components that the argumentative
thread draws together. This includes ‘the empirical data, normative assumptions,
that structure our understandings of the social world, the interpretive judgments
involved in the data collection process, the particular circumstances of a situational
context (in which the findings are generated or the prescriptions applied), and
the specific conclusions’ (Fischer 2003, p 191). Beyond technical issues, the
acceptability of the policy conclusion in such an analysis depends on this full
range of interconnections, including tacit elements withheld from easy scrutiny.
Whereas it is commonplace for empiricists to maintain that their empirical-
analytical orientation is more rigorous and therefore superior to less empirical,
more normative-based methods, ‘this model of critical policy argumentation
actually makes the task more demanding and complex.’ The policy researcher
64
‘still collects the data, but now has to situate or include them in the interpretive
framework that gives them meaning’ (Fischer 2003, p 191).
In Evaluating Public Policy (Fischer 1995; also see Fischer 2012; Fischer and
Gottweis 2012), the author has presented a multi-methodological scheme for
integrating these levels of analysis along concrete case illustration. Drawing on
Toulmin’s (1958) informal logic of argumentation, the approach offers a four-
level logic of interrelated discourses that moves analysis from concrete questions
concerning programmatic efficiency or effectiveness up through the relevant
situational context of the program, through an assessment of the programmatic
implications for the societal system to the abstract normative concerns concerning
the impact of the policy on a particular way of life. More specifically, they range
from the questions of whether a policy program works, whether or not it is
relevant to the particular situation to which it is to be applied, and how it fits
into the structure and process of the existing society system, to a critique of
this system in terms of the basic normative ideals that do or do not undergird it
(Fischer 1995; 2003). Such a critique, from a Habermasian (1987) perspective,
is motivated by the ‘emancipatory interest’ inherent to human inquiry. It is this
additional set of deliberations—contextual, societal, and normative critique—that
defines critical policy inquiry.
Such deliberations across the four levels offer the critical policy analyst a
framework for organizing a dialectic communication between policy analysts
and the participants involved in a deliberation. It is the case, to be sure, that
many of the dominant policy players will not be interested in engaging in such
an extended deliberation as it would run the risk of exposing basic ideological
beliefs to discussion and criticism. But this in no way renders the perspective
irrelevant. Critical policy analysis is advanced to serve a larger public, not just
the immediate decision makers and stakeholders. It can be used, in this regard,
as a probe for testing opposing arguments, thus permitting other parties to a
deliberative struggle—particularly affected parties—to construct better arguments
that both enrich the quality of the communicative exchanges and, in the process,
increase the chances of more effective and legitimate outcomes.
In this approach, good policy advice depends not only on empirical evidence,
but also on establishing understandings that can help to forge consensus. Here the
best decision is frequently not the most efficient one. Rather, it is the one that
has been deferred until all disagreements have either been discursively resolved
or placated, at least enough to be accepted in specific circumstances. Even
though apparently less-than-optimally efficient, such decisions have a greater
chance of being politically usable and thus implementable. Because deliberation
can eliminate or diminish the counter-reactions of those who otherwise seek
to block a particular decision, it also serves to keep the political group together.
The trust and good faith that results from such deliberation carries forward to
future policy decision making.
The policy argument then is never an objective category per se. It is instead
a politically negotiated social construction. As such, we can understand policy
65
knowledge to be a ‘hybrid’ fusion, a kind of ‘co-production,’ of empirical and

normative elements (Jasanoff 2006; Fischer 2009). More than just a matter
of efficiency, policy statements are fundamentally normative constructs about
why and how an action should be done. As Schram and Neisser (1997) have
demonstrated, a policy also conveys directly and indirectly a deeper narrative
about the particular society of which it is part. It is point clearly illustrated
by Schneider and Ingram’s (1993) discussion of the social construction of a
policy’s target population. In particular, they lay out the ways in which such
constructions—for instance, who are the deserving and undeserving recipients
of public monies—are typically employed by political actors to manipulate public
opinion to gain electoral advantage.
Concluding perspective: policy epistemics

The argumentative perspective seeks to move beyond this divide by opening up
to critical scrutiny the practices of science itself. Of essential importance here
is the fact that the normative elements lodged in the construction of empirical
policy research rest on interpretive judgments and need to be made accessible for
examination and discussion. Recognizing that the social meanings underlying
policy research are always interpreted in a particular sociopolitical context—
whether the context of an expert community, a particular social group, or society
more generally—a fully developed deliberative approach focuses on the ways such
research and its findings are themselves built on normative social assumptions
that, in turn, have implications for political decision making. That is, they are
embedded in the very understandings of the objects and relationships that policy
science investigates. Indeed, the very construction of the empirical object to be
measured is sometimes at stake.
For this reason, empirical policy science cannot exist independently of
normative constructions. While introducing deliberation as a complement to
the analytic process is an important advance over a narrow technical orientation,
from the postpositivist argumentative perspective it can only be understood as
a platform from which the next step can be taken. Beyond a complementary
approach, deliberation has to be moved into the analytical processes as well.
Beyond understanding empirical and normative inquiry as separate activities
that can potentially inform one another, they need to be seen as a continuous
process of inquiry along a deliberative spectrum ranging from the technical to the
normative. As deliberation and interpretative judgment occur in both empirical
and normative research, they need to be approached as two dimensions along a
continuum.
It is important to concede that deliberative policy inquiry enters here relatively
uncharted territories. From a range of practical experiments, such as consensus
conferences and citizen juries, it is evident that citizens are more capable of
participating in deliberative processes than generally recognized (Fischer 2009).
But questions about the extent of participation, as well as when and where it
66
might be appropriate, pose complicated issues and thus need to assume a central
place on the research agenda. Particularly challenging is the question of the
degree to which citizens can actually engage in the analytic aspects of inquiry.
Various research projects demonstrate that they can participate in at least parts
of it (Wildavsky 1997), but where should we draw the line? Reconceptualizing
empirical and normative inquiry along an interpretive continuum offers an
approach for pursuing the answers to these questions. Elsewhere the author has
proposed that such investigation be taken up as a component of a new research
specialization called ‘policy epistemics’ (Fischer 2000).
The development of policy epistemics would draw extensively on work in
social constructivism, the newer interpretive sociology of science, and the post-
Kuhnian work in the philosophy of science. However, where the constructivist
approaches in sociology and philosophy focus on the conduct of science, policy
epistemics and the turn to arguments extend the perspective to the goals of an
applied social science—policy analysis—and the task of giving advice to decision
makers. The orientation, as such, stands between professional policy research
and action-oriented policy making. It thus explores the ways in which research
speaks to different normative perspectives in the world of action—what kinds of
knowledge(s) are relevant to particular situations, what the normative implications
of particular empirical findings are, how empirical policy findings and normative
perspectives can brought together in a deliberative process, and the like. These are
questions that draw on a constructivist perspective but, at the same time, require
different practical modes of thought and deliberation.
Policy epistemics, then, brings together the relevant work in interpretive policy
analysis, deliberative experimentation, discursive practices, and policy narration
to explore ways in which the empirical/normative divide can be opened up to
facilitate a closer, more meaningful deliberation among citizens and experts. As
such, policy epistemics focuses on the ways people communicate across differences,
the flow and transformation of ideas across borders of different fields, how different
professionals groups and local communities see and inquire differently, and the
ways in which differences become disputes. Following the lead of Willard (1996),
it takes the ‘field of argument’ as a unit of analysis. As polemical discussion,
argumentation is the medium through which people—citizens, scientists, and
decision makers—maintain, relate, adapt, transform, and disregard contentions
and the background consensus.
Whereas traditional policy analysis has focused on advancing and assessing
technical solutions, policy epistemics investigates the way interpretive judgments
work in the production and distribution of knowledge. In particular, it examines
the social assumptions embedded in research designs, the specific relationships
of different types of information to decision making, the uses of information,
the different ways arguments move across different disciplines and discourses,
the translation of knowledge from one community to another, and the
interrelationships between discourses and institutions.
67
Elaborating our understanding of the epistemic dynamics of public controversies

would allow for a more enlightened understanding of what is at stake in a
particular dispute, including the evaluation of both competing viewpoints and
the effectiveness of policy alternatives. In doing so, the differing, often tacitly
held, contextual perspectives and values could be juxtaposed, the viewpoints
and demands of experts, special interests groups, and the wider public directly
compared, and the discursive dynamics among the participants scrutinized. This
would by no means sideline or exclude scientific assessment; it would only situate
it within the framework of a more comprehensive evaluation.
These are questions that underlie the kinds of concerns and problems that
confront policy decision makers and citizens. There is no shortage of literature
illustrating the ways in which failures in policy making are often attributable to
simplistic technocratic understandings of the relationship between knowledge and
politics. It is just to these sorts of differing rationalities underlying citizens’ and
experts’ responses to each other that policy epistemics would turn our attention.
While such work would not need to absorb the efforts of the field as a whole, it
could become one of the important specializations in policy analysis. Those who
engage in it could work to facilitate democracy while at the same time studying
more specially the processes that are involved in attempting to extend it. Short
of providing policy solutions per se, it would go a considerable distance toward
making an applied policy discipline more relevant to the needs and interests of
both political decision makers and the citizens. And no less important, as Rorty
(1979) has put it, it would help us ‘keep the conversation going.’Indeed, keeping
the political conversation about policy going has emerged as one of the crucial
issues of the day, given the rise of distorted forms of communication such as ‘fake
news’ and ‘post-truth.’ For this reason, when we look at policy argumentation
in the United States today, we can only conclude that the tasks of deliberative
policy analysis are as important—even more important— than they have ever
been. Although argumentation is part of the policy process generally, as outlined
here, the focus has become even more salient and thus important in contemporary
times. In the current political context of extremely polarized political contestation,
argumentation has begun to spin out of control. Given that politics in the United
States is currently divided into two extremes—Right and Left—fueled by talk
radio and think tanks on both sides of the divide, argumentation has become more
rhetorical, aggressive, and accusatory. Our ability to understand the logic of valid
argumentation is thus crucial to analytically sorting out the competing claims.
Given the very detrimental impact that these new realities have on American
political culture and the ability of both citizens and politicians to meaningfully talk
to one another about the many problems plaguing the country, an understanding
of the nature of deliberation, in particular the relation of normative ideological
claims to empirical evidence, is crucial to the future of democratic governance.
68
Notes
1 This chapter is a revised version of two earlier essays: Fischer, F. (2015), as well as an article that
appeared in French in the French Political Science Review.
2 In line with the enthusiasm of the era, leading University of Chicago sociologist Lester Ward
even argued that legislators should have training in the social sciences to be qualified for office,
if not be social scientists themselves.
3 Postpositivism refers here to a long established tradition in the social sciences that approaches
the social world as uniquely constructed around social meanings inherent to social and political
action. Explanation cannot meaningfully proceed as context-independent or value-free. Thus the
pursuit of methods based on the epistemologies of the ‘hard science’ lead to a misrepresentation
and thus misunderstanding of the social objects under investigation.
4 For theoretical discussion of the postpositivist perspective in policy analysis, see Hawkesworth
(1988), Forester (1999) Yanow (2000), Hajer and Wagenaar (2003), Fischer (2003), Lejano
(2006), Wagenaar (2011), and Fischer and Gottweis (2012). For empirical applications, see
Fischer (1993), Hajer (1995), Schram and Neisser (1997), Bevir and Rhodes (2003), Gabrielyan
(2006), Mathur (2008), Needham (2009), Dubois (2009), Orsini and Smith (2010), Plehwe
(2011), and Sandercock and Attili (2012).
5 Like all concepts, the concepts of ‘positivism’ and ‘neopostivism’ have their limitations.
Nonetheless these concepts have a long tradition in epistemological discussions in the social
sciences. The use of the term ‘neopositivist’ is employed to acknowledge that there have
been a number of reforms in the ‘positivist’ tradition that recognize the limitations of earlier
conceptions of the approach, taken to refer to the pursuit of an empirically rigorous, value-free,
causal science of society. That is, there is no one neopositivst approach. The term is employed
as a general concept to denote an orientation that continues to strive for empirically rigorous
causal explanations that can transcend the social context to which they apply, but recognizes the
difficulties encountered in achieving such explanations. Neopositivist policy analysts (Sabatier,
for example) typically argue that while policy research cannot be fully rational or value-free,
the analysis should nonetheless takes these to be standards toward which they should strive.
For general references to these debates see Bernstein (1971), Hawkesworth (1988), and Fischer
(2009).
References
Bevir, M. and Rhodes, R. (2003) Interpreting British Governance, London:
Routledge.
Collingridge, D. and Reeve, C. (1989) Science Speaks to Power: The Role of Experts
in Policy Making, London: Frances Pinter.
deLeon, P. (1989) Advice and Consent: The Development of the Policy Sciences, New
York: Russell Sage Foundation.
Dewey, J. (1927) The Public and Its Problems, New York: Swallow.
Dubois, V. (2009) ‘Toward a Critical Policy Ethnography: Lesson from Fieldwork
on Welfare Control in France,’ Critical Policy Studies 3(2): 221–39.
Farmbry, K. (ed) (2014) The War on Poverty: A Retrospective, Lexington, MA:
Lexington Books.
Fischer, F. (1990) Technocracy and the Politics of Expertise, Newbury Park, CA:
Sage Publications.
Fischer, F. (1993) ‘Policy Discourse and the Politics of Washington Think Tanks’,
in F. Fischer and J. Forester (eds), The Argumentative Turn in Policy Analysis and
Planning, Durham, NC: Duke University Press, 21–42.
69
Fischer, F. (1995) Evaluating Public Policy, Belmont CA, Wadsworth.

Fischer, F. (2000) Citizens, Experts, and the Environment: The Politics of Local
Knowledge, Durham: Duke University Press.
Fischer, F. (2003) Reframing Public Policy: Discursive Politics and Deliberative Practices,
Oxford: Oxford University Press.
Fischer, F. (2009) Democracy and Expertise: Reorienting Policy Inquiry, Oxford:
Fischer, F. (2012) ‘Debating the Head Start Program: The Westinghouse Reading
Scores in Normative Perspective’, in M. Hill and P. Hupe (eds.), Public Policy,
London: Sage Publications.
Fischer, F. and Gottweis, H. (2012) The Argumentative Turn Revisited: Public Policy
as Communicative Practice, Durham, NC: Duke University Press.
Fischer, F. (2015) ‘In Pursuit of Usable Knowledge: Critical Policy Analysis and
the Argumentative Turn’ in F. Fischer, D. Torgerson, A. Durnova and M. Orsini
(eds), Handbook of Critical Policy Studies, Cheltenham: Edward Elgar Publishing.
Forester, J. (1999) The Deliberative Practitioner, Cambridge, MA: MIT Press.
Gabrielyan, V. (2006) ‘Discourse in Comparative Policy Analysis: Privatization
in Britain, Russia and the United States’, Policy and Society 25(2): 47–76.
Gottweis, H. (2006) ‘Argumentative Policy Analysis’, in J. Pierre and B. G. Peters
(eds), Handbook of Public Policy. Thousand Oaks, CA: Sage, 461–80.
Graham, O. L. (1976) Toward a Planned Society: Roosevelt to Nixon, New York:
Habermas, J. (1987) The Theory of Communicative Action, Cambridge, MA: Polity.
Halberstram, D. (1972) The Best and the Brightest, New York: Random House.
Hajer, M. (1995) The Politics of Environmental Discourse, Oxford: Oxford University
Press.
Hajer, M. and Wagenaar, H. (2003) Deliberative Policy Analysis: Understanding
Governance in the Network Society, Cambridge: Cambridge University Press.
Hawkesworth, M. E. (1988) Theoretical Issues in Policy Analysis, Albany, NY:
SUNY Press.
Jasanoff, S. (2006) ‘Ordering Knowledge, Ordering Society’, in S. Jasanoff
(ed), States of Knowledge. The Co-production of Science and Social Order, London:
Routledge, 13–45.
Karl, B. (1975) ‘Presidential Planning and Social Science Research’, in Perspectives
in American History, 3, Cambridge, MA: Charles Warren Center for Studies in
American History.
Lasswell, H. (1951) ‘The Policy Orientation’, in H. Lasswell and D. Lerner (eds),
The Policy Sciences, Stanford, CA: Stanford University Press.
Law, J. (2004) After Method: Mess in Social Science Research, New York: Routledge.
Lejano, R. (2006) Frameworks for Policy Analysis: Merging Text and Context, New
York: Routledge.
Levy, D. (1985) Herbert Croly of the New Republic: The Life and Thought of an
American Progressive, Princeton, NJ: Princeton University Press.
70
Lindblom, C. E. and Cohen, D. (1979) Usable Knowledge: Social Science and Social
Problem Solving, New Haven, CT: Yale University Press.
Majone, G. (1989) Evidence, Argument, and Persuasion in the Policy Process, New
Haven, CT: Yale University Press.
Mathur, M. (2008) Citizen Deliberation in Urban Revitalization: Discursive Policy
Analysis of Developmental Planning Processes in Newark, Saarbrucken: VDM Verlag.
McNamara, R. (1996) In Retrospect: The Tragedy and Lessons of Vietnam, New
York: Vintage Books.
Nagel, S. (1979) ‘Policy Analysis Explosion’, Transaction 16(6): 9–10.
Needham, C. (2009) ‘Interpreting Personalization in England’s National Health
Service: A Textual Analysis’, Critical Policy Studies 3(2): 204–20.
Nelson, R. H. (1977) The Moon and the Ghetto, New York: W.W. Norton.
Ney, S. (2009) Resolving Messy Problems: Handling Conflict in Environment, Transport,
Health and Aging Policy, London: Earthscan.
Orsini, M. and Smith, M. (2010) ‘Social Movements, Knowledge and Public
Policy: The Case of Autism Activism in Canada and the U.S’, Critical Policy
Studies 4(1): 38–57.
Peters, B.G. (2004) Review of ‘Reframing public policy: Discursive politics and
deliberative practices’, Political Science Quarterly, 119(3): 566–67.
Plehwe, D. (2011) ‘Transnational Discourse Coalitions and Monetary Policy:
Argentina and the Limited Powers of the Washington Consensus’, Critical Policy
Studies 5(2): 127–48.
Rittel, H. W. and Webber, M. D. (1973) ‘Dilemmas in a General Theory of
Planning’, Policy Sciences 4: 155–69.
Rorty, R. (1979) Philosophy and the Mirror of Nature, Princeton NJ: Princeton
University Press.
Sandercock, L. and Attili, G. (2012) ‘Multimedia and Urban Narratives in the
Planning Process: Film as Policy Inquiry and Dialogue Catalysts’, in F. Fischer and
H. Gottweis (eds), The Argumentative Turn Revisited: Public Policy as Communicative
Process, Durham, NC: Duke University Press.
Schlesinger, J. (1969) Testimony to U.S. Congress, Senate Subcommittee on
National Security and International Operations, Planning-Programming-
Budgeting Hearings, 91st Cong. 1st session.
Schneider, A. and Ingram, H. (1993) ‘Social Construction of Target Populations:
Implications for Politics and Policy’, The American Political Science Review 87(2):
334–47.
Schram, S. and Neisser, P. T. (eds) (1997) Tales of State: Narrative in Contemporary
U.S. Politics and Public Policy, New York: Rowman and Littlefield.
Schultz, C. L. (1968) The Politics and Economics of Public Spending, Washington,
D.C.: Brookings.
Toulmin, S. (1958) The Uses of Argument, Cambridge: Cambridge University Press.
Veblen, T. (1933) The Engineers and the Price System, New York: Viking.
Wagenaar, H. (2011) Meaning in Action: Interpretation and Dialogue in Policy Analysis,
New York: M.E. Sharpe.
71
Weber, M. (1978) Max Weber: Selections in Translation, W. G. Runciman (ed),

Cambridge: Cambridge University Press.
Weiss, C. (1977) ‘Research for Policy’s Sake: The Enlightenment Function of
Social Research’, Policy Analysis 3(4): 531–45.
Wiebe, R. H. (1967) The Search for Order, 1877–1920, New York: Hill and Wang.
Wildavsky, A. (1997) But Is It True? A Citizen Guide to Environmental Health and
Safety, Berkeley, CA: University of California Press.
Willard, C.A. (1996) Liberalism and the problem of knowledge: A new rhetoric for
modern democracy, Chicago, IL: University of Chicago Press.
Wilson, W. (1987) ‘The Study of Administration’, Political Science Quarterly 2: 1–23.
Wood, R. C. (1993) Whatever Possessed the President? Academic Experts and
Presidential Policy, 1960–88. Cambridge, MA: MIT Press.
Yanow, D. (2000) Constructing race and ethnicity in American: category-making in public
policy and administration, New York: M.E. Scharpe.
72
FOUR
Reflections on 50 years of policy advice

in the United States
Laurence E. Lynn, Jr.
For which of you, intending to build a tower, sitteth not down first,
and counteth the cost, whether he have sufficient to finish it? … Or
what king, going to make war against another king, sitteth not down
first, and consulteth whether he be able with ten thousand to meet
him that cometh against him with twenty thousand? (Luke 14: 28–31)
Birth of a profession
On August 25 ,1965 President Lyndon B. Johnson sent a directive to the heads of
all federal departments and agencies. They were to use a planning, programming,
and budgeting (PPB) system modeled on the system Secretary of Defense Robert
S. McNamara had installed at the Department of Defense in 1961 as a basis for
their agency’s annual budget recommendations. PPB’s beating heart was policy
analysis. Using PPB is simply good management, LBJ suggested (Lynn 2015).
More important was that LBJ’s mandate was good politics. Launching a well-
publicized effort to demonstrate financial discipline and restraint in pursuing
the administration’s policy priorities would increase the prospects that Congress
would support the president’s pending budget requests.
At the time, Lyndon Johnson was caught in a political vice. He had committed
his administration to a War on Poverty and a Great Society and, as well, to military
success in Vietnam. Congressional leaders of his party were, however, refusing to
consider tax increases to pay for these priorities. Brookings economist Charles
Schultze, then head of the U.S. Bureau of the Budget, had already directed agency
heads to create task forces to identify possible budgetary savings to make room
for higher-priority activities. Two months later, Johnson’s PPB mandate abruptly
superseded Schultze’s earlier directive.
This pivot was a product of bureaucratic politics. Schultze and Joseph Califano,
formerly an aide to McNamara, now on the White House staff, proposed the PPB
mandate to head off a move by other Johnson aides that the two men thought
unwise: committing LBJ to the promulgation of a comprehensive set of long-
term goals for America. Schultze and Califano had discussed introducing PPB
to domestic policy planning on a pilot basis, but they pushed instead for going
all-in. In a fraught political environment, LBJ embraced their bold proposal.
73
Three decades later, Peter deLeon claimed that public policy analysis in its
various guises, along with the emergent ‘policy sciences,’ ‘have become prevalent,
… virtually ingrained in the woof and warp of government’ (deLeon 1997, p 7).
Five decades on, David Weimer says in his overview essay in this volume that policy
analysis is now a recognized profession in the United States, an institutionalized
field of scholarship, teaching, and practice.
The birth and coming of age of this profession have been brilliantly assayed
in the magisterial writings of Beryl Radin and Weimer in this volume and in
many other publications, and the field’s literature adds insights and detail to the
story. I will not attempt to improve on these accounts. Instead, because the five-
decade period from LBJ’s mandate to now coincides with my own career in the
profession—I have participated in it as a scholar, teacher, and practitioner1—I
will offer here personal reflections on the field’s emergence and its raison d’être
in American governance, illustrated here and there with a few personal anecdotes.
Unlike many fields of study and practice, ours does not systematically
incorporate the history of its ideas and its seminal literature into the training of
its professionals, probably because there is no consensus on such things. In what
follows, I will suggest what I think the seminal events, ideas, and literatures are
and why their importance merits greater prominence in our training.
Policy based on calculation? It’s what Jesus would do.

The notion that rulers should sitteth down, counteth, and consulteth before
making consequential policy and operational decisions has biblical origins (and
surely even more ancient ones—Bronze age Babylon is inconceivable without
calculation). Doing so is also simply common sense, the exercise of which has,
as Weimer notes, a long history in the United States.
In my reckoning, the field’s American history includes a number of political,
intellectual, and institutional milestones in its evolving toward where we are now.
• Thomas Jefferson commissioned Merriweather Lewis and William Clark and

a team of scientifically trained experts, through the agency of the American
Philosophical Society, to map the lands included in the Louisiana Purchase
and gather information to inform American policies for settling the west.
• President William Howard Taft created a Commission on Economy and
Efficiency to study ways to improve the efficiency of proliferating government
organizations, including the notion of a national budget.
• Beginning in 1912, municipal research bureaus began informing local
governments about the needs of a growing, urbanizing, industrializing America,
including the concept of a municipal budget and a wide range of new public
services.
• Theodore Roosevelt inaugurated social and economic regulation and expanded
the delegation of administrative authority to administer public services.
74
Reflections on 50 years of policy advice in the United States
• The Budget and Accounting Act 1921 interposed the president and a new
bureau of the budget between federal spending agencies and the Congress,
laying the foundation for a powerful executive branch role in policy making
(Schick 1970). It also authorized creation of the General Accounting Office
(GAO) to audit federal financial accounts.
• In 1927, three private research and training organizations were amalgamated
into the Brookings Institution (whose scholars were, ironically, to oppose the
New Deal as bad fiscal policy).
• President Herbert Hoover’s Research Committee on Social Trends surveyed
America’s development using the scientific mood and method ‘to study
simultaneously all of the fundamental social facts which underlie all our social
problems’ in order to correct the ‘undiscriminating emotional approach and
… insecure factual basis in seeking for constructive remedies of great social
problems’ (Hoover 1933).
• The New Deal inaugurated an American welfare state with sharply expanded
roles for the federal government in economic and social affairs. From 1933 to
1943, when it was abolished by Congress, the National Resources Planning
Board (NRPB), eventually located in the newly created Executive Office of
the President, conducted an ambitious policy planning program.
• Through the 1940s and 1950s the Department of the Interior developed a
tradition of policy planning and analysis initiated during the long tenure of
Harold Ickes as Secretary. (I became head of the successor to the original
Program and Policy Planning Office in 1973.)
• In 1945 Project RAND was instituted at the Douglas Aircraft Company
under contract to the War Department’s Office of Scientific Research and
Development to analyze the potential relationship between military planning
and weapon systems development. Project RAND came under the aegis of
the U.S. Air Force in 1948 as the RAND Corporation which, over the next
several years, developed the Planning-Programming-Budgeting System (PPBS)
that McNamara would later install at the Pentagon.
• In 1946 the Council of Economic Advisers was added to the Executive Office
of the President to provide expert economic analysis and advice to the president
regarding a wide range of economic policy issues in the aftermath of the Great
Depression and World War II.
• An Office of Policy Planning that was to become a fixture in the Office of the
Secretary of State was created by George F. Kennan in 1947.
• In 1951 Daniel Lerner and Harold Lasswell published a seminal edited volume
entitled The Policy Sciences.
• During the 1950s and 1960s, according to then-Comptroller General Elmer
Staats, the Bureau of the Budget undertook numerous activities that prefigured
PPB: review of the cost-benefit analysis in water resource programs required
by Congress; long-range budget projections featuring high, low, and most
likely estimates for use in policy making; a systematic budget preview process
conducted by the Bureau of the Budget, a mission-related functional framework
75
for budget preview and special analyses of selected programs; lawfully

sanctioned performance- and cost-based budgeting; selective development of
formal agency program planning (Staats 1968).
• In 1961 Secretary of Defense Robert S. McNamara instituted the RAND-
Corporation-conceived PPBS, which would become the model for LBJ’s 1965
mandate (discussed above).
• The Congressional Budget and Impoundment Control Act 1974 authorized
creation of the Congressional Budget Office (CBO), which, along with the
GAO and the Congressional Research Service (CRS, formerly the 1914-vintage
Legislative Reference Service) enabled Congress to be a strong counterweight
to the U.S. Office of Management and Budget (OMB, formerly the Bureau
of the Budget) in providing information and analysis for policymakers.
• Congress initiated and enacted the Government Performance and Results
Act (GPRA) in 1993, accelerating the movement of government toward
performance managements at all levels.
In this selective recounting, several important facts (which will be amplified

below) emerge.
1) Contemporary policy analysis institutions have evolved over the course of

American history. LBJ’s 1965 mandate was classic incrementalism: evolution,
not revolution.
2) Far from being antithetical to Madisonian politics, as critics have claimed,
institutionalized policy analysis is a consequence of Madisonian politics—
of ambition checking ambition—and one of its legitimizing elements.
Demonstrably efficient government is good politics at every level, and being
efficient requires analysis.
3) Strengthening the capacity of governments to act reasonably and effectively by
affording their executives (and legislators) direct access to good information
and analysis enables the separation of powers and republican governance.
What were they thinking?

One Saturday morning in 1968, I and some members of my staff were finishing
some tables of calculations that McNamara had asked my boss to have prepared
for him. “I told him,” my boss told me, “to call you directly if he wants to see
them.” The call came, and off I went with the table he wanted to see.
McNamara was alone in his office, standing behind the enormous desk once
used by General John J. Pershing. When I approached, he spread his arms wide
and grinned, calling my attention to the chaos of documents in front of him.
He sat down and I stood beside him as he looked at the table I’d brought. He
scanned the dozens of numbers for several moments then stabbed one of them
with his finger. “That doesn’t look right,” he said. I looked at the number he
was pointing to. I had no idea what might be wrong. “That’s OK,” he said, “just
76
check it.” He thanked me and I left to instruct my staff to check the numbers.
McNamara was right. It was a typo, and we corrected it.
I was a 31-year-old policy analyst working in the Systems Analysis Office, and
I had, in Alice Rivlin’s memorable phrase, ‘a place at the table.’
A great debate
That McNamara’s reliance on PPBS instigated the emergence of policy analysis as a
profession is generally acknowledged. The why’s and the how’s of that emergence
are not so well known, but they are important to understanding why, 50 years
later, that profession is integral to American public governance.
There was a clear and compelling model for the kind of advice-giving illustrated
by the anecdote above: providing policymakers with independent, real-time
access to analytic resources.
“At this time,” Winston Churchill said at the outset of World War II, “each
department presented its tale on its own figures and data. [Some departments],
though meaning the same thing, talked different dialects.”2 So, he went on:
One of the first steps I took … was to form a statistical department

of my own … I now installed my friend and confidant of many years
[P.M.S. Blackett, who was to win the Nobel Prize for physics in 1948]
... with half a dozen statisticians and economists whom we could trust
to pay no attention to anything but realities. This group of capable
men, with access to all official information, was able … to present me
continuously with tables and diagrams.. They examined and analyzed
with relentless pertinacity … and also pursued all the inquiries which
I wanted to make myself (Enthoven and Smith 1971).
The analysis of military operations by Blackett and his colleagues (collectively

known as Blackett’s Circus) gave impetus to the emerging field of operations
research and, through the RAND Corporation, to the founding ideas of policy
analysis.
McNamara took the same Churchillian step, and in the same spirit, in 1961. It
is noteworthy that policy analysis in McNamara’s Pentagon was the responsibility
of the Office of Systems Analysis. The word ‘systems’ was chosen deliberately to
convey the idea that military forces should be analyzed and planned as systems of
interacting components designed to produce a specific, and measurable, type of
capability. Seeing complex policies and programs as ‘systems’ is also at the heart
of domestic policy analysis.
The first question I took up as a policy analyst was this: What is the most efficient
way to create the capability to move military personnel and equipment to theaters
of combat rapidly enough to forestall long-drawn-out hostilities? Approximate,
and reasonable, answers could be obtained using linear programming models
that generated least-cost mixes of aircraft, ships, and pre-positioned equipment
77
to meet military planners’ rapid deployment schedules. Use of these models had
the added advantage of generating ‘shadow prices’ (values [analogous to marginal
costs] assigned to resource constraints designed into them). One of these resource
constraints was overseas airfields and ports needed to facilitate deployment.
At the time, the Department of State was negotiating overseas base rights with
some of our foreign allies. State’s senior officials sought McNamara’s endorsement
of the large appropriations needed to ‘buy’ these rights from our allies. Their view
seemed to be: We must pay whatever is necessary. To McNamara, that sounded
like: Base rights are priceless. Really?
No, not really. We could calculate shadow prices on base rights to support
our war plans using alternative ways of procuring and using ships, aircraft, pre-
positioned equipment, and foreign bases. No single overseas facility was priceless
because there were alternatives.
One afternoon my boss, Alain Enthoven, and I entered a small inner sanctum
at the State Department where we were to brief two-time presidential candidate
W. Averell Harriman, then head of State’s iconic Office of Policy Planning, and
key members of his staff. We explained why, if, say, Okinawa were unavailable,
we could still meet military objectives by buying more or more capable aircraft
and ships or using other bases more intensively. We suggested the costs of doing
that, in effect, were the value of Okinawa. Our message to a giant of American
diplomacy and Democratic politics: think about this issue in a different way;
there are alternatives.
We left guessing that one of our auditors was already calling a military colleague
at the Pentagon and saying, “Now I know what you’re up against!”
A bit of bravado
Based on her experiences as policy analyst and planner in the U.S. Department of
Health, Education, and Welfare in the 1960s, Alice Rivlin noted unapologetically
that ‘a bit of bravado is necessary to overcome the inertia of government, to
get attention, and to win a place at the decision table’ (1971, p 4, italics added).3
Policy analysts had learned, however, that ‘educators, doctors, civil servants, …
even generals’ are ‘knowledgeable, necessary, and not always wrong.’ As to their
contributions, Rivlin believed that ‘analysts have probably done more to reveal
how difficult the problems and choices are than to make the decisions easier.’
Aaron Wildavsky, having founded the Berkeley School of Public Policy,
described ‘the art of policy analysis’ as follows:
Policy analysis must create problems that decision-makers are able to

handle with the variables under their control and in the time available
... by specifying a desired relationship between manipulable means and
obtainable objectives. … Unlike social science, policy analysis must
be prescriptive; arguments about correct policy, which deal with the
78
future, cannot help but be willful and therefore political (Wildavsky

1979, p 16).
Policy analysis? Or muddling?

In his curiously overlooked 1968 book The Politics and Economics of Public Spending,
Charles Schultze, the policy analysis movements progenitor, articulated its goals
(Schultze 1968, pp 19–24):
• ‘careful identification and explanation of goals and objectives of policies and

programs’ in order to ‘widen but not homogenize’ agency perspectives on
their programs and on alternative ways of achieving them;
• analysis of the outputs of public programs;
• measurement of total social as well as program costs projected into the future;
• an analysis of the costs and effectiveness of alternative ways to achieve public
policy objectives; and
• conducting these types of analysis as a systematic part of the budget review process.
To further clarify PPB’s intent, Schultze compared it to Charles Lindblom’s 1959

articulation of ‘The Science of Muddling Through’ (Lindblom 1959).4 Lindblom
emphasized, said Schultze, the deep, subtle, complex, and difficult to specify
conflicts among our social values. ‘We discover our objectives and intensity that we
assign to them only in the process of considering particular programs or policies’ (Schultze
1968, p 38, italics in original). In effect, ends and means are, as a practical and
political matter, chosen simultaneously through processes such as successive
limited approximations and partisan mutual adjustment by which we deliberate
on concrete policy proposals (Lindblom 1979). These choices, moreover, may
change with each round of deliberations.
Schultze conceded that Lindblom’s depiction of political choice seemed
antithetical to the problem-solving rationality of decision making informed by
policy analysis. He insisted, however, that it was not. ‘It is not really important,’
he said, ‘that the analysis be fully accepted by all participants in the bargaining
process’ (1968, p 75). Analysis is intended to help—the word ‘help’ is important—
sort out arguments about the facts and, far from depoliticizing choice, to steer
debate toward issues where there are real differences of value and, therefore,
where political judgment is most needed.
Schultze recognized, furthermore, that it may be necessary to ‘guard against the
naiveté [of policy analysts] who ignore political constraints … and [as well] the
naiveté of the decision maker’ who ignores resource constraints and believes that
nobility of purpose produces good policy outcomes. Schultze’s point is simply
that the scarcity of resources available to governments requires policymakers to
explore how to get the most out of those resources. Policy analysis assists that
exploration, particularly in the context of budget making. If ‘muddle’ we must,
let’s do it with good analysis.
79
One of the first muddles McNamara’s policy analysts confronted was why all
military personnel rotations between the United States and our overseas bases were
by the Navy’s World War II-vintage troopships. The analysts quickly showed—
you could do it on the backs of envelopes—that it would be cheaper to rotate
the troops by air; that way, DoD could save the cost of the military personnel
needed to fill the sealift pipeline.
A no brainer, no? No! A bureaucratic battle was necessary to get the point
across, but it was an early policy analysis success. Policy analysis wasn’t magical
thinking or an affront to politics; it was analytical thinking of potentially significant
practical value.
It caught on. When I joined the systems analysis staff in 1965, all three military
services had their own think tanks to enable them to match wits with McNamara’s
‘whiz kids.’ A decade later, when I joined the faculty of the expanding Kennedy
School of Government at Harvard, numerous contract research firms were assisting
the government with policy analysis and research and program evaluation.
It’s ‘metaphysical madness’! Seriously?

Despite the (to me) evident common sense of the experience-based rationale
for policy analysis of Schultze and Rivlin, and its evident political popularity,
controversies began to erupt in academia. Critics piled on. Policy analysts were
not only ignorant of how the political world works, they posed a danger to
democracy, to community values, and the rule of law.5
• Harvard law professor Lawrence Tribe (1972) criticized policy ‘science’ for
its overemphasis on economic assumptions and instrumental rationality and
for neglecting ‘distributive ends, procedural and historical principles, and the
values … associated with personal rights, public goods, and communitarian
and ecological goals’ (p 105).
• Richard Nelson (1977) criticized what he called ‘traditional’ policy analysis
for being ‘relatively blind to exactly the kind of disagreements, and conflicting
interests which need to be perceived in order to guide search for solutions that,
over the long run, do not harm significant values or groups’ (pp 75–6). He
proposed extending the traditional model of policy analysis, de-emphasizing
‘choice and decision’ and paying greater attention to
• the saliency of organizational structure ... and ... the open-ended
evolutionary way in which policies and programs do and should unfold. …
[C]entral high-level decision making ... [is] immune to ... social sciences’
standard research methods. … There is little hope to comprehend basic
patterns of governance and policy making within the thin slices of time
typically considered by contemporary social and policy sciences’ (p 79).
• Observing with cold contempt the ‘metaphysical madness’ of ‘policy scientists’
aiming to supplant politicians and statesmen (and inadvertently conceding the
premise of the policy analysis movement, viz., that analysts are influential),
80
Edward Banfield sniffed that the popularity of social-science-based policy

analysis in both the executive and legislative branches could not be justified
by ‘the existence of a body of knowledge about how to solve social problems’
(Banfield 1977, p 6). ‘Professionals,’ he observed, ‘because of their commitment
to the ideal of rationality, … are chronically given to finding fault with
institutions ... and, by virtue of their mastery of techniques of analysis, to
discovering the almost infinite complexity and ambiguity of any problem
[until problems are viewed as] too complicated for ordinary people to deal
with ...’ (p 18).
In retrospect, I can identify three diagnoses of what the fuss was all about.
It was the economists
To non-economists, it is in the DNA of economists to be hegemonic and, in

their assertiveness, obnoxious. Their ‘metavalue’ of efficiency, which they seem
to regard as tantamount to ‘the welfare of society’ or ‘value maximization’ is
regarded in turn as absurdly reductive (Radin 2013, p 150).
An economist—I, for example—might retort, “if you’re not for ‘value
maximization’, what are you for?” Or “Muddling is not a value.” But that’s not
fair. The legitimate issue is whether economic thinking is, because of its inherent
limitations—yes, its reductiveness—inimical to good public policy. Perhaps societal
values are more likely to be maximized by other processes and ways of thinking.
It helps the economists’ case that efficiency has terrific political salience. If
you build an efficient government, they will vote for you. The important point,
however, is that, though economic reasoning is central to policy analysis, 50
years of analysis has made it clear that the exploration of alternatives to identify
those most likely to be practical and effective requires knowledge and insights
from disciplines and cognate fields ranging from anthropology to zoology. Even
economists have grasped that.
It was the ‘policy scientists’
In 1951, Lerner and Lasswell’s The Policy Sciences laid out a framework for the
emergence of what would become ‘a discernible professional activity’ that some
regarded as a ‘discipline’ (Brewer 1974). Two decades later, a new journal, Policy
Sciences, appeared, edited by Edward Quade, the RAND mathematician behind
PPBS and author of the 1975 text, Analysis for Public Decisions (Quade 1975),
publicized as providing ‘analytic methods as alternatives to traditional public policy
decision-making methods’ (italics added).
Because ‘systems analysis’ was one of the ‘policy sciences’ and because Quade
was closely associated with this new field, seeing what Nelson called ‘traditional
policy analysis’ as the same thing as ‘the policy sciences’ (as Tribe and Banfield,
81
quoted above, and perhaps even Lindblom seemed to do) was not a stretch. Both
were oriented toward introducing systematic analysis into policy-making processes.
But conflating the two and labeling it all ‘metaphysical madness’ is ridiculous.
Contrast Schultze’s and Rivlin’s language for the ambitions of policy analysis—
resources are scarce, analyze alternatives, get to the table—with Brewer’s 1974
declaration, ‘There is an obligation to alert society’s participants to an impending
problem; the policy scientist has a right to clarify and elucidate the competitiveness
and complementarity of interests and objectives of those acting on behalf of society
and of the society being acted upon ...’ (Brewer 1974, p 11, italics added). Some
might consider that quote a better example of metaphysical madness than the
more circumspect claims of the policy analysts.
It was, and is, academia
All things considered, I’ve come to believe that the ‘problem,’ if there is one, with
the policy analysis profession is its having become too entwined with academia and
its values and incentives. Radin apparently agrees. ‘[T]here are some disturbing
signs,’ she writes, ‘that the field has retreated into the ivory tower of assumed
technocratic postures that limit its use by decision makers’ (Radin 2013, p 224).
During the decade I spent practicing policy analysis in the federal government
from 1965 to 1974—this is how I was socialized into the profession—I
encountered many policy-oriented academics. I recall that I found most of them
out to lunch. When I became an academic myself, I learned why.
The intellectual environments in government had been focused on public policy
issues of real political importance. Policy and deliberations could be intense. The
higher the stakes, the greater the intensity. In our desire to speak truth to power,
the question we policy analysts had to confront was, ‘What is the truth we need
to speak?’ That is, what claims are we willing to make to policy makers based
on our analysis? What do we want them to think and why? Will our claims and
reasons stand up to partisan scrutiny? McNamara famously told his policy analysts,
“You do the analysis. I’ll do the politics.” He was an exception.
In academia, discussions of actual public policy issues were mainly staged
affairs: seminars, colloquiums, guest lectures. The intensity, such as it was, was
about academic politics, and the contending interests were factions within the
disciplines, fields of study, and methodologies. The stakes weren’t allocations of
public resources of policy outcomes but allocations of academic recognition. The
rewards weren’t policy achievements but academic achievements and awards.
I didn’t fight them. I joined them. I got pretty good at it. But it wasn’t the
same thing. Academia isn’t the public sector. Policy analysis isn’t policy research.
My argument here starts with a story about another field. In the early postwar
years, the established (also for about five decades) field of public administration
came under withering attack, not for its ‘metaphysical madness’ but for its lack
of intellectual rigor and its propensity to embrace simplistic formulas for how
to govern, its inattention to the findings of academic research. Anthony Bertelli
82
and I reconsidered those criticisms and came to realize that the half-century-
old profession had gained considerable intellectual insight into how American
government actually works and real wisdom concerning the strengths and
weaknesses of our Madisonian republic (Bertelli and Lynn 2006). So why the
sudden smackdown?
The answer is that the training of public administrators and the creation of
knowledge to inform practice, once the responsibility of private research and
training institutions—the Institute for Public Administration, the Brookings
Institution, the bureaus of municipal research—had been appropriated during the
New Deal by elite universities like Columbia, Harvard, and Syracuse, which had
finally grasped that Big Government was creating opportunities for universities to
provide credentials and legitimacy to its administrators. In no time, a profession
that enjoyed healthy respect in the real world of governing was plunged into an
identity crisis as Dwight Waldo, Herbert Simon, Robert Dahl, and other scholars
laid waste to its training and traditions and unscientific ‘principles’ (later to be
rehabilitated as intellectually quite respectable).
Much good has resulted from that takeover. Universities do help legitimize
the profession by credentialing its practitioners. The application of social science
concepts and methods beginning in the 1930s enriched the intellectual resources
available to public administrators as students and practitioners. Professional
associations became forums for intellectual exchange between practitioners and
professors.
In our profession, something similar happened. Ford Foundation largesse led to
the establishment of professional schools of public policy and public affairs at the
nation’s elite institutions and brought the whole system of academic incentives
and rewards to bear on what constituted knowledge for practice (Lynn 1996).
The policy analysis paradigm of Schultze and Rivlin became subject to robust
skepticism from the scholars quoted above. The distinction between ‘policy
analysis’ as an administrative technology of governance and ‘policy research’ as
its offline counterpart became blurred.
Arguably the policy analysis profession has also enjoyed benefits from university
association. Public policy faculties are interdisciplinary, their journals and
professional associations are healthy, and the professional degrees have earned
respect in public personnel systems. At the graduate professional degree level,
students strongly identify with public policy and the skills needed to use it as a
tool of governance. The public management contingent among public policy
professionals has spawned professional organizations that arguably have eclipsed
their public administration counterparts in their influence on practice.
While the field’s academic foundations are a source of strength, it is fortunate, and
essential, that the field’s intellectual life is leavened by government organizations
such as the GAO, CRS, and CBO, private policy research institutions such as
Brookings, the Urban Institute, the Center on Budget and Policy Priorities, and
the American Enterprise Institute, client-oriented policy research firms such as
83
MDRC and Mathematica Policy Research, and innumerable counterparts to

these organizations at state and local levels of government.
Academia nonetheless tends to create distractions from and disincentives to
engage in intellectual activity grounded in the gritty work of governing in real
time. Incorporating practitioners in the profession’s governance and professional
forums and publications is an effective antidote to academic decadence.
On reflection: some verities

Beryl Radin’s chapter in this volume, as well as her other work, highlights
transformation: of the institutions that demand and supply policy analysis and
advice and of the character of practice, all wonderfully illuminated by the personal
experiences of participants. I believe, nonetheless, that there are some verities
that can link the field’s past, present, and future. Five decades ago, here was
an excitement, an animating spirit, a belief that policy analysts had something
important to contribute to the quality of American governance and public policy.
That founding spirit must be sustained. It can be summarized in the following
propositions.
Policy analysis must see itself as committed to enhancing the performance
of America’s Madisonian scheme of politics and governance. No politics–
administration dichotomy for us, no running government like a business, no
‘neutral competence.’ Policy analysts are conduits of knowledge and useful ideas
to those who govern us at all levels of society.
This is emphatically not a commitment to Lasswell’s ‘policy science of democracy.’
Madison argues in The Federalist No. 51 that ‘the great security against dangerous
concentrations of power consists in giving to each branch of government’ the
necessary constitutional means and personal motives to resist encroachments of
the others. … [A]mbition must be made to counteract ambition.’6
By ambition, Madison meant: ‘each department should have a will of its own.’
It follows that efforts by policy analysts to strengthen the capacity of the executive
branch of government are not tantamount to promoting tyranny, as the field’s
postpositive critics have alleged. Those same policy analysts are also engaged in
strengthening the capacity of legislatures to check the ambitions of executives
by evening the analytic playing field (which was distinctly uneven when I was a
practitioner). Courts have privileged policy analysis by endorsing it when finding
that governments have shown they are acting reasonably and, therefore, deserve
judicial deference.
In the same vein, though American politics is seldom regarded as anything
close to rational, the fact is that our political processes depend far more on the
giving of reasons than is commonly acknowledged. Rulemaking requires it.
Budget making requires it. Legislative oversight requires it and so do electoral
politics. The media demand it. So do citizens in their personal transactions with
their governments. The transcendent question in political discourse is: Why?
The business of policy analysis is to address such questions.
84
The rule of law requires reasoning. Managerial judgments and actions are
subject to judicial review for their compliance with constitutional and statutory
law and other lawful acts. According to the doctrine of judicial deference, in
general, U.S. Appellate Judge David Tatel puts it:
Congress delegates authority to administrative agencies not to

authorize any decision at all, but to permit agencies to apply their
expertise. The reason-giving requirement allows courts to determine
whether agencies have in fact acted on the basis of that expertise. … At
a minimum, the requirement of reasoned justification ensures that the
best arguments that can be marshaled in favor of a policy are enshrined
in the public record, so that anyone seeking to undo the policy will
have to grapple with and address those arguments. In this way, the
reason-giving requirement makes our government more deliberative.
And even when it slows agency action, it makes government more
democratic (Tatel 2010, pp 7–8).
Resources to serve public and community interests through the instrumentalities

of government are limited. That means that budget making is the backbone of
American public governance. It follows that the raison d’être of policy analysis
must include, as Schultze emphasized, consideration of how limited resources
can be deployed most effectively in the service of public interests. That in turn
requires analysis of alternatives. That requires somebody to figure out what we’re
trying to accomplish, if not Lindblom’s ‘muddlers’ then the administrators who
are charged with getting the job done and answering to lawmakers and judges
for what they’ve done.
Ideologues of whatever stripe are the most likely to be offended by policy
analysis being a basis for advice to policy makers (unless, of course, they can rig
analysis in their favor). Policy analysis is conducted in a scientific spirit in which
all claims are submitted to the possibility of disconfirmation based on evidence.
Reasons and evidence supporting them are the mother’s milk of policy analysis.
The policy analysis profession in the United States encompasses both client-
oriented policy analysis and social-science-based public policy research, as
Weimer points out. But the imperatives of front-line policy analysis in a world of
Madisonian politics are markedly different from the imperatives of university-based
policy research and from the imperatives governing policy research in soft-money-
driven organizations. The norm of intellectual integrity must be respected in all
of these settings but the inevitable differences among their methods and their
outputs must also be taken into account. So, too, must the pathologies that are
inevitable in each of them. Policy analysis is not a religion. It’s invincibly secular,
multi-ethnic human activity.
It follows that the institutions of professional policy analysis, such as the
Association for Public Policy Analysis & Management (APPAM), must, as
mechanisms of professional accountability, be diverse. While credentialing is the
85
province largely of universities, the profession’s intellectual ethos and content must
reflect the robust involvement of executive agencies of government, think tanks,
institutes, and research centers partially dependent on soft money and donations,
consultancies, legislative entities such as the CBO, the CRS, the GAO, and similar
entities in state and local governments.
Policy makers need honest brokers, people who will see to it that all responsible
possibilities are carefully and honestly considered, that uncertainties are identified,
that assumptions are tested, that ‘what if?’ questions are raised and addressed. It
can be dangerous work. It can be exhilarating. It can be a waste of time. That it
is essential is now established.
Finally, our profession must continue to acknowledge that counting and
calculation are an important part of what we do. Perhaps this goes without saying,
but I’ve often heard groans when serial number crunchers are at the podium. I
understand that; at times, the Journal of Policy Analysis and Management (JPAM) has
seemed more like a journal of multiple regression analysis than a forum with the
intellectual vitality associated with essays and case analyses. But policy deliberation
needs numbers. The era of ‘big data’ and data analytics is here, and that means
new opportunities and challenges for policy analysts. We must be able to enlighten
policymakers on the magnitude and extent of things. Sometimes that can still be
done on the backs of envelopes, but more often it cannot.
A word on our profession’s literature must be added to round out these
reflections. As a doctoral student in economics, reading the discipline’s classics was
mandatory so that we understood the history of its ideas and how and why they
evolved the way they did. It is equally important for the public policy doctoral
students I’ve taught lately, whether they aspire to academic careers or to service
in governments. For the bookshelf of classics for our profession, I nominate these
11 books as seminal to understanding the conceptual, organizational, political,
and administrative dimensions of policy analysis:
Friedmann, J. (1987) Planning in the Public Domain: From Knowledge to Action,
Princeton, NJ: Princeton University Press.
Hitch, C. J. and McKean, R. N. (1967) The Economics of Defense in the Nuclear
Age, Cambridge, MA: Harvard University Press.
Kingdon, J. W. (1995) Agendas, Alternatives, and Public Policies (2nd edn), New
York: Harper Collins (republished in 2003 in the series ‘Longman’s Classics in
Political Science’, with a new foreword.)
Majone, G. (1989) Evidence, Argument, and Persuasion in the Policy Process, New
Haven, CT: Yale University Press.
Morone, James A. (1990) The Democratic Wish: Popular Participation and the Limits
of American Government, New York: Basic Books.
Pressman, J. and Wildavsky, A. (1979) Implementation (2nd edn), Berkeley, CA:
University of California Press.
Rivlin, A. M. (1971) Systematic Thinking for Social Action (The 1970 H. Rowan
Gaither Lectures in Systems Science), Washington, DC: The Brookings
Institution.
86
Schultze, C. L. (1969) The Politics and Economics of Public Spending (The 1968 H.
Rowan Gaither Lectures in Systems Science), Washington, DC: The Brookings
Institution.
Simon, H. A. (1997) Administrative Behavior: A Study of Decision-Making Processes
in Administrative Organizations (4th edn), Glencoe, IL: Free Press.
Wilson, J. Q. (1989) Bureaucracy: What Government Agencies Do and Why They
Do It, New York: Basic Books.
Vickers, S. G. (1983) The Art of Judgment: A Study of Policymaking, New York:
Harper & Row.
In retrospect, much of the intellectual Sturm und Drang that seems to have been
ignited by LBJ’s 1965 mandate seem almost inadvertent satires of reality by those
who either never had to wrap their minds around how to design and administer
a public program so that it works or who were distracted by esoteric arguments
about ‘policy science.’ As I noted above, when you stir too much academic self-
absorption into the bricolage that is policy analysis, a tasty buffet can become
over-seasoned with ideological rancor, interdisciplinary rivalries, turf battles, and
narrowly academic ambition. Not to worry: 50 years on, policy analysis has a
place at the tables where policies are made and implemented. We have reached
the same point, as the medical profession once had to do: we do a lot more good
than harm.
Notes
1 These reflections draw on Lynn (1969, 1980, 1989, 1999, 2001, 2015).
2 Quoted by Enthoven and Smith (1971, pp 87–8). Of this group of experts, it has been noted
that ‘several of [them were] more than mildly eccentric and frequently irritating to the naval
and military commanders they advised’ (Budiansky 2013). Not a few of McNamara’s ‘whiz
kids’ were similarly irritating. I, however, was not.
3 This section closely follows a discussion in Lynn (1999).
4 At: http://urban.hunter.cuny.edu/~schram/lindblom1959.pdf.
5 The following section is adapted from Lynn (1999).
6 Madison quotations are from The Federalist, Modern Library College Editions (softcover),
published by Random House, n/d.
References
Banfield, E. C. (1977) ‘Policy Sciences as Metaphysical Madness’, in R. C.
Goodwin (ed) Statesmanship and Bureaucracy, Washington DC: American
Enterprise Institute, pp 1-35.
Bertelli, A. M. and Lynn Jr., L. (2006) Madison’s Managers: Public Administration
and the Constitution, Baltimore, MD: Johns Hopkins University Press.
Brewer, G. D. (1974) ‘The Policy Sciences Emerge: To Nurture and Structure
a Discipline’, The RAND Paper Series, P-5206, p iii. https://www.rand.org/
content/dam/rand/pubs/papers/2008/P5206.pdf.
Budianski, S. (2013) Blackett’s War: The Men Who Defeated the Nazi U-Boats and
Brought Science to the Art of Warfare, New York: Alfred A. Knopf.
87
deLeon, P. (1997) Democracy and the Policy Sciences, Albany, NY: University of
New York Press.
Enthoven, A. C. and Smith, K. W. (1971) How Much Is Enough? Shaping the
Defense Program, 1961–1969, New York: Harper & Row.
Hoover, H. (1933) ‘Statement on the Report of the-President’s Research
Committee on Social Trends,’ 2 January. Online by Gerhard Peters and John
T. Woolley, The American Presidency Project, http://www.presidency.ucsb.edu/
ws/?pid=23396.
Lindblom, C. (1959) ‘The Science of Muddling Through’, Public Administration
Review 19(2): 79–88.
Lindblom, C. (1979) ‘Still Muddling, Not Yet Through’, Public Administration
Review 39(6): 517–26.
Lynn Jr., L. E. (1969) ‘System Analysis—Challenge to Military Management’, in
D. I. Cleland and W. R. King (eds), Systems, Organizations, Analysis, Management:
A Book of Readings, New York: McGraw-Hill, 216–31.
Lynn Jr., L. E. (1980) ‘Policy Analysis: The User’s Perspective’, in G. Majone and
E. S. Quade (eds), Pitfalls of Analysis, John Wiley for the International Institute
for Applied Systems Analysis.
Lynn Jr., L. E. (1989) ‘Policy Analysis in the Bureaucracy: How New? How
Effective?’ Journal of Policy Analysis and Management, 8(3): 373–77.
Lynn Jr., L. E. (1996) Public Management as Art, Science, and Profession, Chatham,
NJ: Chatham House.
Lynn Jr., L. E. (1999) ‘A Place at the Table: Policy Analysis, Its Postpositivist
Critics, and the Future of Practice’, Journal of Policy Analysis and Management
18(3): 411–25.
Lynn Jr., L. E. (2001) ‘The Making and Analysis of Public Policy: a Perspective
on the Role of Social Science’, in D. Featherman and M. Vinovskis (eds), Social
Science and Policymaking: A Search for Relevance in the Twentieth Century, Ann
Arbor, MI: University of Michigan Press, 187–217.
Lynn Jr., L. E. (2015) ‘Reform of the Federal Government: Lessons for Change
Agents,’ in R. Wilson, N. Glickman, and L.E. Lynn, Jr. (eds), LBJ’s Neglected
Legacy: How Lyndon Johnson Reshaped Domestic Policy and Government. Austin,
TX: University of Texas Press.
Nelson, R. R. (1977) The Moon and the Ghetto: An Essay on Public Policy Analysis.
New York: W.W. Norton.
Quade, E. (1975) Analysis for Public Decisions, Waltham, MA: American Elsevier.
Radin, B. (2013) Beyond Machiavelli: Policy Analysis Reaches Midlife, Washington,
DC: Georgetown University Press.
Rivlin, A. M. (1971) Systematic Thinking for Social Action, Washington, DC: The
Brookings Institution.
Schick, A. (1970) ‘The Budget Bureau That Was: Thoughts on the Rise, Decline,
and Future of a Presidential Agency’, Law and Contemporary Problems 35: 519–39.
Schultze, C. L. (1968) The Politics and Economics of Public Spending, Washington,
DC: The Brookings Institution.
88
Staats, E.B. (1968) ‘Perspective on Planning-Programming-Budgeting’, The GAO

Review (Summer): 12.
Tatel, D. (2010) ‘The Administrative Process and the Rule of Environmental
Law’, Harvard Environmental Law Journal 34: 1–8
Tribe, L. H. (1972). ‘Policy Science: Analysis or Ideology?’ Philosophy and Public
Affairs, 2: 66–110.
Wildavsky, A. (1979) Speaking Truth to Power: The Art and Craft of Policy Analysis,
Boston, MA: Little, Brown.
89
Part Two
Policy analysis by governments
91
FIVE
The practice and promise of policy analysis

and program evaluation to improve decision
making within the U.S. federal government
Rebecca A. Maynard
This chapter draws on a 40-year history of patchwork efforts to use data and
data analysis to inform the development of public policy and to shape its
implementation. The chapter begins with a description of the evolution of the
policy process in the United States, drawing largely on experiences within the U.S.
Departments of Health and Human Services (DHHS), Education (DOED), and
Labor (DOL). All three agencies have been major supporters of and contributors to
advances in the methods of policy analysis and the generation and use of program
evaluation to guide decision making. The chapter draws on the roles of these
agencies in laying the groundwork for the current emphasis on evidence-based
policy making, in part because of their leadership roles and in part because of
the author’s first-hand experience working with these agencies. The chapter
ends with a discussion of the movement over the last decade to create and use
credible evidence on the impacts and cost-effectiveness of programs, policies,
and practices as the foundation for more efficient and effective government and
reflections on the potential implications of the recent change in the administration
for the future of that movement.
Evolution of policy development in federal agencies

Serious attention to program evaluation and public policy analysis in the United
States dates to the late 1960s as the federal government, under the leadership of
President Lyndon B. Johnson committed to the ‘War on Poverty.’ Interest in and
commitment to using the power of government to combat poverty and its root
causes created an imperative for more and better evidence to understand the social,
economic, and education needs of society. President Johnson’s commitment in
1964 was ‘... not only to relieve the symptoms of poverty, but to cure it and,
above all, to prevent it’ (Johnson 1964).
The War on Poverty included policies aimed at three areas of concern:
(1) economic security and health care for the elderly, disabled, and poor; (2)
economic opportunity through employment for all; and (3) equity in educational
opportunity (Matthews 2014). These commitments were underscored by four
major legislative actions that, together, resulted in unprecedented levels of public
investment in the health and welfare of American society flowing from legislative
93
actions (Johnson 1964). Key among these actions were amendments to the Social
Security Act 1965, which created Medicare and Medicaid and expanded social
security benefits coverage; the Food Stamp Act 1964, which institutionalized
what had been a pilot program to improve nutrition among poor families; the
Economic Opportunity Act 1964, which established Head Start and programs
to improve workforce preparation of young people (for example, Job Corps,
VISTA, and the federal work-study program); and the Elementary and Secondary
Education Act (Title 1), which provides financial support to school districts
serving populations with high concentrations of poor children (Klein 2015; U.S.
Department of Health Education and Welfare 1969).
Rising demand for policy analysis and program evaluation

These expansions in federal support for and management of public policies
created a demand for monitoring and evaluation work and, thus, for a cadre
of professionals to carry it out—namely program evaluation and policy analysis
professionals. This new class of professionals needed skills in using data for routine
monitoring of this growing number of large, nationally dispersed programs and
policies. They also needed analytic skills and methods to support simulations of
the likely response to modifications in policies or their contexts of implementation
(Bourguignon and Spadaro 2006; Citro and Hanushek 1991) and to support
rigorous evaluations of the net impacts and cost-effectiveness of the programs
and policies.
Policy analysis and evaluation tools are applied for a variety of purposes,
including to generate evidence on the benefits (or lack thereof) of public
policies; to develop effective program monitoring systems; to generate credible
estimates of fiscal impacts of policies and policy changes and their distributions
across stakeholder groups; and to estimate take-up rates and costs of program and
policy changes. The policy analyst’s job is to generate ‘professionally provided
advice relevant to public decisions and informed by social values’ (Weimer and
Vining 2017). In contrast, the job of program and policy evaluators is to generate
empirical evidence to inform the policy process. Thus, it prioritizes information
on the effectiveness (or lack thereof) of programs, policies, and practices (Scriven
1993; Madaus et al. 1983).
Over the 40 years following the ‘birth’ of the field, a small band of ‘warriors’
from the philanthropic, policy, and research communities fought and won many
battles to increase and improve the availability and use of reliable evidence
about what does and does not work to address the needs of poor and vulnerable
populations (Gueron and Rolston 2013). As a consequence of the breadth of
governmental policy intervention, there are many areas of overlap across agencies,
creating zones of shared responsibility for the generation and use of monitoring
and evaluation. For example, in the areas of health, welfare, and education,
much of the responsibility rests with the DHHS, DOED, and DOL. Moreover,
each of these agencies has primary responsibility for initiatives that also align
94
The practice and promise of policy analysis and program evaluation ...
with the missions of the other agencies—such as Head Start (DHHS), Dropout
Prevention Initiatives (DOED), and Job Corps (DOL). While the missions and
leadership of the federal agencies are largely controlled by the Executive Branch
of the government, the agencies are responsible to Congress for carrying out
mandated policies and programs, they depend on Congress for financial resources
to support mandated and discretionary policies and programs, and they are called
on routinely to advise both Congress and the Executive Branch in deliberations
about national priorities and responses to them.
Evolving structures for policy analysis and program evaluation

As the programs growing out of the War on Poverty took shape, the foundation
for the current level and structure of government oversight was laid. Initially, many
of the analysis and evaluation needs were localized within a single government
agency and tightly focused on monitoring particular programs and policies.
For example, Congressional champions for programs like the Elementary and
Secondary Education Act, which provided supplemental educational resources
to districts, tied that support to specific accountability and evaluation provisions
(Hogan 2007; Weiss 1998).
Similarly, heavy reliance on social scientists in research universities can be traced
to the War on Poverty. For example, the template for Head Start emerged from
the work of a panel of experts that included prominent pediatrician Robert
Cook, and prominent child psychologist Ed Zigler (Administration for Children
and Families 2015), while the template for the negative income tax experiments
(see further below) is credited to economists from the Massachusetts Institute of
Technology and the University of Wisconsin (Ross 1970; Levitt and List 2009).
Evaluations of these programs in their early years were funded through the public
agencies responsible for their management. Although as academic researchers
were drawn into this applied research, they began tapping into private sources
of funding to supplement the federally funded work. For example, throughout
its history, the Head Start program has been evaluated periodically through both
federally supported studies and through field-initiated research carried out by
academics or fledgling research organizations (McKey 1985; Currie and Thomas
1993, 1998).
By the late 1960s, we had also witnessed the birth of large-scale demonstrations
to test the implementation feasibility, assess the impacts, and estimate the cost-
benefits of bold new public policies not yet rooted in legislation. The sizes of
these evaluation enterprises were such that they spawned the birth of large social
science research centers, both within the academy and within free-standing
research organizations. Two prominent examples are the National Health
Insurance Experiments (Newhouse and RAND Corporation 1993) and the
negative income tax experiments (Kershaw and Fair 1976; Watts and Rees 1977a,
1977b). These large public investments in program demonstration projects to test
bold new ideas were accompanied by serious efforts to vet their promise. In so far
95
as the foundational arguments for some of these ideas were economic, the early
empirical methods used in their vetting drew heavily on economic theory and
early versions of predictive analytics, commonly referred to as microsimulation
(Ballas and Clarke 2001; Citro and Hanushek 1991). Notably, these early modeling
efforts took advantage of the emerging access to high-capacity computers and
large-scale, nationally representative data sets from the Census Bureau and
Bureau of Labor Statistics, and administrative records of federal programs like
Aid to Families with Dependent Children (AFDC), Medicaid, Unemployment
Insurance, and Food Stamps. The majority of the evaluations were conducted
through emergent research organizations housed within major universities (for
example, the University of Wisconsin, in the case of the Negative Income Tax
Experiments) and in the emerging research firms and Institutes (for example,
RAND in the case of the National Health Insurance Experiment).
Resources to support policy analysis and program evaluation

For more than 100 years, the federal government has had agencies responsible
for monitoring federal policies and programs and for providing policy analysis
support—namely, the forerunner of the current Government Accountability
Office (GAO) and the Congressional Research Service. ‘The U.S. Government
Accountability Office (GAO) is an independent, nonpartisan agency that works
for Congress’ (Government Accountability Office n.d.). Founded in 1921 under
the Budget and Accounting Act, the agency became independent of the Executive
Branch in 1941. Its primary function has been and continues to be monitoring
how the federal government invests taxpayer dollars, including examining the
effectiveness and cost-effectiveness of federal policies and programs.
‘The Congressional Research Service (CRS)[, which dates to 1914 and is housed
in the Library of Congress,] serves as shared staff to Congressional committees
and Members of Congress’ (Congressional Research Service n.d.). CRS staff
engage in all stages of the legislative process, from the early considerations of
issues under consideration by the legislature to oversight of enacted laws and a
range of agency activities. In essence, it serves as a reference service to provide
Congress with information, including monitoring and evaluation results, to guide
its policy development and evaluation processes.
The expanded role of the federal government in direct services and financial
support for programs that occurred in the latter half of the 20th century created
a demand for rigorous monitoring and evaluation of government policies and
practices across government agencies, as well as branches within them. This
changed the scale and scope of the roles of GAO and CRS and led to the
creation of two new agencies—the Congressional Budget Office (CBO) and the
Office of Management and Budget (OMB). The CBO, which was created in
1974 as a nonpartisan agency, employs a staff of ‘economists and budget analysts
charged with producing dozens of reports and hundreds of cost estimates for
proposed legislation’ (Congressional Budget Office n.d.). The OMB, which was
96
established in 1970, ‘serves the President of the United States in implementing

his vision across the Executive Branch’ (Office of Management and Budget,
n.d.). In addition to having responsibility for developing the President’s budget
proposal to Congress, CBO also has responsibility for measuring the quality of
agency programs, policies, and procedures, including inter-agency coordination
of initiatives, where warranted.
Increased demand for more and better evaluations of federal

programs
The mandates of these federal agencies, together with increases in the number
and scale of social welfare and anti-poverty programs, contributed to increased
demand for more and better evaluation of federal programs, as well as for
improvements in the infrastructure to support them. Much of this demand arose
from within the federal government and, thus, was supported by the government.
Furthermore, during this period, there were, for the first time, robust planned
investments in program and policy development and testing by the federal
government that engaged the nonprofit and academic sectors. For example,
although the government convened the expert panel that conceived Head Start
as implemented under the War on Poverty and subsequently supported extensive
monitoring and evaluation of the program, many other organizations have also
supported evaluations of Head Start over the years (Zigler and Valentine 1979;
Zigler and Muenchow 1992; McKey 1985).
In the 20 years following the launch of the War on Poverty, the increase in social
service programs across government was subjected to many evaluations—some
funded by the federal government, but increasingly often supported in other ways.
Importantly, these evaluations varied widely in their focus, scale, and quality. For
example, most of the more than 600 publications reporting on evaluations of Head
Start within its first 20 years were based on small-scale studies supported by state
and local governments, philanthropies or other nonprofit entities, or conducted
by academic researchers with no or very limited external support (McKey 1985).
As a result, many were not especially well focused and few generated evidence
on the effectiveness of the program.
Similarly, the expansion of government investments in youth employment
programs following the launch of the War on Poverty (for example, Job Corps
and various programs funded under the Youth Employment Demonstration
Projects Act) was accompanied by funding for evaluations of the programs (Betsey
et al. 1985). However, as with Head Start, the vast majority of the published
findings from the evaluations were judged by a National Academy Panel to lack
credibility (Boruch 1985). Moreover, the quality of the evidence was not widely
different among studies funded under different auspices. The fraction of studies
judged to have credible evidence (about 30%) was only slightly above that in a
prior review of studies funded by the Office of Economic Opportunities (Rossi
1969)—a finding that contributed to the National Academy Committee’s call
97
for major improvements in standards for and approaches to program evaluation

(Boruch 1985).
Evolution of the policy analysis and evaluation professions

The War on Poverty and the ensuing more active role for the federal government
in social, economic, and education policy sparked the birth of professional
organizations and journals focused on policy analysis and program evaluation. For
example, the Eastern Evaluation Research Society (EERS), the Association of
Public Policy Analysis and Management (APPAM), and the Network of Schools of
Public Policy, Affairs, and Administration (NASPAA) all emerged in the 1970s to
serve program evaluation and policy analysis professionals. At about the same time,
universities began offering courses in evaluation methodology (Stufflebeam 2001).
These professional organizations and the post-secondary education institutions
they serve have played critical roles in shaping the training and support of policy
analysts and program evaluation experts. These organizations also are professional
homes for many of the policy analysts and evaluators who perform their work
in various market sectors, including the graduate schools of public policy and
administration, policy offices in all levels of government, research and evaluation
experts in major philanthropic organizations, and the large number of evaluation
researchers in for-profit and nonprofit research organizations.
The work of policy analysts and program evaluators

Until recently, policy analysts relied primarily on published research and readily
accessible program administrative data to judge existing policies and practices,
craft new options for consideration, and estimate likely benefits and costs of
program and policy alternatives. In contrast, program evaluators have tended to
immerse themselves in a wide range of field-based evaluation projects, large and
small. These studies often rely on experimental or quasi-experimental methods
to estimate impacts (or lack thereof) and they quite often include complementary
descriptive and qualitative components (Creswell 2003; Maynard 2006; Gueron
and Rolston 2013).
Up until the early to mid 1980s, the policy analysts’ work tended to be driven by
the interests of specific constituent groups, such as a Congressional Committee or
a policy office within a federal agency, such as the DOL or DHHS. Moreover, up
until the mid 1980s, policy analysts were fairly trusting of the quality of evidence
reported in accessible publications.
However, by the late 1980s, the profession had matured and there were several
well-established research organizations specializing in program evaluation and
policy analysis—for example, Abt Associates, Mathematica Policy Research,
MDRC (formerly, the Manpower Demonstration Research Corporation),
RAND, SRI International (formerly Stanford Research Institute), and Westat.
Moreover, faculty scholars at prominent universities including Chicago, Harvard,
98
Wisconsin, and Stanford, were actively involved in large-scale, randomized,

controlled trials of health, welfare, employment and training, and education
programs. Some of these academic researchers also were heading major university-
based research centers that were capable of supporting modest to large-scale
program evaluations and carrying out sophisticated policy analyses. Importantly,
by this time, philanthropic organizations, such as the Ford and Rockefeller
Foundations, were joining the ranks of those demanding and supporting large-
scale field tests of program and policy innovations.
Change was slow, but steady, due to the vision, commitment, and grit of a
relatively few individuals who were intent on moving public policy analysis and
evaluation from a profession dominated by academics whose work prioritized
personal interests and professional incentives for publishing in peer-reviewed
journals, to one dominated by individuals focused on producing useful, credible
evidence to inform policy and practice, irrespective of their home base. Indeed,
this dominance by academic researchers was viewed by some as the root cause of
the dearth of credible evidence to guide policy decisions, such as how the DOL
should respond to the disappointing early results of the Job Training Partnership
Act 1982 (Doolittle et al. 1993).
Greater acceptance of social experiments

By the late 1980s the culture was shifting toward acknowledgement that claims
of evidence of effectiveness based on randomized controlled trials generally had
stronger warrants than did claims derived from evaluations based on other sample
designs and analysis methods (Greenberg et al. 1999). The reason is simple:
Having a control group generated by a lottery provides the basis for knowing
what would have happened to individuals in the study sample who were subjected
to the focal program or policy had they, instead, continued under the otherwise
prevailing conditions (Gueron and Rolston 2013). By this time, several notable
social experiments had demonstrated quite convincingly the benefits of having
experimental as opposed to quasi-experimental or non-experimental evidence
to support policy decisions, particularly when policies are controversial and/or
the stakes are high. For example, it proved to be very useful to have ‘irrefutable’
evidence from experiments to support the highly controversial issue of whether
welfare recipients should be subjected to mandatory job search as a condition of
their eligibility for benefits (Bloom and Blank 1990), whether teenage mothers
with infants should be required to work or be in school to receive their benefits
(Maynard 1995), and whether it is cost-effective for the government to invest
in alternative schools and other forms of employment training for high school
dropouts (Dynarski and Gleason 1998; Christenson and Thurlow 2004).
Importantly, by the late 1980s, experimentation had gained a strong, if not
widespread, standing among important segments of the evaluation, policy, and
practitioner communities as a feasible, responsible, and ethical way to incentivize
states to come up with cost-effective strategies for administering federally funded
99
welfare, and employment and training programs (Rossi 1987; Myers and Dynarski
2003; Fraker and Maynard 1987). This was facilitated by the fact that the evaluation
community had become more ‘adept at ... convince[ing] staff that ... [random
assignment] would avoid unknown and incalculable uncertainty’ (Gueron and
Rolston 2013, pp 256–7).
Over the ensuing years, the fraction of funded program and policy evaluations
in the welfare, education, and employment and training policy spaces relying
on experimental designs increased substantially. For example, the first major
experimental design evaluation funded by the DOE was a large-scale, multi-
site, multi-program evaluation of dropout prevention programs (Dynarski et al.
1998); the DOL funded experimental evaluations of the Job Training Partnership
Act (Bloom et al. 1987); and the DHHS supported many experimental design
evaluations of welfare policy changes, including welfare policies for teenage
mothers (Kelsey et al. 2001; Maynard 1995) and Welfare-to-Work waiver
experiments of the 1990s (Freedman et al. 2000). In fact, the Welfare-to-Work
programs were among the first federal initiatives to grant waivers from federal
policy guidelines conditional on generating experimental evidence to substantiate
(or not) the cost neutrality of the state-favored policies (Harvey et al. 2000).
Paradigm shift
By the turn of this century, we had entered a new playing field. As early as
1981, President Reagan issued an executive order requiring impact analysis of all
major federal policies and programs ‘to reduce the burdens of existing and future
regulations, increase agency accountability for regulatory actions, provide for
presidential oversight of the regulatory process, minimize duplication and conflict
of regulations’ (Federal Register 1981). Not only have these requirements persisted
(U.S. Department of Health and Human Services 2016), more noteworthy is the
fact that it is becoming commonplace that policymakers request and expect to
use evidence to guide their planning and monitoring efforts, even though they
recognize that the available evidence is often lacking in depth, breadth, or direct
applicability. It also has become more routine to expect systematic tracking of
inputs and outputs associated with public policies and programs.
A byproduct of this monitoring has been the creation of rich data sets, albeit
of varying degrees of utility and accessibility, for purposes other than that for
which they were commissioned. But, the most notable change has been the
increasing interest in use of evidence quite explicitly in policy decisions about
how to allocate resources, how to monitor and reward performance, and where
and how to invest evaluation dollars (Haskins and Baron 2011; Mervis 2013).
These changes would not have been possible had there not also been decisions
across federal agencies and offices to create clearinghouses to collect, review,
and make publicly available the findings of prior evaluations they judge to have
generated credible evidence of the likely impacts of programs, policies, and
practices in their sphere of responsibility. Although these efforts began nearly
100
20 years ago (Reidy et al. 1998), it was well into the Obama administration
before they reached a level of maturity that makes them integral to the work
of Congressional committees, OMB, and the Program and Planning Offices of
the DHHS, DOED, and DOL. It remains to be seen whether this momentum
will continue with the transition to the Trump administration and its promise
to adopt a more free-market approach to decision making that relies on a quite
different definition of ‘evidence’ (Gordon & Haskins 2017).
The Institute of Education Sciences under the leadership of its first Director,
Grover (Russ) Whitehurst, set the tempo for the movement to create evidence
platforms with its launch of the What Works Clearinghouse in 2002 (Whitehurst
2003). Importantly, numerous other public and nonprofit agencies have followed
suit. Across the federal government, there are now a number of evidence
clearinghouses. These include clearinghouses focused on education, teen
pregnancy prevention, child development and welfare, and criminal justice, as
well as more generic platforms like the Campbell Collaboration created by the
research and evaluation communities (Table 5.1). For the evidence-based policy
movement as we know it today to continue, it will be important that these types
of infrastructure investments be continued—something that may or may not fit
with the Trump administration’s view of the role of evidence in the policy process.
These platfor ms have made it more practical for government and
nongovernmental agencies to incorporate evidence criteria into their program
funding decisions. For example, the Every Child Succeeds Act 2015 (ESSA)
includes provisions regarding allowable uses of federal education monies that
depend on the amount and rigor of the evidence on expected impacts (Gross
2016). Similarly, the Office of Adolescent Health’s Teen Pregnancy Prevention
Program is now linking much of its program funding to evidence requirements
(Office of Adolescent Health 2016).
Needed improvements in the evidence base

Despite the widespread recognition of the importance of credible evidence to
guide public policy development, the clear majority of policy decisions are still
being made with little credible supporting evidence to guide how we disburse
the more than $6 trillion spent annually in the United States on health, welfare,
economic security, and defense annually (Kettl 2015). There are three primary
reasons: (1) it takes time and resources to accumulate credible evidence on the
broad range of issues relevant to health, social welfare, and education; (2) it takes
time and resources to integrate evidence into decision-making processes; and (3)
sometimes, other considerations do and should trump cost considerations, such
as values and competing priorities.
Not only is the current evidence base lacking in coverage, but much of it is
obsolete and a shamefully large share of it is methodologically flawed—particularly
much of that based on studies conducted prior to the turn of the century. For
example, major public policy issues in education revolve around teacher training
101
Table 5.1: Selected evidence clearinghouses 2017

Name Sponsoring agency Weblink
Best Evidence Encyclopedia Johns Hopkins University, Center for Data www.bestevidence.org/overviews/
Driven Reform in Education overviews.htm
Campbell Collaboration Campbell Collaboration www.campbellcollaboration.org/lib/
Clearinghouse for Labor Evaluation U.S. Department of Labor http://clear.dol.gov/
and Research (CLEAR)
Crime Solutions Office of Justice Programs www.crimesolutions.gov/default.
aspx
Systematic Reviews I-3: International Initiative for Impact www.3ieimpact.org/en/evidence/
Evaluations systematic-reviews/
Evidence for ESSA Johns Hopkins University http://evidenceforessa.org
Find Youth Info The Interagency Working Group on Youth https://youth.gov/evidence-
Programs innovation/program-directory
Clearinghouse on Families and Youth National Commission on Families and Youth http://ncfy.acf.hhs.gov/
Home Visiting Evidence of Effectiveness U.S. Department of Health and Human http://homvee.acf.hhs.gov/
(HomVEE) Services, Administration for Children and
Families, Office of Planning, Research, and
Evaluation
Teen Pregnancy Prevention Review U.S. Department of Health and Human http://tppevidencereview.aspe.hhs.
Study Database Services gov/StudyDatabase.aspx
Benefit-cost Results Washington State Institute of Public Policy www.wsipp.wa.gov/BenefitCost

What Works Clearinghouse (WWC) U.S. Department of Education, Institute of http://ies.ed.gov/ncee/wwc/
Education Sciences
Workforce Systems Strategies U.S. Department of Labor, Employment and https://strategies.workforcegps.org/
Training Administration
Source: This table builds on Maynard et al. (2016), Table 1.
and compensation policies. A recent review of evidence by the What Works

Clearinghouse of four widely touted strategies to improve the teaching force—
Teach for America (TFA), the Teacher Advancement Program, the New Teacher
Induction Model, and My Teaching Partner—identified 49 potentially relevant
studies. While seven of 24 studies of TFA met the Clearinghouse’s evidence
standards, there was only one study meeting standards for each of the other three
programs (author’s review of the What Works Clearinghouse Intervention Report
website). This is in spite of the fact that the standards of evidence applied by the
Clearinghouse are not especially stringent, allowing inclusion of studies using
either randomization or matching to create the comparison group (Institute of
Education Sciences 2016).
More generally, the vast majority of the published literature the What Works
Clearinghouse identified as relevant to particular topic areas it has reviewed fails
to meet its standards for credibility. For example, an examination of the website
identified 9,292 citations for studies of relevance to topics the Clearinghouse had
102
reviewed. However, only 10% of the studies yielded evidence on effectiveness

of the interventions of interest that was judged to be sufficiently credible to
warrant reporting (Table 5.2). Most of the studies did not measure impacts
and, of those that did, a high fraction used data and/or analytic methods that
resulted in estimates of questionable credibility. However, on the positive side
of the equation, over half of the studies that met evidence standards were well
designed and implemented randomized controlled trials. Notably, these findings
on the prevalence and quality of evidence in education are similar to those for
evaluations in the areas of social welfare and employment and training (Coalition
for Evidence Based Policy 2013).
Reliance on evidence in funding decisions

It is easy to be disheartened by these statistics. However, the good news is that
these evidence platforms exist to help policymakers, practitioners, and the public
learn what is and is not known about the effectiveness of various policies and
programs that are in place or being considered. Indeed, where credible evidence
does exist, it has become increasingly important and more widely used to
inform policy. For example, the decision by Congress to appropriate $1.5 billion
over four years to support the Maternal, Infant, and Early Childhood Home
Visiting Program was driven by a 30-year history of research on one program
model, Nurse–Family Partnership, and all of that research had been produced
by the program’s developer (Goodman 2006; Schmit et al. 2015; Haskins and
Margolis 2014; Olds et al. 2013). But, the focus on evidence did not stop with
congressional action to fund the programs. The legislation that authorized the
program required that 75% of the funds be spent on one of 11 programs that had
scientific evidence of their effectiveness (U.S. Department of Health and Human
Services, n.d.a). As of January 2017, the list of ‘effective programs’ had grown to
19 programs (U.S. Department of Health and Human Services, n.d.b). Reflecting
the growing commitment to rigorous evidence for monitoring and future policy
Table 5.2: Sample snapshots of evidence on the effectiveness of education programs and
practices reviewed by the What Works Clearinghouse
Number of studies Percentage of studies identified
Total Teacher excellence topic area Total Teacher excellence topic area
Studies identified 9,292 179 100% 100%
Relevant method and designs 2,501 67 27% 37%
Met evidence standards 944 34 10% 19%
without reservations 601 21 6% 12%
with reservations 343 13 4% 7%
Source: Tabulations by author of data from the What Works Clearinghouse website’s Review of Individual Studies Tool: https://ies.ed.gov/ncee/
wwc/ReviewedStudies
103
decision making, Congress also appropriated funding to continually evaluate four

of the most prevalent of the approved models (Michalopoulos et al. 2015; U.S.
Department of Health and Human Services, n.d.c). And, states and philanthropic
organizations are funding evaluations of many of the less well-known program
models, including some evidence-based models that have been adapted to local
conditions (U.S. Department of Health and Human Services (n.d.d).
The most recent shifts in the policy arena have entailed integrating evidence use
throughout the design, development, testing, and institutionalization processes.
We see this featured in evidence frameworks developed to guide federal funding
decisions, such as the Common Evidence Guidelines for Education Research
issued by the National Science Foundation and the Institute of Education Sciences
(Earle et al. 2013) and the Framework for Design Based Implementation Research
(Means and Harris 2013; Fishman et al. 2013). These frameworks, together with
strong encouragement from the OMB’s Office of Evidence-based Initiatives
(Mervis 2013; Nussle and Orszag 2014) have increased substantially familiarity
with and buy-in from the policy analysis community (Khagram and Thomas
2010; Thyer and Meyers 2011). Indeed, across the board, increasing numbers of
funders and shares of funding for public programs and policies are being shaped
by considerations related to evidence on expected impacts, returns on investment
(benefits-costs), and efficiency (cost-effectiveness). Across government, the level
and strength of evidence on the likely impact of public policies, practices, and
programs is shaping what is funded, the level of funding, and requirements for
program monitoring and evaluation. Many of the evidence-based programs
recently launched in the DHHS, DOE, and DOL, as well as programs overseen by
the Corporation for National Community Service within the Executive Branch,
carry with them evidence requirements for award of funds, as well as requirements
for rigorous evaluations to substantiate (or not) their effectiveness (Table 5.3).
Looking forward
Over the past 40 years, the fields of policy analysis and program evaluation have
evolved from craft occupations, tailored to the local employment setting and
available resources, to mature professions with more explicit expectations about
formal training requirements and approaches to the design, implementation,
conduct, and dissemination of the work. Professionals in these fields now face
more demanding preparation for their jobs. However, they also can expect to enjoy
a more central role in development, enactment, improvement, and evaluation of
the programs, policies, and practices. Forty years ago, most individuals entering
these fields lacked formal training in policy analysis or program evaluation; rather,
most had training in a discipline such as sociology, economics, political science,
or statistics. Today, policy analysts and evaluators are more likely to have had
formal training in a school of public policy or public administration. Many also
will have an advanced degree in applied economics, quantitative sociology, or
104
Table 5.3: Illustrative evidence-guided funding streams

Program, website Description
Maternal, Infant, Health Resources and Services Administration (HRSA), in close partnership with the
and Early Childhood Administration for Children and Families (ACF), funds states, territories and tribal entities to
Home Visiting develop and implement voluntary, evidence-based home visiting programs using models that
Program are proven to improve child health and to be cost-effective. These programs improve maternal
http://mchb.hrsa.gov/ and child health, prevent child abuse and neglect, encourage positive parenting, and promote
programs/homevisiting/ child development and school readiness.
Social Innovation Authorized by the Edward M. Kennedy Serve America Act in April of 2009, the Social Innovation
Fund Fund is a program of the Corporation for National and Community Service (CNCS). The Social
http://www.nationalservice. Innovation Fund’s (SIF) portfolio represents $241 million in federal grants and more than $516
gov/programs/social- million in non-federal match commitments. To date, the SIF’s Classic program has awarded
innovation-fund/about-sif grants to 27 grant-making organizations and 189 nonprofits working in 37 states and the
District of Columbia. Through the SIF’s Pay for Success program, over 30 jurisdictions across the
United States are engaged in testing and implementing Pay for Success projects.
First in the World The FITW program funded by the U.S. Department of Education to support the development,
http://www2.ed.gov/ replication, and dissemination of innovative solutions and evidence for what works in
programs/fitw/index.html addressing persistent and widespread challenges in postsecondary education for students who
are at risk of not persisting in and completing postsecondary programs, including, but not
limited to, adult learners, working students, part-time students, students from low-income
backgrounds, students of color, students with disabilities, and first-generation students.
Investing in The Investing in Innovation Fund, established under section 14007 of the American Recovery
Innovation and Reinvestment Act 2009 (ARRA). The purpose of this program is to provide competitive
http://www2.ed.gov/ grants to applicants with a record of improving student achievement and attainment in order
programs/innovation/ to expand the implementation of, and investment in, innovative practices that are demonstrated
index.html to have an impact on improving student achievement or student growth, closing achievement
gaps, decreasing dropout rates, increasing high school graduation rates, or increasing
college enrollment and completion rates. These grants encourage eligible entities to work in
partnership with the private sector and the philanthropic community, and to identify and
document best practices that can be shared and taken to scale based on demonstrated success.
Trade Adjustment In 2009 the American Recovery and Reinvestment Act amended the Trade Act 1974 to
Assistance authorize the Trade Adjustment Assistance Community College and Career Training (TAACCCT)
Community College Grant Program. On March 30, 2010 President Barack Obama signed the Health Care and
and Career Training Education Reconciliation Act, which included $2 billion over four years to fund the TAACCCT
http://doleta.gov/taaccct/ program. TAACCCT provides community colleges and other eligible institutions of higher
education with funds to expand and improve their ability to deliver education and career
training programs that can be completed in two years or less, are suited for workers who are
eligible for training under the TAA for Workers program, and prepare program participants for
employment in high-wage, high-skill occupations. The Department is implementing the TAACCCT
program in partnership with the Department of Education.
Workforce The Workforce Innovation Fund Grant Program is authorized by the Full-Year Continuing
Innovation Fund Appropriations Act, 2011 (P.L. 112-10). These funds support innovative approaches to the design
https://www.doleta.gov/ and delivery of employment and training services that generate long-term improvements in
workforce_innovation/ the performance of the public workforce system, both in terms of outcomes for job seekers
and employer customers and cost-effectiveness. The program has three goals: (1) provide
services more efficiently to achieve better outcomes, particularly for vulnerable populations and
dislocated workers; (2) support both system reforms and innovations that facilitate cooperation
across programs and funding streams in the delivery of client-centered services to job seekers,
youth, and employers; and (3) build knowledge about effective practices through rigorous
evaluation and how effective strategies to disseminating the effective programs to the broader
workforce system.
Source: Adapted from Maynard et al. (2016), Table 1.
105
political science. Increasingly, they are also likely to be grounded in principles of

both quantitative and qualitative research methods, rather than one or the other.
Up through the end of the Obama administration, all signs had been that the
trends toward more and smarter use of evidence of various forms to inform the
policy process would persist, with a premium placed on highly credible evidence
of expected impacts from alternatives under consideration. And, even should the
recent change in administrations result in a roll-back in the demands for evidence,
it will do so in the context of a nation that has much greater capacity now than
it did even 10 years ago to generate credible evidence due to the explosion
of program administrative data and data warehouses that make extraction and
sharing cheaper and quicker. It also is a world in which liberal and conservative
policymakers have become much more accustomed to using evidence.
Importantly, the field of evaluation now operates on higher standards for
evidence. It has made methodological advances and advances in evaluation tools
to routinize many complex and tedious analytic tasks in the design and conduct
of rigorous research—for example, tools to conduct complex power analyses
(Liu et al. 2006; Dong and Maynard 2013; Dong et al. 2016); do hierarchical
modeling (Raudenbush and Bryk 2002); conduct randomization of sample
members (Schochet 2015); perform complex data mining (Baker and Inventado
2014); and do complex processing and analysis of qualitative data sets (Ritchie et
al. 2013). Thus, there is higher overall quality in the evaluations being conducted.
Continued expansion and improvement of the policy analysis and program
evaluation enterprise requires investment in the data infrastructure. We now live in
a world where massive amounts of data are generated at all levels of government,
most of which stays within a limited network of offices or agencies. The speed and
quality of evaluations could be greatly enhanced by improvements in local data
systems and by strategic integration across data systems (Fantuzzo and Culhane
2015). However, a major barrier to sharing and integration is a lack of trust in
the process (Manzi 2012), which may require the promulgation of principles and
practices for evaluation, much as we have for statistical agencies (Citro 2014).
We are approaching consensus on what constitutes credible evidence, we have
strong methods for generating that evidence, and we have a pipeline of skilled
evaluators. We also have templates for the types of integrated data systems needed
to support quick turnaround and ongoing analyses, and we have several strong
platforms for identifying, synthesizing, and disseminating evidence.
On the flip side, government agencies and the publics they serve have
increasingly been demanding to see the evidence before they lend their support
to new policies, programs, or practices. However, it remains to be seen how
enduring these trends will be across administrations. As of this printing, there are
serious efforts by leaders in the evidence movement on both sides of the political
divide to advise the Trump administration on ways the advances in evidence-based
policy making can also be useful to their goals (for example, Feldman 2017).
Although we have seen some shifts in the emphasis in terms of evidence use in
policy development and management, we have seen no evidence that recent trends
106
toward more and better use of evidence will not continue. Strong public policy
depends on the vigor and vigilance of the policy analysts and program evaluators
and their commitment and capacity to work with policymakers and the public
to promote sharing of the evidence we have and noting the gaps. The biggest
unknown is how the quite different values held by different administrations will
influence policy design and implementation processes.
References
Administration for Children and Families (2015) ‘History of Head Start’,
Washington, DC: Office of Head Start, U.S. Department of Health and Human
Services, https://www.acf.hhs.gov/ohs/about/history-of-head-start.
Baker, R. S. and Inventado, P. S. (2014) ‘Educational Data Mining and Learning
Analytics’, in Learning Analytics, New York: Springer, 61–75.
Ballas, D. and Clarke, G. P. (2001) ‘Modelling the Local Impacts of National
Social Policies: A Spatial Microsimulation Approach’, Environment and Planning
C: Government and Policy 19(4): 587–606.
Betsey, C., Hollister Jr, R. G., and Papageorgiou, M. R. (eds) (1985) Youth
Employment and Training Programs: The YEDPA Years, National Academies of
Sciences.
Bloom, B. and Blank, S. (1990) ‘Bringing Administrators into the Process’, Public
Welfare 48(4): 4–12.
Boruch, R. (1985) ‘Implications for the Youth Employment Experience for
Improving Evaluation Research and Applied Social Policy’ in C. Betsey, Hollister
Jr, and Papageorgiou (eds), Youth Employment and Training Programs: The YEDPA
Years, National Academies of Sciences.
Bourguignon, F. and Spadaro, A. (2006) ‘Microsimulation as a Tool for Evaluating
Redistribution Policies’, Journal of Economic Inequality 4(1): 77–106.
Christenson, S. L. and Thurlow, M. L. (2004) ‘School Dropouts: Prevention
Considerations, Interventions, and Challenges’, Current Directions in Psychological
Science 13(1): 36–9.
Citro, C. F. (2014) ‘Principles and Practices for a Federal Statistical Agency: Why,
What, and to What Effect’, Statistics and Public Policy 1(1): 51–9.
Citro, C. F. and Hanushek, E. A. (1991) Improving Information for Social Policy
Decisions: The Uses of Microsimulation Modeling Volume II Technical Papers,
Washington, DC: National Academy Press.
Coalition for Evidence-Based Policy (2013) ‘Randomized Controlled Trials
Commissioned by the Institute of Education Sciences since 2002: How Many
Found Positive versus Weak or No Effects’, http://coalition4evidence.org/
wp-content/uploads/2013/06/IES-Commissioned-RCTs-positive-vs-weak-
or-null-findings-7-2013.pdf Google Scholar.
Congressional Budget Office (n.d.) ‘Congressional Budget Office: About Us’,
https://www.cbo.gov/about/overview.
107
Congressional Research Service (n.d.) ‘Congressional Research Service:

Overview’, Congressional Research Services, https://www.loc.gov/crsinfo/
about/.
Creswell, J. W. (2003) ‘Research Design: Quantitative, Qualitative and Mixed
Methods Approaches’, Thousand Oaks, CA, Sage Publications, Inc.
Currie, J. and Thomas, D. (1993) Does Head Start Make a Difference? No. w4406,
New York: National Bureau of Economic Research.
Currie, J. and Thomas, D. (1998) School Quality and the Longer-term Effects of Head
Start, No. w6362, National Bureau of Economic Research.
Dong, N. and Maynard, R. (2013) ‘PowerUp!: A Tool for Calculating Minimum
Detectable Effect Sizes and Minimum Required Sample Sizes for Experimental
and Quasi-experimental Design Studies’, Journal of Research on Educational
Effectiveness 6(1): 24–67.
Dong, N., Reinke, W. M., Herman, K. C., Bradshaw, C. P., and Murray, D. W.
(2016) ‘Meaningful Effect Sizes, Intraclass Correlations, and Proportions of
Variance Explained by Covariates for Planning Two-and Three-Level Cluster
Randomized Trials of Social and Behavioral Outcomes’, Evaluation Review 40(4):
334–77.
Doolittle, F., Bell, S., Bloom, H., Cave, G., Kemple, J., Orr, L., Traeger, L., and
Wallace, J. (1993) ‘Summary of the Design and Implementation of the National
JTPA Study’, New York: Manpower Demonstration Research Corporation.
Dynarski, M. and Gleason, P. (1998) How Can We Help?: What We Have
Learned from Evaluations of Federal Dropout-Prevention Programs, Princeton, NJ:
Mathematica Policy Research.
Dynarski, M., Gleason, P., Rangarajan, A., and Wood, R. (1998) ‘Impacts of
Dropout Prevention Programs: Final Report (School Dropout Demonstration
Assistance Program Evaluation Research Report)’, Princeton, NJ: Mathematica
Policy Research.
Earle, J., Maynard, R., and Neild, R. (2013) ‘Common Evidence Guidelines
for Education Research’, Washington, DC: Institute of Education Sciences and
National Science Foundation, http://ies.ed.gov/pdf/CommonGuidelines.pdf.
Fantuzzo, J. and Culhane, D. P. (eds) (2015) Actionable Intelligence: Using Integrated
Data Systems to Achieve a More Effective, Efficient, and Ethical Government, New
York, NY: Palgrave Macmillan.
Federal Register (1981) Regulatory Impact Analysis Requirements: Executive
Order 12291, https://www.archives.gov/federal-register/codification/
executive-order/12291.html.
Feldman, A. (2017) Strengthening Results-focused Government, Washington,
DC: Brookings Institution, https://www.brookings.edu/wp-content/
uploads/2017/01/es_20170130_evidencebasedpolicy.pdf.
Fishman, B. J., Penuel, W. R., Allen, A. R., Cheng, B. H., and Sabelli, N. (2013)
‘Design-based Implementation Research: An Emerging Model for Transforming
the Relationship of Research and Practice’, National Society for the Study of
Education 112(2): 136–56.
108
Fraker, T. and Maynard, R. (1987) ‘The Adequacy of Comparison Group Designs

for Evaluations of Employment-related Programs’, Journal of Human Resources,
22(2): 194–227.
Freedman, S., Knab, J. T., Gennetian, L. A., and Navarro, D. (2000) ‘The Los
Angeles Jobs-First Gain Evaluation: Final Report on a Work First Program in
a Major Urban Center,’ New York, NY: Manpower Demonstration Research
Corporation.
Gordon, R. and Haskins, R. (2017) ‘How the New Budget Could Hurt One of
the Smartest Ideas in policymaking, Politico, March 31, http://www.politico.
com/agenda/story/2017/03/the-trump-administrations-misleading-embrace-
of-evidence-000385.
Government Accountability Office (n.d.). ‘Government Accountability Office:
About Us’, http://www.gao.gov/about/.
Greenberg, D., Shroder, M., and Onstott, M. (1999) ‘The Social Experiment
Market’, The Journal of Economic Perspectives, 13(3): 157–72.
Gross, B. (2016) ‘What Works? Evidence-Based Policymaking Under ESSA’,
Maximizing Opportunities Under ESSA, SEA of the Future (6): 25–37.
Gueron, J. M. and Rolston, H. (2013) Fighting for Reliable Evidence, New York:
Russell Sage Foundation.
Harvey, C., Camasso, M. J., and Jagannathan, R. (2000) ‘Evaluating Welfare
Reform Waivers under Section 1115’, The Journal of Economic Perspectives, 14(4):
165–88.
Haskins, R. and Baron, J. (2011) ‘Building the Connection Between Policy and
Evidence’, in R. Puttick (ed), Using Evidence to Improve Social Policy and Practice,
Alliance for Useful Evidence, 25–51.
Haskins, R. and Margolis, G. (2014) Show Me the Evidence: Obama’s Fight for
Rigor and Results in Social Policy, Washington, DC: Brookings Institution Press.
Hogan, R. L. (2007) ‘The Historical Development of Program Evaluation:
Exploring Past and Present’, Online Journal for Workforce Education and
Development, 2(4): 5.
Institute of Education Sciences (2016) What Works Clearinghouse Procedures and
Standards Handbook, Version 3, Washington, DC: Institute of Education Sciences.
Johnson, L.B. (1964) ‘Remarks Upon Signing the Economic Opportunity Act’,
August 20, online by Gerhard Peters and John T. Woolley, The American Presidency
Project, http://www.presidency.ucsb.edu/ws/?pid=26452.
Kelsey, M., Johnson, A., and Maynard, R. A. (2001) The Potential of Home Visitor
Services to Strengthen Welfare-to-work Programs for Teenage Parents on Cash Assistance,
Princeton, NJ: Mathematica Policy Research, Incorporated.
Kershaw, D. and Fair, J. (1976) The New Jersey Negative Income Tax Experiment
(Vol. 1): Operations, Surveys, and Administration. New York: Academic Press.
Kettl, D. F. (2015). ‘The Job of Government: Interweaving Public Functions and
Private Hands’, Public Administration Review 75(2): 219–29.
109
Khagram, S. and Thomas, C. W. (2010) ‘Toward a Platinum Standard for

Evidence-Based Assessment by 2020’, Public Administration Review 70(s1):
S100–6.
Klein, A. (2015) ‘ESEA Reauthorization: The Every Student Succeeds Act
Explained’, Education Week, November 30.
Levitt, S. D. and List, J. A. (2009) ‘Field Experiments in Economics: The Past,
the Present, and the Future’, European Economic Review 53(1): 1–18.
Liu, X., Spybrook, J., Congdon, R., Martinez, A., and Raudenbush, S. (2006)
Optimal Design for Multi-level and Longitudinal Research, Ann Arbor, MI: University
of Michigan, Survey Research Center of the Institute of Social Research.
Madaus, G. F., Stufflebeam, D., and Scriven, M. S. (1983) ‘Program Evaluation’,
in Evaluation Models, Viewpoints on Educational and Human Services Evaluation.
Boston, MA: Kluwer-Nijhoff, pp 3-22.
Manzi, J. (2012) Uncontrolled: The Surprising Payoff of Trial-and-error for Business,
Politics, and Society, New York: Basic Books.
Matthews, D. (2014) ‘Everything You Need to Know about the War on
Poverty’, Washington Post, January 8, https://www.washingtonpost.com/news/
wonk/wp/2014/01/08/everything-you-need-to-know-about-the-war-on-
poverty/?utm_term=.0f4881a0575b.
Maynard, R. (1995) ‘Teenage Childbearing and Welfare Reform: Lessons from a
Decade of Demonstration and Evaluation Research’, Children and Youth Services
Review 17(1–2): 309–32.
Maynard, R. A. (2006) ‘Presidential Address: Evidence-based Decision Making:
What Will It Take for the Decision Makers to Care?’ Journal of Policy Analysis
and Management 25(2): 249–65.
Maynard, R., Goldstein, N. and Nightingale, D. S. (2016) ‘Program and Policy
Evaluations in Practice: Highlights from the Federal Perspective’, in L. Peck (ed)
New Directions for Evaluation, San Francisco, CA: Jossey-Bass/Wiley, pp 109–
34, DOI: 10.1002/ev.20209, http://onlinelibrary.wiley.com/doi/10.1002/
ev.20209/
McKey, R. H. (1985) ‘The Impact of Head Start on Children, Families and
Communities’, Final Report of the Head Start Evaluation, Synthesis and Utilization
Project, http://files.eric.ed.gov/fulltext/ED263984.pdf.
Means, B. and Harris, C. J. (2013) ‘Towards an Evidence Framework for Design-
Based Implementation Research’, Yearbook of the National Society for the Study of
Education 112(2): 350–71.
Mervis, J. (2013) ‘An Invisible Hand Behind Plan to Realign US Science
Education’, Science 341(6144): 338–41.
Michalopoulos, M., Lee, H., Duggan, A., Lundquist, E., Tso, A., Crowne, S.,
Burrell, L., Somers, J., Filene, J. H., and Knox, V. (2015) ‘The Mother and
Infant Home Visiting Program Evaluation: Early Findings on the Maternal,
Infant, and Early Childhood Home Visiting Program. OPRE Report 2015-11’,
Washington, DC: Office of Planning, Research and Evaluation, Administration
for Children and Families, U.S. Department of Health and Human Services.
110
Myers, D. and Dynarski, M. (2003) ‘Random Assignment in Program Evaluation

and Intervention Research: Questions and Answers’, http://files.eric.ed.gov/
fulltext/ED478982.pdf.
Newhouse, J. P. and RAND Corporation, Insurance Experiment Group
(1993) Free for All?: Lessons from the RAND Health Insurance Experiment,
Cambridge, MA: Harvard University Press.
Nussle, J. and Orszag, P. (eds) (2014) Moneyball for Government, Washington, DC:
Disruption Books.
Office of Adolescent Health (2016) ‘U.S. Department of Health and Human
Services (2016). Teen Pregnancy Prevention Program’, https://www.hhs.gov/
ash/oah/oah-initiatives/tpp_program/about/#.
Office of Management and Budget (n.d.) ‘Office of Management and Budget:
About Us’.
Olds, D., Donelan-McCall, N., O’Brien, R., MacMillan, H., Jack, S., Jenkins, T.,
and Pinto, F. (2013) ‘Improving the Nurse–family Partnership in Community
Practice’, Pediatrics 132(2): S110–17.
Raudenbush, S. W. and Bryk, A. S. (2002) Hierarchical Linear Models: Applications
and Data Analysis Methods (vol. 1), Thousand Oaks, CA: Sage.
Reidy, M., Goerge, R., and Lee, B. J. (1998) Planning for Human Service Reform
Using Integrated Administrative Data, Chicago, IL: Chapin Hall at the University
of Chicago.
Ritchie, J., Lewis, J., Nicholls, C. M., and Ormston, R. (eds) (2013) Qualitative
Research Practice: A Guide for Social Science Students and Researchers, Thousand
Oaks, CA: Sage.
Ross, H. (1970) ‘An Experimental Study of the Negative Income Tax’, Child
Welfare 49(10): 562–69.
Rossi, P. (1987) ‘The Iron Law of Evaluation and Other Metallic Rules’, Research
in Social Problems and Public Policy, 4: 3–20.
Rossi, P. H. (1969) ‘Practice, Method, and Theory in Evaluating Social Action
Programs’, in On Fighting Poverty, New York: Basic Books, 217–34.
Schmit, S., Schott, L., Pavetti, L. and Matthews, H. (2015) ‘Evidence-Based
Home Visiting Programs in Every State at Risk if Congress Does Not Extend
Funding’, Center for Budget and Policy Priorities, http://www.cbpp.org/sites/
default/files/atoms/files/3-10-14hv.pdf.
Schochet, P. Z. (2015) ‘Statistical Theory for the RCT-YES Software: Design-
based Causal Inference for RCTs, Second Edition (NCEE 2015–4011)’,
Washington, DC: U.S. Department of Education, Institute of Education
Sciences, National Center for Education Evaluation and Regional Assistance,
Analytic Technical Assistance and Development, http://ies.ed.gov/ncee/edlabs.
Scriven, M. (1993) ‘Hard-won Lessons in Program Evaluation’, New Directions
for Program Evaluation, 58: 1–107.
Stufflebeam, D. (2001) ‘Evaluation Models’, New Directions for Evaluation 89: 7–98.
Thyer, B. A. and Myers, L. L. (2011) ‘The Quest for Evidence-based Practice:
A View from the United States’, Journal of Social Work 11(1): 8–25.
111
U. S. Department of Health and Human Services, (n.d.a) ‘Health Resources

Services Administration, Administration for Children and Families, Home
Visiting’, https://mchb.hrsa.gov/maternal-child-health-initiatives/home-
visiting-overview.
U.S. Department of Health and Human Services (n.d.b) ‘Evaluating Home
Visiting Programs: State and Local Examples’, https://www.childwelfare.
gov/topics/preventing/prevention-programs/homevisit/evaluate-programs/
examples/.
U.S. Department of Health and Human Services, Administration for Children and
Families (2013) ‘Summary of Child Welfare Waiver Demonstrations’, https://
www.acf.hhs.gov/sites/default/files/cb/waiver_summary_final_april2013.pdf.
U.S. Department of Health and Human Services, (n.d.c) ‘The Maternal, Infant,
and Early Childhood Home Visiting Program Partnering with Parents to
Help Children Succeed’, https://www.acf.hhs.gov/sites/default/files/ecd/
home_visiting_issue_brief_2015.pdf.
U.S. Department of Health and Human Services (2016) ‘Regulatory Guidelines
for Impact Analysis. Washington, DC: U.S. Department of Health and
Human Services’, https://aspe.hhs.gov/sites/default/files/pdf/242926/HHS_
RIAGuidance.pdf.
U.S. Department of Health, Education and Welfare (1969) ‘History of Title I
ESEA. U.S. Department of Health, Education and Welfare’, http://files.eric.
ed.gov/fulltext/ED033459.pdf.
Watts, H. W. and Rees, A. (eds) (1977a) New Jersey Income Maintenance
Experiment (vol. 2), Labor Supply Responses, New York: Academic Press.
Watts, H. W. and Rees, A. (eds) (1977b) New Jersey Income Maintenance
Experiment (vol. 3), Expenditures, Health and Social Behaviors; and the Quality
of Evidence, New York: Academic Press.
Weimer, D. and Vining A. (2017) Policy Analysis: Concepts and Practice, 6th Edition,
New York, NY: Routledge.
Weiss, C.H. (1998) Evaluation: Methods for Studying Programs and Policies (2nd edn),
Upper Saddle River, NJ: Prentice Hall.
Whitehurst, G. J. (2003) The Institute of Education Sciences: New Wine, New Bottles,
http://files.eric.ed.gov/fulltext/ED478983.pdf.
Zigler, E. and Valentine, J. (1979) Project Head Start: A Legacy of the War on Poverty,
New York: Free Press.
Zigler, E. and Muenchow, S. (1992) Head Start: The Inside Story of America’s Most
Successful Educational Experiment, New York: Basic Books.
112
SIX
Policy analysis in the states

Gary VanLandingham
Most assessments of policy analysis in the United States focus on the federal
government and mention activity in the 50 American states only in passing.
This oversight is unfortunate as a large volume of policy analysis occurs at the
state level, both within these governments and by a wide range of nonprofit and
private organizations including think tanks, research institutes, and advocacy
organizations. This activity supports the high volume of innovation that occurs
at the state level, a portion of which is subsequently adopted at the federal level.
States have been called, with good reason, ‘laboratories of democracy’ (Osborne
1990). For example, Minnesota was the first state to authorize charter schools,
which since have been authorized across the country. In 2006 Massachusetts
developed a universal health insurance program that served as the model for
‘Obamacare’—the 2010 federal Affordable Care Act. More recently, in 2010,
Wyoming became the first state to require energy companies to identify the
chemicals used in hydraulic fracturing drilling (fracking), California and other
western states have established a cap-and-trade system to reduce carbon emissions,
Wisconsin has eliminated collective bargaining rights for most of its public
employees, and Kansas has enacted major income tax cuts in the hopes of spurring
economic growth (California Environmental Regulation Agency 2015; Samuels
2015; Vockrodt 2015; Moncrief and Squire 2013). Whether one supports or
opposes such actions, it cannot be argued that these innovations by states have
been inconsequential. Given the ongoing political gridlock at the federal level,
states and other sub-national governments are likely to continue to be the drivers
of policy innovation in the United States for the foreseeable future. Neglecting
the state-level policy analysis that often supports such activity would thus provide
a limited and misleading view of the field.
The history of policy analysis within the states has been greatly shaped by the
institutional changes that have occurred within these governments over the past
50 years (see the chapter by Lynn in this volume.) Given the diversity among
the states, it is not surprising that their policy analysis organizations and activities
also vary widely. While state-level policy analysis has grown rapidly, it has also
fragmented, and many policy analysis organizations face important challenges.
This chapter discusses these trends and the potential future of policy analysis in
the states.
113
Development of policy analysis in the states
While policy analysis in the states is robust, this is a relatively recent development.
In fact, before the mid 20th century most states had very limited policy analysis
capacity and were considered to be backwaters. While there were notable
exceptions, such as the actions taken by some states early in the 20th century to
regulate corporations, by the 1930s through the 1960s the federal government
served as the primary driver of policy innovation through its response to national
challenges such as the Great Depression, World War II, the Cold War, the Civil
Rights era, and the War on Poverty (Moncrief and Squire 2013; Radin 2006).
In part, the federal government had assumed its leading role because state
governments had significant institutional weakness. As noted in a recent text:
If the national government was usurping some of the policy-making

roles traditionally left to the states and their localities, it was largely
due to the states’ intractability in exercising their responsibilities, and
inability to meet the needs of their citizens … And when states did
change, it was often too little, too late’ (Moncrief and Squire 2013,
pp 5–6).
A key problem was that many states had outmoded political systems dominated
by local interests that failed to address larger problems facing their jurisdictions.
For example, until the mid 1960s, Florida’s legislature was among the least
representative in the country. Each of the state’s 67 counties was guaranteed at
least one seat in the legislature, despite the fact that their populations ranged
from less than 10,000 to over 1 million. As a result, less than one fifth of the
state’s voters could elect a majority of both chambers of the state legislature and
it was controlled by the ‘Pork Chop Gang’—a coalition of legislators from rural
counties who were primarily known for supporting segregation, minimizing
taxation, and resisting government reforms. The Florida constitution, written
in the 1880s in reaction to the post-Civil War Reconstruction era, also placed
significant limitations on the legislature, which was to meet only once every
other year for a maximum of 60 days. This situation was far from unique—most
state legislatures operated under severe restrictions on their operations and lacked
the capacity needed to serve as independent and equal branches of government
(Coburn and deHaven-Smith 2002; Kurtz 2010; NCSL (National Conference
of State Legislatures) 2000).
This situation began to change in the 1960s, when major reforms to state
governments, coupled with growing interest in institutionalizing analysis within
government, brought substantial changes to the level and quality of policy analysis
within the states. A key stimulus to these reforms was a series of U.S. Supreme
Court decisions (Baker v Carr 1962; Reynolds v Simms 1964) that found that the
malapportionment of state legislatures was so great that it violated the equal
protection requirements of the federal constitution. States were required to create
114
legislative districts that represented more equal populations, which, in many states,
had the effect of shifting power from rural to urban areas and opening legislative
membership to a new generation of lawmakers who were interested in addressing
the problems facing their states (NCSL 2000).
Concurrently, a major effort was initiated to reform state legislatures, led by
influential groups including the Ford Foundation, the Carnegie Corporation, the
Citizen’s Conference on State Legislatures, and the Eagleton Institute of Politics
at Rutgers University. These groups highlighted the institutional weaknesses of
state legislatures and recommended actions to increase the professionalism of these
bodies, including holding more frequent and longer sessions, raising legislator pay
to enable a wider range of citizens to serve, and significantly increasing legislative
staffing. In response, many states shifted from biennial to yearly sessions and some
adopted year-round cycles where legislators met throughout the year (Moncrief
and Squire 2013; CSG (Council of State Governments) 2010; Kurtz 2010).
As part of these reforms, most states substantially increased their internal supply
of policy analysts. By 1974, two thirds of states reported that their central budget
office conducted program effectiveness analyses (Mushkin 1977) and almost
all states expanded their legislative staffing to include attorneys, budget and
policy experts, researchers and program evaluators, information technology and
communications specialists, and political advisors. These staff were often located
in central agencies that served members in both chambers and parties, such as
fiscal offices that analyzed budget proposals and program evaluation offices that
examined the operation of executive branch agencies. By 2009, state legislatures
reported employing approximately 28,000 professional staff, which enabled
members to assess and generate policy and budget proposals independently of
the executive branch (Hird 2005; Kurtz 2010; NCSL 2000, 2015b).
Actions by the federal government also supported the development of policy
analysis within the states. Responding to the increased desire to incorporate
analysis into government processes, the Government Accountability Office
promoted staff training within state legislative and executive branches to better
ensure that federal dollars were being used effectively and efficiently. The federal
government also spurred the development of several large data initiatives—
including the Uniform Crime Reporting system, the Medicaid Management
Information System, the Family Support Information System, the Social Services
Information system, and the Client Information System—to support analysis
capacity in the states (Mushkin 1977).
Additionally, several multi-state organizations were established or were expanded
to become important policy analysis resources for state governments. These
included the National Conference of State Legislatures, which serves legislators
and staff; the National Governor’s Association, which represents these chief
executives; and the Council of State Governments, which serves officials in all
three branches. In addition, similar organizations were formed to serve specialized
groups of officials such as the National Association of State Fiscal Officers and
the National Association of Chief Information Officers. These organizations
115
disseminate policy-relevant information through meetings and research reports,

and play a key role in spurring the diffusion of innovations among states (Berry
and Berry 2007; Moncrief and Squire 2013).
In recognition of this growing policy analysis capacity of states as well as national
political trends, the federal government has devolved significant responsibility
for many programs to the states, such as job training, public assistance, and
environmental regulation. These changes increased the range and complexity of
issues facing states, but also served to support their high level of policy innovation
(NCSL 2000).
Current status of policy analysis in states

A high level of policy analysis now occurs throughout state governments, although
this activity varies greatly across the nation in part due to states’ substantial
differences in wealth, level of taxation, number of employees, party control,
political culture, and balance of power between their governors and legislatures
(Hird 2005; Kraft and Furlong 2007; Lieske 1993; Weimer and Vining 2011).
Key nodes of activity within most states include legislative units, executive branch
offices, and external bodies such as nonprofit think tanks and university centers
(Mayne et al. 1999).
Policy analysis by state legislatures
State legislatures have made a substantial investment in staffing to provide

their members with access to research and data analysis services. In addition
to committee and caucus staffing, many legislatures have established units that
perform policy analysis studies. As noted by Hird (2005), these units fall into three
major categories: central research offices that respond to specific questions posed
by legislators and committees; committees that provide bill and fiscal analyses
services during legislative sessions but perform more substantive policy analyses
during the interim between sessions; and specialized offices that have the full-time
core mission of conducting substantive policy analyses (Carter-Yamauchi 2014;
NLPES (National Legislative Program Evaluation Society) 2015).
A 2014 survey of this third set of legislative units, conducted by the National
Legislative Program Evaluation Society and completed by offices from 39 states,
showed that they share similarities but also exhibit important differences. While
a few have existed for over 50 years, most were created in the 1970s when
legislatures were rapidly expanding their research capacity. However, three states
(Alaska, Maine, and North Carolina) have created legislative policy analysis
offices since 2000. The specific responsibilities of the offices also vary. In addition
to conducting qualitative and quantitative policy analyses, some also perform
investigations, and a few have a role in developing, assessing, or validating their
states’ performance measurement systems. In most states, the offices’ research
agenda is primarily set by the legislature, either through provisions in statute or
116
appropriations proviso or by directives from legislative leadership or governing

committees. However, in some states individual legislators or committees may
request studies. Somewhat reflecting the population size of their parent states,
the offices’ staffing varies substantially. Most have between 10 and 25 staff, but
five have 50 or more analysts. Similarly, the number of reports published by these
offices ranges from under 10 to over 50 policy analyses annually (NLPES 2015).
The activities of these legislative policy analysis offices can be portrayed by
profiling units in two states: the Florida Office of Program Policy Analysis and
Government Accountability (OPPAGA) and the New Mexico Legislative Finance
Committee (LFC).
The Florida OPPAGA is one of the larger state legislative policy analysis
offices. It has the mission of supporting the state’s legislature and ‘provides
data, evaluative research, and objective analyses to assist legislative budget and
policy deliberation’ (OPPAGA (Florida Office of Program Policy Analysis and
Government Accountability) 2015). The Office was created in 1994 after the
state’s Auditor General had resisted repeated legislative requests to broaden its
mission beyond its traditional financial and compliance auditing. OPPAGA was
staffed by transferring 88 positions from the Auditor General and it was given
the directive to conduct program evaluations and policy studies and support the
development of a statewide performance measurement system (Roth 2015).
OPPAGA is authorized to conduct studies on the full range of state government
activities. In 2014, its reports addressed a wide array of issues, including options
for increasing lottery revenues, the effectiveness of Florida’s system for detecting
Medicaid fraud and abuse, and the status of efforts to remove barriers to business
creation and expansion. In addition to these long-term studies, the office also
conducts a large number of short-term analyses that are not published but are
transmitted to legislators as confidential memoranda. In 2010, the last year for
which full data are available, the office published 65 policy analysis and evaluation
reports and projected that it would issue 165 confidential research memoranda
(OPPAGA 2009, 2010, 2015). In recognition of its work, OPPAGA has received
awards from the American Society for Public Administration’s Center for
Accountability and Performance and the National Legislative Program Evaluation
Society for the quality of its analyses and impact on public policy (NLPES 2015).
The New Mexico LFC was created in 1957 as the fiscal and management arm
of the New Mexico legislature. In 1991, the LFC also assumed responsibility
for conducting program evaluations from the Office of the State Auditor. LFC’s
mission is to ‘provide the Legislature with objective fiscal and public policy
analyses, recommendations and oversight of state agencies to improve performance
and ensure accountability through the effective allocation of resources for the
benefit of all New Mexicans’ (LFC (New Mexico Legislative Finance Committee)
2015a). Among the LFC’s responsibilities are to develop a budget proposal that the
legislature considers as an alternative to the governor’s budget recommendations.
The LFC also issues an annual policy analysis report that examines key issues
facing the state and actions that the legislature may wish to address during its next
117
session. In addition, the LFC issues quarterly reports that analyze the performance
measures reported by each agency and recommends actions that should be taken to
address problem trends. For example, in 2015 the LFC noted that the percentage
of juvenile justice clients who subsequently enter the adult corrections system
was increasing and it recommended specific steps that the Children, Youth and
Families Department should take to address this problem (LFC 2015b).
The LFC conducts in-depth policy analyses of state issues and programs and
released 11 such reports in 2014. For example, one report noted that the state’s
graduation rate was substantially below the national average and used a benefit-
cost analysis model to assess the fiscal and social impact of this problem; this
analysis found that successful actions to graduate 2,600 more students annually
would generate about $700 million in net benefits over the students’ lifetimes
(LFC 2014). In recognition of its high quality of work, the LFC was awarded
the 2015 Excellence in Evaluation Award by the National Legislative Program
Evaluation Society (NLPES 2015).
Policy analysis by states’ executive branch agencies
A variety of executive branch organizations produce policy analysis studies. A

2015 survey of executive budget offices by the National Association of State
Budget Officers found that 35 reported conducting program evaluations using
performance data, and 19 offices reported performing economy and efficiency
studies (NASBO 2015). Several states also have established units, such as Colorado’s
C-Stat, Maryland’s ‘StateStat’ and Washington State’s ‘Results Washington’, that
seek to improve outcomes and operational efficiency by tracking performance
indicators and using these metrics in regular discussions with agency managers
(Behn 2014).
Individual executive branch agencies within states often have research units that
conduct policy analyses and evaluations. While there are no comprehensive data
available on these offices, established networks of these units in some policy areas
demonstrate their proliferation. For example, Statistical Analysis Centers operate
in every state and the District of Columbia to collect, analyze, and disseminate
criminal and juvenile justice data, and just under half the states have established
Sentencing Commissions, which analyze court data and make recommendations
to improve the consistency of penalties assessed for criminal violations (Justice
Research and Statistics Association 2015; National Association of Sentencing
Commissions 2015).
As noted by Mayne et al. (1999), the focus of policy analyses by executive
branch organizations tends to be operational in nature, focusing on strategies to
improve efficiency and effectiveness through changes to existing programs. Due
to political considerations, few such studies assess program outcomes or consider
issues that agency leadership does not wish to raise.
118
Policy analysis by state universities and think tanks
Another key source of policy analysis in the states is the large number of university-
based and private organizations that generate policy studies intended to influence
the policy-making process (Hird 2005). The number of such state-based think
tanks has grown substantially in recent years. While a 1996 study of think tanks
identified 187 such entities in the 50 states and the District of Columbia (Rich
2006), the author’s 2014 update that used the same methodology found that this
number had increased to 659 state-level organizations (see the chapter by Rich
in this volume on think tanks.)
Many of these organizations combine their policy research activities with an
advocacy focus. In fact, many state think tanks are formal affiliates of national
organizations that have explicit ideological aims. For example, Americans for
Prosperity—a national organization founded by conservative/libertarian billionaire
businessman David Koch that supports free markets and entrepreneurship and
advocates for lower taxes, reduced government spending, and limited regulation—
has affiliated think tanks in 35 states that advocate for similar actions (FactCheck.
org 2015; AFP (Americans for Prosperity) 2015). Such ideologically focused think
tanks are often highly political and operate as part of policy networks that combine
policy research with lobbying, fundraising, and electoral support for candidates
who support their agendas. Other affiliated groups of think tanks support similar
causes but are not as politically focused. For example, the Government Research
Association, a network of citizen and business-oriented think tanks which seeks
‘improvement of government and the reduction of its costs’, lists affiliates in 19
states that share similar aims (GRA 2015). On the other side of the political
spectrum, the U.S. Public Interest Research Group, ‘a consumer group that stands
up to powerful interests whenever they threaten our health and safety, our financial
security, or our right to fully participate in our democratic society,’ has affiliated
think tanks in 47 states (USPIRG (U.S. Public Interest Research Group) 2015).
In contrast, other state-level think tanks, particularly those located in public
universities, seek to operate as politically neutral information centers. These
organizations typically have a specific policy area focus. For example, the LeRoy
Collins Institute at the Florida State University focuses on analyzing financial
issues facing governments within Florida. In recent years the Institute has issued a
series of reports examining challenges in the state’s tax system, the funding status
of Florida’s state and local public pension systems, and financial and accountability
issues relating to the state’s community development special districts (Florida State
University LeRoy Collins Institute 2015).
Some university-based think tanks provide policy analysis services that extend
beyond their home states. For example, the Center for Evidence-based Policy
within the Oregon Health and Science University works with policymakers in
more than 20 states to provide evidence intended to guide health policy decisions.
The Center’s Medicaid Evidence-based Decisions Project supports a network
of health agencies in 17 states by providing comprehensive research reviews
119
on the effectiveness of health-related interventions; the participating states use

this information to inform their Medicaid benefit design and coverage policy
choices such as whether to fund different angiography procedures for diagnosing
obstructive coronary artery disease. The Center’s Drug Effectiveness Review
Project provides similar reviews of medical literature to inform the drug coverage
policy decisions of 13 states’ Medicaid and public pharmacy programs (Center
for Evidence-Based Policy 2015).
Likewise, the Washington State Institute for Public Policy (WSIPP), which is
housed within the Evergreen State College, carries out non-partisan research at
the direction of the state legislature and WSIPP’s Board of Directors. The Institute
evaluates state programs and conducts comprehensive research reviews to identify
‘what works’—those programs that high-quality evaluations have shown to be
effective in achieving desired outcomes such as increased high school graduation
rates and reduced recidivism of persons released from prison. WSIPP also has
developed a benefit-cost model that analyzes the long-term return on investment
that the state could achieve through investments in a broad range of program
alternatives across five social policy areas, including criminal justice, substance
abuse, mental health, child welfare, and Pre-K to secondary education. These
reports have been influential in informing the Washington legislature’s policy
choices, and, as discussed below, the benefit-cost model has been adopted by 23
other states that participate in the Pew-MacArthur Results First Initiative (WSIPP
(Washington State Institute for Public Policy) 2015; Pew 2017).
The relative strengths and weaknesses of the different sources of state-level
policy analysis are summarized in Table 6.1, with the major differences being
the types of issues each source tends to focus on, their ability to obtain state data,
their degree of organizational independence, and their access to policymakers.
Impact of state-level policy analysis organizations

A discussion of the role of policy analysis in state government would be incomplete
without considering its use by key stakeholders. Predictably, available research
suggests that state-level policy analyses are most likely to have impact when
conducted within the organization that is making the examined policy choice
(for example, when a legislative research unit conducts an analysis to inform
an upcoming legislative budget decision), in part because internal researchers
have the advantage of direct access to policymakers, which aids communication
of results and helps build confidence in the reliability of the analysis (Mayne
et al. 1999). Similarly, research indicates that state-level analyses that include
readily actionable recommendations have a relatively high degree of use, as do
studies that incorporate benefit-cost analysis (VanLandingham 2011; White and
VanLandingham 2015).
120
Table 6.1: Relative strengths and weaknesses of state-level policy analysis organizations
Organization Relative strengths Relative weaknesses
Legislative • Organizational separation from state • May have incomplete knowledge of
agency operations can increase agency operations
objectivity of their analysis of executive • Results may be seen as political,
branch particularly if legislative and executive
• Ready access to legislative policymakers branches are controlled by different
and understanding of legislative process parties
can enhance relevance of studies • Focus on legislative priorities may limit
• Typically have access to a wide range utility to other stakeholders
of agency data • May avoid issues that are politically
• Studies often focus on fiscal impact, sensitive to the legislature
which can be useful in budget process
Executive and • Location within executive branch and • May avoid inquiries or reporting
agency understanding of agency processes may findings that are politically sensitive
increase relevance of studies to agency to executive branch; findings may be
and executive leadership challenged as self-serving
• Ready access to agency data and • Focus on meeting agency or executive
officials may enhance research quality information needs may narrow scope of
• Studies often focus on operational analysis to operational matters
issues that are within agency control, • Analysis units may be subject to
generating actionable recommendations disproportionate cutbacks when
for program improvement executive leadership changes or budget
shortfalls arise
External think • More likely to focus studies on program • May have limited access to government
tank and social outcomes rather than data and incomplete knowledge of
operational issues program operations
• May have access to specialized expertise • Results may not be timed to decision-
not readily available to executive and making windows
legislative agencies • May have limited access to
• Research not as subject to time policymakers, hindering communication
constraints of results
• Separation from state entity may • Findings may be seen as political or
increase objectivity and perspective self-serving if generated by ideological
• May use communication strategies advocacy organization
(press conferences, social media) that • Topics may not be attuned to political
are not as available to government viability or relevance
analysts to promote findings
Source: Adapted from Mayne et al. (1999, p 32).
Hird (2005) examined the impact of analyses by legislative research offices by

surveying state legislators; this analysis concluded that the offices were moderately
influential in policy outcomes. Larger offices and those that performed more
analytical and longer-term studies were more highly rated by the surveyed
legislators than smaller offices and those that provided mostly descriptive analyses.
121
Interestingly, this study also found that legislative research units in states that were
controlled by a single political party tended to be smaller and focused primarily
on fact gathering, possibly because the dominant parties in such states wished to
ensure that the research offices did not produce politically inconvenient analyses
that the minority party could use to challenge their policy agendas.
VanLandingham (2011) similarly examined legislative policy analysis offices
and found that perceptions of impact were closely related to these offices’
organizational placements. Units that were located in legislative committees or
independent offices were viewed more positively by senior legislative staff than
those housed in legislative auditing offices, largely because the former worked
more closely with legislative stakeholders and provided more readily actionable
reports. Offices located in states that had experienced recent changes in party
control were viewed as less influential than those in states with more stable
legislative environments, likely because changes in party control disrupted the
offices’ relationships with key legislators and staff.
The impact of policy analyses generated by external entities appears to vary
by source. As noted by VanLandingham (2006), a nationwide survey of senior
legislative staff assessed the perceived credibility and usefulness of alternative
sources of policy analysis to their state legislatures. The National Conference
of State Legislatures was rated as ‘credible’ by 92% of survey respondents and
as ‘useful’ by 87% of these senior staff. Over half of these persons also viewed
information provided by federal agencies, other national organizations such as
the Council of State Governments, and industry groups as highly or somewhat
credible and useful. In contrast, less than half of the survey respondents viewed
information provided by partisan offices, think tanks, and the media as highly
or somewhat credible and useful. Clark and Little (2002) found that legislators
from states where legislators frequently transition to higher offices, those serving
in part-time legislatures, and those from states without term limits were most
likely to rely on information from national organizations.
State agency use of policy analysis has not been as widely studied. Lester (1993)
found that agency officials reported relying primarily on policy information from
their peers in other agencies, newspapers, counterparts in federal agencies, and
the governor’s office; policy analysis from research organizations and universities
was not relied on heavily. A 2015 study on state agencies’ use of information to
improve their operations found that use of analysis varied by type of agency—
natural resource, substance abuse, and mental health agencies reported the highest
use of such analyses (Jennings and Hall 2015).
Challenges facing state-level policy analysis

While policy analysis in the states is currently robust, it faces several challenges.
Increased political polarization, legislative term limits, and changes in party
control have fundamentally changed the environment in which the work occurs
in many states. State-level policy analysts also operate in an increasingly crowded
122
information marketplace, making it more difficult for them to gain the attention
and support of policymakers. These challenges have negatively impacted many
well-respected policy analysis organizations in recent years.
Limits on legislative service, commonly called term limits, have fundamentally
changed the policy environment in many states. These limitations, which prohibit
elected officials from serving in the state legislature for more than a specified
numbers of years, are in effect in 15 states (NCSL 2015a). These restrictions
have typically been created through citizen initiatives, generally in reaction to a
desire to increase the diversity of candidates and out of a belief that long-serving
legislators have too much power and/or are under the control of special interests
(Bowser 2005). These limits vary by state, with six instituting lifetime limits on
service—the most restrictive being Michigan, with six years for its house and
eight for senate (NCSL 2015a). While research on the impact of term limits
is limited, it is generally agreed that they reduce institutional knowledge and
professionalism in legislatures (Squire 2012; Percival 2015). Mirroring the national
trend, the political environment of many states has also become more ideologically
polarized over time (Shor 2014). Many states have also experienced changes in
party control of their executive and legislative branches in recent years (NCSL
2015a). Collectively, these trends can make it harder for state policy analysts to
build and maintain relationships among influential policymakers.
Overall, these changes have had significant implications for policy analysis
organizations, which can find themselves in much less supportive political
environments. While policy analysis offices have a mission of ‘speaking truth to
power’ as neutral and objective honest brokers, several have reported that they now
operate in a ‘shoot the messenger’ environment in which they became suspect to
interest groups and political leaders that wish to only receive analysis results that
support their desired actions (Radin 2006; Denhardt 1981).
These challenges have affected many state policy analysis organizations. For
example, term limits became effective in Florida in 2000 and resulted in a growing
deference by the legislature to the governor in policy and budget decisions as
experienced legislators were increasingly forced out of office. Beginning in the
mid 2000s, the governor repeatedly proposed eliminating the OPPAGA after it
had issued several reports critical of his policy proposals. The state’s term limits
took full effect in 2010 when the last of the long-serving members (those who
had moved to the Senate in the early 2000s after being termed out of the House
of Representatives) were forced to leave office. The 2011 Legislature substantially
reduced funding for its research units, substantially changed the environment
in which they work, and eliminated some units altogether. OPPAGA’s budget
was cut by approximately 25% and its staffing was reduced from 80 to 48 analyst
positions. OPPAGA’s organizational independence was significantly reduced,
eliminating the requirement that its director be confirmed by the full legislature
for a four-year term and instead specifying that they serve at the pleasure of the
Speaker of the House and President of the Senate. Much of OPPAGA’s work
is now no longer publically available but is transmitted directly to legislative
123
leadership and is exempt from public record disclosure. The Auditor General,
which conducts financial audits, experienced a similar staffing cut and its state
agency oversight duties were reduced, with the frequencies of routine operational
audits of agencies decreased to once every three years. Two other legislative analysis
offices were defunded—the Technology Review Workgroup, which examined
the state’s information technology systems, and the Legislative Committee on
Intergovernmental Relations, which examined the impact of state policy on local
governments (NCSL 2010; Roth 2015).
Other well-regarded state policy analysis units have experienced similar negative
outcomes in recent years. For example, the Oregon Progress Board, which tracked
progress towards the economic, social, and environmental goals established in the
state’s strategic plan, was eliminated in 2009 (Weber et al. 2014), and the Kentucky
Long-Term Policy Research Center, which examined policy issues facing that
state, was defunded by that state’s legislature in 2010 (Lexington Herald-Leader
2010). Characterizing the political nature of such actions, the Ohio Legislative
Office of Education Oversight was eliminated in 2005, reportedly after an
incoming House Speaker objected to a report the office had issued on charter
schools, an issue of personal interest to the new presiding officer (Zajano 2005).
Whither policy analysis in the states?

State-level policy analysts are facing much more complex environments than they
did in the past. In the field’s infancy, there was a general consensus that more
information and greater transparency was needed to improve policy making,
and that analysts who faithfully filled this information gap would be heard and
supported. The experiences of the past 40 years have lifted the veil of naïveté on
which this perspective was based, but have also provided much greater insights
into the roles that analysts can successfully play within the policy process.
Strategic use of these insights is increasingly important because the environment
in which state policy analysts must operate is becoming more challenging. Many
analysts are now facing legislative term limits that erode institutional memory
and capacity, much greater political polarization that focuses on winning policy
debates and has little concern for protecting independent sources of information,
and increased competition from other sources of often ideologically based policy
research. State policy analysis offices operate in minefields in which missteps can
be fatal. They must be able to quickly prove their worth as honest information
brokers to new actors who quickly shift in and out of political positions, and they
can easily be seen as political impediments to be removed if the information they
provide is unwelcome. Moreover, where they once were one of the only sources
of analysis available to state policymakers, state policy analysts now operate in
a crowded analytical space, where information is available from many sources
including federal and state agencies, university think tanks, and national and state
advocacy groups (Radin 2000, Hill 1997) and a growing discordant stew of new
media websites, bloggers, and opiners that often have little regard for facts.
124
The growing diversity in policy analysis providers may be viewed in alternative

ways—as a positive outcome that supports the democratic process in the states by
enabling groups of citizens to compete in the policy arena, or as an outsourcing
of policy analysis to interest groups. A meta-analysis of the role of think tanks
in the policy process (Results for Development 2014) found that think tanks
encourage demand for policy alternatives and support open, democratic political
systems. However, others have argued that many of these outside groups have
explicit policy agendas and blur the line between research and advocacy, resulting
in disputes over both facts and conclusions as well as contributing to ideological
polarization (Radin 1997).
At least three alternative futures for policy analysis in the states appear feasible.
In the first scenario, the current status quo would continue, with state-level policy
analysis provided by both a rich mixture of organizations within the governments
and a wide range of external entities, including university research centers and
advocacy-oriented think tanks. In this future, policy analysts working within
state governments would continue to experience the challenges created by messy
politics; their ride would be bumpy at times but their offices would generally
survive in something closely resembling their current form. As exemplified
by New Mexico’s LFC, such offices would need to be flexible and politically
sensitive, continually seeking to build trust, understanding, and reciprocity with
key stakeholders while preserving their independence and ability to ‘speak truth
to power’ (Radin 1997).
In a second scenario, the current negative trends would continue and policy
analysis offices within states would be increasingly marginalized and politicized. As
exemplified by the experiences in Florida, Kentucky, and Oregon, many internal
policy research offices may find their role increasingly limited to providing short-
term and descriptive research that is politically non-threatening (Hird 2005);
substantive analysis would be largely outsourced to favored external entities that
are trusted to deliver the conclusions desired by political elites. In this future,
policy analysts would often find themselves in the role of providing justifications
for the preferred policy choices of leaders rather than serving as trusted advisors
who provide objective information to inform these choices.
In a third scenario, policy analysis within states would survive but evolve into
new forms. In this future, the mission of policy analysis units would broaden to
include serving as knowledge curators in addition to their traditional function
as research generators. This new role would seek to address one of the inherent
challenges facing policy analysts—the decision-makers they serve often require
quick answers to many questions, yet original research to address these matters
is very time-intensive. While a huge amount of information that could help
to inform these policy choices can be quickly accessed via the web, much of
this material is of dubious quality, including, unfortunately, many published
policy analysis studies. To address this problem, policy analysts can use resources
developed by a growing number of research intermediaries to curate relevant
materials and quickly provide valid and useful information to decision makers.
125
These intermediaries include a large number of research clearinghouses

that have been created to conduct systematic reviews of available research and
develop lists of ‘what works’—those policy choices that rigorous analyses and
evaluations have found to be effective. The clearinghouses include entities such
as Crimesolutions.gov, operated by the U.S. Department of Justice; Blueprints for
Healthy Youth Development, operated by the University of Colorado, Boulder;
and the aforementioned Coalition for Evidence-Based Policy. Using structured
protocols, the clearinghouses assess and rate interventions based on the level
of rigorous evidence that exists about the outcomes they generate. The Pew-
MacArthur Results First Initiative, a partnership of The Pew Charitable Trusts and
the John D. and Catherine T. MacArthur Foundation, has further consolidated
this information by creating a centralized database that enables policy analysts to
quickly identify the ratings issued by eight clearinghouses that examine social
policy interventions. The Results First Initiative has also developed a benefit-cost
model that enables analysts to use this evidence base to predict the long-term
return on investment that their governments could achieve by funding various
social programs.
Such curation resources can enable state policy analysts to expand the range of
services that they can provide to their policymakers in relatively short time frames.
For example, legislative and executive branch policy analysis offices in 20 states,
including the New Mexico LFC, have used tools provided by Results First to
analyze the effectiveness of their states’ investments in criminal justice and other
social program areas. New Mexico has used the LFC’s analyses to inform decisions
to direct over $80 million to evidence-based programs that were predicted to
generate strong returns on investment to state residents (VanLandingham and
White 2015). The ability to provide such analysis has contributed to LFC’s strong
reputation as well as the ongoing support it enjoys from policymakers. Other state
policy analysis organizations may wish to adopt such a curation function to help
them address the challenging environments in which they operate.
References
Americans for Prosperity (2015) ‘States’, http://americansforprosperity.org/.
Behn, R. (2014) The PerformanceStat Potential: A Leadership Strategy for Producing
Results, Washington, DC: Brookings Institution Press.
Berry, F. and Berry, W. (2007) ‘Innovation and Diffusion Models in Policy
Research’, in P. A. Sabatier (ed), Theories of the Policy Process (2nd edn), Boulder,
CO: Westview.
Bowser, J. (2005) ‘The Effects of Legislative Term Limits’, in The Book of the States
2005 114–15. Lexington, KY: The Council of State Governments.
California Environmental Protection Agency (2015) ‘Cap-and-trade program’,
www.arb.ca.gov/cc/capandtrade/capandtrade.htm.
Carter-Yamauchi, C. (2014) ‘The History and Evolution of Legislative Committee
Systems’, RACCS Newsletter, Fall 2014. Denver, CO: NCSL.
126
Center for Evidence-Based Policy (2015) ‘About Us’, www.ohsu.edu/xd/

research/centers-institutes/evidence-based-policy-center/evidence/index.
cfm 1.
Coburn, D. and deHaven-Smith, L. (2002) Florida’s Megatrends—Critical Issues in
Florida, Gainesville, FL: University of Florida Press.
Council of State Governments (2010) The Book of the States 2010, Lexington,
KY: The Council of State Governments.
Denhardt, R. (1981) ‘Towards a Critical Theory of Public Organization’, Public
Administration Review 41(6): 633.
FactCheck.org. (2015) ‘Americans for Prosperity’, http://www.factcheck.
org/2014/02/americans-for-prosperity-3/.
Florida Office of Program Policy Analysis and Government Accountability (2009)
Business Plan 2009–10, Tallahassee, FA: OPPAGA.
Florida Office of Program Policy Analysis and Government Accountability (2010)
‘Reports, 2010’, www.oppaga.state.fl.us/ReportsYearList.aspx?yearID=22.
Florida Office of Program Policy Analysis and Government Accountability
(2015) ‘About OPPAGA’, www.oppaga.state.fl.us/shell.aspx?pagepath=about/
about.htm.
Florida State University LeRoy Collins Institute (2015) ‘Welcome to the LeRoy
Collins Institute’, http://collinsinstitute.fsu.edu/.
Government Research Association (2015) ‘About GRA’, www.graonline.org/
about-the-governmental-research-association/.
Hird, J. (2005) Power, Knowledge, and Politics: Policy Analysis in the States,
Washington, DC: Georgetown University Press.
Jennings, E. Jr. and Hall, J. (2015) ‘State Agency Attention to Scientific Sources
of information to Guide Program Operations’, in A. Shillabeer, T. Buss, and D.
Rousseau (eds), Evidence-Based Public Management: Practices, Issues, and Prospects
(1st edn), Armonk, NY: M.E. Sharpe.
Justice Research and Statistics Association (2015) ‘Statistical Analysis Centers’,
www.jrsa.org/sac/.
Kraft, M. and Furlong, S. (2007) Public Policy: Politics, Analysis, and Alternatives
(2nd edn), Washington, DC: CQ Press.
Kurtz, K. (2010) ‘75 Years of Institutional Change in State Legislatures’, in The
Book of the States 2010, Lexington, KY: The Council of State Governments
85–91.
Lester, J. (1993) ‘The Utilization of Policy Analysis by State Agency Officials’,
Science Communication 14(3): 267–90.
Lexington Herald-Leader (2010) ‘State Research Center’s Demise Comes as
Surprise’, April 7.
Lieske, J. (1993) ‘Regional Subcultures of the United States’, Journal of Politics
55(4): 888–913.
127
Mayne, J., Divorski, S., and Lemaire, D. (1999) ‘Locating Evaluation: Anchoring
Evaluation in the Executive or the Legislature, or Both or Elsewhere?’ in R.
Boyle and D. Lemaire (eds), Building Effective Evaluation Capacity: Lessons from
Practice, New Brunswick, NJ: Transaction Publishers 38.
Moncrief, G. and Squire, P. (2013) Why States Matter: An Introduction to State
Politics, Plymouth, UK: Rowman and Littlefield.
Mushkin, S. (1977) ‘Policy Analysis in State and Community’, Public Administration
Review 37(3): 245–53.
National Association of Sentencing Commissions (2015) ‘Sentencing
Commission/Council Contact List’, http://thenasc.org/images/
NASCComCouncilContactList20150807.pdf.
National Association of State Budget Offices (2015) ‘2015 Budget Processes in the
States’, https://higherlogicdownload.s3.amazonaws.com/NASBO/9d2d2db1-
c943-4f1b-b750-0fca152d64c2/UploadedImages/Reports/2015%20
Budget%20Processes%20-%20S.pdf.
National Conference of State Legislatures (2015a) ‘Ensuring the Public Trust
2015: Public Policy Evaluation’s Role in Serving State Legislatures’, Denver,
CO: NCSL.
National Conference of State Legislatures (2015b) ‘Size of State Legislative
Staff—1979, 1988, 1996, 2003, 2009’, www.ncsl.org/research/about-state-
legislatures/staff-change-chart-1979-1988-1996-2003-2009.aspx.
National Legislative Program Evaluation Society (2015) ‘NLPES Awards
Program’, www.ncsl.org/legislators-staff/legislative-staff/program-evaluation/
nlpes-awards.aspx.
New Mexico Legislative Finance Committee (2014) Cost-Effective Options for
Increasing High School Graduation and Improving Adult Education, Santa Fe, NM:
LFC.
New Mexico Legislative Finance Committee (2015a) ‘About LFC’, www.nmlegis.
gov/lcs/lfc/lfc_about.aspx.
New Mexico Legislative Finance Committee (2015b) Children, Youth and Families
Department, Key Quarterly Performance Measures Report 1st Quarter 2015, Santa
Fe, NM: LFC.
Osborne, D. (1990) Laboratories of Democracy, Boston, MA: Harvard Business
School Press.
Percival, D. (2015) Smart on Crime: The Struggle to Build a Better American Penal
System, Boca Raton, FL: CRC Press.
Radin, B. (1997) ‘Presidential Address: The Evolution of the Policy Analysis Field:
From Conversation to Conversations’, Journal of Policy Analysis and Management
16(2): 204–18.
Radin, B. (2000) Beyond Machiavelli: Policy Analysis Comes of Age, Washington
DC: Georgetown University Press.
Radin, B. (2006) ‘Professionalization and Policy Analysis: The Case of the United
States’, in H. K. Colebatch (ed), The Work of Policy: An International Survey,
Lanham, MD: Lexington Books.
128
Results for Development (2014) Linking Think Tank Performance, Decisions, and
Context, Washington DC: Results for Development.
Rich, A. (2006) ‘State Think Tanks’, Database from Andrew Rich, received from
personal communication, April 3.
Roth, T. (2015) Personal Interview, October 10.
Samuels, R. (2015) ‘Walker’s Anti-union Law has Labor Reeling in Wisconsin’,
Washington Post, February 26.
Shor, B. (2014) ‘How U.S. State Legislatures Are Polarized and Getting More
Polarized (in 2 Graphs)’, Washington Post, January 14, www.washingtonpost.
com/blogs/monkey-cage/wp/2014/01/14/how-u-s-state-legislatures-are-
polarized-and-getting-more-polarized-in-2-graphs/.
Squire, P. (2012) The Evolution of American Legislatures: Colonies, Territories, and
States 1619–2009, Ann Arbor. MI: The University of Michigan Press.
The Pew Charitable Trusts (2017) ‘Where We Work’, www.pewtrusts.org/en/
projects/pew-macarthur-results-first-initiative/where-we-work.
U.S. Public Interest Research Group (2015) ‘About Us’, www.uspirg.org/page/
usp/about-us.
VanLandingham, G. (2006) A Voice Crying in the Wilderness—Legislative Oversight
Agencies’ Efforts to Achieve Utilization, unpublished doctoral dissertation, Florida
State University, Tallahassee, FL.
VanLandingham, G. (2011) ‘Escaping the Dusty Shelf: Legislative Oversight
Agencies’ Efforts to Promote Utilization’, American Journal of Evaluation 32(1):
85–97.
VanLandingham, G. and White, D. (2017) ‘New Mexico Results First: Lessons
From a Long-Term Evidence-Based Policymaking Partnership’, in J. Owen and
A. Larson (eds), Researcher–Policymaker Collaboration: Strategies for Launching and
Sustaining Successful Partnerships,New York, NY: Routledge.
Vockrodt, S. (2015) ‘Brookings Institution: Kansas Income Tax Cuts Were Not
a Good Idea’, http://www.pitch.com/news/article/20562140/brookings-
institution-kansas-income-tax-cuts-were-not-a-good-idea
Washington State Institute for Public Policy (2015) ‘Benefit Cost Results’, www.
wsipp.wa.gov/BenefitCost.
Weber, B., Worcel S., Etuk, L., and Adams, V. (2014) Tracking Oregon’s Progress:
A Report of the Tracking Oregon’s Progress (TOP) Indicators Project, Corvallis, OR:
Oregon State University.
Weimer, D. and Vining, A. (2011) Policy Analysis: Concepts and Practice (5th edn),
New York: Routledge.
White, D. and VanLandingham, G. (2015) ‘Cost-benefit Analysis in the States:
Status, Impact, and Challenges’, Journal of Cost-benefit Analysis 6: 369–99.
Zajano, N. (2005) Personal Interview, March 15.
129
SEVEN
Policy analysis and evidence-based decision

making at the local level
Karen Mossberger, David Swindell, Nicholet Deschine Parkhurst,
and Kuang-Ting Tai
Since the municipal reform movement at the turn of the 20th century, data and
analysis have been tools for change and innovation at the local government level.
With the advent of new technologies and sources of data, local governments
are actively exploring how to use these effectively for evidence-based decision
making. While there is a lack of systematic research on how policy analysis is
being used at the local level, local governments have some special opportunities
for experimentation and learning in this area.
Local governments deliver many core services in the United States, ranging
from public safety to human services delivery. They are also engaged in many
different policy areas, including land use planning for purposes of economic
development, health and safety, and even morality policy (Sharp 2005). As the
level of government closest to citizens, local governments may involve citizens
in processes for collecting, sharing, and discussing data and analysis. So, as
VanLandingham notes, in his chapter in this volume, that states are often thought
of as laboratories of democracy in the federal system, cities represent an even
broader range of experimentation, with varied tax and service preferences as
well as managerial and governance approaches nested within the frameworks of
their home states.
While this chapter highlights the uses of the traditional tools of analysis at
the local level, we will also consider performance management data, open
data, and predictive analytics using ‘big data.’ After examining some case studies
demonstrating local applications of policy analysis and data, we consider how
these recent trends may influence the use of analysis at the local level.
Context for local policy making in the United States

In order to understand the use of policy analysis and evidence at the local level, we
first offer a basic description of the policy and institutional landscape. Substantial
variation in policy responsibilities and capacities exists across local governments,
based on several factors: type of government, state law, and city size.
Local government is an important provider of services in many policy areas
in the United States, accounting for just under one third of non-defense
spending (Wolman 2012). This includes public safety, housing and community
amenities, education, transportation, recreation, and culture as domains where
131
local government spending accounts for a significant portion of total public

expenditures; but local governments are also engaged in many other policy
areas as well (Wolman 2012). Because local governments in the United States
depend to a great extent on own-source revenues (in comparison with other
developed nations), they also have strong incentives to undertake local economic
development, encouraging investment and employment within their boundaries
(Wolman 2012; Mossberger and Stoker 2001).
In the 2012 Census of Governments, there were 90,056 local governments
in the United States. This total included approximately 3,000 counties, 20,000
municipalities, and 16,000 townships, all of which had general-purpose local
governments with diverse policy responsibilities. In terms of special purpose
governments, the United States was home to approximately 13,000 independent
school districts. Additionally, there were 38,000 special districts that focused on a
single policy area, which could be economic development, transportation, parks,
sewer-sanitation, water conservation, mosquito abatement, or library districts, to
name but a few (Hogue 2013).
The 2012 Census of Governments also shows that the numbers of local
governments vary considerably by state. Illinois is home to almost 7,000 local
governments while Hawaii has only 21 (the national average is 1,801 per
state). Controlling for population shows considerable variance as well. North
Dakota turns out to love local governments quite a bit with 383.8 units of
local government per 100,000 population, and Hawaii has the fewest with only
1.5 units per 100,000 population. While local government employment far
outstrips federal employment, this varies significantly from a high of 286 local
government employees per 10,000 population in Wyoming, to a low of 63 per
10,000 population in Delaware. Local governments, therefore, vary greatly in
size and capacity.
This array of local governments exists within a constellation of what are
essentially 50 systems of local government, regulated by state laws that affect
the structures, powers, and finances of local entities (Frug and Barron 2008).
Municipalities and counties exist throughout the United States (whereas towns
and townships tend to cluster in New England and the Midwest). While a court
decision in 1868 known as Dillon’s Rule established that local governments are
created and regulated by the states, state governments often allow a fair amount of
discretion to local governments for policy decisions. Dillon’s Rule mandates that
local governments can only exercise authority specifically granted by state statute.
But in states that have adopted ‘home rule’ legislation, eligible local governments
can undertake activities as long as they are not specifically prohibited from doing
so by state law (Krane et al. 2001; Frug 1980). The majority of states—36 of 50—
provide for some level of home rule, most often for municipalities (Richardson et
al. 2003, cf. Kubler and Pagano 2012), especially in states with larger populations.
Historically, the municipal reform movement at the turn of the 20th century
introduced the council-manager form of government to promote professional
management and to counter corrupt machine politics at the local level. Today,
132
Policy analysis and evidence-based decision making at the local level
there are varied forms of local government in the United States, even within
a single state. The council-manager form has a professional chief executive
appointed by an elected city council, modeled after corporate forms of governance
(Judd and Swanstrom 2015). In mayor-council cities, the elected mayor is the
chief executive. In 2011 council-manager governments accounted for 54% of
municipalities with populations between 5,000 and 249,999. However, they
comprised only about 38% of cities over 250,000 (ICMA (International City/
County Management Association) 2011). There has also been some blending of
models over time, with some mayor-council governments providing positions
for chief administrative officers (Frederickson et al. 2004).
Growing professionalization of local government management has increased
capacity to utilize analyses and evidence. There are also a number of professional
organizations that spread ideas across local governments and their officials,
including groups of mayors, managers, cities, counties, and townships. These
are represented by such organizations as the National League of Cities, the
International City/County Management Association (ICMA), the National
Civic League, the Alliance for Innovation, and the Engaging Local Government
Leaders organization. Furthermore, several recent developments have emerged
among funding entities focusing on evidence-based decision making at the local
level of government, including initiatives at the Arnold Foundation and the
Bloomberg Philanthropy’s ‘What Works Cities’ efforts. As the above description
demonstrates, local government is a complex and varied phenomenon in the
United States, embedded within a decentralized system of federalism with varying
degrees of autonomy for cities. This means that the functions, responsibilities, and
capacities of local government vary. Adoption of new practices often depends on
voluntary diffusion rather than federal or state requirements. Yet, there have been
organizations supporting the adoption of analytical and data-driven practices, and
some local governments that have been at the forefront in promoting such uses.
Historical development of evidence-based policy making at the

local level
The rise of council-manager government at the beginning of the 20th century
focused on core values of professionalism and efficiency (Wollman and Thurmaier
2012). Concurrent with the city manager movement was the establishment of new
municipal research organizations to promote the spread of ‘scientific management’
based in part on the collection of data to inform policy and practice. The most
well-known example was the New York Bureau of Municipal Research, which
influenced public administration as a field (Stivers 1995, 1997; see also Weimer
and Peters in this volume). Municipal research bureaus spread to other cities as
well, and the empirical methods and training programs spread through teaching in
public administration at universities (McDonald 2010). The federal government
supported the Advisory Commission on Intergovernmental Relations (ACIR)
for almost three decades to help identify promising local government practices.
133
Observation, data collection, and analysis were important to these organizations’

efforts to affect management and policy.
While the municipal reform movement set the stage for the use of analysis
and data, the specific techniques used today in policy analysis spread to local
governments in the 1960s and 1970s. Ideas often diffuse across local and state
governments because of federal incentives or emulation of federal programs
(Eyestone 1977), and this was the case with policy analysis. The use of the
Planning-Programming-Budgeting System (PPBS) by the federal government
in the 1960s encouraged the adoption of policy analysis at the local level (see
Weimer and Peters in this volume for a history of PPBS at the federal level).
Initially, local governments employed policy analysis in budgeting even as the
use of PPBS declined. Federal grants during the 1960s, such as the Community
Renewal Program, required some use of systematic analysis by local governments
(Kraemer 1973; Mushkin 1977).
Diffusion also occurs as adopters take cues from leading innovators and
professional associations (Walker 1969; Balla 2001; Mossberger 2000; Shipan
and Volden 2012). Larger cities, notably New York and Los Angeles, were early
proponents of policy analysis, working with think tanks such as Rand (Kraemer
1973). Professional associations, such as ICMA, also promoted the use of policy
analysis by setting professional standards and norms through their publications and
national training (see as an example the ICMA publication by Kraemer 1973).
Even with these influences, however, one survey conducted by ICMA in 1971
reported that only 16% of local governments with populations between 5,000
and 100,000 had staff units doing analytical work (Mushkin 1977).
There are other sources for local policy analysis, however. Local governments
rely on universities or consultants to fill many of their analytic needs, at least on
major issues. For example, many cities hire university researchers to help project
economic impact for large-scale public investments (see Rosentraub and Swindell
2009). Other local actors, including unions, chambers of commerce, local think
tanks, or advocacy groups may also conduct analysis to persuade policymakers (or
voters in the case of referenda). For example, there have been multiple analyses
of pension reforms in Phoenix, with differing views of the costs and benefits of
ballot proposals from locally based think tanks and university research institutes
(Chieppo 2013; Reilly 2016; Gilroy et al. 2016). Ex post evaluations also may
be funded or undertaken by local organizations, such as Duke Power contracting
for an external study measuring the satisfaction, economic impact, and market
impact of hosting the 2012 Democratic National Convention in Charlotte,
North Carolina (Heberlig et al. 2017). Analysis that originates outside of local
governments may be formally or informally used by local officials, by the media,
and others in policy debates.
Use of policy analysis at the local level may also take advantage of resources
outside of the locality. National think tanks produce research and policy analysis
with implications for local governments, including some think tanks focused on
urban policy and local government, such as the Urban Institute that President
134
Johnson founded in 1968 to produce research on federal antipoverty programs. In

subsequent years, the Urban Institute has conducted research on tax policy, low-
income housing, health care reform, welfare reform, and urban change, among
other topics (Urban Institute n.d.b). Another national research organization geared
toward local issues is the Brookings Institution’s Metropolitan Policy Program,
which makes recommendations on policies related to regional economic growth,
innovation, and demographic change.
Professional associations also distribute findings from research and policy analysis
to their members, such as ICMA and the Alliance for Innovation (discussed in
more detail in the following sections). Other professional associations disseminate
research as well. The National League of Cities offers applied research and
analysis on a number of topics, including economic development, finance,
governance, housing, infrastructure, the sharing economy, sustainability, and urban
development (NLC (National League of Cities) 2016). The National Association
of Counties includes on its website research on transportation and infrastructure,
justice, civic engagement, community and economic development, health, and
more (National Association of Counties 2015). All four of these organizations
are dues-based organizations in which individuals or jurisdictions join and pay
a fee. Other organizations have corporate sponsors to support their efforts. One
example of this model is the Smart Cities Council, which advocates for cities to
adopt smart technologies in order to improve service delivery and efficiencies
(Smart Cities Council, n.d.). They produce research to help cities and provide
these through multiple networks, conferences, and webinars.
In addition to internal expertise, professional associations, and think tanks, many
cities have built close working partnerships with local universities. Sometimes,
these relationships are at an individual level with specific faculty members.
Other institutions have built research centers dedicated to working with local
governments in a wide range of capacities and policy domains. Arizona State
University is home to the Center for Urban Innovation. The University of
Pennsylvania has the Institute for Urban Research. Rutgers University is home
to the Center for Urban Policy Research. These represent only a small sample
of the university-based research centers bridging academic research with the
applied needs of practitioners trying to solve local problems. Still, significant
challenges remain for closing the gap between university-based expertise and
local government needs (Gordon 2016; Wang, Bunch, and Stream 2013).
Data-driven policy making

Professional associations, philanthropies, and think tanks have promoted greater
use of data for decision making. Local governments now confront more
information and evidence than ever before, with performance data, open data,
and big data. How is this related to the traditional practice of policy analysis or
to new forms of analysis at the local level?
135
Performance management has increased in local governments in the past few

decades (Ejersbo and Svara 2012). Contracting out and managed competition
have increased demands for measuring outputs and outcomes (Swindell and Kelly
2005). Calls for ‘reinventing government’ and more strategic management in the
1990s likewise advocated for measuring results at the local level (Ejersbo and Svara
2012; Osborne and Gaebler 1992). Performance measurement has been a focus
of the Urban Institute (Urban Institute n.d.a) and ICMA created their Center
for Performance Analytics (ICMA n.d.) to provide technical assistance for local
governments to upload and analyze standardized performance data. Advice on
performance management is also available through other professional associations,
such as the National League of Cities (NLC 2014) and the Government Finance
Officers Association (Government Finance Officers Association 2016).
While the aims of performance measures are often managerial, the data collected
may have implications for larger debates about programs and policies. As Radin
points out in her chapter, the need to implement programs prevents the drawing
of sharp boundaries between management and policy (see also Moynihan 2014).
Mayors recognize the need for this information too. The National League of
Cities conducted a content analysis of 100 mayoral ‘state of the city’ speeches
in 2016 and found that ‘[M]ayors are helping their cities see the value of using
technology and data to drive decisions and make their city governments more
efficient and effective’ (Langan et al. 2016, p 1).
In recent years, some cities have put their performance data online for citizens
as well, often in the form of dashboards or in open data portals. Open data refers
to governmental information collected by public agencies that is available to the
public without restrictions (Janssen et al. 2012; Veljković et al. 2014; Meijer
et al. 2014). Just as policy analysis initially spread at the local level due in large
part to federal initiatives, the open data movement accelerated after the Obama
administration’s 2009 Memorandum on Transparency and Open Government, which
required federal agencies to make diverse datasets available online (White House
2009).
Portland and San Francisco were local government pioneers in this area at the
end of 2009, but by November 2015, 49 of the 100 largest cities had some type
of open data portal (Tai and Mossberger 2015). Chicago, for example, has an
extensive open data portal with more than 1,000 datasets, including performance
metrics (City of Chicago, n.d.). The most accessed datasets include information
on city employees and salaries, crime, building permits, affordable housing
developments, and business licenses.1 The site allows users to search, filter data,
and visualize it with charts or maps.
Portals vary widely in terms of the information provided and usability (Robinson
et al. 2010; Tai and Mossberger 2015). The literature on open data has discussed
uses by citizens and software developers but one question that has not been raised
is whether the greater availability of data enhances policy analysis at the local level.
This is a new resource for local governments to do comparative analysis of data
and policy options across departments and across cities. Open data is also a source
136
of data for civic organizations, local universities, or others who want to conduct
useful analyses on local issues. Bloomberg Philanthropies (headed by the former
New York City mayor) has launched a national network called ‘What Works
Cities’ to ‘accelerate cities’ use of data and evidence’ to ‘improve services, inform
local decision-making and engage residents’ (Bloomberg Philanthropies 2016).
Assistance and best practices for open data are prominent topics in the network.
Some cities have turned to big data and predictive analytics to identify trends
and solutions for managing traffic, crime, neighborhood services, and more. Big
data has been defined as ‘large, diverse, complex, longitudinal and/or distributed
datasets generated from instruments, sensors, Internet transactions, email, video,
click streams, and/or all other digital sources available’ (White House 2014, pp
2–3). While big data use has become common in the private sector, utilization
in the public sector is more recent and limited (DeSouza 2014).
Contemporary policy analysis practices among local governments

Research focused on policy analysis at the local level is limited. Much of the
existing literature emerged in the 1970s as policy analysis was beginning to
spread through local governments (Schneider and Swinton 1979; Mushkin 1977;
Kraemer 1973; Dror 1976). In an effort to develop a better understanding of local
government usage of policy analysis, data, and partnerships today, ASU’s Center
for Urban Innovation joined with ICMA and the Alliance for Innovation over
the summer of 2016 to conduct two studies of local government administrators.
The larger survey targeted a sample of 603 local managers on a range of emerging
innovation issues (hereafter called the Innovation Survey). Municipal governments
accounted for 81.3% of respondents and county governments comprised 18.7%
of the Innovation Survey. Their jurisdictions ranged in population from 69 up
to 3.8 million. The smaller project interviewed 56 local administrators from
among the approximately 400 member jurisdictions of the Alliance, regarding
their use of policy analysis, data, and analytical partnerships (hereafter called the
Policy Survey).
The smaller Policy Survey had the advantage of addressing a range of analytical
tools and data, asking the 56 city managers: ‘… thinking about the last two years,
how often does your organization utilize information from any of the following
tools?’ The survey provided nine (9) options for respondents to score on a four-
point scale: never, rarely, sometimes, and often. Figure 7.1 illustrates the percentage
of the respondents that indicated ‘sometimes’ or ‘often.’ The most commonly
used tool city managers use from this list is performance measurement (94%). In
conversations with the respondents (reinforced in the literature), public sector
usage of performance measurement tends to be primarily for internal budgeting
and planning. At the other end of the scale, only 38% of the managers indicate
that they utilize ‘big data’ as a tool in their organizations. Given that these
respondents were sampled from the membership of the Alliance for Innovation
and are self-selecting into that organization as managers committed to cutting-edge
137
techniques, this low usage rate of big data is somewhat surprising. It suggests that
the usage among the wider array of jurisdictions (those without the innovation
proclivities) is likely even lower. Overall, however, the use of various forms of
data and analysis was common among Alliance for Innovation respondents, with
over 70% who reported that they used each of the tools other than big data at
least sometimes.
Figure 7.1: Frequency of policy analysis tool usage (% sometimes or often)
"Big Data"
Formal Program Analysis
Citizen Surveys
Formal Policy Analysis
Cost-Effectiveness Analysis
Economic Impact Analysis
Benchmarking
Cost-Benefit Analysis
Performance Measurement
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Source: Policy Survey, Alliance for Innovation members.
The results in Figure 7.1 illustrate usage of these tools rather than sources
for policy or data analysis. The Policy Survey of the city managers explored
external partners and sources as well as internal capacity. The survey first asked
respondents: ‘Thinking back over the past two years, have you utilized any of the
following resources to help inform you on a policy or program issue confronting
your organization?’ Respondents answered on a four-point scale: not at all, to a
very limited extent, to a fair extent, or to a large extent. The survey then asked
the same question (with the same answer set) but focusing on utilization relative
to organization or management decisions confronting their organizations. Figure 7.2
highlights the results on these two questions.
The results illustrate that Alliance managers rely most heavily on internal staff
for research on both policy and management decisions. Notably, professional
associations are the second most relied on resource for management decisions
and third for policy issues. While between one quarter and one third of the
respondents turned to universities as a resource on one or both of these areas,
15.4% say they never use universities as a resource on policy issues and 27.5%
138
say they never use universities on managerial decisions. These are almost the
identical numbers for managers who never use academic journals. This highlights
untapped opportunities for universities to help their neighboring communities
address real issues confronting them.
Figure 7.2: Resource utilization for decisions (% to a fair or large extent)
Think Tanks
Academic Journals
Local University
Private Consultants
Professional Assns
Other Govt Agencies
Internal Staff
0% 20% 40% 60% 80% 100%

Mgmt Decision Policy Issue
A critical issue is the usage of analysis and data in decision making. The Policy
Survey also asked city managers: ‘When you have used information from these
sources, to what extent did the information from that source typically influence
the outcome of a policy or management decision?’ Figure 7.3 illustrates the results
for those that use the various sources. Again, and unsurprisingly, information
provided by internal staff was not only the most common, but staff information
was also the most influential. Influence among the remaining sources follows
the same pattern as Figure 7.2 with private consultants, other governments, and
professional associations clustering together. Information from local universities
did better relative to usage, but think tanks and academic journals exhibited the
least influence on decision outcomes, perhaps because those sources were more
general and not tailored to the specific contextual situations facing individual
jurisdictions.
The Innovation Survey, with its broader sample, explored several aspects of
how local governments ‘learn’ and utilize information. For instance, almost
two thirds of the local administrators (63.7%) said they agree or strongly agree
that their organization regularly obtains information on successful new practices
from other local governments, indicating a robust network of policy and
administrative practice diffusion. This result also suggests there is significant room
139
for improvement in which the remaining one third of organizations can be better
integrated into these knowledge networks. A slightly smaller majority (57.7%)
reported agreement or strong agreement that their organization regularly shares
information on successful new practices with other local governments.
Figure 7.3: Influence on outcome of policy/management decision (% to a fair or large extent)
Academic Journals
Think Tanks
Local University
Professional Assn
Other Govts
Private Consultant
Internal Staff
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
The Innovation Survey focused on four emerging or prevalent policy areas in

which local governments are currently engaged: performance analytics, public
engagement, regulation of the sharing economy, and infrastructure financing.
Table 7.1 reports the jurisdictions with executives that utilize any of a range of
potential sources of research on these policy areas. The results highlighted several
interesting patterns. First, on the initial three policy areas, local government
executives rely most heavily on research provided by their professional associations,
but rely most heavily on external consultants and internal staff regarding
infrastructure financing. International examples provide the least input, which
is perhaps not surprising given the unique structure of local governance in the
United States. Local government executives tend not to rely significantly on
academic publications.
Use of performance data is nearly universal among respondents in the Policy
Survey of Alliance members and is a topic that the larger ICMA Innovation Survey
of Local Governments addressed in some detail. The Innovation Survey asked
respondents whether their jurisdiction collects performance data to help assess
the quality of service provision. Only about two in five (41.7%) indicated that
140
they do. The remaining 58.3% do not collect such data. Jurisdiction population
is related to the likelihood of collecting performance data. The mean population
of the communities that do collect the data was 119,421 and only 17,855 for
those that do not (t= 5.05; p<0.0001).
Table 7.1: Local government sources for research on new policies and practices (%)
Source Performance Public Regulation of Infrastructure
Data Analytics Engagement Sharing Economy Financing
Internal Staff Expertise 36.2 48.4 13.3 32.2
External Consultants 27.9 30.7 11.8 32.5
Other Local Governments 31.3 47.1 21.6 24.9
State Government 17.1 23.6 17.3 23.6
Federal Government 11.9 18.4 12.8 15.1
International Examples 8.6 13.3 5.8 1.7
Professional Associations 40.1 48.6 25.0 26.9
Academic Publications 24.7 27.7 11.3 9.0
News Media 9.1 39.0 17.1 5.6
Conferences and Webinars 39.1 45.8 20.9 26.5
Source: ICMA Innovation Survey of Local Governments.
For those jurisdictions reporting that they do collect performance data, the
survey asked respondents to rate the ‘extent to which performance data are used
for each of the following purposes in your organization.’ Respondents scored
each of the nine purposes on a four-point scale: not used, used very little, used
moderately, or used considerably. Figure 7.4 illustrates that there is not as much
variation among the extent of uses over these purposes. Confirming other studies
(see for instance Folz et al. 2009 and Hatry 2007), performance measures play
the most extensive role in helping justify and formulate budget requests (but see
Kavanagh 2013 for a counter view). Even the least extensive use (monitoring
contractors) has over half of respondents indicating moderate or extensive usage
for that purpose.
Examples of analysis and evidence: Alliance for Innovation and

ICMA case studies
Case studies shared by professional organizations demonstrate some uses of analysis
and data-driven practices across policy areas. A review of these case studies also
demonstrates that local governments often collect and share evidence with citizens,
and are experimenting with new forms of data.
We draw the examples discussed in this section from case studies of innovative
practices collected by the Alliance for Innovation and available to its members
141
(Alliance for Innovation n.d.). The Alliance for Innovation is part of a partnership
with Arizona State University and ICMA. This partnership brings together
researchers and practitioners to discuss, experiment with, and study new
approaches to local government management and the issues confronting local
governments. The cases available in the database and cited here include submissions
from member cities and from the ICMA Center for Performance Measurement,
which has collected case studies in a variety of policy and program areas (for
example, fleet maintenance, parks and recreation, information technology, and
so on).
Figure 7.4: Extent of performance data usage by purpose (% used moderately or extensively)
Monitoring Contractors
Motivating Personnel
Building Public Trust
Long-term Planning
ID Problem Areas
Informing Public
Service Improvements
Informing Elected Officials
Justifying Budget Requests
0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
Source: ICMA Innovation Survey of Local Governments.
Several examples of policy analysis appear in the Alliance for Innovation and
ICMA cases, and they are associated with major initiatives. Such analyses tend to
be undertaken to help with comprehensive or strategic planning, or for capital
investment decisions. However, there are examples where local governments want
to know about the cost-effectiveness of an innovation or alternatives to solve a
problem they have identified through data.
• Seattle, WA has developed comprehensive planning for solid waste recycling,

using economic modeling. The city maintains in-house expertise to track,
evaluate, and revise the comprehensive plan over time.
• Bernalillo County, NM uses policy analysis to prioritize its capital investment.
County employees conduct analysis, scan for best practices nationally and
internationally, and conduct public meetings to share the process for evaluating
142
priorities. They make public the project scores and explanations for these scores
online, and allow citizens to use the website to share ideas, see results, and track
projects. They use a consultant to help design criteria and measurements but
the project is supported by two full-time staff who conduct analyses.
• Bellevue, WA uses analytical techniques from program evaluation to promote
budgeting for outcomes. City analysts use logic models ‘to establish and explain
a chain of logic between resources, activities, and results in each outcome area’
(ICMA 2015, p 1).
• As part of a strategic planning process, the Maryland National Parks and
Planning Commission and Prince George’s County Parks and Recreation
Department employed a national consulting firm to conduct a level-of-
service analysis of its facilities and to create projections and decision matrices.
This analysis was accompanied by other data-gathering activities, including a
random sample mail survey in English and Spanish, focus groups, and public
and stakeholder meetings.
• Las Vegas, NV uses data and research on alternatives in other cities for its
street-sweeping operations. By investigating practices in other cities, Las
Vegas employees found that posting street-sweeping schedules and then
ticketing cars was more effective for generating revenues than for promoting
cleanliness. Subsequently, they found by tracking data in Las Vegas that it was
more effective to work around parked vehicles as long as sweeping occurred
frequently. Learning from data on customer complaints, they also improved
other operations using techniques such as concentrating activity in one
geographic area at a time. The use of GPS on the street sweepers also provides
accurate data administrators can use for monitoring productivity.
• Santa Monica, CA introduced the use of alternative fuels in its vehicle fleet
and found that it could save money as well as lower environmental impact. By
tracking performance and doing a cost-effectiveness analysis, the city discovered
they could recapture the average $5,000 cost for conversion with less spending
for maintenance (due to cleaner burning fuels) and the use of cheaper fuels
such as Compressed Natural Gas.
Interacting with citizens

Local governments often involve citizens in evidence-based decision making
through citizen surveys, public meetings, and sharing of data online through
dashboards, scorecards, and open data portals. Given that citizen participation
is most likely at the local level (Berry et al. 1993; Briggs 2008; Fung 2004), the
public role in the use of data is perhaps strongest at this level as well. The ICMA
case studies on performance data often mention the use of such data by community
councils, citizen advisory panels, neighborhood teams, and community groups.
Surveys are an important tool for local governments and are a source of data
for performance management, planning, and analysis.
143
• Carlsbad, CA collaborates with the Social and Behavioral Research Institute

at California State University-San Marcos, for example, to conduct their
annual citizen surveys and a staff survey every five years. The city’s State-of-
Effectiveness Report presents the findings. This example highlights how local
universities may be partners in such survey efforts.
• Kansas City, KS has revised its citizen satisfaction survey to align with
performance measures gathered on city services so that citizen perceptions
and priorities can be taken into account in decision making. Additionally, the
city took the data to public meetings and engaged the ‘hacker community’ to
make the data available online in an interactive way.
Laboratories for innovation

Although it is the states that have famously been called the ‘laboratories of
democracy’,2 local governments are far more numerous than states, and provide
diverse opportunities for learning due to variations in local autonomy across states
that facilitate experimentation over different rules configurations. Furthermore,
benchmarking and performance data collected across cities and counties provides
an opportunity for local governments to conduct comparative analyses and to
learn from each other. Several communities reported using the ICMA data for
this purpose.
• Bellevue, WA tracks jurisdictions with similar population size and compares

trends in performance levels over a four-year period. Bellevue also produces
a report on the city’s vital signs, which ‘highlights 16 key citizen satisfaction
and effectiveness measures that cross department lines and give the reader a
sense of the overall well-being of the city. A summary chart of these vital signs
compares the city’s actual performance during the fiscal year against the targets
set the year before’ (ICMA 2002, p 17).
• Austin, TX produces a Community Scorecard with measures for public safety,
youth, family, neighborhood vitality, sustainability, and affordability, both over
time and compared with other communities.
• Troutdale, OR and Carlsbad, CA have used the comparative data from
ICMA to benchmark their own cities and to contact jurisdictions with high
performance.
• The 11 largest municipalities in the Phoenix, AZ metropolitan area participate
in the Valley Benchmark Cities project. This group voluntarily comes together
to determine a range of performance benchmarks and collect the necessary
comparable data, which are brought together in an annual report aimed at
informing city councils and citizens. The exercise also highlights those high-
performing jurisdictions from which the other jurisdictions can learn and
improve (Stockwell and Swindell 2015).
144
Experimentation with big data is also part of the data landscape for local
governments.
• Las Vegas, NV uses big data to prioritize capital improvement projects and
promote renewable energy. The city collected data on solar performance,
production, building energy consumption, and cost from 40 facilities, and
analyzed the data to identify high-impact solar projects. The city has saved tens
of millions of dollars and now derives 20% of its power from solar sources after
installing solar equipment at 25 sites, including at the wastewater treatment
facility (ICMA 2016).
• The Orange County (CA) Sheriff’s Department is aggregating existing
employee performance data to look for trends and to use predictive analytics
to alert managers when officers develop potentially harmful behavior patterns.
The department integrated an index of risk indicators on commendations
and complaints, citizen claims, traffic collisions, use of force, Internal Affairs
investigations, and worker compensation claims. The department has also used
this big data to track trends in use of force or relationships between use of force
and other factors, such as employee injuries. The department has learned that
the results are often counterintuitive, challenging the department’s assumptions.
Conclusion: past, present, and future of analysis and evidence at

the local level
From the municipal reforms at the beginning of the 20th century onward, data
and analysis have been important instruments for local governance. Techniques
like policy analysis spread because of federal influence and the efforts of
professional associations, and such factors are at work now in the diffusion of new
developments such as open data, big data, and evidence-based decision making.
Today, local governments are in the forefront of experimenting with new forms
of data-driven governance.
As illustrated by the two surveys and the case studies discussed in this chapter,
local governments are undertaking policy analysis for a wide range of purposes:
as part of a larger planning process, comprehensive or strategic planning,
prioritization of capital projects, and identifying emerging issues for further
analysis. In-house professional staff often conduct formal analysis, but many of
these cities and counties reported working at times with partners from outside
the local government organization.
The ICMA Innovations Survey illustrates the variation in capacity for
undertaking and using policy analysis at the local level. Despite an environment
that is rich with many types of policy evidence, it is not at all clear that all local
governments base decisions on data and analysis. While use of performance data
is one of the most common types of evidence-based decision making at the
local level, the Innovation Survey indicated that only about two fifths of local
governments even collect such data. This variation appears tied to jurisdictional
145
size, which correlates with the adoption of many leading practices at the local
level (Trautman 2016). This suggests that it is more likely to be larger local
governments that generate their own analyses. Yet, there are other possible sources
of policy analysis for local governments, from consultants, think tanks, professional
associations, and local research or advocacy organizations.
Another partner that may be underutilized, however, is the local university.
While not all local governments are close to one, many can draw on neighboring
or state institutions. Many universities have faculty with a range of expertise who
would be excited by the prospect of working in and with local communities
(Gordon 2016; Morgan and Mulligan 2014). The challenge is overcoming the
barriers to such usage. Wang et al. (2013) conducted a panel discussion with a
group of local government managers and found some of these barriers to include:
limited access to academic journals, irrelevant topics, overreliance on quantitative
methods and advanced statistical analyses (including how the articles are written),
and the lack of interactions between practitioners and local government officials.
The Policy Survey asked city managers what they thought the main barriers are
that inhibit the usefulness of academic research in journals for local officials. This
quantitative survey reinforces much of Wang and her colleagues’ work. The most
significant barrier is relevancy (too much focus on policy and not enough on
management). They note that such research is too theoretical and disconnected
from their operational needs, and that academic writing is too dense with too
many words and too much technical jargon.
Policy analysis presents an opportunity to interact with local governments in
a more pragmatic way, especially if the analysis includes recommendations for
implementation. Clearly, the results from the surveys highlight the limited current
uses of academic research and university faculty as partners. But they also point
to many opportunities for greater connections. Such partnerships also provide
excellent learning opportunities for students, practitioners, and faculty willing
to innovate in new ways to link research to evidence-based decision making.
One clear need for the future is more extensive and more systematic research
on policy analysis at the local level. The surveys and case studies presented here
suggest further avenues for research, such as more extensive surveys and more
systematic comparison of trends and outcomes. The Alliance for Innovation survey
included a self-selected group of proactive city managers, and the larger and more
representative ICMA survey did not include as many types of data and analysis.
With the emergence of new analytical tools, and a movement for evidence-based
decision making in many quarters, the timing is right for renewed research on
data-driven practice in local governments.
A research agenda on local government use of policy analysis and data would
be uniquely positioned to examine the difference that transparency and citizen
participation make in the use of analysis and resulting decisions. As the case studies
and survey data illustrate, local governments have often consulted citizens for their
views and used performance measures in dashboards and reports, sharing them
with the public and with citizen councils, advisory groups, and neighborhood
146
organizations. The potential for participatory forms of policy analysis is strongest

at the local level (Fischer 2003). Will the diffusion of open data portals drive
local governments further in this direction? That may be optimistic, as the types,
relevance, quality, and usability of data vary. There is not much research yet on
how open data gets used for analysis or other purposes.
Still, research on the local use of analysis and data can also foster learning
from comparison. Predictive modeling associated with big data can improve
knowledge of general trends. Open data could foster internal comparisons across
city departments, as well as cross-city comparative analysis. Performance data
has allowed cities to compare their outcomes with others and to identify other
jurisdictions for ideas about policy or program alternatives. All of these types
of data could support more and better policy analysis. It could also increase
learning across jurisdictions and contribute to scholarly policy research through
comparison.
One of the more encouraging aspects of policy analysis at the local level is
that cities offer great hope as crucibles from which innovations and governance
improvements will emerge. Mayors and city managers are rising up to counter the
growing cynicism and anti-empirical attitudes of federal and state political figures.
This may best be personified in the recent efforts of the late Benjamin Barber
in helping establish the new Global Parliament of Mayors, which has brought
together mayors in the United States and around the world to address problems
that traditional nation states are failing to resolve (Global Parliament of Mayors,
n.d.). And, as Atlanta mayor Kasim Reed has noted, 330 U.S. mayors endorsed
the Paris Climate Accord in the weeks following President Trump withdrawing
the nation from that agreement, citing the clear scientific evidence of the threat
climate change presents to the nation and the world (Reed 2017).
City managers and mayors currently continue to value pragmatic evidence
to aid them in their decision making over the more politically charged rhetoric
increasingly common at the federal and state levels. But local governments are
not immune to this possibility, and future research on local government policy
analysis must take into account not only the development of technology and
technique, but also changes to the political environment in which data and
analyses are interpreted and utilized.
Notes
1 http://data.cityofchicago.org/, sites sorted by ‘most accessed’ and ‘this year’ as well as the list
of topics displayed at the bottom left of the home page.
2 Supreme Court Justice Louis Brandeis in New State Ice Co. v Liebmann.
References
Alliance For Innovation (n.d.) Case studies & Awards [Database of case studies],
http://transformgov.org/en/news/case_studies_and_awards.
Balla, S. J. (2001) ‘Interstate Professional Associations and the Diffusion of Policy
Innovations’, American Politics Research 29(3): 221–45.
147
Berry, J. M., Portney, K. E., and Thomson, K. (1993) The Rebirth of Urban
Democracy, Washington, D.C.: Brookings Institution.
Bloomberg Philanthropies (2016) About What Works Cities, https://
whatworkscities.bloomberg.org/about/.
Briggs, X. S. (2008) Democracy as Problem Solving: Civic Capacity in Communities
Across the Globe, Cambridge, MA: MIT Press.
Chieppo, C. (2013) ‘Can Phoenix Find a Way Out of Its Pension Mess?’ Governing,
www.governing.com/blogs/bfc/col-phoenix-police-firefighter-public-pension-
spiking.html.
City of Chicago (n.d.) Data Portal, https://data.cityofchicago.org.
DeSouza, K.C. (2014) Realizing the Promise of Big Data: Implementing Big Data
Projects, IBM Center for the Business of Government.
Dror, Y. (1976) ‘Policy Analysis for Local Government’, Local Government Studies
2(1): 33–46.
Ejersbo, N. and Svara, J. H. (2012) ‘Bureaucracy and Democracy in Local
Government’, in K. Mossberger, S.E. Clarke, and P. John (eds), The Oxford
Handbook of Urban Politics, New York: Oxford University Press, 152–78.
Eyestone, R. (1977) ‘Confusion, Diffusion and Innovation’, American Political
Science Review 71(2): 441–7.
Fischer, F. (2003) Reframing Public Policy: Discursive Politics and Deliberative Practices,
New York: Oxford University Press.
Folz, D., Abdelrazek, R., and Chung, Y. (2009) ‘The Adoption, Use, and
Impacts of Performance Measures in Medium-sized Cities,’ Public Performance
and Management Review 33(1): 63–87.
Frederickson, H.G., Johnson, G.A., and Wood, C.H. (2004) The Adapted City:
Institutional Dynamics and Structural Change, Armonk, NY: M.E. Sharpe.
Frug, G. (1980) ‘The City as a Legal Concept’, Harvard Law Review 93(6):
1059–154.
Frug, G. and Barron, D. (2008) City Bound: How States Stifle Urban Innovation,
Ithaca, NY: Cornell University Press.
Fung, A. (2004) Empowered Participation, Princeton, NJ: Princeton University
Press.
Gilroy, L., Constant, P., and Randazzo, A. (2016) ‘Arizona Voters Overwhelmingly
Approve Public Safety Pension Reform’, Reason Foundation, http://reason.
org/blog/show/arizona-pension-reform-safety.
Global Parliament of Mayors (n.d.) About Us, https://globalparliamentofmayors.
org/about-us/.
Gordon, E. (2016) ‘The Iron Cage and the Ivory Tower,’ Governing, www.
governing.com/cityaccelerator/blog/The-iron-cage-and-the-ivory-tower.html.
Government Finance Officers Association (2016) Government Finance Officers
Association, www.gfoa.org.
Hatry, H. (2007) Performance Measurement: Getting Results, Washington, DC:
Urban Institute Press.
148
Heberlig, E., Leland, S., and Swindell, D. (2017) American Cities and the Politics
of Party Conventions, Albany, NY: SUNY Press.
Hogue, C. (2013) Government Organization Summary Report: 2012. Government
Division Briefs, U.S. Department of Commerce, Economics and Statistics
Administration, U.S. Census Bureau. September 26, 2013, G12-CG-ORG,
www2.census.gov/govs/cog/g12_org.pdf.
International City/County Management Association (n.d.) Center for Performance
Analytics, http://icma.org/en/results/center_for_performance_measurement/
home.
International City/County Management Association (2002) What Works 2002:
Performance Data, ICMA Center for Performance Measurement and Analytics.
International City/County Management Association (2014) Form of Government
Statistics – Municipalities, https://icma.org/documents/form-government-
statistics-municipalities-2014
International City/County Management Association (2015) ‘Using Logic Models
to Help Deliver Strategic Outcomes in Bellevue, WA’, ICMA Center for
Performance Measurement and Analytics.
International City/County Management Association (2016) ‘Data-driven
Decision Making: Las Vegas Gets Results’, http://icma.org/en/icma/
newsroom/highlights/Article/106966/DataDriven_Decision_Making_Las_
Vegas_Gets_Results.
Janssen, M., Charalabidis, Y., and Zuiderwijk, A. (2012) ‘Benefits, Adoption
Barriers and Myths of Open Data and Open Government’, Information Systems
Management 29(4): 258–68.
Judd, D. and Swanstrom, T. (2015) City Politics (9th edn), New York: Routledge.
Kavanagh, S. (2013) Lessons from Performance Measurement Leaders: A Sample of
Larger Local Governments in North America, Chicago, IL: Government Finance
Officers Association.
Kraemer, K.L. (1973) Policy Analysis in Local Government: A Systems Approach to
Decision Making, Washington, D.C.: International City Management Association.
Krane, D., Platon, R., and Hill, M., Jr. (2001) Home Rule in America: A 50-state
Handbook, Washington, D.C.: Congressional Quarterly Books.
Kubler, D. and Pagano, M.A. (2012) ‘Urban Politics as Multilevel Analysis’, in
K. Mossberger, S.E. Clarke, and P. John (eds), The Oxford Handbook of Urban
Politics, New York: Oxford University Press, 114–32.
Langan, T., McFarland, C., and Rainwater, B. (2016) ‘State of the Cities 2016’,
Washington, DC: National League of Cities, www.nlc.org/find-city-solutions/
city-solutions-and-applied-research/state-of-the-cities-2016.
McDonald, B.D. (2010) ‘The Bureau of Municipal Research and the Development
of a Professional Public Service’, Administration & Society 42(7): 815–35.
149
Meijer, A., Hoog, J., Twist, M., Steen, M., and Scherpenisse, J. (2014)
‘Understanding the Dynamics of Open Data: From Sweeping Statements
to Complex Contextual Interactions’, in M. Gasco-Hernandez (ed), Open
Government: Opportunities and Challenges for Public Governance, New York:
Springer.
Morgan, J. and Mulligan T. (2014) ‘Economic Development’, in F. Bluestein
(ed) County and Municipal Government in North Carolina (2nd edn), Chapel Hill,
NC: UNC Press.
Mossberger, K. (2000) The Spread of Enterprise Zones and the Politics of Ideas,
Mossberger, K. and Stoker, G. (2001) ‘The Evolution of Urban Regime Theory:
The Challenge of Conceptualization’, Urban Affairs Review 36(6): 810–35.
Moynihan, D. P. (2014) ‘Between Policy and Management: Common Ground
and Shared Opportunities’ 2014 Kershaw Award Lecture, presented at the
annual meeting of the Association for Public Policy Analysis and Management,
Albuquerque, November 2014. https://www.lafollette.wisc.edu/images/
publications/facstaff/moynihan/kershaw.pdf
Mushkin, S.J. (1977) ‘Policy Analysis in State and Community’, Public
Administration Review 37(3): 245–53.
National Association of Counties (2015) Reports and Publications, www.naco.org/
resources/reports-and-publications.
National League of Cities (2014) Performance Management: A Guide for City Leaders,
Center for City Solutions and Applied Research, www.nlc.org/Documents/
Find%20City%20Solutions/City-Solutions-and-Applied-Research/CSAR%20
Performance%20Management%20Report%202014.pdf.
National League of Cities (2016) Find City Solutions, www.nlc.org/find-city-
solutions.
Osborne, D. and Gaebler, T. (1992) Reinventing Government: How the Entrepreneurial
Spirit Is Transforming the Public Sector, Reading, MA: Addison-Wesley.
Reed, K. (2017) 6th Annual Municipal Finance Conference (Keynote), www.
brookings.edu/events/6th-annual-municipal-finance-conference/#.
Reilly, T. (2016) Prop 124 – Understanding Arizona’s Propositions: Changes to the
Public Safety Personnel Retirement System (PSRS), Phoenix, Arizona: Morrison
Institute for Public Policy, Arizona State University, https://morrisoninstitute.
asu.edu/products/prop-124-%E2%80%93-changes-public-safety-personnel-
retirement-system-psrs.
Richardson, J. Jr., Zimmerman Gough, M., and Puentes, R. (2003) ‘Is Home
Rule the Answer? Clarifying the Influence of Dillon’s Rule on Growth
Management’, Washington, DC: Brookings Institution Center on Urban and
Metropolitan Policy, www.brookings.edu/research/is-home-rule-the-answer-
clarifying-the-influence-of-dillons-rule-on-growth-management/.
Robinson, D. G., Yu, H., and Felten, E. W. (2010) ‘Enabling Innovation for Civic
Engagement’, in D. Lathrop and L. Ruma (eds), Open Government: Collaboration,
Transparency, and Participation in Practice, Sebastopol, CA: O’Reilly.
150
Rosentraub, M. and Swindell, D. (2009) ‘Doing Better: Sports, Economic Impact

Analysis, and Schools of Public Policy and Administration’, Journal of Public
Affairs Education 15(2): 219–42.
Sharp, E.B. (2005) Morality Politics in American Cities, Lawrence, KS: University
Press of Kansas.
Shipan, C. and Volden, C. (2012) ‘Policy Diffusion: Seven Lessons for Scholars
and Practitioners’, Public Administration Review 72(6): 788–96.
Smart Cities Council [n.d.] http://smartcitiescouncil.com/.
Stivers, C. (1995) ‘Bureau Men, Settlement Women: Constructing a Usable Past
for Public Administration’, Public Administration Review 55: 522–9.
Stivers, C. (1997) ‘Settlement Women and Bureau Men: Rejoinder to Schacter’,
Public Administration Review 57: 370.
Stockwell, B. and Swindell, D. (2015) ‘Joining Forces: Arizona Cities Collaborate
to Improve Performance’, PM 97(11): 12–14.
Swindell, D. and Kelly, J. (2005) ‘Performance Measurement versus City Service
Satisfaction: Intra-city Variations in Quality?’ Social Science Quarterly 86(3):
704–23.
Tai, K. T. and Mossberger, K. (2015) ‘Comparative Analysis of Open Data
Portals in 100 Largest Cities of the U.S.’, presented at the Annual Meeting of
the Association for Public Policy Analysis and Management, Miami, November.
Trautman, R. (2016) ‘Small-town Policy Makers’, Public Administration Review
76(2): 221–4.
Urban Institute. (n.d.a) Performance Measurement and Management, www.urban.
org/policy-centers/cross-center-initiatives/performance-measurement-and-
management.
Urban Institute. (n.d.b) The Urban Institute Through the Years, www.urban.org/
about/our-history.
Veljković, N., Bogdanovic-Dinic, S., and Stoimenov, L. (2014) ‘Benchmarking
Open Government: An Open Data Perspective’, Government Information Quarterly
31(2): 278–90.
Wang, J., Bunch, B., and Stream, C. (2013) ‘Increasing the Usefulness of Academic
Scholarship for Local Government Practitioners’, State and Local Government
Review 45(3): 197–213.
Walker, J. L. (1969) ‘The Diffusion of Innovations among the American States’,
American Political Science Review 63(3): 880–99.
White House (2009) Transparency and Open Government: Memorandum for the Heads
of Executive Departments and Agencies, https://obamawhitehouse.archives.gov/
the-press-office/transparency-and-open-government.
White House, Executive Office of the President (2014) Big Data: Seizing
Opportunities, Preserving Values, https://obamawhitehouse.archives.gov/sites/
default/files/docs/big_data_privacy_report_may_1_2014.pdf
Wolman, H. (2012) ‘What Cities Do: How Much Does Urban Policy Matter?’ in
K. Mossberger, S.E. Clarke, and P. John, The Oxford Handbook of Urban Politics,
New York: Oxford University Press, 415–41.
151
Wollmann, H. and Thurmaier, K. (2012) ‘Reforming Local Government

Institutions and the New Public Management’, in K. Mossberger, S.E. Clarke,
and P. John (eds), The Oxford Handbook of Urban Politics, New York: Oxford
University Press, 179–209.
152
EIGHT
Committees and legislatures

Philip Joyce
Many people may not immediately associate Congress and policy analysis. The
well-documented failure of the Congress to enact policies to solve the nation’s
problems may lead observers to believe that there must not be much policy
analysis going on in the Congress. In point of fact, however, the Congress has
great analytical capacity. As a full-time legislature with a large staff, Congress has
a great deal of both ex ante and ex post policy analysis available. The problem is
not so much the availability of this information, but the rather episodic use that
the Congress makes of this information, or the tendency of individual members
or committees to make it fit some preconceived policy preferences.
Congressional policy analysis includes, to some extent, the analysis that is done
by staff of the House and Senate themselves—particularly by committee staff—
but the main congressional analytical capacity lies in its support agencies—the
Congressional Budget Office (CBO), the Government Accountability Office
(GAO), and the Congressional Research Service (CRS). After a brief look
at committee staff capacity, the chapter will focus on the work of the support
agencies and then conclude with a look forward, discussing issues related to the
future of policy analysis for the Congress.
Congressional committees
The Congress is organized by committee. These ‘little legislatures’ are the entities
within the Congress in which expertise on particular substantive issues resides
(Deering and Smith 1997). While committees certainly have analytical capacity,
it is a stretch to call even what most committees do ‘policy analysis.’ It is policy
analysis to the extent that it involves the assessment of policy alternatives and
that congressional committees, in performing their oversight role, hold hearings
where they receive input on the performance of programs from other analytical
organizations, including the support agencies discussed below. It tends not to have
some other characteristics that typify policy analysis, such as objectivity and depth.
It is the authorizing committees, rather than the appropriations committees,
who are normally the major players as they are tasked with understanding
programs in greater depth as a part of their need to periodically authorize federal
programs and agencies. Since ‘authorization’ (if done right) involves setting goals
for programs and assessing what has worked and what has not, it is an appropriate
locus of policy analysis. Authorizing committees are also tasked with periodically
reviewing federal programs and agencies, and policy analysis could be quite useful
153
in making adjustments to existing programs, creating new ones, or eliminating

those that are not working.
While committee work can be a useful policy input, recent trends have made
it less likely that the Congress will conduct policy analysis in committees. This
is particularly true because staffing levels for congressional committees have
declined substantially in recent years. Table 8.1 shows data on staffing levels
for the Congress as a whole, and also divided into committee staffing, personal
staffing, and staffing for support agencies. Support agency staffing will be discussed
below, but it is clear from the table that the trend has been away from staffing for
congressional committees in recent years. The most dramatic example of this is
in the House, where committee staffing, which had risen by 10% from 1979 to
1989, was cut almost in half between 1989 and 1999. This reduction, in fact,
occurred in one fell swoop when the House was taken over by the Republicans
after the 1994 midterm election. It has been maintained at that lower level since.
The Senate committee staff has also declined precipitously; in 1999 it was almost
40% lower than in 1979. In the case of the Senate, however, it was more gradual
and consistent. The large reduction in committee staffing levels stands in stark
contrast to personal staffing levels, which remained relatively constant over this
30-year period.
The nature of committee work, the existence of the support agencies listed
below, and the reduction in staff capacity means that very little policy analysis
is done by committee staffs. There are some notable exceptions, such as some
work by joint committees such as the Joint Committee on Taxation and the Joint
Economic Committee, but much committee work is focused on producing and
Table 8.1: Congressional staffing, 1979–2009

1979 1989 1999 2009
House Committee 2,027 2,267 1,267 1,324
House Personal 7,067 7,569 7,216 6,907
Senate Committee 1,410 1,116 910 913
Senate Personal 3,593 3,867 4,272 3,884
GAO 5,303 5,016 3,275 3,191

CRS 847 860 703 675
CBO 207 226 232 238
OTA 145 143 0 0
Other 6,169 5,881 5,729 4,230
Total 26,768 26,495 23,604 21,362

Source: Vital Statistics on Congress
154
explaining legislative proposals, rather than on analysis. Moreover, the increasing

partisan polarization of the Congress has carried with it a decreased demand
for analysis in favor of a taste for partisan ammunition. To the extent, therefore,
that committees ever had a desire to focus on policy analysis, this incentive has
declined in recent years.
Congressional oversight of Executive Branch agencies, of course, is also
conducted in committee. Whether one considers this oversight to fit under the
umbrella of ‘policy analysis’ depends on how broad a definition one is comfortable
with. Some oversight is conducted as a part of the authorization or reauthorization
process mentioned above. Other oversight is conducted specifically in an attempt
to embarrass the Executive Branch or to score political points. Decades ago,
McCubbins and Schwartz (1984) identified this latter kind of oversight as ‘fire
alarm’ oversight because it was intended to focus on some high-profile alleged
misdeeds as opposed to the more routine (and more difficult) ‘police patrol’
oversight, which attempts to determine how well agencies or policies are working.
This kind of ‘gotcha’ oversight does not fit very well with traditional notions of
policy analysis. All forms of oversight are less likely to occur when the Congress
and the Presidency are controlled by the same political party. Purely partisan
exercises like the Benghazi investigation have been less likely to occur during
the early part of the Trump administration than they were in the latter part of
the Obama administration.
One other issue is worth mentioning. Given the reduction in committee
staffing, and the lack of incentives suggested above, the analytical gap is often
being filled by lobbyists and other organized interests, most of whom have their
own incentives, which are largely to protect the status quo. (See the chapter
by Godwin, Godwin, and Ainsworth in this volume on the use of analysis by
industry associations.) In a system that is already rigged in favor of established
interests, this creates additional impediments to policy change. Drutman (2016)
vividly illustrates this phenomenon:
More and more, offices are staffed primarily by inexperienced (but

enthusiastic) twentysomethings, who are expected to be expert on
two dozen issues, as of yesterday. Of course, they can’t be that, because
nobody can. Instead, scrapped for time, they usually operate in reactive
mode, responding to information as it comes to them. Almost all
of that information is from interested parties. By an overwhelming
margin, that information comes from businesses and trade associations,
who are looking out for their own bottom lines. They usually
provide detailed justifications for their perspectives, often presented
by extremely smart and experienced lawyers and lobbyists, many of
whom are very well-connected (Drutman 2016, p 5).
155
The Congressional Budget Office
The CBO was established by the same law—the Congressional Budget and
Impoundment Control Act—that created the other institutions (the budget
committees) and rules of the new congressional budget process in 1974. The
law stated that CBO was to be headed by a Director and Deputy Director, who
would be appointed by the Speaker of the House and President Pro Tempore of
the Senate, on recommendation of the budget committees. All other CBO staff
were to serve at the pleasure of the Director.
From the standpoint of policy analysis, CBO, which is a relatively small agency
(about 200 at its founding, and roughly 250 by 2016) has three different kinds
of functions—helping the Congress to set overall fiscal policy, providing cost
estimates of proposed legislation, and conducting more in-depth analyses of
proposed policies (Joyce 2011).1
Setting overall fiscal policy
The annual budget resolution—a blueprint for congressional budget action— is

heavily influenced by the development of the CBO baseline—a projection of the
outlook for the federal budget over the following 10 years (Congressional Budget
Office 2016). This baseline gives the Congress a starting point for deliberations
on the budget. Within CBO, perhaps the most important input into the baseline
is the development of an economic forecast since much of revenue and spending
is heavily influenced by factors such as economic growth, personal income, and
unemployment. The baseline generally assumes a continuation of current law (for
revenues and mandatory spending, including entitlements) or current policy (in
the case of discretionary, or appropriated, spending). It is, in effect, what would
happen if the budget were put on automatic pilot.
Cost estimating
CBO’s Budget Analysis Division (BAD), in addition to helping to develop the

baseline, provides cost estimates of proposed legislation. Section 403 of the
Budget Act requires that each bill reported out (eligible to be considered on the
floor) by a congressional committee (excluding appropriation bills) to the floor
of the House or Senate include a CBO cost estimate. The preparation of a cost
estimate involves three separate steps beyond the construction of the baseline:
first, discovering that a bill has, in fact, been ordered reported—that is, that a
cost estimate will be necessary; second, determining what that bill actually does;
third, estimating how much the bill would cost (or save) relative to current law
(the baseline).
In addition to doing estimates of the effects of legislation on federal costs, CBO
also separately estimates the costs of legislation on state and local governments,
156
and the private sector. This requirement came about as a result of the passage of
the Unfunded Mandates Reform Act 1995.
In-depth analysis
CBO’s first director, Alice Rivlin, made a series of decisions in the first year that
became crucial to CBO’s ability to perform its functions as she had envisioned
them. These included selecting staff whom she believed could perform in an
objective, nonpartisan manner, and organizing the agency in a way that protected
the longer-term policy analysis work from being forced out by the short-term
budget work (Joyce 2011, pp 22–5). Perhaps the most important decision,
however, was that CBO would make no policy recommendations. There was
nothing in the statute that prevented such recommendations from being made
but Rivlin decided immediately that making recommendations would quickly
align CBO with one side or the other in partisan debates and would put it on
a slippery slope toward being viewed as partisan (Joyce 2011, p 29). (See the
chapter by Lynn in this volume for more about Rivlin.)
She also asserted, in early contacts with the budget committees, that CBO
worked for the whole Congress (not only the budget committees) and that CBO
needed to maintain its ability to set its own agenda. By protecting the agenda-
setting role, Rivlin enabled CBO to become—and remain—independent. One
very specific manifestation of this was a decision that CBO made that it would
not respond to requests from individual members unless they had been signed
off by chairs or ranking members of committees (Joyce 2011, p 27). She also
decided that she would make the agency’s reports and cost estimates public; this
enhanced the transparency and credibility of the products.
Rivlin’s dream of having analytical work as a significant part of CBO’s portfolio
got a substantial boost from a high-profile analysis done in 1977 by CBO about
President Jimmy Carter’s energy policy. This was, however, only one issue of
hundreds of potential policy initiatives affecting the federal budget. Director
Rivlin, in response to these potential targets for analytical work, began to
focus on organizing CBO to do policy analysis in a number of areas, including
macroeconomics; natural resource and commerce; tax policy; health and human
resource issues; and national security. To a far greater extent than in BAD,
these policy analysis divisions are staffed by PhD economists, whose job it is to
produce analyses to inform the future legislative agenda for the Congress. To
take health policy as a recent high-profile example, the studies issued by CBO
between 2000 and 2008 represented issues on which CBO had done significant
work in advance of the Obama health reform effort, including studies focused
on the costs of health care, the impact of various policies on coverage, and the
cost-effectiveness of various health strategies (Joyce 2011, p 181).
While the examples provided above (the Carter energy policy and health
reform) represent examples where CBO’s broader policy analysis work was quite
influential, there are many more cases where the influence of this in-depth policy
157
analysis work on the development of policy is less apparent. CBO can control
what it studies, the questions that it asks, and the timing of the release of its reports
but it cannot force the Congress to pay attention to the analyses. Often the way
that the results of these analyses are communicated to the Congress is through
the congressional hearing process, where the Director or CBO staff who have
worked on the report may be asked to testify.
CBO budget work is more consistently influential, largely because it is part of a
routine process (the budget process) that requires the information that it provides.
Since the budget committees need the CBO baseline in order to prepare the
budget resolution, and since congressional committees are required to include
cost estimates in committee reports, these analyses are much more likely to be
paid attention to. A particularly important recent example of this influence was
CBO’s estimate of the budgetary effect of what became the Affordable Care Act.
Since President Obama had pledged not to sign a bill that would ‘add one dime
to the deficit’ he elevated the budgetary considerations over others (White House
2009). When he did that, he (in the words of one long-time congressional staff
person) ‘just handed … CBO the keys’ to health care reform (Joyce 2011, p 209).
Still, it is a stretch to call these narrow budgetary analyses ‘policy analysis’ since
they are really focused on the narrow question of federal budgetary cost and do
not focus on the effects of policies much, if at all.
During the first six months of the Trump administration, CBO has had a very
high profile because of its responsibility to estimate the effects of the various
bills that were considered by the House and Senate to repeal and/or replace the
Affordable Care Act. The major focus has been on CBO’s estimates of how many
people would lose coverage over a 10-year period under Republican proposals to
gut President Obama’s signature policy accomplishment. Estimates have ranged
from 22 million to 32 million, depending on the specific proposal. These estimates
have been credited (or blamed, depending on your point of view) for killing (as
of this writing) Republican efforts to fulfill a major campaign promise. At the
same time, Trump administration officials have attacked the credibility and even
the existence of CBO; The Director of the Office of Management and Budget
(OMB), Mick Mulvaney, mused that ‘the time of CBO has probably come and
gone’ (Joyce 2017). This is probably wishful thinking, but all eight former directors
of CBO penned a letter to the leadership of the Congress strongly objecting to
the attacks on the organization. It is likely to survive this latest barrage, as it did
earlier attacks in the Reagan and Clinton administrations (to name two).
The majority of CBO’s professional employees have graduate degrees. In the
policy analysis divisions, most of the analysts have PhDs, mainly in economics.
More of the employees in the BAD have Master’s degrees; many of these are in
public policy or public administration. While some CBO employees are hired
directly out of school, others may have significant experience in government
or think tanks. The thing that they all have in common is an ability to perform
their work in a nonpartisan manner.
158
The Government Accountability Office
The GAO was established, as the General Accounting Office, as part of the 1921
legislation that also created the predecessor to OMB. GAO is a much larger agency
than CBO; in fiscal year 2015 it employed almost 3,000 full-time equivalent (FTE)
staff (United States Government Accountability Office (GAO) 2016b). Even
though it is much larger than CBO, Table 8.1 indicates that it has seen a substantial
reduction in staffing since the 1970s, when it had more than 5,000 employees.
Unlike CBO, GAO routinely does work for individual members of Congress in
addition to congressional committees. Much of its work focuses on reviewing
the implementation of programs and policies, rather than explicitly feeding into
policy development. GAO’s portfolio is also much broader than CBO’s, in that
it extends far beyond consideration of economic and budgetary issues.
According to the agency’s strategic plan:
GAO is an independent, nonpartisan professional services agency in

the legislative branch of the federal government. Commonly known
as the audit and investigative arm of the Congress or the ‘congressional
watchdog’, we examine how taxpayer dollars are spent and advise
lawmakers and agency heads on ways to make government work
better (GAO 2014).
GAO is headed by the Comptroller General of the United States, who is

nominated by the President and confirmed by the Senate, for a single 15-year
term. The Comptroller General, despite being nominated by the President, is
a congressional employee, subject to removal only by the Congress. Since its
founding in 1921, eight individuals have served as Comptroller General.
GAO’s work has evolved substantially over its history. It was no accident that
GAO was created as a part of the same law that created the President’s budget
office, as ‘Congress needed a counterweight in the form of an enhanced ability
to see how the money that it appropriated was being spent’ (Havens 1992, p
201). Initially, this was limited to a ‘preaudit’ role. GAO would review individual
requests for payment of ‘vouchers’, with a ‘highly formal, legalistic review of each
voucher, with … approval … being dependent on conformity with an elaborate
set of rules governing the use of public funds’ (Havens 1992, p 202). Especially
with the increase in the size and scope of government after the New Deal and
the Second World War, this was a tremendously time-consuming task, even for
GAO’s staff of more than 14,000 employees, and resulted in a backlog of more
than 35 million unaudited vouchers (Havens 1992).
As a result of this mismatch between responsibilities and resources, a series of
laws and actions in the mid 1940s had the result of substantially, and permanently,
changing the focus of GAO activity. The Legislative Reorganization Act of 1946
explicitly directed the GAO to ‘make an expenditure analysis of each agency
in the executive branch of the government’ and generally to focus on efficient
159
management of funds rather than simply on questions of legality (Mosher 1979).

This change in focus also resulted in a substantial reduction in GAO staffing; by
the mid 1950s, GAO staffing had been reduced to around 6,000 (Havens 1992).
Studies of the history of GAO give greatest credit to Comptroller General Elmer
Staats (whose term ran from 1966 to 1981) in terms of transforming the agency
to its current focus on program evaluation (Havens 1992, pp 204–5). Clearly
Staats had an interest in moving GAO in the direction of doing more program
evaluations, and the Congress itself made the mandate for program evaluation
more explicit in the 1970 Legislative Reorganization Act, which directed the
Comptroller General to ‘review and analyze the results of Government activities
and programs under existing law’ (Mosher 1979, pp 176–7). The change in GAO’s
name, which did not occur until 2004, was an official acknowledgment (albeit a
somewhat delayed one) that its focus was no longer primarily on accounting but
rather on accountability.
Much of GAO’s work is not policy analysis in the classic sense, but could be
more appropriately described as either program evaluation or ‘performance
auditing.’ Given GAO’s mandate, and the size of its staff, it is not surprising that its
activities touch virtually everything that the federal government does. According
to GAO, in fiscal year 2016, GAO received requests for work from 95% of the
standing committees of the Congress and ‘testified 119 times before 69 separate
committees or subcommittees that touched virtually all major federal agencies’
(GAO 2017, p G-1).
The breadth of GAO’s activities is reflected in its organization. GAO is organized
in a way that generally reflects the scope of activities of the federal government,
with approximately 70% of GAO staff at its Washington, DC headquarters, and
with employees serving on one of 14 ‘teams’ (GAO 2016a), focusing on areas such
as defense, acquisition management, health, natural resources, financial markets.
GAO staffing levels, having declined substantially in the 1940s and 1950s,
experienced another substantial cut during the 1990s, directly following the
Republican takeover of the Congress after the 1994 elections. Referring back to
Table 8.1, the number of GAO staff declined from more than 5,300 to fewer than
3,300 between 1979 and 1999; about 80% of that reduction came between 1993
and 1999. While the impetus for some of this reduction was political (GAO had
been criticized by some Republicans for policy advocacy) and thus was directly
related to its expanded program evaluation mission (Grasso and Sharkansky 2001),
it is also undoubtedly true that part of the reason for this large staff reduction
stemmed from the Republican pledge to decrease congressional staffing; given
the size of the GAO, it was an attractive target. More than 60% of GAO staff, as
of 2016, had either a Master’s degree or PhD. The average GAO employee was
44 years of age, and had spent over 15 years in federal service and over 13 years
working for the agency (GAO 2016b).
GAO tracks three performance measures in an attempt to demonstrate its past
effectiveness. First, it claims that its work saved the federal government more than
$63 billion in fiscal year 2016, after saving $75 billion in 2015. Second, it argues
160
that more than 1,200 separate program and operational benefits occurred as a
result of its recommendations. Third, it estimates that, annually between 2012
and 2016, between 73% and 79% of its recommendations have been implemented
(GAO 2017, p G-10).
Perhaps the highest-profile example of the scope of GAO’s work can be seen
in its ‘high risk’ list, which has been produced since 1990 (and is now produced
every other year at the beginning of a Congress), and as of 2017 includes 34
different items. Table 8.2 summarizes some of these items, listing first all of the
items that have been on the list for all of the 25 years that it has been produced,
and a selected set that have been added more recently. The table shows that there
are six items that have been on the list since the beginning; the original 1990 list
had 14 items, with the other eight having been resolved over the past 25 years
(GAO 1990). The other 28 have been added since, with three items being added
for the first time in the last report.
Table 8.2: Selected items from GAO high risk list

Item Year added
DOD Supply Chain Management 1990
DOD Weapon System Acquisition 1990
DOE Contract Management for the National Nuclear Security Administration and Office of Environmental 1990
Management
NASA Acquisition Management 1990
Enforcement of Tax Laws 1990
Medicare Program 1990
DOD Financial Management 1995
Ensuring the Security of Federal Information Systems and Cyber Critical Infrastructure and Protecting the 1997
Privacy of Personally Identifiable Information
Managing Federal Real Property 2003
Medicaid Program 2003
Restructuring the Postal Service to Achieve Sustainable Financial Viability 2009
Improving the Management of IT Acquisitions and Operations 2015
U.S. Government’s Environmental Liabilities 2017
Source: Government Accountability Office, Fiscal Year 2017 Performance Plan, p G-5.
Congressional Research Service

The CRS is organizationally part of the Library of Congress and is the oldest of
the congressional support agencies, having been created in 1914. In number of
staff it is between GAO and CBO, with 675 staff in 2009 (see Table 8.1). While
CRS has not experienced quite the magnitude of staff reductions as has GAO, it
161
also saw its staff and budget cut after the 1994 Republican congressional takeover;
CRS lost approximately 20% of its staff positions between 1989 and 1999.
After creating a capacity for members of Congress to do research by creating a
research library in the Capitol in the early 1800s, the Congress in 1914 created a
Legislative Reference Service (LRS), which was later codified in the Legislative
Reorganization Act 1946 (Robinson 1992). Later, the 1970 amendments to the
Legislative Reorganization Act changed the name of the LRS to the Congressional
Research Service and, as a part of ‘the congressional declaration of analytical
independence from the executive branch,’ explicitly invited the newly renamed
CRS to conduct policy analysis (Robinson 1992, p 184):
(d) It shall be the duty of the Congressional research service, without

partisan bias
• (1) Upon request, to advise and assist any committee of the Senate or the
House of Representatives or any joint committee of Congress in the
analysis, appraisal and evaluation of legislative proposals within that
committee’s jurisdiction, or of recommendations submitted to Congress
by the President or any executive agency, so as to assist the committee in:
a. Determining the advisability of enacting such proposals
b. Estimating the probable results of such proposals and alternatives thereto;
and
c. Evaluating alternative methods for accomplishing these results. …
This is practically a recipe for doing policy analysis. Doing this kind of work
also required CRS to hire staff at a somewhat more advanced and experienced
analytical level. The earlier tasks involved a relatively more junior individual
to answering relatively simple, discrete questions. Once the work expanded to
require more detailed knowledge of policy areas and more sophisticated analytical
expertise, this implied hiring a different kind of analyst.
At present, CRS activities are organized across the following five functional
areas, each of which is the responsibility of one of CRS’s Research Divisions
(descriptions taken from Congressional Research Service (CRS) 2016):
• American Law: This division’s work addresses all legal questions that arise
in a legislative context or are otherwise of interest to Congress. Some issues
relate to the institutional prerogatives of Congress under the Constitution.
Other questions involve constitutional and legal principles of statutory analysis
that cross legislative policy areas, such as federalism, commerce powers, and
individual rights.
• Domestic Social Policy: This division’s work includes analyses of domestic
policy and social program issues. These include education; labor and worker
safety; health care insurance and financing; health services and research; aging
policy studies; social security, pensions and disability insurance; immigration,
162
homeland security, domestic intelligence and criminal justice; and welfare,

nutrition, and housing programs.
• Foreign Affairs, Defense and Trade: The work of this division emphasizes
analyses of security, political, and economic developments at a global level as
well as in every region of the world. Experts examine political, economic, and
security relations between the United States and other nations; development and
foreign assistance; human rights; international organizations; and transnational
issues such as terrorism, refugees, crime, human rights, and nonproliferation.
• Government and Finance: The division studies the organization, structure,
operations and management of Congress and its support agencies, the executive
and the judicial branches; the congressional budget and appropriations process,
the legislative process and congressional history; and issues related to homeland
security and emergency management, community development, elections,
and American federalism.
• Resources, Science and Industry: The division focuses on issues and policy
developments concerning the nation’s natural resources and environmental
management, scientific advances and technology applications, and industry
and infrastructure.
CRS performs a wider variety of services than the other two support agencies
and does a great deal of work for individual members of Congress. Just examining
the portfolio of CRS products in fiscal year 2015 alone gives one a feel for its
range of activities. In that fiscal year, CRS (CRS 2016):
• provided 5,296 in-person briefings, consultations, and testimonies;

• wrote 3,166 confidential memoranda;
• responded to more than 50,000 telephone and email inquiries;
• published 8,210 bill summaries; and
• wrote 1,264 new CRS reports and updated almost 2,500 other reports and
products.
In its annual report to the Congress, CRS also summarized the main policy
areas that it had assisted the Congress with in 2015. As would be expected,
CRS assistance to the Congress in any one year matches the issues that are on
the congressional policy agenda. This means that annually CRS is quite active in
the budget and appropriations arena; in fact, the most popular CRS product is its
summary of the status of appropriations. CRS also was active in 2015 in advising
the Congress concerning issues related to the federal budget deficit, the debt limit,
comprehensive energy legislation, defense reform, revisions to the Elementary
and Secondary Education Act, and health legislation. In addition CRS assists
congressional committees more generally in performing their oversight function.
In a sense CRS is not as well known as the other two support agencies by
design. While some CRS reports are publicly available, CRS differs from CBO
and GAO in that the majority of CRS products are confidential, and are not
163
shared with anyone besides the requestor. If, for example, a member of Congress
requested a question related to the recognition of same-sex marriages in the states,
as happened frequently in 2015 in the wake of the Supreme Court decision in
Obergefell v Hodges (CRS 2016), the response of the CRS analyst to the request
by a member or committee would not be released by CRS, and would not
become public unless the requester released it themself. For this reason, much less
information is available on the substance of CRS work than for CBO or GAO.
Key analytical and institutional issues surrounding policy analysis

in the Congress
The Congress, then, has created substantial analytical capacity. Three issues stand
out when considering the future of policy analysis in the Congress: the effect of
analytical ability of the Congress on the balance of power between the legislative
and executive branches; analytical factors that may compromise the production
of high-quality analyses; and the challenges of producing such information in
a political environment that grows seemingly more partisan by the year. This
chapter will conclude with a discussion of each of these three issues.
Congressional policy analysis and the balance of powers
The statement above that the 1970 Reorganization Act represented a statement
of congressional analytical independence has been borne out in fact, through
the work of these three support agencies. This is particularly true because the
description of the U.S. policy-making process as one where ‘the President proposes
and the Congress disposes’ tends to be accurate much of the time.
This has clear implications for the support agencies. Many of the questions that
members of Congress ask CRS analysts are focused on assisting the Congress in
responding to presidential initiatives. CBO has also provided analytical support
for congressional policy clashes with presidents. There are at least two dimensions
to this. First, CBO has promoted ‘honest numbers’ by serving as a check on
the tendency by the Executive Branch to mold the numbers to fit presidential
preferences or assumptions about the positive effects of administration policies
(Williams 1998). It has created the capacity to do nonpartisan cost estimates
and has generally supplanted OMB (at least for the press) as the place to go for
credible budget numbers. Second, CBO analyses have enabled the Congress
to challenge policy proposals that come out of the Executive Branch, such as
the aforementioned Carter energy policy and the twin health care initiatives of
President Clinton in 1993 and 1994, and President Obama in 2009 and 2010.
GAO, because much of its work comes on the back end of the policy process,
is more likely to assist the Congress in exercising oversight of the Executive
Branch’s implementation of policies, rather than assisting the Congress directly
in responding to proposed policies. Every time the GAO testifies before
Congress concerning a report that it has prepared, or makes recommendations
164
for management improvements in federal programs or agencies, it is empowering

the Congress.
This analytical independence may be under threat, however. As Table 8.1 and
the discussion above have documented, the Congress has systematically reduced
the staffing of two of the three support agencies since the early 1990s, with only
CBO escaping reductions in staff levels. GAO and CRS, which together had
almost 6,000 staff as recently as 1989, have fewer than 4,000 today. Add to that
the elimination of a fourth support agency—the Office of Technology Assessment
(OTA) in 1995, and the reduction in committee staff documented above—and
there is a risk of the Congress engaging in analytical disarmament. In addition
if the Republican majority in Congress responded to criticisms of CBO from
the Trump administration by weakening the agency it would invariably result
in weakening the Congress—any Congress—relative to the Executive Branch.
Potential compromises to analytical quality
If the analytical products produced by support agencies are to have a positive

impact on the conduct of public policy, it is important that they are not only
produced, but that they are credible, timely, and thorough. Two notable issues
will illustrate this point. First, the need to produce information in response to
the congressional calendar may lead to shortcuts that compromise the quality of
information. Second, the questions that are addressed are often overly narrow,
and this may lead to misleading conclusions.
(1) Compromising accuracy and quality for timeliness and thoroughness
Academics doing policy analysis can consider evidence carefully, without regard
to how long it might take to produce a given study or report. The pressures to
produce analyses prior to the consideration of legislation (in the case of CBO
and CRS), to satisfy a desire to know more, or to meet an oversight schedule (in
the case of GAO) can lead to a tension between producing a ‘great’ answer (that
is, the definitive research based on unassailable evidence) when it is too late for
the information to enter the policy debate and a ‘pretty good’ answer (research
that is the best that can be produced in the time available, or given analytical
constraints) that is produced in a timely manner.
Consider CBO’s two main calendar-driven products—the budget baseline (a 10-
year projection that is produced three times a year) and cost estimates of proposed
legislation. In the former case, the act of doing multi-year budget forecasting is an
error-filled endeavor and those errors become magnified the further out into the
future that the projections extend. One of the best examples of this was in 2001,
when projections by CBO and OMB indicated that the federal government might
be experiencing cumulative surpluses of $5.6 trillion over 10 years. This paved
the way for the (in retrospect) unaffordable Bush tax cuts (Joyce 2011, pp 79–80).
CBO, in particular, has been transparent about the uncertainty surrounding its
165
estimates and has encouraged the Congress to recognize this uncertainty before
responding to any numbers. This also extends to cost estimates, where the time
horizon has been extending and now extends, in most cases, to 10 years, and
where cost estimates must be produced under tight deadlines in order for the
estimate to be available when the Congress considers the legislation.
Similarly, in order to be useful, GAO reports must respond to the congressional
calendar. Indeed, an apparently valid criticism of GAO in the mid 1990s was
that GAO reports took too long to be completed; in 1993 and 1994 the on-
time completion rates were 40% and 43%, respectively. These delays apparently
resulted at least in part from ‘an extensive internal review process that was part
of the agency’s culture … GAO was very concerned to get the details correct,
which required multiple internal reviews before release’ (Rubin 2003, p 126).
GAO’s response to this was to re-engineer its project completion process, in
particular to emphasize shorter jobs over longer ones. In 1994 the average GAO
job took 10.4 months; that had dropped to 5.4 months by 1997 (Rubin 2003, p
129). The improvements in timeliness of products that began during the 1990s
has been maintained since; in 2015, 98% of reports were completed on time.
(2) Analyses may be overly narrow
One of the criticisms that one often hears of analysis in the Congress is that it
rarely focuses comprehensively on the effects of policies. If the CBO considers
costs to the federal government, it does not focus on the costs to others, or
the benefits of a policy. If the GAO focuses on benefits, conversely, it does not
emphasize the cost of producing those benefits.
Two issues can illustrate this. The first is dynamic scoring, which has been
debated in budget circles for more than 30 years. The notion of dynamic scoring is
that the macroeconomic feedback effects of policies should be taken into account
when estimates are made. At the level of individual policies, this means that if
economic behavior changes as a result of a policy (for example, if an increase
in taxes leads to less of the taxed behavior) that behavioral change would be
accounted for. In a broader sense, however, the dynamic scoring debate has to do
with estimating the positive and negative effects of policies on macroeconomic
outcomes. In an effort to build these macroeconomic effects into at least the
most significant cost estimates, the House of Representatives in January of 2015
adopted a rule that requires both CBO and the Joint Committee on Taxation
to provide estimates of the macroeconomic effects of selected major legislation
(United States House of Representatives, Committee on the Budget 2015).
The second issue concerning broader effects has to do with the criticism that
analyses do not connect costs and benefits. In the case of CBO, this translates
into a charge that cost estimating places too much weight on costs to the federal
government and not enough on the benefits of legislation. The work of the
CBO policy divisions represents an exception to this but, by and large, CBO cost
estimates tend to ignore the benefit side of the equation. There is no question,
166
to cite the key example, that the Affordable Care Act would not look the way
it does now if there had been less focus on the federal cost. If successful, many
health reforms may provide many benefits in the form of improved health care.
People may be healthier, and live longer, than they would have without the benefit
provided. However, a health benefit to an individual, or even a lot of people,
is not at all the same thing as a reduction in costs to the federal government. In
fact, when people live longer they may spend more time receiving government-
funded health care (Medicare, Medicaid, or veterans’ health, for example), thus
increasing the federal government’s cost. In recognition of this fact, Westmoreland
(2007) advocated that the Congress evaluate the effects of health legislation on
the benefit side (at least in addition to costs, if not instead of), using a measure
such as Quality-Adjusted Life Years (QALYs).
GAO analyses, conversely, tend to focus on whether programs are working
or not, and only to a lesser extent on their costs. Thus, there are many useful
issues related to program design and implementation that may be raised in GAO
reports but GAO is not generally in the business of either estimating costs (in
part because this would be seen as straying into CBO’s turf) or of systematically
relating costs to policy benefits. Since the Congress has both CBO and GAO at
its disposal, there is nothing to stop the Congress from taking analyses of both
agencies into account and asking questions about costs and benefits, particularly
when considering changes to existing programs.
Policy analysis in a partisan political environment
It is notable that each of these three support agencies embraces a culture of

nonpartisanship. Given that the Congress, as a legislative body, is an institution
filled with partisans, it is perhaps a surprising development that the Congress
would embrace nonpartisan analysis at all. This does not mean, however, that the
Congress is an easy environment in which to do objective analysis in practice.
Grasso and Sharkansky (2001), in an article discussing the ability of independent
auditing to flourish in a partisan political environment, argue (using GAO and
Israel’s State Comptroller as case studies) that ‘a purist view of audit independence
is untenable in a situation in which the consumers as well as the producers of audit
reports seek to inquire into activities that are inherently sensitive and political’
(p 1). They argue that the well-documented move away from financial auditing to
program evaluation in GAO put the agency in a more vulnerable political position.
Both CRS and CBO intentionally chose to avoid making policy
recommendations in their reports and public statements to Congress, in part as
a strategy of self-protection. GAO chose a different strategy; if GAO did not
make recommendations, however, it seems likely that it would be a much less
effective organization. Perhaps, for example, many of the documented savings
and program changes cited by the agency would not have occurred.
The question of analytical objectivity raises other interesting questions, in
particular about possible tensions between objectivity and neutrality. The latter
167
could become an excuse for not saying much of anything in an attempt to present
both sides of a given story. This is all well and good when there are credible
cases to be made on both sides of an issue, but if the weight of the evidence is
overwhelmingly on one side, objectivity would require saying so; ‘neutrality’
might lead to giving too much weight to analytically discredited arguments.
CBO’s unofficial motto is ‘on the one hand, on the other’ as an acknowledgment
of its mandate to look at both sides of an argument. CBO, however, got itself
in trouble in 2013 when its immigration cost estimate said that a less restrictive
immigration policy would have positive economic and budgetary effects. CBO
did not say ‘some people think the bill will cost money and others think it will
save money’; rather, its estimate reflected its conclusions based on the evidence.
Louis Fisher (2011) documents the debate between objectivity and neutrality
in great detail, particularly with respect to what he views as a shift in the CRS
norms away from a standard of objectivity to one of neutrality. Fisher had a front
row seat from which to observe this shift as someone who worked for CRS for
more than three decades. Fisher himself was disciplined by CRS leadership for
failing to adhere to the ‘neutrality’ standard. This is a fascinating case study that
makes for gripping reading for those who are interested. For now, I will simply
say that there is a difference between objectivity and neutrality, and that it is
worth thinking about whether the support agencies are the most supportive to the
Congress when objectively reporting results, rather than trying to consistently
trying to occupy some analytical middle ground. It is, of course, ultimately up
to the Congress to decide whether it simply wants a summary of pros and cons,
or expert analysis and conclusions.
We cannot leave a discussion of the political environment for analysis without
noting the elephant (and donkey as well) in the room. That is—policy analysis in
the Congress is now being conducted in a much more partisan environment and
this makes it much more difficult for the producers of these analyses, and for the
public which may seek to learn something from them. Members of Congress, and
congressional committees, are often not looking for analysis, but ammunition.
That is, they already know the answer and they are looking for ‘evidence’ to
back it up. Thus, the demand for high-quality analysis is in decline. To the extent
that this weakens the support agencies, it will ultimately weaken the Congress.
Conclusion
Lienert (2005) in a study of 28 mostly industrialized countries focused on the
distribution of budgetary power between the executive and the legislature, using a
number of measures to attempt to determine the level of independent budgetary
power possessed by the legislature in each country. Among these are the ability
to amend the budget, the amount of time available for budgeting, the ability
to restrict activities in execution and (most relevant for the current study) the
‘technical support (available to) the legislature.’ Lienert’s total index ranks the
United States Congress as the legislature with the most independent budgetary
168
power out of all of the countries studied, and (with Mexico and Korea) as the
legislature with the greatest level of technical support. Since the focus was on
budgetary power, it appears that Lienert was most influenced by CBO, but as
this chapter has demonstrated, GAO and CRS each are substantial sources of
analytical support. The key question going forward seems to be not whether the
Congress has the analytical capacity, but whether it makes good use of it. Policy
analysis in the Congress is unlikely to have much benefit in the absence of the
desire of the Congress to engage in evidence-based decision making. Given the
extent of political polarization that currently exists, the continued demand for
such evidence seems very much in doubt.
Notes
1 These functions are discussed in Joyce (2011) in detail in chapter 3 (pp 53–92), chapter 4 (pp
93–121) and chapter 5 (pp 122–53).
References
Congressional Budget Office (2016) The Budget and Economic Outlook: Fiscal Years
2015–2026, Washington, DC: U.S. Government Printing Office.
Congressional Research Service (2016) Annual Report: Fiscal Year 2015.
Deering, C. and Smith, S. (1997) Committees in Congress, Washington, DC: CQ
Press.
Drutman, L. (2016) ‘A Congress Without Its Own Knowledge Is a Depending
Congress’, in K. Kosar, et. al (eds), Restoring Congress as the First Branch, R Street
Policy Brief #50.
Fisher, L. (2011) Defending Congress and the Constitution, Lawrence, KS: University
Press of Kansas.
Grasso, P. G. and Sharkansky, I. (2001) ‘The Auditing of Public Policy and the
Politics of Auditing: the U.S. GAO and Israel’s State Comptroller’, Governance
14(1): 1–21.
Havens, H. (1992) ‘The Evolution of the General Accounting Office: From
Voucher Audits to Program Evaluations’, in C. Weiss (ed), Organizations for
Policy Analysis, Newbury Park, CA: Sage Publications.
Joyce, P. (2011) The Congressional Budget Office: Honest Numbers, Power and
Policymaking, Washington, DC: Georgetown University Press.
Joyce, P. (2017). ‘The Trump Administration Detests the Congressional Budget
Office. Here’s Why It’s Important’, The Monkey Cage (Washington Post), June
14, www.washingtonpost.com/news/monkey-cage/wp/2017/06/13/the-
trump-administration-detests-the-congressional-budget-office-heres-why-its-
important/?utm_term=.439966d4c6cd.
Lienert, I. (2005) ‘Who Controls the Budget: The Legislature or the Executive?’
IMF Working Paper.
McCubbins, M. and Schwartz, T. (1984) ‘Congressional Oversight Overlooked:
Police Patrols versus Fire Alarms’, American Journal of Political Science 2(1): 165–79.
169
Mosher, F. (1979) The GAO: The Quest for Accountability in American Government,
Boulder, CO: Westview Press.
Robinson, W. (1992) ‘The Congressional Research Service: Policy Consultant,
Think Tank, and Information Factory’, in C. Weiss (ed), Organizations for Policy
Analysis, Newbury Park, CA: Sage Publications.
Rubin, I. (2003) Balancing the Federal Budget: Trimming the Herds or Eating the Seed
Corn? New York: Chatham House Publishers.
United States Government Accountability Office (1990) ‘Letter to Congressional
Committees Identifying GAO’s Original High Risk Areas’, January 23.
United States Government Accountability Office (2014) Strategic Plan: 2014–
2019, Washington, DC: U.S. Government Printing Office.
United States Government Accountability Office (2016a) ‘Our Teams’, www.
gao.gov/about/workforce/teams.html.
United States Government Accountability Office (2016b), email correspondence
from Janice Morrison providing statistics on GAO employees, November 9.
United States Government Accountability Office (2017), GAO Performance Report:
Fiscal Year 2018, Washington, DC: U.S. Government Printing Office.
United States House of Representatives, Committee on the Budget (2015) ‘House
Rules and Macroeconomic Scoring’, http://budget.house.gov/uploadedfiles/
house-rules-and-macroeconomic-scoring.pdf.
United States Senate, Committee on the Budget (1976) Congressional Budget
Reform: Public Law 93-344, Enacted July 12, 1974, 94th Congress, 2nd sess.,
August.
Westmoreland, T. (2007) ‘Standard Errors: How Budget Rules Distort
Lawmaking’, Georgetown Law Review, 95: 1555–1610.
White House, Office of the Press Secretary (2009) ‘Remarks by the President to
a Joint Session of Congress on Health Care’, September 9.
Williams, W. (1998) Honest Number and Democracy: Social Policy Analysis in the
White House, Congress, and in Federal Agencies, Washington, DC: Georgetown
University Press.
170
Part Three
Policy analysis outside of
government
171
NINE
Policy advisory committees:

an operational view
Michael Holland and Julia Lane
Introduction
Advisory committees are part and parcel of policy making for nearly 50 federal
agencies ranging from the vast Department of Defense to the tiny Administrative
Council of the United States. Advisory committee input is used at many levels
ranging from the development of high-level strategic initiatives, as vehicles for
research community input, to program prioritization efforts, as instruments of
program evaluation, all the way down to the provision of input on very narrowly
defined decisions, such as the peer review of individual research proposals.
Regulatory agencies, particularly those with responsibility for human health
and the environment, rely heavily on advisory committees for the production of
scientific assessments that inform the setting of standards (Jasanoff 2009).
In practice, there are many reasons that advisory committees are used by the
federal bureaucracy. Fleisher provides a conceptual framework which situates
federal advisory committees in the broader policy context whereby government
agencies use input from the advisory committees to mediate the pressures they
feel from other government policy makers, such as Congress or the Executive
Office, and advocacy groups as they run their policy formulation, implementation,
and policy revision processes (Fleisher 2015). Moffitt provides an excellent and
sweeping overview, arguing that advisory committees are used to manage agency
reputation, avoid ‘embarrassing situations,’ to compensate for deficiencies in
agency knowledge, and to redress agency weakness that derives from instability
or tumult inside the agency (Moffitt 2010).
In the United States the process whereby advice is sought from advisory
committees is strictly regulated by the Federal Advisory Committee Act (FACA) of
1972 (Bybee 1994; Ginsberg 2009), with oversight of these committees provided
by the General Services Administration’s (GSA) Committee Management
Secretariat. The GSA characterizes advisory committees according to the
following uses: (i) national policy issues, (ii) non-scientific programs, (iii) scientific
technical programs, (iv) grant recommendations, (v) regulatory negotiations, (vi)
special emphasis areas, and (vii) other. The Federal Advisory Committee Act
Amendments of 1997 (P.L. 105-153) amended FACA to govern agency use of
173
reports provided by the National Academy of Sciences and the National Academy
of Public Administration.
Advisory committees are numerous and engage a large number of people in
providing advice to government. In 2016 there were approximately 1,000 active
FACA committees used by about 50 federal agencies, costing $339.6 million
(General Services Administration 2016). The cost per committee averages about
$400,000 (Ginsberg 2015). More than 57,000 separate individuals served on
these committees, with nearly 30% of those members serving on more than one
committee, for a total roster of more than 73,000 advisory committee members
reported by the GSA.1 National Institutes of Health (NIH) peer review and
special emphasis panels accounts for the overwhelming majority of those multiple
appointments. In addition to the FACA committees established by the federal
agencies, another 6,000 individuals serve as volunteers annually on National
Research Council (NRC) study panels established under the auspices of the
National Academies of Science. Most of those NRC study panels provide advice
to the federal government. Much smaller then is the National Academy of Public
Administration, which produces an average of 10 reports per year.
This chapter focuses on the use of advisory committees by science agencies since
scientific technical programs, grant review, and special emphasis panels account
for more than 75% of FACA committee meetings reported by the GSA (Ginsberg
2015). Science policy, somewhat counterintuitively, is one of the least data-driven
of any policy arena (Marburger 2005), with a strong case to be made that science
policy—in the United States at least—is little more than science budget policy
(Sarewitz 2007). In fact, post-World War II U.S. science policy is designed to
rely on expert opinion (Brooks 1996). Thus, advisory committees in this science
agency context provide a useful counterpoint to the policy analysis described
elsewhere in this volume. The focus of the chapter is on the United States.
The operation of advisory committees is overwhelmingly influenced by the
norms of the research community and the research agencies rather than by any
formal principles of policy analysis or program evaluation. While the process
of establishing committees is highly formalized under FACA, the operation of
advisory committees—at least in the science agency context—is quite ad hoc.
With the exception of the small number of advisory committees established in
statute at the direction of Congress, agencies are given substantial flexibility in
the membership, roles, and tasks they assign to their advisory committees.
Only a small minority of those tens of thousands of advisory committee
members have any formal training in policy analysis—or in any formal process
of scientific assessment. As Susan Cozzens notes in her discussion of science and
technology policy professionals, individuals who have established a degree of
prominence at the forefront of research in a field of science or engineering embark
on their ‘committee careers,’ sitting on peer review panels, advisory boards, and
National Academy panels (Cozzens 2009). The apogee of this path is membership
on the National Science Board or the President’s Council of Advisors on Science
and Technology (PCAST). Cozzens notes that while these committee members
174
Policy advisory committees
may possess great experience, great wisdom, and great vision, they are ‘amateurs’
in the best sense of the word. They move between participating in the policy
process and their day jobs.
Among those engaged in science policy full time, Cozzens notes most enter
the field with a background in a scientific or technical field rather than with any
formal training in policy analysis, science policy, or technology policy. Efforts
to bolster the formal analytic rigor of science or technology policy have had
limited enduring institutional impact. In 1995 Congress eliminated its Office
of Technology Assessment (OTA), which pioneered technology assessment
methodology. Although OTA produced reports that were highly regarded by
the research community, Congress claimed the reports were almost always too
late to inform its deliberations. In the Executive Office of the President, the
Science & Technology Policy Institute (STPI) was established in the belief that
the President’s Science Advisor and the Office of Science & Technology Policy
were in need of greater analytic support. In the fast-paced context of the White
House, STPI has proven not to be as useful as hoped. They are as slow moving
as the advisory committees, which are available to the Office of Science and
Technology Policy (OSTP) through the agencies or the National Academies, and
STPI analytical products do not come with the reputational stamp of approval of
the advisory committee reports.
Background
Policy making is complex and in areas at the forefront of science and technology,
answers are rarely clear cut—even to those most expert in the field. As a result,
science agencies often turn to external sources for input and advice, advice that
is more frequently reliant on expert opinion than formal analysis. FACA was
established to ensure that the process was open and transparent; that advisory
bodies were accessible to the public. What this means in practice is not particularly
clearly defined, but the Sunshine Foundation identifies a set of core principles:
‘completeness, primacy, timeliness, ease of physical and electronic access, machine
readability, non-discrimination, use of commonly owned standards, licensing,
permanence and usage costs’ (Ginsberg et al. 2012).
Advisory committees can also be created by Congress. In this case, it is argued
that the committees’ structures are created to ensure that the bureaucracy is
responsive to a broad array of institutional interests (Balla and Wright 2001).
Advisory committees for program prioritization and evaluation

Research agency leadership can skillfully use advisory committees to coalesce
community input, identify programmatic priorities, build political support, and
apply managerial pressure to the powerful research institutions they fund. For
example, the Department of Energy’s Office of Science has repeatedly engaged
advisory committees—both FACA committees and NRC panels—to justify and
175
oversee a multi-billion-dollar suite of major scientific user facilities for materials

science. In one case in the early 1980s, only a few short years after materials
scientists had demonstrated they could harness ‘waste’ x-rays produced by high-
energy physics’ accelerator rings to extend the frontiers of materials research, the
Office of Basic Energy Sciences commissioned a National Academy study to
identify the highest priorities for the construction of new scientific user facilities
for material research and related disciplines (National Research Council 1984).
Seitz-Eastman identified four high-priority facilities: a soft x-ray source, a hard
x-ray source, a pulsed neutron source, and a steady-state neutron source. In 1986
Al Trivelpiece, the Senate-confirmed head of the Office of Science, resolved the
distributional political tensions among his intensely competitive and politically
well-connected national labs by assigning the soft x-ray source to Berkeley
National Lab on the west coast, the hard x-ray source to Argonne National Lab
in the Midwest, the steady-state neutron source to Oak Ridge National Lab in
the south, and a relativistic heavy ion collider, a nuclear physics facility prioritized
in the Nuclear Sciences Advisory Committee’s long-range plan, to Brookhaven
National Lab in the north-east (Salgado 1986). This ‘Trivelpiece plan’ focused each
of the labs, the two research communities, and, in the case of materials science, the
microelectronics, pharmaceutical, steel, and other materials-dependent industries
into a stable coalition that was able to build broad political support for building
out all four of the facilities. When the reactor-based steady-state neutron source
was no longer politically viable after the Three Mile Island accident, the Office
of Science returned to the Seitz-Eastman plan and assigned the pulsed neutron
source to Oak Ridge. The four facilities were built out over a span of more than
20 years, with Oak Ridge’s Spallation Neutron Source turning on only in 2007.
Once its suite of three x-ray light sources (the two identified in Seitz-Eastman and
Brookhaven’s National Synchrotron Light Source, which was under construction
when the Seitz-Eastman panel began its work) were up and running, the Office
of Basic Energy Sciences assigned its Basic Energy Sciences Advisory Committee
(BESAC) the task of developing a systematic, comparative evaluation scheme
to assess both the scientific productivity and the operational efficiency of the
facilities. This Birgeneau-Shen report showed that Berkeley’s Advanced Light
Source (ALS) was significantly underperforming (Birgeneau and Shen 1997).
The Department of Energy (DOE) ordered the lab to replace the facility director
and docked the facility’s funding. After allowing the new director a period to
reform operations at the facility, BESAC was again tasked with applying the
Birgeneau-Shen methodology to the ALS operations and scientific productivity
(Petroff 2000). With quantitative evidence of improved performance in hand, the
Office of Basic Energy Sciences reversed the funding cut, thus creating a strong
narrative about the exercise of its own oversight responsibility.
Oversight staff, primarily those in the White House Office of Management
& Budget and the Congressional appropriations committees, are often the
primary audience, customer, and enforcer of advisory committee reports. The
National Academy’s decadal survey in astronomy and astrophysics, now in its sixth
176
iteration, is the primary vehicle for establishing a ranked listing of the facilities
construction priorities for ground-based astronomy at the National Science
Foundation (NSF), space-based astronomy at NASA, and cosmology-related
accelerator-based experiments at DOE (National Research Council 2010). The
Office of Management and Budget (OMB) enforces the prioritization by refusing
to provide construction funds for any facility failing to make the decadal survey’s
cut, and appropriators tend to justify increases to project budgets by referencing
decadal surveys.2
Oversight staff also use advisory committees to call into question the underlying
premise or core design elements of a program proposal. In the case of the
Department of Energy’s Genomes to Life (GTL) User Facilities, the Office
of Biological & Environmental Science proposed four GTL user facilities as a
follow-on to DOE’s investment in the Human Genome project, each focused
on a single systems biology task: protein production and characterization,
characterization and imaging of biomolecular machines, proteomic analysis of
microorganisms, and modeling of microbial community cellular systems. OMB
forced a National Academy review, which endorsed the concept of large-scale
biology user facilities and suggested that four facilities were warranted. The
committee, however, recommended a complete redesign of the program so that
each facility would be equipped to carry out all of the original four tasks but
focused on a single application area, for example biofuels, carbon cycling, and
so on (National Research Council 2006).
Advisory committee reports as political vehicles

Advisory committee reports, even those issued with the prestige of the National
Academies, can have a purely political purpose. One of the most famous examples
of such a report is ‘Rising Above the Gathering Storm’ (National Academy of
Sciences 2007), which was self-funded by the Academies in response to bipartisan
requests by the chairs and ranking members of the National Science Foundation’s
House and Senate authorization committees. The committee, Committee on
Prospering in the Global Economy of the 21st Century, included presidents of
major universities, Nobel laureates, CEOs of Fortune 100 corporations, and
former presidential appointees. The report is a hefty collation of prior Academy
report recommendations, designed for ease of clearance through the Academies
report clearance process—and to produce a satisfying thud when landing on the
desks of members of Congress. The most technical sources for the high-impact
‘Rising Above the Gathering Storm’ report was a New York Times reference, and
many of the claims of scientific mediocrity made in the report did not hold up to
external scrutiny (Benderly 2007). However, the report played a significant role in
building support for the Bush Administration’s 2006 American Competitiveness
Initiative (which would have doubled the budgets of the National Science
Foundation, the Department of Energy’s Office of Science, and the labs of the
National Institutes of Science and Technology) and for passage of the America
177
COMPETES Act of 2007. In fact the chair of the Academy panel, Norm
Augustine, testified before the House Science Committee on October 20, 2005
before any substantial work on the report had begun (Augustine 2005).
Alternative model
The advisory system established in Germany—the Wissenschaftsrat—is quite
different. The structure is similar to that of the U.S. National Academies of
Science in that the members are top scientists but the funding structure is very
different—there is a direct funding line from the federal government and the
Länder. The development of plans of work is not devolved to individual agencies,
but to a Plenary Assembly. which is made up of both a scientific committee
and an administrative committee. The former consists of 24 scientists proposed
by key scientific institutions and eight public figures proposed by federal and
state governments. The latter consists of 22 members appointed by federal
and state governments. Possibly in response to different financial incentives,
the Wissenschaftsrat produces fewer reports but much more attention is paid
to them. And, possibly in response to different organizational structures, their
recommendations are, in the experience of one of the authors, taken more
seriously by government science agencies.
Peer review panels

The workhorses of both agencies are the scientific review committees, or peer
review panels. The operational aspects of the committees is much more opaque
than advisory boards—at NSF, for example, panelists are not identified, although
they are at NIH. The selection at NSF is entirely up to the program officer; the
selection at NIH is completely separate from the program officer. However, the
way in which peer review operates is essentially the same—panelists receive the
full proposal, meet in person or virtually to discuss, and then rank the proposals
in terms of quality (Russell et al. 2008).
The value of peer review is that it does allow the science to be evaluated
by experts, and ideally identifies and funds the best science. This is in marked
contrast to newly emerging countries, where there is often concern that research
funds can be allocated in corrupt ways. One of the authors served on a science
and technology advisory committee in Kazakhstan, where the existence of the
committee was explicitly to reassure the scientific community that there was
external oversight over the allocation of research funds.
There is, unfortunately, little scientific evidence about the value of peer review;
a recent meta analysis found only 10 articles addressing the topic in general,
and none evaluating the impact of different approaches on the quality of the
outcomes (Demicheli and Di Pietrantonj 2007). While some suggestions have
been promulgated to improve the process (Marsh et al. 2008), there is substantive
concern that the panel approach is essentially conservative in nature (Bornmann
178
2011). The time burden of peer review resulting from explosive growth in proposal
submissions and the excessive conservatism of the peer review process has been
identified as a systemic flaw of NIH by no less than a former president of the
National Academies and a former NIH director (Alberts et al. 2014).
One author’s personal experience is that the NSF peer review process, although
hailed as a gold standard for scientific review, is subject to multiple failures. The
result is a reinforcement of existing networks, making it difficult for new ideas
and new approaches to be heard.
The first of these is technological. At NSF, for example, program managers are
forced to rely on an IT system developed in the 1990s. The reviewer database
is so difficult to use that program managers often resort to word of mouth to
identify potential panelists and reviewers. Lists of conflicts of interest are manually
generated, and enforcement depends on the willingness of proposers to reports
conflicts, the capacity of program managers to manually cull information from
proposal submissions, and the dedication of staff to work through and track
information on excel spreadsheets.
The second is the variation in committee composition (Lee et al. 2013). There
is a social dynamic that works in any peer review process (Bohannon 2013) and
the authors have personally seen how panels will downrate very good proposals
simply because one of the dominant members of the panel doesn’t like one small
piece of the proposal. Of course, any human process will have random error, but
there is a great deal of evidence that unknown individuals, or individuals who
are not part of the ‘good old boy’ network are systematically likely to do less well
than those who are (Ginther et al. 2011).
The third is the failure of NSF (and other scientific agencies, for that matter)
to apply the same standards of science to its own processes. Azoulay makes a
convincing argument for the value of randomized control trials in panel review
(Azoulay 2012), yet repeated attempts to encourage NSF program officers to use
the scientific method in testing different committee structures and evaluation
methods have met with failure.
Despite this, the NSF approach is an international gold standard—but primarily
because of the quality of their program managers rather than their advisory
committee structures (by contrast, the Research Councils UK and the Australian
Research Council have largely relied on civil servant administrators, to the
detriment of their science).
Notes
1 www.gsa.gov/portal/content/100800.
2 As an example, House Report 114-130, accompanying the Fiscal Year 2016 Commerce,
Justice, Science, And Related Agencies Appropriations Bill, includes the following language:
‘An increase of $35,800,000 is recommended to develop capabilities within the Exoplanet
Exploration program to directly image exoplanets on the Wide-Field Infrared Survey Telescope
(WFIRST) mission with a coronagraph and to develop technology for future potential missions,
consistent with the priorities in the Astrophysics Decadal Survey for WFIRST and exoplanet
technology.’
179
References
Alberts, B., Kirschner, M. W., Tilghman, S., and Varmus, H. (2014) ‘Rescuing
US Biomedical Research from Its Systemic Flaws’, Proceedings of the National
Academy of Sciences 11(16): 5773–7.
Augustine, N. R. (2005) ‘Rising Above the Gathering Storm: Energizing and
Employing America for a Brighter Economic Future’, Testimony before the
U.S. House of Representatives Committee on Science, October 20.
Azoulay, P. (2012) ‘Research Efficiency: Turn the Scientific Method on
Ourselves’, Nature 484(7392): 31–2.
Balla, S. J. and Wright, J. R. (2001) ‘Interest Groups, Advisory Committees, and
Congressional Control of the Bureaucracy’, American Journal of Political Science
45(4): 799–812.
Benderly, B. L. (2007) ‘Rising Above “The Gathering Storm”’, Science Careers:
14 December.
Birgeneau, B. and Shen, . (1997) ‘Report of the Basic Energy Sciences Advisory
Committee Panel on DOE Synchrotron Radiation Sources and Science’,
Washington, DC: US Department of Energy’s Office of Science’s Office of
Basic Energy Science.
Bohannon, J. (2013) ‘Who’s Afraid of Peer Review’, Science 342(6154): 60–65.
Bornmann, L. (2011) ‘Scientific Peer Review’, Annual Review of Information
Science and Technology 45(1): 197–245.
Brooks, H. (1996) ‘The Evolution of US Science Policy’, in B. Smith and C.
Barfield (eds) Technology, R&D, and the Economy, Washington, DC: Brookings
Institution and American Enterprise Institute; p. 15–48.
Bybee, J. S. (1994) ‘Advising the President: Separation of Powers and the Federal
Advisory Committee Act’, The Yale Law Journal 104(1): 51–128.
Cozzens, S. E. (2009) ‘Science and Technology Policy Professionals: Jobs, Work,
Knowledge, and Values’, Encyclopedia of Life Support Systems: 92.
Demicheli, V. and Di Pietrantonj, C. (2007) ‘Peer Review for Improving the
Quality of Grant Applications’ The Cochrane Library.
Fleisher, L. (2015) The Role of Science Advisory Boards in US Federal Health
Policy (doctoral dissertation, Johns Hopkins University, Baltimore, MD).
General Services Administration (2016) ‘Federal Advisory Committee Act
(FACA) Management Overview’, www.gsa.gov/portal/content/104514.
Ginsberg, W. R. (2009) ‘Federal Advisory Committees: An Overview’,
Washington, DC: Congressional Research Service, R40520.
Ginsberg, W. R. (2015) ‘The Federal Advisory Committee Act: Analysis of
Operations and Costs’, Washington, DC: Congressional Research Service,
R44248.
Ginsberg, W. R., Carey, M. P., Halchin, L. E., and Keegan, N. (2012)
‘Government Transparency and Secrecy: An Examination of Meaning and
Its Use in the Executive Branch’, Washington, DC: Congressional Research
Service, R42817.
180
Ginther, D. K., Schaffer, W. T., Schnell, J., Masimore, B., Liu, F., Haak, L. L.,
and Kington, R. (2011) ‘Race, Ethnicity, and NIH Research Awards’, Science
333(6045): 1015–19.
Jasanoff, S. (2009) The Fifth Branch: Science Advisers as Policymakers, Cambridge,
MA: Harvard University Press.
Lee, C. J., Sugimoto, C. R., Zhang, G., and Cronin, B. (2013) ‘Bias in Peer
Review’, Journal of the American Society for Information Science and Technology
64(1) 2–17.
Marburger, John H. (2005) ‘Wanted: Better Benchmarks’ Science 308(5725): 1087.
Marsh, H. W., Jayasinghe, U. W., and Bond, N. W. (2008) ‘Improving the
Peer-review Process for Grant Applications: Reliability, Validity, Bias, and
Generalizability’, American Psychologist 63(3): 160.
Moffitt, S. L. (2010) ‘Promoting Agency Reputation through Public Advice:
Advisory Committee Use in the FDA’, The Journal of Politics 72(3): 880–93.
National Academy of Sciences (2007) Rising Above the Gathering Storm: Energizing
and Employing America for a Brighter Economic Future, Washington, DC: The
National Academies Press, www.nap.edu/openbook.php?record_id=11463.
National Research Council (1984) ‘Major Facilities for Materials Research
and Related Disciplines: Presentations to the Major Materials Facilities
Committee’, Washington, DC: The National Academies Press, https://doi.
org/10.17226/19349.
National Research Council (2006) ‘Review of the Department of Energy’s
Genomics: GTL program’, Washington, DC: The National Academies Press,
https://doi.org/10.17226/11581.
National Research Council (2010) ‘New Worlds, New Horizons in Astronomy
and Astrophysics’, Washington, DC: The National Academies Press, https://
doi.org/10.17226/12951.
Petroff, Y. (2000) ‘Report of the Basic Energy Sciences Advisory Committee
Subpanel Review of the Advanced Light Source at Lawrence Berkeley National
Laboratory’, Washington, DC: US Department of Energy’s Office of Science’s
Office of Basic Energy Science, DOE/SC-0018.
Russell, I., Bailey, D., Cooper, S., and Moskal, B. M. (2008) ‘Panel Session-
NSF Panels: A Reviewer’s Perspective’, 2008 38th Annual Frontiers in Education
Conference, p F1G–1.
Salgado, J. F. (1986) ‘Secretarial Site Selection Decisions on Specific Energy
Research Projects’, Washington, DC: US Department of Energy’s Office of of
the Under Secretary.
Sarewitz, D. (2007). ‘Does Science Policy Matter?’, Issues in Science and Technology
23(4): 31–8.
181
TEN
Public opinion and public policy in the

United States
Saundra K. Schneider and William G. Jacoby
In a properly functioning democracy public opinion should not only be correlated

with, but also a major determinant of, public policy. Is that the case in the United
States? In this chapter we will address that question by covering the major lines
of empirical research on the relationship between American public opinion and
public policy. Our primary focus will be substantive in nature but we also need
to address methodological issues about the measurement of both opinion and
policy. Our general conclusion is cautiously optimistic: Policy generally does
follow the contours of citizen preference, but elites also have ample opportunity
to shape manifestations of public opinion.
Early research
Since the very beginnings of empirical research in the social sciences, scholars have
recognized the critical importance of the opinion–policy linkage to the effective
functioning of a democratic society. For example, V. O. Key stated that ‘Unless
mass views have some place in the shaping of policy, all the talk about democracy
is nonsense’ (1961, p 7). Similar statements can be found in the writings of such
influential theorists as Lasswell (1941), Schumpeter (1950), and Easton (1965).
Indeed, the development and growth of public opinion polling was motivated
in part by a desire to increase democratic responsiveness. Prior to the inception
of mass surveys, the influential journalist and commentator Walter Lippmann
charged that public opinion was a fiction created by political elites to justify their
policy activities (1922, 1925). Hence public opinion was basically irrelevant and
even detrimental to any realistic sense of governmental responsiveness. Early
pollsters such as George Gallup strongly believed that survey research would
enable governmental officials to learn what the public really wanted, and lead
them to act accordingly (Gallup and Rae 1940).
Unfortunately, empirical social scientific research found that the world did
not live up to such optimistic perspectives. For example, sociologists such as
C. Wright Mills (1956) argued forcefully that public policy was generated by a
‘power elite’ that channeled governmental resources and benefits to their own
ends. At the same time, other public opinion researchers were providing evidence
that ordinary citizens did not seem to possess the kinds of orientations that were
believed to be necessary for a properly functioning democratic political system
(for example, Stouffer 1955; Prothro and Grigg 1960; McClosky 1964). Large
183
segments of the mass public displayed marked intolerance toward, and a willingness
to deny basic civil rights and liberties to, groups with whom they disagreed. In
the same vein, very few people appeared to understand or adhere to the basic
‘rules of the game’ for democratic decision making. In a set of complementary
findings, other scholars produced convincing evidence that most citizens were
unable to ‘translate’ meaningful policy preferences into electoral decisions and
that mass attitudes about political issues were disorganized, inconsistent, and often
non-existent (Campbell et al. 1960; Converse 1964).
For all of the preceding reasons, the early evidence suggested that public
opinion could not serve as a meaningful guide for public policy making. Instead,
scholars formulated a number of alternative theories to account for the existence
of seemingly democratic governance despite the lack of effective mass input. For
example, pluralism holds that competition among elite groups is the primary
dynamic underlying public policy (for example, Dahl 1972). Since resources
are inherently limited, group representatives engage in a constant process of
bargaining, accommodation, and compromise that tends to balance diverse
interests. And, in so doing, the resultant policies approximate societal preferences.
From a somewhat different perspective, theories of democratic elitism suggested
that democratic practices and values survived precisely because most of the mass
public are effectively disengaged from the political system, giving them little
influence and leaving room for elites (who do possess democratic inclinations)
to determine policy (Bachrach 1967).
Such pluralist and elite theories seem to produce democratic outcomes
without involvement of the mass public but critics pounced on these ideas almost
immediately. For example, E. E. Schattschneider (1960) asserted that interest
groups subverted democracy by shifting the scope of conflict in ways that benefited
their specific constituencies. Bachrach and Baratz (1962; 1963) made a similar
point by showing that powerful interests could simply prevent important issues
from ever being raised or placed on the public agenda. In a broad and powerful
critique, Theodore Lowi (1969) argues that a system based entirely on interactions
among elite groups cannot provide the benefits of a true functioning democracy.
In summary, social science research through the 1970s painted a fairly pessimistic
picture about the potential for democratic responsiveness in American society.
The public policy perspective

In the field of public policy, research on democratic responsiveness has tended
to focus on the policy agenda. This is the general process through which social
problems become defined as political issues worthy of governmental attention
and perhaps action. The concept of a policy agenda first appeared in Lasswell’s
stage heuristic model of policy development (1951). This perspective holds that,
regardless of substantive area, policies evolve through a rational series of distinct
steps. In one of the earliest stages of this model, society (that is, the mass public
and elites together) recognizes the importance of a particular issue and pushes it
184
Public opinion and public policy in the United States
to a prominent position in the media and political discussion. At later stages of

the process, government officials respond to the issues. The stage heuristic model
thus implies a rather direct connection between what the public wants and the
government provides.
While the state heuristic model is distinctive in its attempt to rationalize the
policy process, critics rejected it as too mechanical and apolitical (for example,
Lindblom 1968; Nakamura 1987; Sabatier and Jenkins-Smith 1993). To provide
a more realistic view of the process, Cobb and Elder (1983) and Eyestone
(1978) produced similar models of policy development that focused directly on
agenda setting. Fundamental to their view is the idea that many social problems
never make it onto the governmental agenda. In order to do so, a problem
must be promoted by an active and informed segment of the population. This
forces it to the attention of public officials, who are then more likely to take
action accordingly. Through this process, public policies are responsive to the
constituencies that are most active and involved in each substantive area.
Anthony Downs (1972) next provided an important and influential theoretical
contribution by arguing for the existence of an issue attention cycle. This is
defined as a recurrent process in which the salience of issues before the public
ebbs and flows. The rise of an issue to prominence depends on the substantive
nature of the problem and the groups that it affects. The salience of the issue
tends to fall when the government takes action. Thus, the central focus of the
issue attention cycle is the pattern of responsiveness by the government and the
public. And, despite the decline in public attention over time, there is a legacy
of administrative mechanisms that remain in place to address the concerns that
initiated the cycle.
John Kingdon (2003) expanded on Downs’ framework by arguing that policy
making involves interactions between three components or, as he called them,
streams. The problem stream consists of the myriad of issues confronting society.
The policy stream encompasses public officials, administrators, and policy
specialists. The political stream is composed of pressure groups, public opinion,
and electoral incentives. The policy agenda is formed when these three streams
coincide and merge around a common problem and proposed solution. Thus,
Kingdon’s theory allows for governmental responses to public opinion. But, it
also demonstrates why there are limits to the connection between policy and mass
preferences since the government is attentive to other actors in the respective
streams as well as its own preconceived solutions.
Baumgartner and Jones’ (2009) influential punctuated equilibrium model
ascribes a fairly limited role to public opinion in determining the content of
the policy agenda. During the long periods of stability that characterize the
equilibrium phase, policy making is dominated by established interests with an
immediate stake in governmental outputs. The public devote very little attention
to the everyday mechanisms and their concerns are effectively bypassed. Public
preferences are only met to the degree that they coincide with those of more
immediate participants in the policy-making process. During the shorter, more
185
intense, punctuation periods, public opinion is more likely to have an effect in

modifying the course of policy responses. But here, too, they are only one of
many actors involved in this process so the eventual agenda that enters the next
equilibrium period may not match citizen preferences very closely.
The advocacy coalition approach proposed by Sabatier and Jenkins-Smith
(1993) holds that policy subsystems are the basic unit that defines governmental
activity. These subsystems are composed of actors from a variety of public and
private organizations concerned about specific problems. Thus, the approach is
very similar to pluralism. And as such, the public’s role remains in the background.
Public opinion can be used as a resource by groups within a subsystem but it
has little direct effect of its own on policy making. Instead, the dynamics of
governmental activity are determined by cohesion and conflict within and
between subsystems.
A very different perspective from the policy literature holds that there is a
connection between the public and governmental elites but the predominant
line of influence runs from the latter to the former—the opposite direction from
that usually suggested by democratic responsiveness. Social construction theory
(Schneider and Ingram 1993) and related approaches (Edelman 1964, 1977; Dery
1994; Stone 2011) argue that policymakers present symbolic representations of
social phenomena to the mass public in attempts to attract and maximize popular
support for their preferred courses of action. Thus elites drive the process by
using language or images that they assume will receive positive reactions from
citizens. This perspective has been used to describe policy activity in a variety
of contexts, including efforts to encourage home ownership (DeNeufville and
Barton 1987), economic policy through the Federal Reserve (Abolafia 2004),
maintaining Social Security (Stone 2011), health care (Schlesinger and Lau 2000),
and disease prevention (Brown 1990).
Thus, the theoretical policy literature really does provide a wide variety of
perspectives on the degree to which governmental policy responds to citizen
preferences. On the empirical side, policy studies have tended to focus on
particular substantive areas. The general finding is that there is at least some
consistency between what the public wants and what the government does. For
example, this has been shown with respect to welfare reform (Schneider and
Jacoby 2005), health care (Grogan and Rigby 2008), environmental protection
(Konisky 2010), and disaster response (Schneider 2008). But, since the work has
been focused on single substantive topics, there is very little sense of cumulation
in the findings. So it is reasonable to conclude that the policy literature has not
provided a general consensus on the nature or extent of democratic responsiveness
in the United States.
Democratic representation at the national level

In the political behavior subfield, empirical research on policy representation
really began with Miller and Stokes’ classic article, ‘Constituency Influence in
186
Congress’ (1966). These scholars used a unique set of attitudinal data obtained
from members of Congress and a survey of respondents sampled from their home
congressional districts. This information was supplemented with the congressmen’s
roll call votes.
Miller and Stokes employed a multiple-equation model, which was quite
sophisticated for the time, to show not only the consistency between constituents’
attitudes and their representatives’ roll call votes, but also the intervening influences
of the congressmen’s own attitudes and their perceptions of constituency attitudes.
They examined these connections in three policy domains: social welfare, foreign
policy, and civil rights. Their main finding was that the degree of representation
varied markedly by issue area.
The representatives’ attitudes and their perceptions of constituency attitudes
always showed a robust correlation with roll call votes, regardless of policy domain.
But, the linkage between representatives’ attitudes and actual constituency attitudes
varied; it was quite weak for social welfare concerns, only slightly stronger for
foreign policy, and quite a bit stronger for civil rights. The correlations of the
representatives’ perceptions of constituent attitudes and actual constituent attitudes
showed the same pattern. Thus it appears that members of Congress believed that
their own votes mirrored the preferences that dominated their home districts.
And that, in turn, suggests an effective system of democratic representation. But,
the accuracy with which representative beliefs and attitudes reflected constituency
opinion varied markedly depending on the salience of the issue domain. In the
late 1950s, when the data were collected, economic prosperity made social welfare
concerns less salient. And, the complexities of foreign policy were, literally, quite
distant from the perspectives of ordinary citizens. As a result, governmental
officials did not receive clear signals from the public. In contrast, civil rights for
African Americans was quickly evolving as an issue that affected all aspects of
American society. Accordingly, it is likely that the public communicated their
preferences more loudly and that representatives paid closer attention when they
did. So, communication breakdowns between the public and their representatives
weaken, but do not eliminate, the opinion–policy linkage. These results provided
an optimistic counterpoint to some of the more pessimistic theories about mass
influence discussed earlier.
The Miller-Stokes study was highly influential but very difficult to replicate
precisely because it required detailed information about both constituency and
representative opinions. However, Achen (1977, 1978) pointed out that the
estimates from the Miller-Stokes model were based on correlational measures
of consistency between mass and elite preferences and behaviors. He shows that
the correlation coefficient is unsuitable for this purpose. Achen states bluntly
that ‘[l]arge correlations can occur when representatives are distant from their
constituents; small correlations can happen when they are near. Correlations
should be abandoned in the study of representation’ (1977, p 805). Using
alternative measures that show how close and responsive representatives are to their
constituents (rather than the covariance between the two), Achen demonstrates
187
that the degree of representation was lower than Miller and Stokes estimated,
and that differences across issue domains were very slight at most.
While social scientists generally recognized the validity of Achen’s criticisms,
it was difficult to incorporate them into subsequent research efforts because
measures of proximity required placements of the public and of political elites on
common policy scales. Such information was not generally available. But several
researchers proceeded by operationalizing his concept of policy responsiveness;
that is, the degree to which elites take actions that are substantively consistent with
the direction of constituency opinion. For example, Monroe (1979) compared
public opinion survey responses on questions that asked about specific policies
to subsequent governmental actions on those same policies, focusing on citizens’
preferences for the status quo or for change in various policy areas. He found that
policy action was consistent with the direction of public opinion in about two
thirds of the cases. Considering specific substantive areas, consistency was highest
in foreign policy (92% of governmental actions consistent with opinion) and lowest
in defense (44% consistent). Consistency also was higher on more salient issues
than on issues that the public did not consider to be as important. So, speaking in
very general terms, the direction of governmental action usually conforms to the
direction of public preferences. Again, however, the directionality of preferences
only points toward change or the status quo. Therefore, the substantive content
of policy representation still could be quite limited.
Page and Shapiro (1983) examine congruence in opinion change and policy
change over time. They focus on public opinion survey questions from a variety
of sources that ask about a given specific policy at two time points. They also
assess governmental action over a similar, but lagged time interval. Page and
Shapiro find that, when government policy does change, those changes mirror
movement in the distribution of public opinion about two thirds of the time.
Policy changes follow opinion changes most closely when the latter are large
and consistent in direction. Policy responsiveness is attenuated when the amount
of opinion change is small and the direction fluctuates. But, opinion–policy
consistency shows similar levels across domestic and foreign policies and it does
not vary depending on which national institution carries out the policy change.
The level of consistency between opinion and policy actually increases when the
policy change occurs at the state level.
But, what about cases in which policy does not move in a manner consistent
with changes in public opinion? Page and Shapiro show that about one third of
the total number of issues they examined involve cases where opinion changes but
policy does not—that is, governmental actions do not change over time. Many
of these examples involve situations where there is a ‘floor’ or ‘ceiling’ effect that
makes policy change impossible. In other cases, congruent policy change occurred
eventually but took a longer time to do so.
Turning again to only those issues for which policy change did occur, Page
and Shapiro were left with one third of the cases in which policy change moved
in the opposite direction to opinion change. They attribute some of these cases
188
to imperfect measures of policy and show that most of the other inconsistencies
occurred when opinion change was small and when opinions seemed to be
fluctuating within the public. They go on to conclude that ‘[w]hen Americans’
policy preferences change by a substantial amount, without reversal, public policy
(if it changes at all) overwhelmingly tends to move in the same direction’ (Page
and Shapiro 1983, p 180). While Page and Shapiro admit that there are important
caveats that must precede any generalizations, their results do appear to show an
impressive degree of policy responsiveness to citizen preferences.
One of the most important developments in the study of democratic
responsiveness is the thermostatic model of the connection between opinion
and policy (Wlezien 1995). This approach predicts the existence of coordinated
cyclical patterns in both mass preferences and governmental activity. The cycles
arise because public opinion on policy is always expressed relative to what
the government currently is doing with respect to that policy. The essential
relationship can be shown with the following simple equation:
Rt =Pt−At
In the equation, Pt represents the public’s preference for the ‘amount’ of

government activity (for example, the level of spending) on a policy at time t
and At represents the actual amount of governmental activity on that policy at
the same time point. The Rt term on the left-hand side of the equation gives
the public’s relative preference, which is expressed in responses to public opinion
survey questions (for example, ‘Should the government spend more or less on this
policy?’). If the public wants more than the government is currently providing,
then Rt will be positive and opinion surveys should show preferences for more
policy activity. If the government is doing more than the public believes it should,
then Rt will be negative and polls will reveal greater public preference for less
policy activity. This process is inherently dynamic since both of the terms on the
right-hand side change over time. If democratic responsiveness really exists, then
At should increase over time when Rt is positive. But, as soon as At surpasses
Pt then the public’s relative preferences reverse course. When their desire for
less activity becomes sufficiently pronounced, the government will also change
direction and begin reducing the outputs contained in At. This results in the
coordinated cycles, with Rt decreasing over time while At increases, and vice
versa. In this manner, government responds to public preferences but the public
also responds to what government does.
Soroka and Wlezien (2010) provide a fairly comprehensive empirical test of
the thermostatic model. Their results show that it comprises a powerful and
seemingly accurate representation for temporal patterns in public opinion and
national government spending. They do find that the thermostatic process works
most effectively on salient issues (for example, welfare and health care) but is
weaker in areas that draw less public attention (for example, foreign aid and
space exploration). Responsiveness transcends sophistication strata within the
189
mass public, with thermostatic cycles appearing among less-educated subgroups

as well as for citizens who are more sophisticated and involved in the political
system. And taking cross-national evidence into account, Soroka and Wlezien
show that the clarity of the connection between opinion cycles and spending
patterns is itself dependent on governmental institutions. For present purposes,
the relevant finding is that ‘a Madisonian presidential system encourages policy
representation in a way that is not evident in parliamentary systems’ (p 173). It
is difficult to interpret such a finding in any other way than as strong evidence
of democratic responsiveness.
Erikson, MacKuen, and Stimson (2002) build on the thermostatic model
to develop their conception of a macropolity. This broad and powerful model
describes not only the correspondence between public opinion and policy, but
also the mechanisms through which democratic representation occurs. The
macropolity perspective relies on a new way of looking at public opinion in terms
of the public’s policy mood (Stimson 1999). As noted earlier, individual citizens’
policy orientations are generally fragmented, incomplete, and disorganized. But,
the process of aggregating individual attitudes to the macro level smoothes out
most of the inconsistency, providing a more coherent sense of overall public
opinion. Furthermore, macro-level opinions across multiple issues show similar
trajectories over time. Stimson (1999) uses this to generate a concept of broad
‘mood’ that transcends the details of discrete policy positions. He then goes on
to show that this public mood is highly consistent with the broad contours of
American politics since the 1950s (Stimson 2004). Public mood is important
in large part because it is the manifestation of public opinion toward which
governmental officials are most attentive.
Erikson et al. (2002) argue that democratic representation can arise through two
different processes. On the one hand, electoral replacement accounts for policy
change by turnover in governmental personnel. With different people making
decisions, governmental institutions will generate different outputs. On the other
hand, rational anticipation of future sanctions can produce policy change despite
ongoing incumbent officeholders and administrative staffs. Instead, public officials
want to keep their own positions so they make decisions that are intended to
maintain public support. Both of these processes should result in governmental
policies that are responsive to public opinion. Erikson et al’s analysis demonstrates
that the two mechanisms exist simultaneously in the different branches of
American national government. The presidency and the Senate are responsive
largely through electoral replacement. But, rational anticipation provides the best
explanation for responsiveness in the House of Representatives. As one probably
would expect, the Supreme Court shows the lowest level of responsiveness across
the branches. Nevertheless the justices do appear to make decisions that take
anticipated public reactions into account.
Taken together, the evidence indicates that, at the national level, democratic
responsiveness is alive and well in the United States. There are some important
limitations to this general conclusion that have to be recognized. For one thing,
190
the close connection between opinion and policy emerges most clearly on broad
ideological indicators of these two concepts. The links between opinions on
specific programs and policy steps on the one hand, and governmental activity
within discrete policy areas on the other, is a bit more tricky to characterize. At
the same time, responsiveness varies across different economic strata within the
overall mass public. Stated simply, governmental decision makers are more attentive
to the wealthier segments of American society than to the poor (for example
Bartels 2008; Gilens 2012; McCall 2013). There are also ongoing concerns that
elites may be manipulating public opinion through issue framing and related
tools like issue priming (for example, Druckman 2004; Druckman et al. 2013).
While we certainly cannot dismiss such possibilities entirely, the evidence still
seems to show that national policy follows mass opinion to a greater extent than
public preferences are shaped by governmental action (Page and Shapiro 1983,
1992; Erikson et al. 2002).
Democratic representation at the state level

Compared to the national government, states have proven to be leaders in policy
innovation and experimentation. Accordingly, the American states are frequently
called ‘laboratories of democracy.’ This is convenient for social scientists, who can
exploit the wide variability in state political, economic, and social characteristics
to examine democratic responsiveness. In fact, there has been a great deal of
research aimed at explicating the links between public opinion and public policy
in the states.
One of the unique challenges in this subfield is the measurement of relevant
concepts. Unlike the assessment of public opinion at the national level, there are
very few surveys that provide comparable evidence of citizen preferences across
the states. Similarly, empirical manifestations of policy outputs across the states
have been problematic. Working out these issues has been a necessary precondition
for effective research on state-level policy responsiveness.
In order to overcome the lack of direct public opinion, early researchers relied on
qualitative assessments of state political culture (Key 1949; Elazar 1966; Sharkansky
1969). A number of other researchers developed indirect but empirically grounded
measures of state culture based on socioeconomic and demographic characteristics
(Luttbeg 1971; Zelinsky 1973; Gastil 1975; Garreau 1981; Lieske 1993; Hero
1998). In all of these cases, the general assumption is that political culture provides
a broad representation of the preferences of state populations.
There is a sizable literature that shows connections between state political culture
and state policy (Patterson 1968; Johnson 1976; Schlitz and Rainey 1978; Joslyn
1982; Savage 1982; Nardulli 1990; Elazar 1994) but critics have pointed to the
subjectivity inherent in these measures and the problems arise while assuming that
they reflect public opinion (Hero and Tolbert 1996). Bearing out this concern,
direct comparisons of state-level culture measures and attitude measures produce
very weak results (Lowery and Sigelman 1982).
191
A seemingly more suitable alternative is to employ survey-based measures of

state public opinion. While there were some early attempts to exploit state-level
survey research (Kovenock, Prothro, and Associates 1973; Black et al. 1974;
Erikson 1976), these efforts were limited by the unusual datasets on which they
relied. Starting in the 1980s, however, major advances were made by the research
team of Gerald Wright, Robert Erikson, and John McIver (1985, 1987) who
produced estimates of state opinion by aggregating results from CBS News/New
York Times polls of state electorates. These researchers provided the first yearly
estimates of state partisanship (that is, the proportion of Democrats versus the
proportion of Republicans within the state electorate) and ideology (that is, the
proportion of liberals versus the proportion of conservatives within each state),
but similar types of scores have been produced recently by the Gallup organization
(Jones 2009, 2010; Newport 2010). Other researchers have used surveys with
massive sample sizes (the 2000 and 2004 National Annenberg Election Surveys)
to disaggregate responses to the state level and produce similar measures of state
electorate partisanship and ideology, as well as a state-level analog to Stimson’s
public mood variable (Carsey and Harden 2010).
A long-standing, but different, tradition in this field has been to employ indirect
measures of citizen preferences. The general strategy is to use variables that should
be related to, or consequences of, an electorate’s orientations. These include
interparty competition, legislative malapportionment, the partisan composition
of state legislatures, and class support for the respective parties (Dawson and
Robinson 1963; Dye 1966; Hofferbert 1966; Sharkansky 1969; Sharkansky and
Hofferbert 1969; Jennings 1979; Peterson 1995). Still another approach is to assess
state opinion using the ideologies of elected public officials who are assumed to
reflect the views of their voters (Rabinowitz et al. 1984; Holbrook-Provow and
Poe 1987; Barrilleaux and Miller 1988). Probably the best known of these latter
measures is the citizen ideology score developed by Berry et al. (1998). While there
is ongoing controversy over the validity of this measure (for example, Brace et al.
2007; Berry et al. 2007a, 2007b), it does provide a convenient yearly placement of
each state’s population along a liberal–conservative continuum (Berry et al. 2010).
A recent innovation that has proven to be extremely popular in the subfield is
to use simulation-based measures of state public opinion. While the roots of this
strategy go back a long way (Pool et al. 1965; Weber et al. 1972), the modern
approach, known as multilevel regression and poststratification (MRP) employs
recent methodological advances (Park et al. 2004, 2006; Lax and Phillips 2009a).
First, a national survey is used to obtain preferences on a specified policy issue
within each of several sociodemographic subgroups. Second, the sizes of the
respective sociodemographic groups within the states are obtained from reference
sources (for example, the U. S. census). Third, the subgroup preferences obtained
at the national level are weighted by the within-state subgroup sizes and summed
across the subgroups to obtain a single policy preference for each state. While the
methodology depends critically on the relationship between sociodemographic
characteristics and public opinion (for example, Seidman 1975), it does provide
192
what appear to be reliable estimates of state-level public opinion on fairly specific

topics. As a result, MRP is being used very frequently for a variety of research
purposes.
Just as with state public opinion, the measurement of state policy has proven to
be difficult and elusive. For purposes of discussion it is convenient to differentiate
policy-specific measures and composite policy measures. As the name implies,
policy-specific measures focus on a single substantive area. Within this group
there are two major variants: First, many researchers employ program spending
figures as general continuous measures of activity within a given policy area (for
example, Gray 1974; Garand 1985, 1988; Barrilleaux et al. 2002). Others use
program enactments as dichotomous indicators of specific steps in the policy
process; such measures have been used in a very wide variety of substantive
areas (for example, Weber and Shaffer 1972; Gray 1973; Eyestone 1977; Berry
and Berry 1990; Preuhs 2005; Lax and Phillips 2012). Still other scholars create
measures based on idiosyncratic, but important, aspects of specific policies.
For example, Meier et al. (1989) use graduation ratios to measure educational
outcomes and Rom (2012) uses the percentage of the population enrolled within
welfare programs as an indicator of policy coverage. Obviously, it is difficult to
summarize this kind of work beyond giving examples, precisely because each
policy has its own unique aspects.
The composite policy approach is based on the general idea that it is impossible
to capture state policy activity by focusing on a single substantive area. Instead,
data reduction techniques like factor analysis and principal components analysis
are applied to indicators drawn from multiple program areas to recover more
general dimensions of policy making (for example, Hofferbert 1966; Sharkansky
and Hofferbert 1969; Crew 1969; Gray 1974; Plotnick and Winters 1985;
Barrilleaux and Miller 1988; Garand and Hendrick 1991). Other researchers have
used this approach to measure a concept they have named ‘policy liberalism’ (for
example, Klingman and Lammers 1984; Wright et al. 1987). The general idea is
to combine a set of indicators that tap the state’s placement along an ideological
dimension. These measures have proven to be very popular and they have been
replicated a number of times (Gray et al. 2004; Monogan et al. 2009; Gray 2012).
One potential problem with the original policy liberalism measures is that they
combine information across a wide time interval. In order to address this issue,
recent work by Caughey and Warshaw (2016) uses item response theory models
to estimate yearly state policy liberalism scores that extend back to the 1930s.
Additional problems with state policy liberalism measures are that they locate
states on an abstract continuum and that they inherently combine an arbitrarily
selected set of empirical indicators. In order to overcome these issues, Jacoby
and Schneider (2001, 2009) focus specifically on state policy spending across
program areas to produce summary measures of state policy priorities. Rather
than specifying in advance the nature of the dimension (a strategy that could
misrepresent the ways that states spend money), their procedure shows that
spending patterns conform very closely to a bipolar unidimensional continuum
193
that contrasts spending on policies intended to provide particularized benefits

to population groups (for example, welfare and health care) with spending on
policies intended to provide collective goods to the broader population (for
example, infrastructure and education).
Turning from the measures of the two relevant concepts to the question of the
relationship between them, virtually all research on state policy representation
examines some measure of policy as a function of citizen preferences in the
state, plus appropriate control variables. Many specific statistical techniques have
been employed for this purpose but virtually all of them have been based on
regression models or generalizations thereof (for example, logit and probit models;
simultaneous equation models; event history models, and so on). The results vary
widely depending more on the specific variables used in the statistical models
than the estimation strategies, underscoring the importance of the measurement
issues just discussed.
Early studies that relied on political culture or indirect measures of citizen
preference typically found that socioeconomic factors had the strongest effects
on policy, often completely pushing out the impact of more explicitly political
factors (Dawson and Robinson 1963; Dye 1966; Hofferbert 1966). At least two
different interpretations of these findings were offered. On the one hand, Dye
(1966) stated that ‘economic resources were more influential in shaping state
policies than any of the political variables previously thought to be important in
policy determination’ (p 29). On the other hand, several scholars stressed that
socioeconomic factors could, themselves, arise from political demands within
the states (for example, Hofferbert 1974; Godwin and Shepard 1976; Hayes and
Stonecash 1981).
The predominant scholarly view changed as soon as more direct measures of
state public opinion were brought into the models. Erikson, Wright, and McIver’s
influential book Statehouse Democracy (1993) provides convincing evidence that
there is a strong and direct connection between public opinion and public policy
in the American states. There were other scholars who came before Erikson et
al. and reported similar conclusions (for example, Weber et al. 1972; Weber and
Shaffer 1972; Erikson 1976; Nice 1983; Plotnick and Winters 1985). But their
analyses tended to be more limited in scope and their interpretations were more
cautious in tone.
The evidence shows that public opinion exerts a direct impact on policy
outputs as well as an indirect effect, operating through the electoral connection
(Erikson et al. 1993). States with more liberal and Democratic electorates tend
to elect more liberal and Democratic governors and legislators. Of course, these
public officials then promote more liberal policy alternatives. But, even after
taking this into account, the strong relationship between opinion and policy still
holds. Based on this kind of evidence, the predominant interpretation is that an
opinion-sharing model accounts for state-level democratic responsiveness (Hill
and Hinton-Andersson 1995; Soss et al. 2001; Wood and Theobald 2003). It is
important to emphasize that the connection between opinion and policy exists
194
regardless of the specific policy measures that are included in the empirical models
(Erikson et al. 1993; Jacoby and Schneider 2001; Soss et al. 2001; Wood and
Theobald 2003; Schneider and Jacoby 2006; Monogan et al. 2009).
While the overall results of a great deal of research suggest a fairly optimistic
view of democratic responsiveness in the states, there are some qualifications to
the general conclusion that have to be mentioned. Just as discussed at the national
level, correlational evidence about opinion and policy does not necessarily mean
congruence between the two. Lax and Phillips (2009a, 2012) show that there
are nontrivial gaps between citizen preferences and policy outputs, leading to
‘democratic deficits’ in some program areas. Other researchers have reported
similar results (Mooney and Lee 1995; Burstein 2003; Haider-Markel and
Kaufman 2006). It is also important to recognize that institutional features of
state government can mediate the linkages between opinion and policy (Erikson
et al. 1993; Brace et al. 2000; Maestas 2000; Barrilleaux et al. 2002; Yates and
Fording 2005; Lewis et al. 2015a).
There is ongoing controversy about direct democracy institutions and their
impact on public policy. Some researchers argue that ballot initiatives strengthen
democratic representation by clarifying the public’s voice on specific policies (for
example, Zax 1989; Gerber 1996; Matsusaka 2004; Bowler and Donovan 2010;
Lewis 2011; Lewis et al. 2015b). Others say that initiatives do nothing to facilitate
policy responsiveness to opinion (for example, Lascher et al. 1996; Camobreco
1998; Monogan et al. 2009).
Interestingly, there has been relatively little analysis of the degree to which
interest groups might affect citizen influence on state policy. And, the empirical
evidence that does exist is contradictory. Gray, Lowery, and colleagues have
definitely done the most work in this area but they conclude that interest groups
do not decrease the impact of public opinion or direct democracy mechanisms
(Gray and Lowery 1996; Gray et al. 2004; Monogan et al. 2009). In contrast,
Jacoby and Schneider (2001) show that the direct impact of interest groups on
policy priorities far outstrips that of public opinion (also see Schneider and Jacoby
2006; Burstein 2003).
Conclusion
Research on the nature and extent of democratic representation in the United
States comprises a very healthy area of scholarship (Shapiro 2011). This is perfectly
understandable since a vital defining characteristic of a democratic political
system is that there exists a causal connection from citizens’ policy preferences
to elite decisions and governmental programs. Fortunately, that currently seems
to be the case: There is a general consensus among political scientists that public
opinion does have an important impact on public policy making. Of course, that
does not close the book on this topic. We can still refine our understandings of
the representation process in a number of ways. Issues for further consideration
include refinements of measures, the potential for variable responsiveness to
195
different population groups, the potential effects of economic inequality, and the
possibility of nonrecursive, reciprocal connections between opinion and policy.
In addition, the world itself is changing. At the time of this writing, the
American electorate is highly divided such that policy responsiveness of any kind
will generate hostile responses from certain segments of the public. Furthermore,
the current president was elected without majority support and he actively
promotes a number of policy positions that are relatively unpopular within the
American public. So it will be interesting to see how democratic responsiveness
fares under such stressful conditions. Undoubtedly, a great deal of future research
will be devoted to exactly this topic.
References
Abolafia, M. Y. (2004) ‘Framing Moves: Interpretive Politics at the Federal
Reserve’, Journal of Public Administration Research and Theory, 14: 349–70.
Achen, Christopher H. (1977) ‘Measuring Representation: Perils of the
Correlation Coefficient’, American Journal of Political Science 21: 805–815.
Achen, Christopher H. (1978) ‘Constituency Opinion and Congressional
Behavior: A Reexamination of the Miller-Stokes Representation Data’, American
Journal of Political Science, 22: 511–35.
Bachrach, Peter (1967) The Theory of Democratic Elitism: A Critique, Boston, MA:
Little, Brown.
Bachrach, Peter and Morton S. Baratz (1962) ‘Two Faces of Power’, American
Political Science Review, 56: 947–52.
Bachrach, Peter and Morton S. Baratz (1963) ‘Decisions and Nondecisions: An
Analytical Framework’, American Political Science Review, 57: 632–42.
Barrilleaux, Charles, Thomas Holbrook, and Laura Langer (2002) ‘Electoral
Competition, Legislative Balance, and American State Welfare Policy’, American
Journal of Political Science 46: 215–27.
Bartels, Larry M (2008) Unequal Democracy: The Political Economy of the Gilded
Age, Princeton, NJ: Princeton University Press.
Baumgartner, Frank R. and Bryan D. Jones (2009) Agendas and Instability in
American Politics (2nd edn), Chicago, IL: University of Chicago Press.
Berry, Frances Stokes and William D. Berry (1990) ‘State Lottery Adoptions as
Policy Innovations: An Event History Analysis’, American Political Science Review
84: 395–416.
Berry, William D., Richard C. Fording, Evan J. Ringquist, Russell L. Hanson,
and Carl E. Klarner (2010) ‘Measuring Citizen and Government Ideology in
the U. S. States: A Re-appraisal’, State Politics and Policy Quarterly 10: 117–35.
Black, Merle, David M. Kovenock, and William C. Reynolds (1974) Political
Attitudes in the Nation and the States. Chapel Hill, NC: Institute for Research in
Social Science, University of North Carolina at Chapel Hill.
Bowler, Shaun and Todd Donovan (2010) ‘Public Opinion, Direct Democracy
and Fiscal Policy’, Paper presented at the 2010 Annual Meetings of the Midwest
Political Science Association.
196
Brace, Paul, Laura Langer, and Melinda Gann Hall (2000) ‘Measuring the
Preferences of State Supreme Court Judges’, Journal of Politics 62: 387–413.
Brace, Paul, Kevin Arceneaux, Martin Johnson, and Stacy G. Ulbig (2007) ‘Reply
to “The Measurement and Stability of State Citizen Ideology”’, State Politics
and Policy Quarterly 7: 133–40.
Brown, JoAnn (1990) ‘The Social Construction of Invisible Danger’, in Andrew
Kirby (ed), Nothing to Fear, Tucson, AZ: University of Arizona Press.
Burstein, Paul (2003) ‘The Impact of Public Opinion on Public Policy: A Review
and an Agenda’, Political Research Quarterly 56: 29–40.
Camobreco, John F. (1998) ‘Preferences, Fiscal Policy, and the Initiative Process’,
Journal of Politics 60: 819–29.
Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes
(1960) The American Voter, Chicago: University of Chicago Press.
Carsey, Thomas M. and Jeffrey J. Harden (2010) ‘New Measures of Partisanship,
Ideology, and Policy Mood in the American States’, State Politics and Policy
Quarterly 10: 136–56.
Caughey, Devin and Christopher Warshaw (2016) ‘The Dynamics of State Policy
Liberalism, 1936–2014’, American Journal of Political Science 899–913.
Cobb, Roger W. and Charles D. Elder (1983) Participation in American Politics: The
Dynamics of Agenda-Building, Baltimore, MD: Johns Hopkins University Press.
Converse, Philip E. (1964) ‘The Nature of Belief Systems in Mass Publics’, In
David E. Apter (ed), Ideology and Discontent, New York: Free Press.
Crew, Robert E. Jr. (1969) ‘Dimensions of Public Policy: A Factor Analysis of
State Expenditures’, Social Science Quarterly 50: 381–88.
Dahl, Robert A. (1972) Polyarchy: Participation and Opposition, New Haven, CT:
Yale University Press.
Dawson, Richard E. and James A. Robinson (1963) ‘Inter-Party Competition,
Economic Variables, and Welfare Policies in the American States’, Journal of
Politics 25: 265–89.
DeNeufville, Judith I. and Stephen E. Barton (1987) ‘Myths and the Definition
of Policy Problems: An Exploration of Home Ownership and Public-Private
Partnerships’, Policy Sciences 20: 181–206.
Dery, David (1994) Problem Definition in Policy Analysis. Lawrence, KS: University
of Kansas Press.
Downs, Anthony (1972) ‘Up and Down with Ecology—The Issue Attention
Cycle’, The Public Interest 28: 38–50.
Druckman, James N. (2004) ‘Political Preference Formation: Competition,
Deliberation, and the (Ir)relevance of Framing Effects’, American Political Science
Review 98: 671–86.
Druckman, James N., Erik Peterson, and Run Slothuus (2013) ‘How Elite
Partisan Polarization Affects Public Opinion Formation’, American Political
Science Review 107: 57–79.
Dye, Thomas (1966) Politics, Economics, and the Public: Policy Outcomes in the
American States, Chicago, IL: Rand-McNally.
197
Easton, David (1965) A Systems Analysis of Political Life, New York: Wiley.
Edelman, Murray (1964) The Symbolic Uses of Politics, Urbana, IL: University of
Illinois Press.
Edelman, Murray (1977) Political Language: Words That Succeed and Policies That
Fail, New York: Academic Press.
Elazar, Daniel (1966) American Federalism: A View from the States, New York:
Crowell.
Elazar, Daniel (1994) The American Mosaic: The Impact of Space, Time, and Culture
on American Politic, Boulder, CO: Westview Press.
Erikson, Robert S. (1976) ‘The Relationship Between Public Opinion and
State Policy: A New Look Based on Some Forgotten Data’, American Journal
of Political Science 20: 25–36.
Erikson, Robert S., Michael B. MacKuen, and James A. Stimson (2002) The
Macro Polity, Cambridge, UK: Cambridge University Press.
Eyestone, Robert (1977) ‘Confusion, Diffusion, and Innovation’, American Political
Science Review 71: 441–7.
Eyestone, Robert (1978) From Social Issues to Public Policy, New York: Wiley.
Gallup, George and Saul F. Rae (1940) The Pulse of Democracy. New York: Simon
and Schuster.
Garand, James C. (1985) ‘Partisan Change and Shifting Expenditure Priorities
in the American States, 1945-1978’, American Politics Quarterly, 14: 355–91.
Garand, James C. (1988) ‘Explaining Government Growth in the U.S. States’,
American Political Science Review, 82: 837–49.
Garand, James C. and Rebecca M. Hendrick (1991) ‘Expenditure Tradeoffs in
the American States: A Longitudinal Test, 1948–1984’, Western Political Quarterly
44: 915–40.
Garreau, Joel (1981) The Nine Nations of North America, Boston, MA: Houghton-
Mifflin.
Gastil, Raymond D. (1975) Cultural Regions of the United States, Seattle, WA:
University of Washington Press.
Gerber, Elisabeth R. (1996) ‘Legislatures, Initiatives, and Representation: The
Effects of State Legislative Institutions on Policy’, Political Research Quarterly
49: 263–86.
Gilens Martin (2012) Affluence and Influence: Economic Inequality and Political Power
in America, Princeton, NJ: Princeton University Press.
Godwin, Kenneth R. and W. Bruce Shepard (1976) ‘Political Process and
Public Expenditures: A Re-examination Based on Theories of Representative
Government’, American Political Science Review 70: 1127–35.
Gray, Virginia (1973) ‘Innovation in the States: A Diffusion Study’, American
Political Science Review 68: 1174–85.
Gray, Virginia (1974) ‘Expenditures and Innovation as Dimensions of
“Progressivism”: A Note on the American States’, American Journal of Political
Science 18: 693–9.
198
Gray, Virginia (2012) ‘Socioeconomic and Political Context of States’, in Virginia

Gray, Russell L. Hanson, and Thad Kousser (eds), Politics in the American States
(10th edn), Washington, DC: CQ Press.
Gray, Virginia and David Lowery (1996) The Population Ecology of Interest
Representation: Lobbying Communities in the American States, Ann Arbor, MI:
University of Michigan Press.
Gray, Virginia, David Lowery, Matthew Fellowes, and Andrea McAtee (2004)
‘Public Opinion, Public Policy, and Organized Interests in the American States’,
Political Research Quarterly 57: 411–20.
Grogan, Colleen M. and Elizabeth Rigby (2008) ‘Federalism, Partisan Politics,
and Shifting Support for State Flexibility: The Case of the U. S. State Children’s
Health Insurance Program’, Publius: The Journal of Federalism, 39: 47–69.
Haider-Markel, Donald P., and Matthew S. Kaufman (2006) ‘Public Opinion and
Policy Making in the Culture Wars: Is There a Connection Between Opinion
and State Policy on Gay and Lesbian Issues?’ in Jeffrey E. Cohen (ed), Public
Opinion in State Politics, Stanford, CA: Stanford University Press.
Hayes, Susan W. and Jeff Stonecash (1981) ‘The Source of Public Policy: Welfare
Policy in the American States’, Policy Studies Journal 9: 681–98.
Hero, Rodney E. (1998) Faces of Inequality: Social Diversity in American Politics,
Oxford, UK: Oxford University Press.
Hero, Rodney E. and Caroline J. Tolbert (1996) ‘A Racial/Ethnic Diversity
Interpretation of Politics and Policy in the States of the U. S.’, American Journal
Hill, Kim Quaile and Angela Hinton-Andersson (1995) ‘Pathways of
Representation: A Causal Analysis of Public Opinion-Policy Linkages’, American
Hofferbert, Richard I. (1966) ‘Relationship Between Public Policy and Some
Structural and Environmental Variables’, American Political Science Review 60:
73–82.
Hofferbert, Richard I. (1974) The Study of Public Policy, Indianapolis, IN: Bobbs-
Merrill.
Holbrook-Provow, Thomas and Steven Poe (1987) ‘Measuring State Political
Ideology’, American Politics Quarterly 15: 399–416.
Jacoby, William G. and Saundra K. Schneider (2001) ‘Variability in State Policy
Priorities: An Empirical Analysis’, Journal of Politics 63: 544–68.
Jacoby, William G. and Saundra K. Schneider (2009) ‘A New Measure of Policy
Spending Priorities in the American States’, Political Analysis 17: 1–24.
Jennings, Edward T., Jr. (1979) ‘Competition, Constituencies, and Welfare Policies
in American States’, American Political Science Review 73: 414–29.
Johnson, Charles A. (1976) ‘Political Culture in American States: Elazar’s
Formulation Examined’, American Journal of Political Science 20: 491–509.
Jones, Jeffrey M. (2009) ‘Political Party Affiliation: 30 States Blue, 4 Red in ’09
So Far’, Gallup.com’s State of the State Series, www.gallup.com/poll/122003/
Political-Party-Affiliation-States-Blue-Red-Far.aspx.
199
Jones, Jeffrey M. (2010) ‘Party ID: Despite GOP Gains, Most States Remain
Blue’, Gallup.com’s State of the State Series, www.gallup.com/poll/125450/Party-
Affiliation-Despite-GOP-Gains-States-Remain-Blue.aspx.
Joslyn, Richard A. (1982) ‘Manifestations of Elazar’s Political Subcultures: State
Public Opinion and the Content of Political Campaign Advertising’, in John
Kincaid (ed), Political Culture, Public Policy, and the American States, Philadelphia,
PA: Institute for the Study of Human Issues.
Key, V. O., Jr. (1949) Southern Politics in the State and Nation, New York: Knopf.
Key, V. O., Jr. (1961) Public Opinion and American Democracy, New York: Knopf.
Kingdon, John W. (2003) Agendas, Alternatives, and Public Policies (2nd edn), New
York: Addison-Wesley.
Klingman, David and William W. Lammers (1984) ‘The “General Policy
Liberalism” Factor in American State Politics’, American Journal of Political Science
28: 598–610.
Konisky, David (2010) ‘Public Preferences for Environmental Policy
Responsibility’, Publius: The Journal of Federalism 41: 76–100.
Kovenock, David M., James W. Prothro, and Associates (1973) Explaining the
Vote: Presidential Choices in the Nation and the States, 1968 (Part II), Chapel Hill,
NC: Institute for Research in Social Science, University of North Carolina at
Chapel Hill.
Lascher, Edward L., Jr., Michael G. Hagen, and Steven A. Rochlin (1996) ‘Gun
Behind the Door? Ballot Initiatives, State Policies, and Public Opinion’, Journal
of Politics 58: 760–75.
Lasswell, Harold D. (1941) Democracy Through Public Opinion, Menasha, WI: Banta.
Lasswell, Harold D. (1951) ‘The Policy Orientation’, in David Lerner and Harold
D. Lasswell (eds), The Policy Sciences, Palo Alto, CA: Stanford University Press.
Lax, Jeffrey R. and Justin H. Phillips (2009a) ‘Gay Rights in the States: Public
Opinion and Policy Responsiveness’, American Political Science Review 103:
367–86.
Lax, Jeffrey R. and Justin H. Phillips (2009b) ‘How Should We Estimate Public
Opinion in The States?’ American Journal of Political Science 53: 107–21.
Lax, Jeffrey R. and Justin H. Phillips (2012) ‘The Democratic Deficit in the
States’, American Journal of Political Science 56: 148–66.
Lewis, Daniel C. (2011) ‘Bypassing the Representational Filter? Minority Rights
Policies under Direct Democracy Institutions’, State Politics and Policy Quarterly,
11: 198–222.
Lewis, Daniel C., Saundra K. Schneider, and William G. Jacoby (2015a)
‘Institutional Characteristics and State Policy Priorities’, State Politics and Policy
Quarterly, 15: 447–75.
Lewis, Daniel C., Saundra K. Schneider, and William G. Jacoby (2015b) ‘The
Impact of Direct Democracy on State Spending Priorities’, Electoral Studies 40:
531–8
Lieske, Joel (1993) ‘Regional Subcultures of the United States’, Journal of Politics
55: 888–913.
200
Lindblom, Charles E. (1968) The Policy-Making Process, Englewood Cliffs, NJ:

Prentice-Hall.
Lowery, David and Lee Sigelman (1982) ‘Political Culture and State Public Policy:
The Missing Link’, Western Political Quarterly 35: 376–84.
Lowi, Theodore J. (1969) The End of Liberalism: Ideology, Policy, and the Crisis of
Public Authority, New York: Norton.
Luttbeg, Norman R. (1971) ‘Classifying the American States: An Empirical
Attempt to Identify Internal Variations’, Midwest Journal of Political Science 15:
703–21.
Maestas, Cherie (2000) ‘Professional Legislators and Ambitious Politicians: Policy
Responsiveness of State Institutions’, Legislative Studies Quarterly 25: 663–90.
Matsusaka, John G. (2004) For the Many or the Few: The Initiative, Public Policy,
and American Democracy, Chicago, IL: University of Chicago Press.
McCall, Leslie (2013) The Undeserving Rich: American Beliefs about Inequality,
Opportunity, and Redistribution, New York: Cambridge University Press.
McClosky, Herbert (1964) ‘Consensus and Ideology in American Politics’,
American Political Science Review 58: 361–82.
Meier, Kenneth J., Joseph Stewart, Jr., and Robert England (1989) Race, Class,
and Education: The Politics of Second Generation Discrimination, Madison, WI:
University of Wisconsin Press.
Miller, Warren E. and Donald E. Stokes (1966) ‘Constituency Influence in
Congress’, in Angus Campbell, Philip E. Converse, Warren E. Miller, and
Donald E. Stokes (eds), Elections and the Political Order, New York: Wiley.
Mills, C. Wright (1956) The Power Elite, New York: Oxford University Press.
Monogan, James, Virginia Gray, and David Lowery (2009) ‘Public Opinion,
Organized Interests, and Policy Congruence in Initiative and Noninitiative U.
S. States’, State Politics and Policy Quarterly 9: 304–24.
Monroe, Alan D. (1979) ‘Consistency Between Public Preferences and National
Policy Decisions’, American Politics Quarterly 7: 3–19.
Mooney, Christopher Z. and Mei-Hsien Lee (1995) ‘Legislative Morality in the
American States: The Case of Pre-Roe Abortion Regulation Reform’, American
Nakamura, Robert T. (1987) ‘The Textbook Policy Process and Implementation
Research’, Review of Policy Research 7: 142–54.
Nardulli, Peter (1990) ‘Political Subcultures in the American States: An Empirical
Examination of Elazar’s Formulation’, American Politics Quarterly 18: 287–315.
Newport, Frank (2010) ‘More States “Competitive” in Terms of Party
Identification’, Gallup.com’s State of the State Series, www.gallup.com/
poll/141548/States-Competitive-Terms-Party-Identification.aspx.
Nice, David C. (1983) ‘Representation in the States: Policymaking and Ideology’,
Social Science Quarterly 64: 404–11.
Page, Benjamin I. and Robert Y. Shapiro (1983) ‘Effects of Public Opinion on
Policy’, American Political Science Review 77: 175–90.
201
Page, Benjamin I. and Robert Y. Shapiro (1992) The Rational Public: Fifty Years of
Trends in Americans’ Policy Preferences, Chicago, IL: University of Chicago Press.
Park, David K., Andrew Gelman, and Joseph Bafumi (2004) ‘Bayesian Multilevel
Estimation with Poststratification: State-Level Estimates from National Polls’,
Political Analysis 12: 375–85.
Park, David K., Andrew Gelman, and Joseph Bafumi (2006) ‘State-Level Opinions
from National Surveys: Poststratification Using Multilevel Logistic Regression’,
in Jeffrey E. Cohen (ed), Public Opinion in State Politics, Stanford, CA: Stanford
University Press.
Patterson, Samuel C. (1968) ‘The Political Cultures of the American States’,
Journal of Politics 30: 187–209.
Peterson, Paul E. (1995) The Price of Federalism. Washington, DC: The Brookings
Institution.
Pool, Ithiel de Sola, Robert Abelson, and Samuel L. Popkin (1965) Candidates,
Issues, and Strategies, Cambridge, MA: MIT Press.
Preuhs, Robert R. (2005) ‘Descriptive Representation, Legislative Leadership,
and Direct Democracy: Latino Influence on English Only Laws in the States,
1984–2002’, State Politics and Policy Quarterly 5: 203–24.
Prothro, James W. and Charles M. Grigg (1960) ‘Fundamental Principles of
Democracy: Bases of Agreement and Disagreement’, Journal of Politics 22:
276–94.
Rabinowitz, George, Paul H. Gurian, and Stuart E. MacDonald (1984) ‘The
Structure of Presidential Elections and the Process of Realignment, 1944 to
1980’, American Journal of Political Science 27: 611–35.
Rom, Mark Carl (2012) ‘State Health and Welfare Programs’, in Virginia Gray,
Russell L. Hanson, and Thad Kousser (eds), Politics in the American States: A
Comparative Analysis (10th edn), Washington, DC: CQ Press.
Sabatier, Paul and Hank C. Jenkins-Smith (1993) Policy Change in Learning: An
Advocacy Coalition Approach, Boulder, CO: Westview.
Savage, Robert L. (1982) ‘Looking for Political Subcultures: A Rummage Sale
Approach’, Western Political Quarterly 34: 331–36.
Schattschneider, E. E. (1960) The Semisovereign People: A Realist’s View of Democracy
in America, New York: Holt.
Schlesinger, Mark and Richard R. Lau (2000) ‘The Meaning and Measure of
Policy Metaphors’, American Political Science Review 94: 611–26.
Schlitz, Timothy D. and R. Lee Rainey (1978) ‘The Geographic Distribution of
Elazar’s Political Subcultures Among the American Public: A Research Note’,
Western Political Quarterly 35: 410–15.
Schneider, Anne and Helen Ingram (1993) ‘Social Construction of Target
Populations: Implications for Politics and Policy’, American Political Science
Review 87: 334–47.
Schneider, Saundra K. (2008) ‘Who Is to Blame: (Mis)Perceptions of the
Intergovernmental Response to Disasters’, Publius: The Journal of Federalism 38:
715–38.
202
Schneider, Saundra K. and William G. Jacoby (2005) ‘Elite Discourse and

American Public Opinion: The Case of Welfare Spending’, Political Research
Schneider, Saundra K. and William G. Jacoby (2006) ‘Citizen Influences on
State Policy Priorities: The Interplay of Public Opinion and Interest Groups’,
in Jeffrey E. Cohen (ed), Public Opinion in State Politics, Stanford, CA: Stanford
University Press.
Schumpeter, Joseph A. (1950) Capitalism, Socialism, and Democracy, New York:
Harper.
Seidman, David (1975) ‘Simulation of Public Opinion: A Caveat’, Public Opinion
Shapiro, Robert Y. (2011) ‘Public Opinion and American Democracy’, Public
Opinion Quarterly 75: 982–1017.
Sharkansky, Ira (1969) ‘The Utility of Elazar’s Political Culture: A Research
Note’, Polity 2: 66–83.
Sharkansky, Ira and Richard I. Hofferbert (1969) ‘Dimensions of State Politics,
Economics, and Public Policy’, American Political Science Review 63: 867–79.
Soroka, Stuart N. and Christopher Wlezien (2010) Degrees of Democracy: Politics,
Public Opinion, and Policy, New York: Cambridge University Press.
Soss, Joe, Sanford F. Schram, Thomas P. Vartanian, and Erin O’Brien (2001)
‘Setting the Terms of Relief: Explaining State Policy Choices in the Devolution
Revolution’, American Journal of Political Science 45: 378–95.
Stimson, James A. (1999) Public Opinion in America: Moods, Cycles, and Swings,
Boulder, CO: Westview.
Stimson, James A. (2004) Tides of Consent: How Public Opinion Shapes American
Politics, Cambridge, UK: Cambridge University Press.
Stone, Deborah A. (2011) Policy Paradox: The Art of Political Decision Making (3rd
edn), New York: W. W. Norton.
Stouffer, Samuel A. (1955) Communism, Conformity, and Civil Liberties, New
York: Doubleday.
Weber, Ronald E., Anne H. Hopkins, Michael L. Mezey, and Frank Munger
(1972) ‘Computer Simulation of State Electorates’, Public Opinion Quarterly
36: 49–65.
Weber, Ronald E. and William R. Shaffer (1972) ‘Public Opinion in American
State Policy Making’, Midwest Journal of Political Science 16: 633–49.
Wlezien, Christopher (1995) ‘The Public as Thermostat: Dynamics of Preference
for Spending’, American Journal of Political Science 39: 981–1000.
Wood, B. Dan and Nick Theobald (2003) ‘Political Responsiveness and Equity
in Public Education Finance’, Journal of Politics 65: 718–38.
Wright, Gerald C., Robert S. Erikson, and John P. McIver (1985) ‘Measuring
State Partisanship with Survey Data’, Journal of Politics 47: 469–89.
Wright, Gerald C., Robert S. Erikson, and John P. McIver (1987) ‘Public Opinion
and Policy Liberalism in the American States’, American Journal of Political Science
31: 980–1001.
203
Yates, Jeff and Richard Fording (2005) ‘Politics and State Punitiveness in Black
and White’, Journal of Politics 67: 1099–121.
Zax, Jeffrey S. (1989) ‘Initiatives and Government Expenditures’, Public Choice
63: 267–77.
Zelinsky, Wilbur (1973) The Cultural Geography of the United States, Englewood
Cliffs, NJ: Prentice-Hall.
204
ELEVEN
Political parties and policy analysis

Zachary Albert and Raymond J. La Raja
One of the most notable trends in contemporary American politics is the dramatic
increase in the ideological polarization of the major political parties. Partisanship
has always permeated political life, from the way citizens make choices at the
ballot box to congressional roll call voting (for example Bartels 2000; Bond and
Fleisher 2000). It should come as no surprise, then, that the ideologies associated
with each party condition the set of policy prescriptions presented to the public.
However, the increasingly strong connection between partisanship and ideology
has transformed the politics of the policy process so that now policy analysis is
screened less and less for its adherence to standards of objectivity and expertise
and more for the ideological positions embraced by key constituencies within the
parties. Compared with previous eras when the parties reflected greater ideological
heterogeneity among their factions, the current ideological convergence among
intraparty constituencies (particularly within the Republican Party) makes partisan
decision makers less likely to rely on reputable expertise in pursuing policies.
This partisan trend has critical implications for the use of solid policy analysis in
government decision making.
To understand these transformations in the policy process, we propose an
approach to understanding policy making and policy analysis through the
structure of partisan politics. We view current changes in the party system as
shaping the sources and uses of policy analysis. To get a handle on this shifting
process, we conceive of political parties beyond their formal organizations (for
example national committees or party caucuses), as extended networks of interest
groups, think tanks, and other social organizations. These groups carry out the
comprehensive task of analyzing and developing public policies, with the formal
parties and their elected representatives adopting ideas that reflect the consensus
view of the key factions within the party network. In this informational exchange,
policy ideas emerge to the degree they adhere to the goals and interests of the
extended party coalition—a coalition that has shifted over time, in both parties,
to reflect highly ideological positions.
Throughout this chapter we will argue that the electorally oriented party
organizations in the United States are not ‘doers’ of policy analysis. This is
particularly salient in comparison with their European counterparts. Rather,
we claim that interest groups, think tanks, and other activist groups—which
are informally affiliated with one of the two major parties—conduct policy
analysis and develop policy proposals, funneling this information to formal party
actors that include candidates, party officials, and elected officials. The formal
205
party in this relationship acts as a filter through which the ideas of like-minded
organizations within their ‘extended party network’ (EPN) are turned into policy
proposals supported by normative claims that these enhance the national interest.
This informational exchange is not a new phenomenon—politicians have never
been policy analysts—but what has changed is the degree to which politicians’
reliance on expertise has given way to partisan considerations when it comes to
the creation and use of policy analysis. Our conceptual model of extended party
networks makes this transformation clear, as we point out how the changing
nature of party coalitions affects who provides policy guidance to political leaders.
To demonstrate this point, we will first highlight the historical and contemporary
role of parties in the policy process. We will then introduce the concept of the
EPN and argue that it offers important insights into how and why parties engage
with external organizations and activists to develop and justify policies. We will
end with several examples and a discussion of the implications of such an approach
to party policy analysis.
The importance of parties in the policy process

The role of parties in the policy process is an increasingly salient topic as partisan
polarization means that a wide variety of political processes are now characterized
by partisan dynamics. Rising partisanship in the contemporary electorate
(Hetherington 2001) and legislature (McCarty et al. 2006) raises anew questions
about when, how, and with whom parties engage in policy making. If the two
major parties are increasingly distant on a wide range of policy issues (Layman
et al. 2006), it is reasonable to ask what role parties—and partisanship—play in
the actual development and analysis of public policy. Specifically, we might ask:
How do parties identify problem areas in need of policy intervention? How do
they develop and weigh policy alternatives? What role does policy analysis play
in the justification of—or opposition to—particular policies? And, perhaps most
importantly, what is the role of the political party in the development and analysis
of policy relative to other political actors?
Recent theoretical and methodological innovations expand opportunities to
answer these questions. EPN theory points to the importance of understanding
political parties as a network of formal party organizations and informally affiliated
interest groups, think tanks, universities, media outlets, and political activists
(Grossmann and Dominguez 2009; Koger et al. 2010; Bawn et al. 2012). Groups
within the party network have been shown to exert significant influence over
the party nominating process (Cohen et al. 2008; Hassell 2015) and the success
of candidates (Desmarais et al. 2015) as they attempt to further their own policy
goals. However, so far EPN scholars have not provided an account of how affiliated
groups contribute to parties at each stage of the policy process. In this chapter
we offer a theoretical discussion of what this process might look like.
206
Parties in historical context: patronage, progressivism, and the modern party
American political parties have their origins in the first years of the nation, despite
widespread concern about the ‘mischiefs of faction’ (Madison 1787). Importantly,
the earliest parties structured substantive policy votes around the ‘Great Principle’
of how strong the federal government should become (Aldrich 1995). However,
the 19th and early 20th centuries witnessed a decline in the power of political
parties over policy outcomes (Mayhew 1974; Cooper and Brady 1981). By the
mid 19th century, patronage politics and weak congressional parties defined the
political landscape. Essentially, parties were distributors of political goods, often
lacking the will or ability to implement or assess policy programs. They focused
instead on winning elections, a goal better served by patronage politics than
disciplined policy implementation (Aldrich 1995).
This focus on electoral competition rather than coherent policy programs
was observed in Congress by Woodrow Wilson, who wrote that ‘There is
within Congress no visible, and therefore no controllable party organization …
The legislation of a session is simply an aggregate of the bills recommended
by Committees’ (1885, p 80). Wilson, like many political scientists after him,
viewed the British party system more favorably because of what he perceived as
the vigor of partisan action around clear policies in Westminster.1 In contrast,
the self-interested bargaining of the U.S. party system—driven by its effective
distribution of pork and, consequently, its ability to get politicians from both
parties re-elected—left little room for the party as disciplined policymaker.
The Progressive Era changed the nature of party politics. The early 20th century
saw a backlash against the patronage practices of party machines, with calls instead
for an efficient, rational, apolitical bureaucratic state (March and Olson 1983).
Political parties were no longer—if they ever had been—responsible for analyzing
and implementing policy, as the task was now widely assumed to fall under the
purview of non-partisan administrative agencies. Alongside and often driving
this push for rational, fact-based government was the birth of a number of now
highly significant think tanks that provided empirical and normative guidance
to policymakers (See the chapter by Rich on think tanks in this volume). Since
this period, the growth of the administrative state and the concomitant need for
policy analysis and expertise has only expanded opportunities for think tanks,
universities, and interest groups to seek influence over public policy.
The Progressive Era, then, saw an increase in both the demands placed on
government and the size and ability of government to meet them. At the same
time, parties remained focused primarily on electoral politics while think tanks
and emergent issue-groups began to play a larger role in the policy process (Wilson
1973). Lowi (1979) describes this as a process of policy abdication, with members
of Congress increasingly willing to allow bureaucrats to govern in their stead.
These historical trends suggest that American party organizations have been
ill-suited to construct policies in response to significant public problems and
have therefore turned to allied groups—what we call the EPN—for assistance.
207
Today, however, policy analysis in the EPN has become far more politicized.
As partisan coalitions have become more ideologically aligned, the parties
have increasingly turned to analyses conducted by ideological allies rather than
reputable, non-partisan actors. To date, the literature on parties in the policy
process has neglected robust accounts that illustrate group-centered policy making
in a partisan framework.
Prior accounts of parties and policy making
Many in modern political science have adopted Mayhew’s (1974) premise that
politicians are ‘single-minded reelection seekers’, because ‘the electoral goal …
must be achieved over and over if other ends are to be entertained’ (pp 16–17; but
see Fenno 1973). Indeed, some have concluded that the single, defining feature
of modern political parties is competition for office (Downs 1957; Schlesinger
1991). Unfortunately, this focus on parties in the electorate has resulted in the
comparative neglect of parties in government and as organizations that are integral
to the policy development process. Consistent with this conclusion, Reiter (2006)
found that the number of studies on party organizations has declined precipitously
since its peak in the 1930s and 1940s, and studies of parties in government have
always comprised less than 20% of the literature. Of course, this focus on parties
in elections is justified as parties certainly expend the greatest amount of time,
energy, and resources on electoral pursuits. At the same time, however, parties do
more than simply help members get elected. In most accounts, parties are said to be
charged with the ‘selection and aggregation of demands’ and the implementation
of policies consistent with their constituent preferences (Klingemann et al. 1994,
p 8). Parties often turn to policy analyses in order to justify their particular party
program. Nevertheless, the extant literature on parties provides little evidence
regarding the role of parties in this process.
Policy committees and entrepreneurs

There are several minor exceptions to this neglect. For example, some scholars
have examined the role of party policy committees in Congress,2 a logical
place to look for party policy activity. Indeed, the 1946 Joint Committee on
the Organization of Congress had grand visions of formal party committees
which would ‘formulate overall legislative policy of the two parties’ by regularly
coordinating with the Executive (p 13). In reality, however, the committees have at
most played a modest role in implementing policy and little to no role in the actual
formulation and assessment of it. Bone (1956) finds that Senate policy committees
‘have never been “policy” bodies, in the sense of considering and investigating
alternatives of public policy’ (p 352). Subsequent scholars have concurred, noting
that the main task of the policy committees has been to empower party leadership
to coordinate action and structure votes (Jones 1964; Kelly 1995; Crespin et al.
208
2014). Policy committees, then, perform an important function, but one which
tells us little about how parties develop and assess public policies.
Others have posited that individual ‘political entrepreneurs’ best explain the
process of policy development and change (Price 1971; Schickler 2001; Wawro
2001). Entrepreneurs are individuals who frame public issues and shape the terms
of policy debate, provide innovative policy solutions, and work to consolidate
these solutions into institutional or policy change (Sheingate 2003). However,
in our view the policy entrepreneur approach is an incomplete explanation of
the policy process. First, successful entrepreneurship is quite costly and highly
reliant on an amenable institutional and political context, making it quite rare
(DeGregorio 1997; Wawro 2001). Second, these accounts downplay the role of
non-governmental groups in providing information and ideas to entrepreneurs.
As Grossmann (2014, p 6) notes, policy entrepreneurs are embedded within
‘governing networks’ and we believe that these networks—and the partisanship
that increasingly defines them—structure the activities of policy development
and analysis.
It seems, then, that formal party institutions and actors are ill-equipped to attend
to the full process of policy analysis and development on their own. Indeed, the
weakness of American parties in developing public policy is especially noticeable
when compared to the formal, institutionalized role of many European parties
in the policy process. Many of these nations have formal statutes stipulating
who has the authority to be involved in policy making and establishing internal
policy-making processes and institutions (NDI (National Democratic Institute)
2011). For example, in the United Kingdom both the Labour and Conservative
Parties have institutionalized bodies that filter and manage policy proposals before
conference (NDI 2011, p 16). Similarly, in 2007 the French Union for a Popular
Movement formed the ‘Project for France’, a party organization that generated
more than 500 policy analyses and proposals. Equally formal, institutionalized
policy-making structures—often termed ‘corporatist’ arrangements (for example
Cawson 1986)—are found to varying degrees in many European nations. These
corporatist institutions result in stable bargaining over policy among representatives
from capital, labor, and government.
In contrast, American political parties have not established internal organizational
structures to develop, implement, and analyze public policy. Rather, we argue,
parties turn to external and informally affiliated organizations for assistance in
developing public policy ideas. In the absence of institutionalized party policy
processes—and in an increasingly polarized political landscape—partisanship and
ideology are the main factors structuring the use of policy analysis by political
parties. This is a phenomenon that EPN theory could explain well.
Extended parties and the development of public policies
The concept of political parties as extended networks imparts a new theory with
old roots. Early scholars conceived of the American party as ‘umbrella-like’, a ‘big
209
tent’ coalition of many diverse parts sharing similar political values (Key 1964;
Sorauf 1964; Eldersveld 1964). EPN theorists have refined this coalitional view of
parties by more closely examining the relationships and actions of actors within
the big tent. Bawn et al. (2012) argue that modern parties are best described as
extended networks of donors, interest groups, think tanks, media outlets, and
activists, all of whom shape the policy preferences and ideas of the party. Koger
et al. (2010) show that the party system is indeed characterized by two distinct,
polarized networks. Formal party organizations, then, are just one part of an
expansive network of partisan political actors, which may explain why earlier
investigations of formal partisan institutions found little party engagement in the
policy process.
The dynamics of this extended coalition revolve around the tasks of screening,
nominating, and supporting candidates sympathetic to the desires of the coalition,
with the ultimate goal of passing policy that benefits the coalition as a whole.
Bawn et al. write:
[P]arties … are best understood as coalitions of interest groups and

activists seeking to capture and use government for their particular
goals … The coalition of policy-demanding groups develops an
agenda of mutually acceptable policies, insists on the nomination
of candidates with a demonstrated commitment to its program, and
works to elect these candidates to office. In this group-centric view
of parties, candidates will, if the coalition has selected them well, have
as their paramount goal the advancement of the party program (Bawn
et al. 2012, p 571).
In this way, EPN theory provides a group-centered account of party policy

making that is partisan and centered on the provision of resources, both material
and symbolic.
However, despite sound theory, no empirical examination of the impact of
the extended party on the policy process has been conducted. Scholars have
demonstrated the important role informal party actors play in screening and
nominating candidates (Cohen et al. 2008) and helping them win elections
(Dominguez 2011; Desmarais et al. 2015) but have not connected the electoral
efforts of these actors to the achievement of favorable policy outcomes. Policy
analysis represents a potent tool through which EPN actors can influence party
policy.
Policy supply: the case for EPN provision of policy resources

Informal party actors such as interest groups and think tanks should be willing
to supply information, expertise, and resources to policymakers. Most generally,
the provision of information and data can guide both legislators and voters to
support policies favored by the EPN. Interest groups and think tanks in each
210
party’s orbit invest considerable resources in raising awareness of policy problems

and advocating particular solutions. Indeed, the theory of lobbying as a ‘legislative
subsidy’ (Hall and Deardorff 2006) points to the fact that these actors seek to
augment the resources of allies in order to assist them in the fight for favorable
policies. Policy analysis represents one key resource—ostensibly objective data—
that allied legislators can use to create and justify policies. And, for a number of
reasons, the provision of information through policy analysis seems to be a quite
efficient mechanism through which groups can influence policy outcomes.
First, EPN actors can use policy analysis to raise issues to the party’s agenda
and introduce new solutions that ultimately become party policy. As Kingdon
(1984) shows, interest groups are adept at setting the policy agenda and attaching
their own solutions to new problems. For example, the think tank Resources
for the Future was instrumental in elevating the issue of climate change onto the
governmental agenda and then meeting with mainly Democratic members of
Congress to advance their own policy ideas (McGann 2016). In the contemporary
landscape, these processes occur largely along partisan lines, with certain interest
groups and think tanks predictably petitioning Republicans (for example the 60
Plus Association and Heritage Foundation) or Democrats (for example Americans
for Democratic Action and The Brookings Institution). Extended party groups
are able to work with an affiliated party in order to advocate for policy change
and alter the lines of political cleavage (Schattschneider 1960) in order to achieve
favorable policy change (Baumgartner and Jones 1993). The increasingly partisan
nature of issue advocacy and policy analysis has served to politicize and degrade
policy information from the EPN, though it has certainly increased the prestige
of these groups within their respective coalitions.
Secondly, the provision of ideas, data, and even wholesale legislative proposals
represents an extremely effective way to influence the content of legislation. Such
action moves beyond ‘vote buying’ and instead allows groups to insert their own
ideas into legislative proposals. Information and ideas can be directly incorporated
into policy and justified on empirical grounds; monetary donations, on the other
hand, appear to alter the degree of effort legislators commit to the cause of a
particular group (for example Hall and Wayman 1990).
For example, liberal think tanks and interest groups were key providers of ideas
and analysis in the creation of the Patient Protection and Affordable Care Act.
Many of these ideas—like the Progressive Policy Institute’s 2007 proposal for an
individual mandate, income-based assistance, and a marketplace based on choice
and competition—were directly reflected in the final bill (see McGann 2016). In
a more direct attempt to influence the content of legislation, the GOP-affiliated
American Legislative Exchange Council (ALEC) combines policy analysis
with specific policy proposals that they then disseminate to state legislatures for
consideration. One Governing magazine report found that ALEC legislation is
introduced into state legislative bodies at a rate of roughly 1,500 bills per year, with
an average of 200 enacted into law (Greenblatt 2003). In this way, ALEC is able
to provide a much-needed commodity (policy) to affiliated legislators (typically
211
Republicans) and at the same time directly influence legislation in order to benefit
their members. Importantly, in both these examples the flow of information and
its incorporation into legislative proposals is highly conditional on the ideological
and partisan affiliation of the policy analyst. To the degree that these factors are
more important than an organization’s reputation for expertise, the politicization
of policy analysis might undermine effective policy and bias outcomes.
Policy demand: the case for party acceptance of EPN support

If informal actors within the extended party indeed funnel policy information
and ideas to the formal party, why should party members be willing to accept
such assistance? Part of the reason is likely the power informal actors have over
nominations and elections. Research has shown that party-affiliated groups are
willing to selectively challenge politicians who stray from their policy interests
(Masket 2009; Boatright 2014) and so some legislators may accept EPN ideas and
analysis out of fear of electoral reprisal. However, there are reasons to assume that
the relationship between informal and formal actors is more congenial, driven
by common purpose and mutually beneficial exchange rather than threats. Here
we outline three specific ways EPN assistance might serve the needs of formal
party members.
First, politicians are generalists rather than specialists, often lacking in-depth
knowledge of relevant policy debates. Confronted with an endless stream of
policy problems, legislators and parties tend to develop expertise in a select
number of domains (Lehnen 1967). Individual legislators will generally develop
expertise in policy domains relevant to their district, aiding them in their re-
election bids, while virtually ignoring other realms (Fenno 1973; Hall 1996).
The problem of expertise is further exacerbated by the ‘brain drain’ lobbying
firms exert on congressional staffs. As Drutman (2015) notes, ‘In an ever more
complex policy environment, the need for expertise becomes greater, but the
gap between public sector and private sector expertise is wider’ as quality staffers
are drawn to more lucrative private sector work (p 18). Additionally, the cuts
to government bureaucracies, particularly analytical units such as the Office of
Technology Assessment, reduce the amount of information and analysis officials
get from internal government sources.
All this, then, points to the fact that politicians often lack the specific knowledge
they need to perform their job responsibly. And, for the reasons above, members
of the extended network should be willing to subsidize their limited knowledge
and information in order to impact final legislation. In short, the increasing
complexity, specialization, and fragmentation of the federal government means
that policymakers have ‘found themselves unable to realize their activist agendas
without the informational resources and acquiescence or active support of societal
actors’ (Skogstad 2008, p 205).
Second, and relatedly, both parties and legislators have limited time and resources
to devote even to issues they care about. Webb and Kolodny (2006), for instance,
212
find that both parties employ the vast majority of their resources in electoral
rather than policy pursuits. The same is true for individual politicians, especially
following several decades of divestment from congressional staffers (Drutman and
Teles 2015). Indeed, allocations to members of Congress for official representation
duties suggest that staff tend to bear the brunt of reduced congressional allowances
(Brudnick 2015). And, importantly, even a fully staffed office devotes only
limited time to policy analysis and development. In fact, Romzek and Utter
(1997) show that politicians, like parties, generally employ staffers with expertise
in navigating the political world but little policy expertise, thereby prioritizing
political outcomes (for example re-election or coalition building) over policy
outcomes. Thus, given limited time and resources, parties and politicians should
again be viewed as demanders of policy information.
Finally, the points above rely on an overarching assumption that parties prioritize
the allocation of scarce resources according to the logic of the electoral imperative.
As many scholars have noted, winning elections is a necessary prerequisite for
the accomplishment of other political goals, and so:
[P]arty politicians set greater store in the notion of winning elections

than in using election outcomes to achieve a broad range of policy
goals. Candidates have interests and commitments in policy questions,
but … [f]or the most part, they are attracted to a particular party more
because of its promise as a mechanism for moving into government
than as a mechanism for governing itself (Keefe 1994, p 37).
The electoral imperative, then, leads both politicians and parties to prioritize
re-election as their primary goal. This fact is reflected in the types of candidates
who enter public office. According to Ehrenhalt (1991), modern politicians are
full-time professionals in their craft, who increasingly have the skill and resources
to win and retain office. At the same time, he notes, the individualization of
campaigns and politics means that these politicians often lack the requisite skill
and desire to work together to formulate coherent policies and party programs.
Increasing gridlock and poor institutional performance seem to support this
conclusion. Politicians, however, must typically provide some policy outputs in
order to stay in office (Cox and McCubbins 1993). In fact, as Lee (2016) shows,
recent parity in party competitiveness has only increased the need for each party
to offer a distinct policy platform. For this reason, we should once again expect
politicians to rely on extended party members to subsidize their policy efforts.
An extended party policy network

Our theoretical argument can thus be summarized as follows. First, political parties
have been designed to further the electoral goals of the individual legislators in
them. To win office, both parties and candidates make policy commitments and
therefore need to legislate once in office. This requires some degree of policy
213
analysis and the development of defensible policy ideas. Because parties have
underinvested in the development of stable, partisan policy-making institutions,
they can be characterized as demanders of policy information and expertise.
Members of the EPN, however, have inverse goals; they seek policy output and
engage in electoral and partisan politics only to further that goal.
As such, we expect informal party members to subsidize and engage in the
policy-making process by funneling information and ideas to the formal party,
with the party adopting policies that are electorally valuable.3 Members of the
EPN can use information and data generated through policy analysis to raise issues,
advocate solutions, and provide political cover for their legislative allies as they
advance policy proposals. Political parties, though, are not mere pawns. Rather,
they act as a filter through which the proposals of their extended network are
winnowed, selecting ideas and justifications that are both ideologically appealing
and electorally valuable. It is this function, we argue, that best explains the
importance of political parties in developing public policies. Scholars have thus
far failed to find direct evidence of party policy analysis because formal parties are
not policy analysts. Rather, they are decision makers. They largely receive policy
analyses and choose those proposals that are ideologically consistent, supported
by their extended network, and electorally valuable. An understanding of party
policy development thus requires an appreciation of the party as an extended
network and the ways in which this network provides politicized information that
biases policy outcomes toward their interests.
Anecdotal evidence of EPN involvement in the policy process

Several recent examples illustrate this theoretical argument at work. We should
stress, though, that nongovernmental actors have used information to achieve
instrumental policy objectives throughout American history. However, while these
actors were once quite bipartisan in these efforts—acting as the quintessential
‘access seekers’ described in the interest group literature (for example Hansen
1991)—in the contemporary landscape they increasingly do so in a predictable,
partisan fashion. The 1970s and 1980s were in many ways the bridge between
these two styles. Increasing partisan polarization (for example Layman et al. 2006)
and the proliferation of mainly conservative think tanks (Medvetz 2012) during
this period encouraged many organizations to direct their ideas, information, and
resources to a single party. As Barbara Sinclair (2006) notes, non-party groups
increasingly ‘found that one of the parties simply was so uncongenial to their
interests and concerns that they were driven into the arms of the other’ (312).
Thus, today we see major policy analysis and development occurring along
party lines and with the aid of a set of ideologically homogeneous think tanks and
interest groups. For example, we have alluded to the entirely partisan passage of the
Affordable Care Act (ACA) 2010, despite strong opposition from the Republican
Party and conservative organizations. This bill—and candidate Obama’s earlier
healthcare promises—drew heavily on the problems and ideas identified and
214
advocated by liberal groups. Healthcare policy analysis had long been conducted
in universities, think tanks, and interest groups, and these analyses provided
political cover for many of the ACA’s key provisions. The bill’s successful passage
was due mainly to the incorporation of these ideas into the final bill, providing
Democrats with justifications for specific policy provisions as well as a coalition
of supportive stakeholders. Some of these coalition members—like Health
Care for America Now—even conducted campaigns to sway public opinion in
favor of the bill (Oberlander 2010). Many conservative organizations, on the
other hand, conducted and disseminated policy research that argued against the
ACA’s provisions (McGann 2016). In this way, most of the policy information
surrounding the bill flowed through partisan channels. In general, it seems that
building and maintaining such partisan coalitions—by incorporating the policy
ideas of its members—is the main task of the party.
In contrast, a major obstacle for the Trump administration in trying to repeal
and replace the ACA was that the Republican Party did not adequately use its
EPN to develop a viable alternative to the ACA. When the Congressional Budget
Office (CBO) presented analyses that demonstrated the deleterious effects on
insurance coverage for millions of Americans under Republican-sponsored bills,
there was very little analytical support to challenge the findings of the CBO, or
at least frame the goals of the health bill in ways that could broaden appeal across
the Republican electoral coalition.
Still, the influence of extended network actors over partisan policy making
can be seen in other areas of the Trump presidency. In fact, the administration’s
relatively limited political and policy experience suggests that external groups
might play an even larger role than in the past. While the use of policy analysis to
justify proposals has been notably absent, the Trump agenda closely mirrors the
playbook of key conservative organizations. For example, journalists have noted
the strong influence that conservative groups like the Heritage Foundation have
had over early agenda items, especially regulatory and tax reform (Glueck 2016;
Shephard 2017). These domains have been studied extensively by conservative
organizations for several decades and we expect this research will shape future
policy debates and, if incorporated into final legislation, provide Republicans
with a coalition of extended party supporters. Early disagreements with policy
analysts in government (for example the CBO and Environmental Protection
Agency) further suggest that information from outside the bureaucracy will play
a critical role in the Trump presidency.
Policy analysis and the extended party in contemporary politics
Moving forward, scholars should seek to better understand party involvement in

policy analysis. As we have argued, policy development as a party-structured activity
has been thoroughly understudied. This kind of research is especially important
given the contemporary context of highly polarized politics, which seems to be
driving the policy process more than in previous eras. As such, it appears that
215
objective policy analysis has given way to a more politicized, ideological brand
of party policy engagement. The conceptual approach we recommend enables
scholars to identify more clearly and systematically the ways in which partisanship
influences policy analysis.
Political parties have been polarizing since roughly the early 1980s, a fact
reflected in increased party unity in roll call voting (Bond and Fleisher 2000)
as well as the more ideologically consistent—and divergent—stances of the
parties and their constituent members (Layman et al. 2006; Poole and Rosenthal
2012). Polarization also extends beyond Congress. A growing number of studies,
for instance, show that nominations and campaign funding follow a highly
distinctive partisan pattern that was not as salient during the 1950s and 1960s.
Grossmann and Dominguez (2009) find that most interest groups nominate and
donate predominately within one party coalition, while previously there was
more bipartisan support. Similarly, at the mass level, Heaney et al. (2012) find
that party activists are members of distinct political organizations divided along
partisan lines, with very few memberships in ‘bridging organizations’ that work
with both parties.
Given our argument we hypothesize that polarization in nominations, financing,
organizational membership, and congressional voting should be reflected in the
policy-making process as well. Because candidates increasingly receive polarized
support during their campaigns, they should be more likely to draw their ideas
for policy from highly partisan and ideological communities. Indeed, EPN theory
predicts that early screening of candidates by the party network will translate into
policy favoring the interests and values of the supporting coalition (Bawn et al.
2012). The sorting of interest groups and think tanks into partisan coalitions
means that the analyses and information party actors receive is increasingly divided
between two highly divergent ideological approaches.
Because party activists are exposed almost exclusively to policy information
via their highly partisan organizational memberships, they should discount any
policy that does not fit within this ideological framework. Regrettably, we even
observe partisan rivals on policy issues drawing on different ‘facts’ to support their
particular ideological viewpoint, as observed, for example, in national debates
about voter fraud and climate change. Close seat margins in Congress have
exacerbated this problem as parties have been further incentivized to provide
distinctive policies to demonstrate to voters that they are worthy of controlling
the majority (Lee 2016). If polarization continues to advance we expect that
policy ideas and policy analysis will become more cabined within tight partisan
networks. This trend represents a dramatic departure from the vision of neutral
policy experts outlined by Progressive reformers in the early 20th century. In
an era when polarization has made members of both parties increasingly hostile
and distrusting toward their opponents, the providers of party policy ideas are
similarly subject to scrutiny and cynicism by rival camps.
Interwoven in this story of increased partisan polarization is a changing
campaign finance landscape that reinforces the distancing of the two major parties.
216
Changes in campaign finance rules—notably the Bipartisan Campaign Reform

Act 2002 (BCRA) and the Citizens United 2010 decision—have privileged the
role of ideological, partisan groups in the campaign process. These changes have
given outside groups, such as Super PACs, an advantageous position from which
to influence elections while simultaneously reducing the ability of the formal
political parties to screen and support preferred candidates. Importantly, the formal
parties tend to prefer candidates who are more moderate on policy and therefore
closer to the median voter (La Raja and Schaffner 2015). While Super PACs fall
within the EPN, they give additional leverage to the ideological groups within
the coalition who fund politicians and push for more extreme policies within the
parties. Assuming these campaign finance trends continue, especially the increase
in Super PACs, we can once again expect that strongly partisan and ideological
dynamics will characterize policy analysis in the near future.
Should policy analysis that is characterized by extended networks and
partisanship worry Americans? Strongly differentiated parties were normatively
appealing to the 1950 APSA Committee on Political Parties because they provide
clear choices and accountability to voters. On the other hand, prior research
suggests a strong bias, toward wealthy constituencies and away from the most
vulnerable, in the interest group universe that makes up much of the EPN (for
example Hall and Wayman 1990; Strolovitch 2006; Gilens and Page 2014). Indeed,
empirical research has shown that polarizing tendencies have stimulated an increase
in economic inequality (McCarty et al. 2006). The strengthened influence of
informal actors through financing, nominations, and policy analysis might bias
policy outcomes toward the most active, involved, and resourceful members of
the EPN. As a result, ‘ideology and public discourse in general will be dominated
by voices that many voters consider too extreme’ (Bawn et al, 590). Thus, EPN
involvement in the policy process—and especially the politicization of policy
analysis conducted by these actors—might be both a cause and consequence
of greater partisan polarization and could lead to more extreme policy outputs.
If EPN politics is generally more extremist, we might expect that EPN policy
making could lead to greater gridlock in the political system. If the two parties
are drawing on policy ideas from distinct, polarized networks, they will become
increasingly unlikely to support their opponent’s policies, which presents
significant challenges to governance in a political system with separated powers.
This problem is exacerbated by the fact that both parties are able to draw on what
appears to be objective empirical analysis. In addition, because parties may shy
away from acting on issues that divide their respective coalitions, policies that are
passed will likely be sub-optimally designed, as parties need to incorporate the
preferences and interests of the wide range of supporters within their coalitions
rather than serve a broader public. The efforts by Republicans to ‘repeal and
replace’ the ACA is a case in point. The replacement offered by the Republican
Party leadership tends to benefit a small but powerful ideological constituency in
the Republican Party, much to the detriment of rural and working class voters
who supported Donald Trump in the 2016 election. To the degree that partisan
217
and coalitional concerns are conditioning contemporary parties’ usage of policy

analysis, we worry that American public policy will increasingly be defined by
such extremism and gridlock.
The Trump presidency obviously adds another challenge to the process of using
sound policy analysis to make governmental decisions. The populism that Trump
exploits delegitimizes expertise and policy analysis. Experts and reputable analyses
are often ignored. Trump’s claim, for example, that voting fraud is rampant led
him to create the Presidential Advisory Commission on Election Integrity even
though there is an overwhelming consensus among voting scholars and election
administrators in the states (including many Republican officials) that vote fraud
is not a problem. On issues such as climate change, his appointees have repeatedly
denied the findings of academic studies as they seek to remove regulations that
aim to address this problem. The president has refused to sign the widely endorsed
Paris Agreement that tries to get nations to pursue common goals to combat
climate change. These trends are not only restricted to the Republican Party,
though. Many Democrats have ignored empirical evidence highlighting the
negative economic effects of minimum wage laws and the positive educational
impact of charter schools in poor neighborhoods, for instance. On both sides,
then, over the long term, the growing gap in the policy networks that the two
major parties rely on for policy analysis will make compromise and governing
more difficult. This does not mean that objective policy analysis has become a
futile enterprise. In fact, the need for such an analysis seems as important as ever
as a means of holding both political parties accountable. The challenge, however,
is making these analyses relevant in a party system with fewer overlapping policy
networks than previously.
Notes
1 E.E. Schattschneider (1960), for instance, argued that America’s relatively low voter turnout
(compared to Great Britain) was due to the lack of clear, policy-based differences between the
two parties. This mindset was also reflected in APSA’s 1950 report on the state of American
political parties.
2 In the Senate, these are the Republican Policy Committee and Democratic Policy Committees.
In the House, the Democrats have the House Democratic Steering and Policy Committee,
while the Republicans have separate House Republican Policy and House Republican Steering
Committees.
3 In some ways, then, our approach is similar to that of Advocacy Coalition Framework scholars,
who argue that coalitions of groups with similar belief systems battle, in policy subsystems,
over the content of public policy (Sabatier and Jenkins-Smith 1993). However, we frame these
coalitions in partisan terms and explicitly incorporate political parties into battles over policy
ideas.
References
Aldrich, J. H. (1995) Why Parties? The Origin and Transformation of Political Parties
in America, Chicago, IL: University of Chicago Press.
Bartels, L. M. (2000) ‘Partisanship and Voting Behavior, 1952–1996’, American
218
Baumgartner, F. R., and Jones, Bryan D. (1993) Agendas and Instability in American
Politics, Chicago, IL: University of Chicago Press.
Bawn, K., Cohen, M., Karol, D., Masket, S., Noel, H., and Zaller, J. (2012)
‘A Theory of Political Parties: Groups, Policy Demands and Nominations in
American Politics’, Perspectives on Politics 10: 571–97.
Boatright, R. G. (2014) Congressional Primary Elections, London: Routledge.
Bond, J. R. and Fleisher, R. (2000) Polarized Politics: Congress and the President in
the Partisan Era, Washington, DC: CQ Press.
Bone, H. A. (1956) ‘An Introduction to the Senate Policy Committees’, American
Brudnick, I. (2015) Members’ Representational Allowance: History and Usage,
Congressional Research Service.
Cawson, A. (1986) Corporatism and Political Theory, Oxford: Blackwell Publishers.
Cohen, M., Karol, D., Noel, H., and Zaller, J. (2008) The Party Decides: Presidential
Nominations Before and After Reform, Chicago, IL: University Of Chicago Press.
Cooper, J. and Brady, D. W. (1981) ‘Institutional Context and Leadership Style:
The House from Cannon to Rayburn’, American Political Science Review 75:
411–25.
Cox, G. W. and McCubbins, M. D. (1993) Legislative Leviathan: Party Government
in the House, Berkley, CA: University of California Press.
Crespin, M. H., Madonna, A., Sievert, J., and Ament-Stone, N. (2014) ‘The
Establishment of Party Policy Committees in the U.S. Senate: Coordination,
Not Coercion’, Social Science Quarterly 96: 34–48.
DeGregorio, C.A. (1997) Networks of Champions: Leadership, Access, and Advocacy in
the U.S. House of Representatives, Ann Arbor, MI: University of Michigan Press.
Desmarais, B. A., La Raja, R. J., and Kowal, M. S. (2015) ‘The Fates of
Challengers in U.S. House Elections: The Role of Extended Party Networks
in Supporting Candidates and Shaping Electoral Outcomes’, American Journal
Dominguez, C. B. K. (2011) ‘Does the Party Matter? Endorsements in
Congressional Primaries’, Political Research Quarterly 64: 534–44.
Downs, A. (1957) An Economic Theory of Democracy, New York: Harper.
Drutman, L. (2015) The Business of America Is Lobbying: How Corporations Became
Politicized and Politics Became More Corporate, Oxford: Oxford University Press.
Drutman, L. and Teles, S. (2015) ‘Why Congress Relies on Lobbyists Instead of
Thinking for Itself ’, The Atlantic.
Ehrenhalt, A. (1991) The United States of Ambition: Politicians, Power, and the Pursuit
of Office, New York: Three Rivers Press.
Eldersveld, S. (1964) Political Parties: A Behavioral Analysis, Chicago, IL: Rand
McNally.
Fenno, R. F. (1973) Congressmen in Committees, Boston, MA: Little, Brown and
Company.
Gilens, M. and Page, B. I. (2014) ‘Testing Theories of American Politics: Elites,
Interest Groups, and Average Citizens’, Perspectives on Politics 12(3): 564–81.
219
Glueck, K. (2016) Trump’s Shadow Transition Team, POLITICO. Retrieved from

http://politi.co/2fAFfaW.
Greenblatt, A. (2003) What Makes Alec Smart? www.governing.com/topics/
politics/What-Makes-Alec-Smart.html.
Grossmann, M. (2014) Artists of the Possible: Governing Networks and American
Policy Change since 1945, Oxford: Oxford University Press.
Grossmann, M. and Dominguez, C. B. K. (2009) ‘Party Coalitions and Interest
Group Networks’, American Politics Research 37: 767–800.
Hall, R. L. (1996) Participation in Congress, New Haven, CT: Yale University Press.
Hall, R. L. and Deardorff, A. V. (2006) ‘Lobbying as Legislative Subsidy’, American
Hansen, J. M. (1991) Gaining Access, Chicago, IL: University of Chicago Press.
Hassell, H. (2015) ‘Party Control of Party Primaries: Party Influence in
Nominations for the US Senate’, Journal of Politics 78(1): 75-87.
Heaney, M. T., Masket, S. E., Miller, J. M., and Strolovitch, D. Z. (2012) ‘Polarized
Networks: The Organizational Affiliations of National Party Convention
Delegates’, American Behavioral Scientist 56(12): 1654 –1676.
Hetherington, M. J. (2001) ‘Resurgent Mass Partisanship: The Role of Elite
Polarization’, American Political Science Review 95: 619–31.
Joint Committee on the Organization of Congress (1944–1946) (1946) Report
of the 1946 Joint Committee on the Organization of Congress.
Jones, C. O. (1964) Party and Policy-Making: The House Republican Policy Committee,
New Brunswick, NJ: Rutgers University Press.
Keefe, W. J. (1994) Parties, Politics, and Public Policy in America, Washington, DC:
CQ Press.
Kelly, S. Q. (1995) ‘Democratic Leadership in the Modern Senate: The Emerging
Roles of the Democratic Policy Committee’, Congress and the Presidency 22(2):
113–139.
Key, V. O. (1964) Politics, Parties, and Pressure Groups, New York: Crowell.
Klingemann, H. D., Hofferbert, R. I. and Budge, I. (1994) Parties, Policies, and
Democracy, Boulder, CO: Westview Press.
Kingdon, J. W. (1984) Agendas, Alternatives, and Public Policies, New York: Pearson.
Koger, G., Masket, S., and Noel, H. (2010) ‘Partisan Webs: Information Exchange
and Party Networks’, British Journal of Political Science 39: 633–53.
La Raja, R. J. and Schaffner, B. (2015) Campaign Finance and Political Polarization:
When Purists Prevail, Ann Arbor, MI: University of Michigan Press.
Layman, G. C., Carsey, T. M., and Horowitz, J. M. (2006) ‘Party Polarization in
American Politics: Characteristics, Causes, and Consequences’, Annual Review
Lee, Francis E. (2016) Insecure Majorities: Congress and the Perpetual Campaign,
Chicago, IL: University of Chicago Press.
Lehnen, R. G. (1967) ‘Behavior on the Senate Floor: An Analysis of Debate in
the U. S. Senate’, Midwest Journal of Political Science 11: 505–21.
Lowi, T.J. (1979) The End of Liberalism, W. W. Norton and Company.
220
Madison, J. (1787) Federalist No. 10. Daily Advertiser, November 22.

March, J. G. and Olson, J. P. (1983) ‘Organizing Political Life: What Administrative
Reorganization Tells Us about Government’, American Political Science Review
77: 281–96.
Masket, S. E. (2009) No Middle Ground: How Informal Party Organizations Control
Nominations and Polarize Legislatures, Ann Arbor, MI: University of Michigan
Press.
Mayhew, D. R. (1974) Congress: the Electoral Connection, New Haven, CT: Yale
University Press.
McCarty, N. M., Poole, K. T., and Rosenthal, H. (2006) Polarized America: The
Dance of Ideology and Unequal Riches, Cambridge, MA: MIT Press.
McGann, J. G. (2016) The Fifth Estate: Think Tanks, Public Policy, and Governance.
Washington, DC: Brookings Institution Press.
Medvetz, T. (2012) Think Tanks in America. Chicago, IL: University of Chicago
Press.
National Democratic Institute (2011) Political Parties, Public Policy, and Participatory
Democracy: How Internal and External Consultation Helps European Political Parties
to Develop Responsive Public Policy.
Oberlander, J. (2010) ‘Long Time Coming: Why Health Reform Finally Passed’,
Health Affairs 29: 1112–16.
Poole, K. T. and Rosenthal, H. L. (2012) Ideology and Congress (4th edn),
Piscataway, NJ: Transaction Publishers.
Price, D. E. (1971) ‘Professionals and “Entrepreneurs”: Staff Orientations and
Policy Making on Three Senate Committees’, Journal of Politics 33: 316–36.
Reiter, H. L. (2006) ‘The Study of Political Parties, 1906–2005: The View from
the Journals’, American Political Science Review 100: 613–18.
Romzek, B. S. and Utter, J. A. (1997) ‘Congressional Legislative Staff: Political
Professionals or Clerks?’, American Journal of Political Science 41: 1251–79.
Sabatier, P. A. and Jenkins-Smith, H. C. (1993) Policy Change and Learning: An
Advocacy Coalition Approach, Boulder, CO: Westview Press.
in America, Boston, MA: Cengage Learning.
Schickler, E. (2001) Disjointed Pluralism: Institutional Innovation and the Development
of the U.S. Congress, Princeton, NJ: Princeton University Press.
Schlesinger, J. A. (1991) Political Parties and the Winning of Office, Ann Arbor, MI:
University of Michigan Press.
Sheingate, A. D. (2003) ‘Political Entrepreneurship, Institutional Change, and
American Political Development’, Studies in American Political Development 17:
185–203.
Shephard, A. (2017). The D.C. Think Tank Behind Donald Trump, February 22,
https://newrepublic.com/article/140271/dc-think-tank-behind-donald-trump.
Sinclair, B. (2006) Party Wars: Polarization and the Politics of National Policy Making,
Norman, OK: University of Oklahoma Press.
221
Skogstad, G. (2008) ‘Policy Networks and Policy Communities: Conceptualizing

State-Societal Relationships in the Policy Process’, in White, L., Simeon, R.,
Vipond, R., and Wallner, J. (eds), The Comparative Turn in Canadian Political
Science, Vancouver: UBC Press.
Strolovitch, D. Z. (2006) ‘Do Interest Groups Represent the Disadvantaged?
Advocacy at the Intersections of Race, Class, and Gender’, Journal of Politics
68(4): 894–910.
Wawro, G. (2001) Legislative Entrepreneurship in the U.S. House of Representatives,
Ann Arbor, MI: University of Michigan Press.
Webb, P. and Kolodny, R. (2006) ‘Professional Staff in Political Parties’, in R.
Katz and W. Crotty (eds) Handbook of Party Politics, London: Sage Publications,
pp 337–47.
Wilson, J. Q. (1973) Political Organizations, Princeton, NJ: Princeton University
Press.
Wilson, W. (1885) Congressional Government: A Study in American Politics, Boston,
MA: Houghton, Mifflin, Harcourt.
222
TWELVE
Policy analysis by corporations and

trade associations
Erik Godwin, Kenneth Godwin, and Scott Ainsworth1
Compared with other organizations, corporations have a distinct advantage among

those interests that use policy analysis to influence government. Corporations
employ many of the components of policy analysis such as calculating the net
present value, discount rate, internal rate of return of choice alternatives, and
environmental management systems (EMS) when making routine business
decisions. Corporations also employ more lobbyists, contact public officials more
frequently, file more comments on proposed regulations, sit on more government
advisory committees, and hire more policy analysis firms than other interest
organizations (Kerwin 2003; Baumgartner et al. 2009; Godwinet al. 2013).
Stanford business school professor David Baron (2013) argues in Business and
Its Environment that firms choosing to invest in political activities will have an
advantage over other firms. Baron demonstrates that investing in policy analysis and
political action often has a higher rate of return than investing in new technology,
hiring additional workers, or using other mechanisms that increase profits.
There are key differences, however, between policy analysis by corporations and
trade associations (hereafter CTAs) and analyses carried out using the techniques
set forth in textbooks. Policy analysts in the public sector should identify the
policy option that will provide the greatest net benefit to the public; CTA policy
analysts identify the option that will provide the greatest net benefit to their
organization. A benefit-cost analysis prepared by the public sector should contain
all major positive and negative externalities generated by each policy option.
Private sector analyses, unless they are analyses required by governments, typically
exclude externalities that do not affect their organization. Public sector analysts
may weight benefits and costs to different sets of people. The Environmental
Protection Agency (EPA), for example, weights health costs to children more
heavily than health costs to adults. Corporations give the greatest weight to the
factors that will have the greatest impact on the corporation’s profits.
Economic, scientific, and technical analyses such as benefit-cost, cost-
effectiveness, risk assessment, sensitivity analysis, and EMS are one category of
the tools that CTAs employ. These tools are products of underlying models that
predict the impacts of future human behavior on outcomes of interest to the
CTAs. Models, however, are simplifications of the real world and policymakers,
attorneys, and students of public policy must recognize that models used by
policy analysts are subject to the model’s assumptions concerning which variables
to include and to the accuracy of the data utilized. Among these assumptions
223
are such factors as the appropriate discount rate and weighting of different
distributional outcomes (Fisher et al. 2010). On such issues as climate change,
the complexity of the models is so great that it takes supercomputers multiple
days to run a single iteration (Farber 2007; Randall et al. 2007; McGuffie and
Henderson-Sellers 2005).
CTAs also use political and legal policy analyses to assist their organizations
in achieving their desired policy outcomes. Although conceptually distinct
from economic/scientific/technical analyses, the case studies we discuss below
demonstrate that policy analyses are highly interwoven with political and legal
analyses. Studies of regulatory rulemaking, political agenda setting, and court cases
show that the political and legal components of policy analysis have substantial
impacts on the validity of the scientific-technical analyses (West 2009; Shapiro
and Morrall 2012; Godwin et al. 2013).
Policy analysis by corporations has five interrelated objectives: (1) minimize the
firm’s political and economic risks; (2) gain a competitive advantage over other
firms; (3) anticipate and respond to government policy proposals; (4) compare
the impacts of alternative policies on the firm’s production processes and profits,
and (5) challenge regulations in the courts. Corporations also engage in policy
analysis when governments require such analyses. A corporation engaged in natural
resource extraction, for example, may be required to complete an environmental
impact analysis, risk assessment, and sensitivity analysis for a proposed project.
Similarly, local governments may require real estate developers to prepare impact
analyses concerning the effects of these projects on traffic, noise, pollution, and
government services. Corporations often use EMSs and their certification to
select among potential suppliers and to signal to their consumers the quality of
the corporation’s production process (Prakash and Potoski 2006).
Trade associations make extensive use of policy analysis to protect their members
from unwanted regulations, to bolster their members’ claims to governmentally
provided benefits, and to examine how alternative policy implementation
procedures will affect their members. Smaller corporations that do not have the
expertise to analyze the expected effects of new and proposed regulations find
trade association analyses particularly useful. Finally, trade associations use their
political policy analyses to develop issue positions and to build political coalitions.
Policy analyses assist lobbyists when they meet with public officials, testify before
congressional committees, and submit information to congressional staff. The
analyses supply the supporting documentation necessary to respond to legislators’
questions. CTAs often want their analyses to be part of the official record so that
they can use the information from their analyses to challenge a policy in the
courts or to encourage the legislative and judicial branches to require changes in
a regulation or its implementation process.
224
Policy analysis by corporations and trade associations
Approach of this chapter
We use multiple techniques to examine policy analysis by CTAs. First, we analyze

interviews with lobbyists for Fortune 1000 companies.2 These interviews provide
information on 216 political issues that the lobbyists identified as important
to their firms. For each issue, the lobbyists describe their strategy and the
information they provided policymakers. We collected information concerning
the use of policy analysis when engaging in informal contacts before the agency
issued its proposed rule, attending public hearings, providing written comments
concerning the proposed rule, and contacting the agency after the new rule had
been proposed. Our results indicate that lobbyists rate contacts with regulatory
officials before the regulatory agency proposed a rule as the most effective of
these lobbying techniques. Our results are similar to those of Kerwin (2003),
West (2009), and Yackee (2011).
Why is using policy analysis to lobby an agency prior to the issuance of a
proposed rule so critical? Because organizations can affect the model the agency
uses to justify its rule. If, for example, the National Institutes of Health (NIH)
is planning to regulate the uses of nanotechnology in developing new drugs
and medical treatments, informal contacts early in the process can influence the
characteristics of the model the agency will use in estimating the benefits, costs,
and risks of different testing procedures and new drugs. If a pharmaceutical
company can influence NIH to adopt the company’s favored model, this will
provide that company with a substantial advantage over other companies whose
models require different procedures and practices (Subramanian et al. 2015).
Our second method of clarifying the use of policy analysis in regulation
compares the effectiveness of different commenters on a proposed rule to achieve
changes to that rule. When there are public hearings on an issue, we analyze
the testimony to determine if it employs economic/scientific/technical policy
analysis. Our survey shows that large CTAs use policy analysis on a majority of
issues on which they lobby. Smaller CTAs, public interest groups, state and local
governments, and individuals are significantly less likely to employ such analyses.
Our third approach explicates three case studies. The first examines the
politics surrounding the permit request by TransCanada Corporation to build
the Keystone XL pipeline. Our investigation relies on materials available from
governments, publications by interested parties, and information published in the
media. The two remaining cases—the passage and implementation of the Data
Quality Act and the EPA’s regulation of perchlorate—are products of our earlier
research (Godwin et al. 2013). Our study of these cases includes interviews with
proponents and opponents of the policies and with government officials involved
in the decisions. We also analyzed congressional hearings and policy analyses
provided to policymakers by interest organizations.
We next examine the use of policy analysis cross-nationally as firms respond to
voluntary programs to reduce environmental damage. We review the adoption
of EMSs by firms and explain why firms spend millions of dollars to have their
225
production process certified and audited by the International Standardization

Organization (ISO).
The last section of the chapter speculates on the likely impacts of the 2016
elections on regulatory rulemaking. We show that this reorientation of the federal
government toward a smaller regulatory state is likely to increase significantly
CTA policy analysis in the federal regulatory process.
This chapter emphasizes bureaucratic rather than legislative policymaking. (See
the chapter by Joyce on the use of policy analysis by Congress in this volume.)
We do this for several reasons. First, organizations rely more heavily on policy
analysis when attempting to influence the bureaucracy than when attempting
to influence the legislature. Party affiliation and constituency influence typically
dominate congressional policy making. Information, however, is a key resource
in bureaucratic decision making. As Chubb (1983) points out, only organizations
with the resources necessary to build comprehensive policy models typically lobby
the bureaucracy concerning proposed regulations and their implementation. Few
citizen groups have these resources (McFarland 2004; Godwin et al. 2013).3
Citizen groups choose to lobby on highly visible and salient issues as this activity
is likely to increase their membership and donations. In contrast, lobbyists for
CTAs prefer to stay out of the public eye. The media ignore most implementation
decisions. Waivers that corporations receive from regulatory agencies rarely
make headlines. Yet these decisions often have huge implications for the affected
corporations. A change in the allowable parts per billion of arsenic, for example,
can increase dramatically a firm’s production costs and its competitive position.
Similarly, granting of a waiver to a company may mean the difference between
a company’s survival and bankruptcy.
Our final reason for focusing on the bureaucracy and regulatory activity is this
is where our expertise is greatest.4
How do corporations and trade associations orient their analyses?

Successful attempts to use policy analysis to change an existing or proposed
regulation require that the policy analyst understand the structure of government,
the applicable institutional norms, and the places in the policy process where
analysis has the greatest impact. The Office of Information and Regulatory Affairs
(OIRA) in the federal government and state offices that perform similar functions
generally oversees the promulgation of new regulations. When deciding whether
to recommend that the president or the governor approve a major new policy or
regulation, OIRA and state regulatory agencies usually ask the following questions:
1. Does the proposed policy serve a legitimate public purpose?

2. Does the policy address a market failure? If so, can the government fix that
failure by making the market more efficient rather than by increasing the
number of regulations?
226
3. Does the policy decrease efficiency by decreasing competition within the

private sector?
4. Are the causal relationships between the policy change and the expected
outcome clear and scientifically based?
5. Who are the stakeholders? How does the proposed rule change their incentives?
Are there sub-groups of a given category that will respond differently? (for
example, when regulating pesticides, large agribusiness farms may have
concerns that vary from those of micro-farmers).
6. What are the monetary and nonmonetary benefits and costs of the proposed
change?5 What is the baseline for the benefit-cost analysis? Does the analysis
account for existing trends? How will the government evaluate if a policy
change is successful?
7. What is the appropriate discount rate for future benefits and costs?
8. What are the effects of the regulation on equity? Are benefits or costs
concentrated among the more or the less advantaged populations?
9. How probable is it that the bureaucracy will implement the regulation as
intended? What are the barriers to successful implementation? What are the
possible unintended consequences?
10. Is this regulation consistent with the statutory intent of the authorizing law
and is the regulation consistent with the president or governor’s policy agenda?
At the national level, a proposed regulation must meet six additional requirements.
It should: (a) not unduly burden small businesses; (b) not be an unfunded
mandate to the states; (c) address environmental justice considerations, (d) not
harm significantly energy supply, distribution, and use; (e) protect children from
environmental health and safety risks, and (f) be consistent with the agency’s
previously approved data quality guidelines.6 Each of the 10 questions and six
requirements provides policy analysts with a potential lever to influence policy
outcomes.
The effectiveness of policy analysis during the rule-making process

To study policy analysis in the rule-making process, we collected all comments
and government responses on every economically significant rule finalized in 2008
in the Department of Energy (DoE), the EPA and the Department of Transportation
(DoT).7 The comment letters addressed 994 separate issues. Business organizations—
corporations, trade associations, peak associations, and professional associations—
dominate the commenting process. Sixty-four percent of all comments come
from business interests; only 16% come from citizens and citizen action groups.
CTAs are successful on more than 40% of the requests they make in comment
letters while slightly over 10% of citizen action group comments achieve their
goals (Godwin et al. 2013).
Two factors account for these differences. The first concerns the goals of the
comment. A comment can ask for a more stringent rule, a less stringent rule, for
227
no change in the proposed rule, or for a clarification of the proposed rule. Table
12.1 shows that business interests achieve their objective 100% of the time when
they request that an agency not change its proposed rule. In contrast, citizens
and citizen groups are successful on only 14% of their requests that an agency
not change its rule. Why would a corporation argue for no change? Because the
proposed rule gives the firm a competitive advantage, and it wants to keep that
advantage. For example, if EPA’s proposed rule imposes substantially more rigorous
pollution abatement, a corporation that has invested heavily in the latest pollution
abatement technology is likely to prefer that EPA not reduce the stringency of
Table 12.1: Participation levels and likelihood of success of business and citizen organizations
Business organizations (CTAs) Citizen organizations
Number of issues commented on 634 155
Success rate when requesting
a less stringent provision 33% 14%
a more stringent provision 4% 9%
no change in the provision 100% 14%
a clarification of a rule’s provision 70% 29%
Average success rate on all comments 42% 12%
Source: Adapted from Godwin et al. (2013, p 102).
its proposed rule. A corporation that has not made such investments may request
a less stringent rule or a longer time to meet the new requirements. Given the
importance of rulemaking to competition within and between industries, most
conflicts during the comment phase occur between competing firms.
When corporations request the clarification of a rule, their success rate is 70%.
The success rate is high because clarifications often have little impact on the
regulation’s stated goal. Business interests are less effective when they request
a change that will make a rule less stringent. Only one third of these requests
succeed. CTAs are successful in only 4% of the cases where their objective is a
more stringent regulation. The major reason for the low success rate of comments
requesting greater stringency is that granting the request would lead to reduced
competition within an industry (Godwin et al. 2013).
Does the greater success of business interests than citizen groups demonstrate
that business is capturing regulatory agencies? No. It is the quality of a comment
and its fit with the agency’s mission that influence its success. Business and citizen
groups are equally successful when their comments are of similar quality.8 Business
interests win more frequently because they invest more heavily in policy analysis.
228
Case studies of corporate and trade association policy analyses
To provide a more in-depth understanding of CTAs’ use of policy analysis,

we detail four illustrative case studies and provide a discussion of regulatory
waivers. Our first case concerns the Keystone XL pipeline permit request by
the TransCanada Corporation. We use data available from government records
as well as the benefit-cost analyses, risk assessments, and environmental impact
statements prepared by think tanks, trade associations, peak associations, and
large corporations.
Policy analyses involved in the Keystone XL Pipeline decision
The Keystone XL decision was highly salient, partisan, and ideological, and
affected billions of dollars in investments, contracts, and wages. Granting the
permit would be a victory for supporters of the fossil fuel industry and would
set an important precedent for future pipeline proposals. For opponents of the
pipeline, Keystone XL symbolized the degradation of the environment that the
oil from tar sands represented. It also provided an example of the environmental
risks that the fossil fuel proponents were willing to take to achieve their desired
goals. Opponents argued that as the pipeline would cross the Ogallala Reservoir,
the groundwater source that supplies water to much of the Great Plains, a major
spill could have a disastrous impact on agriculture and water quality. Opponents
stressed TransCanada’s terrible safety record, and they maintained that TransCanada
had greatly underestimated the risk the pipeline would pose. TransCanada’s permit
request estimated that its state-of-the art pipeline would average only 1.4 spills
every 10 years, but its most recently completed pipeline had experienced seven
spills in a single year (NRDC (Natural Resources Defense Council) 2012).
Numerous organizations conducted benefit-cost analyses, risk assessments, and
environmental impact analyses of the proposed pipeline. IHS Cambridge Energy
Research Associates (IHS CERA) and the Battelle Memorial Institute conducted
these analyses for TransCanada. The American Enterprise Institute, the American
Petroleum Institute (a petroleum and natural gas trade association), the AFL-CIO,
and the National Association of Manufacturers also completed comprehensive
pro-pipeline studies. These analyses focused on the large number of jobs that the
pipeline would create. In its benefit-cost analysis, IHS CERA estimated that the
pipeline would create 117,000 jobs in North America.
Environmental groups challenged the IHS CERA claims. Rather than 117,000
new jobs, their analyses estimated that the pipeline would create only seven
permanent jobs (NRDC 2012). The environmental groups showed that the
petroleum taken from tar sands was extremely dirty and using it would send the
wrong message about the importance of stemming climate change. The pipeline
was inherently dangerous to agriculture and to water quality as a major spill could
corrupt the major water source in the Great Plains. Finally, the recent increase
229
in natural gas production in the United States meant that the oil the pipeline
carried was not needed in North America.
An important aspect of the Keystone XL case was the participation by different
governments. This is a common pattern in lobbying the legislature and the
bureaucracy (Baumgartner et al. 2009; Godwin et al. 2013). The State of North
Dakota prepared an analysis showing that the pipeline should include a spur to
the Bakken shale formation. This would increase the net social benefits of the
pipeline and decrease the cost of getting Bakken shale oil to refineries. Risk
assessments by the State of Nebraska and the Department of the Interior indicated
that the proposed pipeline route would create unacceptable risks to the Ogallala
Reservoir. Those reports led TransCanada to amend its proposal and to route
the pipeline away from sensitive environmental areas, national parks, and much
of the reservoir. The EPA prepared assessments of the pipeline’s impact on global
warming. The Department of Commerce, DoT, and the Office of Pipeline Safety
provided analyses supporting the pipeline. The final 2,000-plus-page finding of
the Department of State included data from environmental impact assessments,
risk assessments, sensitivity analyses, and benefit-cost analyses from public and
private parties. The Department of State recommended granting TransCanada
the permit to build the pipeline. President Obama, however, chose not to issue
the permit. When Donald Trump became president, he issued four presidential
memorandums and one executive order allowing TransCanada to resubmit its
application to build the pipeline (VanWyhe 2017). On March 24, 2017 the Trump
administration approved the application (Labott and Diamond 2017).
The Data Quality Act
Political analysis is often a critical component to policy analysis. To obtain a

new legislative policy, an interest organization must put together a sufficiently
large coalition to obtain majorities in subcommittee, committee, and floor votes.
Alternatively, the interest can attach its desired policy to an existing bill. The first
approach requires getting the issue on the political agenda and keeping it there.
As Baumgartner et al. (2009) have shown, this process typically requires multiple
years. An alternative approach is to have a policy analyst develop a policy proposal
and identify the legislators who are strategically placed to add it as a rider to an
existing bill during the committee mark-up process or in conference committee.
The Data Quality Act (DQA) provides an example of the second method of
gaining legislative action and successful implementation.9 The DQA began as
a two-paragraph rider to the Consolidated Appropriations Act 2001 (Pub. L.
106-554). The pharmaceutical, pesticide, tobacco, and energy industries shared
the goal of reducing the ability of the regulatory agencies to deny licenses to
new products, to limit the uses of a product, and to remove a product from the
market. CTAs in these industries hired Dr. James Tozzi to accomplish this goal.
Tozzi, who has a B.S. in chemical engineering and a Ph.D. in economics,
became a major player in policy analysis when President Ronald Reagan made
230
him the first director of OIRA, a department within the White House Office
of Management and Budget (OMB). One of Tozzi’s favorite lobbying tactics
is to add a rider to legislation. Once he crafts the language necessary to make
the desired policy change, Tozzi finds a legislator who will insert a seemingly
innocuous rider into a lengthy and difficult-to-veto bill. Tozzi is so effective in
using this tactic that his fellow lobbyists nicknamed him ‘Stealth.’10
At the request of President Reagan, Congress passed the 1980 Paperwork
Reduction Act.11 The ostensible purpose of the Act was to reduce the paperwork
burden federal regulations placed on private businesses. Its actual goal was to make
it more difficult to pass new regulations. Using the authority given him by the
Paperwork Reduction Act, Reagan issued Executive Order 12088. This order
established OIRA and gave it the power to force agencies to prepare a benefit-cost
analysis of all economically significant rules. President Reagan appointed Tozzi as
the first Administrator of OIRA because of his anti-regulation reputation. Tozzi
staffed OIRA with economists who viewed regulations as inferior to market
mechanisms.12 OIRA became the most important weapon that presidents could
use to kill or weaken regulations they opposed.
OIRA’s major evaluation tool is benefit-cost analysis. Unless exempted by
congressional action, any agency’s proposed regulation that has a total impact on
the economy of greater than $100 million falls under OIRA’s authority. OIRA
can require agencies to demonstrate that the benefits of a proposed regulation
outweigh its costs. If OIRA determines that the agency’s analytic model or the
data used in that model contain errors, OIRA can require the agency to revise
the analysis or to withdraw the proposed rule.13
Pharmaceutical and chemical industries have objected to what they considered
unfair practices concerning the evidence regulatory agencies use in their decisions
to license a product or to limit its use. To license a new pesticide or pharmaceutical,
the manufacturer must conduct numerous experimental and epidemiological
studies that demonstrate the product’s safety and effectiveness. The regulatory
agencies decide the types and numbers of tests the producer must fund; the
number of experimental subjects required in tissue, animal, and human tests; the
statistical significance the studies must achieve; and other safeguards to prevent
harmful products from reaching the market. The producer submits the results to
the agency. This process is expensive and often takes years.
Although the EPA and the Food and Drug Administration required the chemical
and pharmaceutical manufacturers to follow rigorous testing procedures, the
agencies did not have similar requirements for studies they used to ban a product
or to restrict its use. This meant that while producers must have multiple peer-
reviewed studies with hundreds of subjects, an agency could delay or prevent
the licensing or reauthorization of a product using a single less rigorous study.
To redress this seeming double standard,14 Tozzi designed a two-paragraph rider
to insert into the 712-page omnibus appropriations bill. The rider directed the
OMB to issue general guidelines ensuring and maximizing the quality, objectivity,
utility, and integrity of information disseminated by federal agencies. The rider
231
also required that agencies develop standardized data guidelines applicable to all
parties when making regulatory decisions. Tozzi convinced Rep. Jo Ann Emerson
(R-MO) to slip the paragraphs into the appropriations bill during House mark-
up, and Senator Richard Shelby (R-AL) agreed to ensure that the rider remained
in the Senate version of the bill. Congress considered the appropriations bill
late in the congressional session, and no legislators other than Emerson and
Shelby appear to have been aware of the rider’s implications. The rider became
the DQA. When Congress passed the appropriations bill and President Clinton
signed it, everyone who might have opposed it was unaware of the rider and its
implications. Yet the rider and its implementation resulted in the most significant
change in regulatory politics since 1980.
The key provision of the DQA was that OMB would establish general data
quality standards that apply to all data an agency uses in its decisions. Tozzi knew
that the task of establishing those standards would fall to OIRA. Given OIRA’s
anti-regulatory bias, Tozzi expected that the implementation guidelines would
make it more difficult for agencies to keep products off the market or to restrict
their use. Tozzi succeeded beyond his wildest dreams.
Notice the steps that political policy analysis involves. It identifies a policy goal,
the political steps to achieve the institutional changes necessary to achieve that
goal, and the best way to implement those changes. In the case of the DQA, the
institutional changes were massive.
John Graham, President G. W. Bush’s Administrator of OIRA, interpreted the
DQA as requiring every federal department or agency engaging in regulation
to draft guidelines to ensure data quality and to submit the guidelines to OIRA
for approval. The guidelines that OIRA approved for the EPA and the Food
and Drug Administration (FDA) reduced substantially the types of studies and
data the agencies could use. Graham also required all departments and agencies
involved in regulation to develop a specific test for each harmful outcome that
might lead to the denial of a product’s registration or to a restriction on its use.
This meant that if an agency discovered that a product harmed people or the environment
but had not developed a specific test for measuring that harm, the agency could not use that
information. The disallowed harms represent a major unmeasured cost of the DQA.
Graham’s interpretation of the DQA raised substantially the cost of preventing
a product’s registration or renewal. Rather than leveling the playing field between
producers and advocates for consumer and worker safety, the DQA slanted the
playing field in favor of producers. Producers have the monetary incentive and
resources to produce the data an agency requires; regulatory agencies and most
opponents of the products do not have the same incentives and resources.
When an agency’s decision harms an organization and that organization believes
that the agency did not follow its OIRA-approved guidelines, the organization
can file a complaint with the White House. The agency must respond to the
complaint in writing. OIRA then decides whether the agency followed its
own guidelines. In the first two years after OIRA approved the EPA and FDA
guidelines, 32 industry groups used the DQA to prevent regulatory agencies
232
from restricting their products. Seven citizen action groups attempted to use the
DQA to strengthen a regulation (Weiss 2004).
Opponents of the DQA hoped that President Obama would name a director of
OIRA who would restrict the scope of the Act. The opposite occurred. President
Obama appointed Cass Sunstein, a professor of law at the University of Chicago,
to head OIRA. Sunstein expanded the scope of the DQA by interpreting the
Act as allowing OIRA to review not only pending regulatory decisions, but also
all past decisions. This interpretation allows OIRA to re-examine any significant
regulation that a president opposes. Sunstein’s interpretation provides part of
the justification for President Trump’s executive order requiring any agency to
remove two regulations for each new regulation it proposes (Davenport 2017).
President Trump named Naomi Rao as the new Administrator of OIRA. Rao
strongly supports reducing regulations, and environmental interest groups expect
that she will roll back as many of President Obama’s environmental regulations
as possible.15
The regulation of perchlorate: collaboration between private and government policy

analysts
Our interviews with lobbyists for Fortune 1000 corporations discovered numerous
instances of government officials using policy analysis to advocate on behalf of
interest organizations. One such case was the regulation of perchlorate. For more
than 30 years, the EPA has attempted to reduce the allowable levels of perchlorate
in drinking water.
Perchlorate, a component in rocket fuels and medical tracers, has two undesirable
characteristics. High levels of exposure to the chemical reduce the ability of
the thyroid gland to metabolize iodine, and an iodine deficiency can cause
developmental disorders in fetuses and goiters in adults. Perchlorate is extremely
persistent in the environment. This increases the toxic exposure of individuals.
Reducing the allowable level of perchlorate would raise the production costs of
corporations in the defense, energy, and aerospace industries and would affect
substantially the work of the Department of Defense (DoD), the National
Aeronautics and Space Administration (NASA), and the DoE.
The Safe Drinking Water Act 1974 (SDWA) requires the EPA to determine the
safe level of exposure to toxic chemicals. That level is the oral reference dose (RfD);
it serves as the baseline for all federal laws regulating a chemical’s production, use,
and cleanup. States may apply more stringent standards than the EPA, but most states
do not. A change in a RfD can increase manufacturing and cleanup expenses by
millions of dollars and can drive some producers out of business. CTAs, therefore,
lobby heavily on RfD decisions. The EPA proposed a RfD for perchlorate of one
part per billion (1-ppb), the equivalent of one grain of salt in an Olympic-size
swimming pool.
Four private companies reacted immediately. Aerojet, American Pacific
Corporation, Kerr-McGee Chemical, and Lockheed Martin formed the Perchlorate
233
Study Group (PSG) to lobby against the proposed standard. The PSG worked
closely with the DoD and NASA to gather technical and safety information
concerning perchlorate. The policy analysis by the PSG demonstrated that neither
the producers of perchlorate nor the EPA could measure perchlorate at the one
ppb level. This meant that EPA could not test its proposed RfD. The PSG analysis
also showed that there were no data indicating that the existing 200 ppb standard
harmed humans or animals. Presidents Reagan and George H. W. Bush pressured
the EPA not to pursue the one ppb standard.
After the election of President Clinton, the EPA again proposed a perchlorate
RfD of 1-ppb. This standard would require the DoD and NASA to spend more
than $40 billion to clean up only the Colorado River Basin. The total costs of
the 1-ppb standard would exceed $100 billion. In response to the EPA proposal,
the DoD spent $30 million on research that demonstrated that perchlorate did
not pose a significant health risk, and it replicated the PSG study indicating that a
RfD of 200 ppb was sufficient to protect public health. After examining the data
submitted by the EPA and the DoD, OIRA opposed the new standard. President
Clinton, however, did not accept OIRA’s finding. He asked the National Academy
of Sciences (NAS) to supervise a study concerning the safe level of perchlorate. The
Academy concluded that the RfD should be no lower than 24.5 ppb, and cited the
exceptional cost of reducing perchlorate below that level. The EPA chose not to
pursue a more stringent RfD.
In response to the EPA’s inaction, California and Massachusetts set their state
standard for perchlorate at 5 ppb. This placed perchlorate producers and users in
those states at a competitive disadvantage. The same firms would have a competitive
advantage, however, if the EPA were to issue a RfD of 5 ppb. In response to pressure
from the legislators from California and Massachusetts, the EPA proposed a 5-ppb
standard. The manufacturers and users of perchlorate outside of California and
Massachusetts used the DQA to challenge EPA. The challenge argued that the
5-ppb standard would reduce competition within the industry and that the data
the EPA used in its benefit-cost and risk analyses did not meet the EPA’s DQA
guidelines. The EPA was unable to refute these challenges, and OIRA ruled that
the EPA had violated its DQA requirements. In 2015 the EPA issued a Health
Hazard Alert for perchlorate (U.S. Environmental Protection Agency 2015), but
it did not propose a new RfD.
The Keystone pipeline and perchlorate issues demonstrate three important
aspects of policy analysis by CTAs. First, successful CTAs often have advocates
within the government. While perchlorate producers spent almost $10 million
on policy analyses and lobbying, the DoD spent over $30 million. Second, high
government officials have better access to the White House than private sector
lobbyists. A crucial period in the rule-making process is when an agency sends a
draft of its proposed rule to other government departments that the rule affects. The
agency also sends the draft to OIRA. The White House encourages an interagency
review of this draft and expects the departments and agencies to work out any
disagreements before the agency publishes its proposed rule in The Federal Register.
234
As the example of perchlorate illustrates, government officials’ access to OIRA and

the White House during the interagency review can be critical in determining the
specific provisions in a proposed rule. Third, even if a benefit-cost analysis were to
result in a more stringent rule, requests for stricter regulatory changes that reduce
competition face tough sledding within the Executive Branch.16
Voluntary EMSs and policy analysis
After the adoption of the United Nations Framework Convention on Climate

Change in 1992, the European Union (EU) chose to use a voluntary approach to
encourage individual corporations to adopt an EMS. Having the same mandatory
standards in all EU countries would be difficult as the countries in the EU have vastly
different resources, technologies, and commitments to reducing environmental
damage. In addition, different types of firms face different environmental issues.
The EMS voluntary system is known as the ISO 14000 family of standards. The
major objective of the ISO 14000 series is to promote a set of norms that encourage
effective and efficient environmental management and to provide tools that are
cost-effective, system-based, flexible, and reflect the best organizational practices
available for gathering, interpreting, and communicating environmentally relevant
information (ISO (International Organization for Standardization) 2016).
The Kyoto Protocol in 1997 made Northern Europe the major region of the
world that agreed to mandatory greenhouse gas reductions. The countries in
this region believed that their efforts to reduce greenhouse gases put them at a
competitive disadvantage to other countries, particularly the United States. The
EU responded by extending its ISO 14000 program to include certification of a
firm’s EMS and an audit of the performance of the firm in achieving environmental
sustainability. A key aspect of these additions was that firms must evaluate the
environmental conduct of their suppliers. This provision forced many multinational
firms and suppliers based in the United States to adopt their own EMS and to
have ISO certify it.
The procedures for developing an EMS, having it certified, and preparing for
an external audit read much like a textbook in policy analysis. The EMS must
create a causal model of the firm’s production process. The firm must conduct
a benefit-cost analysis and risk assessment of each component of that process,
demonstrate the technical expertise to implement the EMS, show that the EMS
does not conflict with national and local environmental regulations, indicate how
the firm will measure the results of using its EMS, and make clear how the firm
and external auditors can evaluate the firm’s performance (ISO 2016).
The ISO in Geneva, Switzerland develops EMS standards. The ISO issues revised
standards every five years and has developed EMSs appropriate to service as well as
manufacturing firms. The ISO created a certification process for EMSs, and audit
procedures to measure a firm’s progress toward environmentally sustainable practices.
By 2015, the ISO 14000 family of voluntary rules, certifications, and audits had
spread to more than 300,000 corporations in 171 countries (ISO 2015, 2016).
235
Why would a firm voluntarily adopt an EMS, pay for its certification, and submit
to external auditing of the firm’s production process? The cost of voluntarily
complying with the ISO standards is a multimillion-dollar expenditure for large
firms. The cost to smaller firms that lack the in-house expertise necessary to perform
the above analyses is proportionately much greater (Johnstone and Labonne 2009).
Businesses develop EMSs for several reasons, regulatory pressure being just one.
As governmental agencies develop and enforce environmental regulations, businesses
have incentives to establish for themselves clear standard operating procedures that
detail how their manufacturing procedures are consistent with environmental
regulations. Indeed, there is evidence that an EMS can stave off legal action and
can bolster a firm’s defense if governments take legal action against it (Carr and
Thomas 1997; King et al. 2005). Of course, governments are not the only audiences
or stakeholders for businesses. Consumers and investors also have expressed interest
in EMS, so a business might establish an EMS to signal limited liability regarding
legal actions to investors and environmental sensitivity to concerned consumers
(for example, Anderson 2002). A principal reason for firms adopting ISO 14000
standards outside of the EU is to gain access to the EU market. In other words,
the use of ISO 14000 standards by the EU cuts down the number of free riders
who lower production costs by using less environmentally sound business practices.
Standardized EMSs allow regulators within a country to use a common measure
when considering which businesses are making the greatest strides toward cleaner
manufacturing processes.17 During the development of an EMS, a manufacturer is
likely to obtain a greater awareness of its own operational procedures (for example,
Melnyk et al. 2003; Rondinelli and Vastag 2000; Bansal and Roth 2000). The in-
depth analysis of the entire manufacturing processes creates an overall awareness
of inefficiencies in manufacturing procedures. Efficiency gains are made across the
board, not just with environmental practices.
Eco-labeling, which comes from certification, can appeal to consumers
interested in purchasing goods that have a positive impact on the environment. The
environmental friendliness of a production product is an example of a credence
good. Credence goods have a characteristic that is not immediately apparent to
the consumer, even though it is still important to the consumer. For instance, a
consumer might prefer to purchase an industrial battery manufactured with little
detriment to the environment. But a battery produced by a manufacturer with an
EMS looks like the one produced by a manufacturer with no EMS. Certification
provides the consumer confidence by bestowing credibility on the manufacturer
with the EMS.
In summary, although the adoption of an EMS is voluntary, the policy analysis
inherent in EMSs can guard against legal liabilities, enhance internal production
efficiencies, and create comparative advantages in the marketplace due to eco-
labeling. More importantly, the in-depth knowledge of manufacturing standards
related to certified EMSs allows businesses to speak with authority to government
officials concerned about environmental degradation and provides those businesses
access to markets that otherwise would be barred to them.
236
Regulatory waivers
Books concerning policy analysis rarely discuss regulatory waivers, but obtaining
waivers constitutes a major aspect of corporate policy analysis. Uniform application
of regulations increases predictability for the regulated. Uniformity, however,
does not enhance equity or efficiency when circumstances make a regulation
unreasonable or harmful. Waivers can provide an important safety valve in the
administrative process. They are particularly important in the regulation of
industries characterized by rapid innovation.
All agencies have standard operating procedures (SOPs) for granting waivers.
The three major justifications for granting a waiver are economic hardship,
competition, and innovation. A waiver application must show that granting
the waiver will not harm the public or reduce competition. Corporate policy
analysts work with their firm’s attorneys to prepare waiver applications that fit
each agency’s guidelines and SOPs. Analysts within the agency then evaluate the
benefits and costs of the waiver.
To take a hypothetical example, the EPA might require that micro-particulate
emissions from power plants not exceed a certain level by the year 2025. Duke
Energy might request a two-year waiver for an older plant while the company
completes the construction of its replacement. Duke Energy would provide a
benefit-cost analysis and risk assessment to the regulating agencies. This analysis
might show that not granting the waiver would increase costs to consumers and
risk supply interruptions. Duke also would provide health-related data to indicate
that allowing the existing plant to operate for two additional years would not
pose a significant risk to public health.
Agencies often grant waivers when a regulation would reduce the competition
in an industry by forcing firms to shut down or by inhibiting the entry of new
firms. Rossi (1995) gives examples of these situations in his case study of the
cogeneration of energy using green technologies. DoE created its rules to
regulate large power plants. These rules were not appropriate for smaller, green
technologies. Recognizing this, the DoE granted numerous waivers to the green
energy firms while the DoE developed rules appropriate to the new technologies.
The SOPs of every agency require that it consider the past behavior of an
organization when deciding its request for a waiver. If a corporation has a history
of fraud or other bad behavior, this should make it significantly more difficult for
the company to receive a waiver. This may not occur, however, if firms in the
regulated industry have undue influence on the regulating agency. Edward Wyatt
of the New York Times chronicled some of the more egregious examples of an
agency not following its own guidelines. From 2008 to 2011, the Securities and
Exchange Commission (SEC) granted more than 200 waivers to Goldman Sachs,
Bank of America, Wells Fargo, and JPMorganChase. The waivers provided the
four banks substantial competitive advantages over other banks and permitted them
to: (a) bypass the SEC in raising money and managing mutual funds, (b) avoid
millions of dollars in fines, and (c) have extensive protection from class action suits.
237
David Ruder, a previous SEC chairperson, stated that these waivers were worth
hundreds of millions of dollars to each of the banks.18 The SEC granted these
waivers despite having prosecuted all four corporations for fraudulent activities
that helped to create the 2008 financial crisis (Wyatt 2012).
In summary, waivers are important policy tools for regulating agencies. Their
use can encourage new technology and greater competition, and reduce economic
hardships. The possible availability of a waiver provides corporations opportunities
for legal and political analyses that may lead to impressive gains for corporations.
Studies of lobbying and policy analysis, however, rarely examine the granting or
denying of waivers and the policy analyses that affect these outcomes.
The effects of major policy shifts on policy analysis

We saw in our case studies that a change of administrations, a change in public
opinion, or an experienced political entrepreneur can lead to dramatic policy
shifts. These shifts often determine policy feasibility and direct CTA lobbyists
and policy analysts toward different goals and strategies. At the suggestion of
an external reviewer, we use as an example of a major policy shift the changes
in environmental policy options that have occurred since the inauguration of
Donald Trump. Candidate Trump and President Trump claimed that climate
change was a hoax. He vowed to eliminate or reduce environmental policies
passed in previous administrations. President Trump fulfilled this promise with
a series of executive orders, political appointments, and proposed budget cuts.
Trump’s actions reverse more than two decades of presidential acknowledgment
of the reality of climate change and the need for policies to reduce the impacts
of U.S. actions that increase global warming.19
In multiple elections the Republican Party has argued that regulations curtail
increases in productivity and economic growth. Scholarly research indicates
that environmental regulations may have costs equal to ten% of the national
government’s budget (Jorgenson and Wilcoxen 1990; Broughel 2017). Donald
Trump, before and after his election, adopted an anti-regulatory perspective
and pursued it more actively than any other president in history (Restuccia and
Cook 2017). The executive orders issued in his first three months in office make
clear the president’s commitment to deregulation. Executive Order 13766, for
example, requires that when a department or agency proposes a new regulation,
it must identify two regulations that it will eliminate. The order does not specify
how agencies will accomplish this or whether eliminating an existing regulation
will require benefit-cost analyses and risk assessments.
According to Andrew Bremberg, director of the White House’s Domestic
Policy Council, the goal of the administration is systemic reform that lasts well
beyond Trump’s presidency. Highest on the agenda is a rollback of President
Obama’s climate change regulations. Trump also signed off on congressional
actions that used a seldom-invoked 1996 law to block four energy-related rules.
Trump ordered the EPA to review—and most likely rescind—two sweeping
238
Obama-era rules, one restricting power plants’ greenhouse gas pollution and one
spelling out Clean Water Act protections for wetlands and waterways (Restuccia
and Cook 2017).
State governments also have worked to reduce regulations. In a paper written
for the American Enterprise Institute in 1998, Oxford University economist
Robert Hahn found that only 10 states required a benefit-cost analysis of proposed
regulations. This situation has changed radically. Currently, almost all states require
such analyses. Rhode Island, for example, had no office to review regulations or
their impact in 1998. There were no requirements for benefit-cost analysis, risk
assessment, or similar policy analysis tools. Today the state has a comprehensive
system that requires the Office of Regulatory Review (ORR) to determine if
a regulation proposed by the legislature, a department, or an agency meets the
general requirements OIRA requires of proposed federal regulations (Hahn 1988,
pp 327–8). The Democratic governor directed ORR to decrease regulations by
15% during 2017.
At the time we are writing this chapter, we cannot predict exactly how these
political changes will affect policy analysis by CTAs. We expect, however, that
the level of policy analysis will increase substantially. Just as a major tax reform
initiates a feeding frenzy by corporate lobbyists to ensure that the reform will
advantage their firm (Birnbaum and Murray 1988), major changes in regulatory
rulemaking are likely to increase the policy analyses that support the lobbying
efforts of CTAs. Rather than having to fight new regulations that raise the costs
of production, however, corporate policy analysts will concentrate on writing
proposals that reduce those costs while simultaneously giving a competitive
advantage to their company. If President Trump is successful in reducing the
budgets of regulatory agencies, the agencies will become increasingly dependent
on policy analyses from interest organizations. As we indicated at the beginning
of this chapter, business interests are in the best position to supply the analyses
that departments and agencies need.
Summary
Corporations have substantial expertise in policy analysis as the tools required are
like those that corporations use in their normal investment and decision-making
processes. More important, investments in policy analysis and lobbying often
increase profits more than alternative investments. Trade associations use policy
analysis to prevent undesirable policy changes and to provide their members
predictability concerning new and upcoming policy changes.
Our interviews with lobbyists for CTAs indicate that policy analysis is critical
to lobbying. Studies of the rule-making process demonstrate that business interests
have a substantial comparative advantage over other private sector organizations.
This advantage stems from CTAs’ greater use of policy analysis to generate
technical information and to evaluate the data that the agencies use in developing
their proposed rules. The case study of the DQA indicates the importance of
239
political policy analysis. Analyzing the SOPs of an agency can provide CTAs with
the information they need to change the process of policy making. The growing
importance of regulatory waivers reveals that economic and technical analyses
often require integration with legal analysis. Our case studies demonstrate the
importance of having advocates within the government. Finally, our case study of
ISO 14000 voluntary environmental actions indicates that policy analysis can be
equally important in these situations as it is in command and control regulatory
policies.
Our assessment of policy analysis by CTAs leads to three recommendations.
First, policy analysts should gain a better understanding of the institutions, norms,
and SOPs that constrain the policy-making process. This would provide analysts
insights concerning where their analytic tools can make the greatest impact. This
also would furnish policy scholars insights concerning those aspects of the policy
process that political science currently neglects. Second, we recommend that
policy scholars give greater attention to the relationship between policy analysis
and lobbying. Understanding this relationship would produce important insights
into the advantages that business organizations have over other interests. This
understanding would move scholars beyond comparing the success of various
organizations in achieving changes in proposed rules and would encourage
them to examine the relationship between policy analysis and policy making
by Congress and the White House. Finally, research in policy analysis and the
policy process should pay greater attention to regulatory waivers. Governments
grant tens of thousands of waivers each year that are worth billions of dollars.
Understanding the role of policy analysis in this process would not only provide
a better understanding of the impact of waivers on the economy, but also would
help us evaluate the quality of American democracy.
Notes
1 The authors gratefully acknowledge the support from National Science Foundation grants
SES-0752245 and SES-0752212. The analyses and conclusions in this chapter do not reflect
the position of the National Science Foundation or the institutions with which the authors are
affiliated.
2 Godwin et al. (2013, pp 96–100) provides the sampling procedures, response rates, and protocols
used in these interviews.
3 There are exceptions to this generalization. Our analysis of regulatory comments showed that
the Natural Resources Defense Council is highly active and effective in the rule-making process.
4 Erik Godwin is Director of Regulatory Reform for the State of Rhode Island. He previously
served as a lobbyist for the EOP Group, a Washington, DC firm that prepares policy analyses to
support corporations lobbying federal agencies. He served in the White House as a policy analyst
in the Office of Information and Regulatory Affairs (OIRA). Kenneth Godwin worked as a
policy analyst for corporations, government agencies, and think tanks, including Westinghouse
Research, the Battelle Memorial Institute, USAID, the U.S. Department of Education, the
University of North Carolina Center for Civil Rights, and Resources for the Future. Scott
Ainsworth specializes in formal models of lobbying and policy analysis. His most recent work
includes the executive and legislative branches of government.
5 Not all regulations must have net social benefits. Congress exempts certain laws. The Endangered
Species Act and the Americans with Disabilities Act have such exemptions.
240
6 For the statutes and executive orders that created these requirements, see Godwin et al. (2013,
pp 83–84).
7 We chose these three agencies because they account for a clear majority of economically
significant rules.
8 We measured the quality of a comment by its inclusion of references to legislation, agency
decisions and data, or court cases. We also measured whether the comment used new policy
analysis results and data. We measured the ‘idea density’ of a comment using software developed
at the University of Georgia.
9 The DQA is also known as the Information Quality Act.
10 http://en.wikipedia.org/wiki/Jim_Tozzi.
11 Pub. L. No. 96-511, 94 Stat. 2812, codified at 44 U.S.C. §§ 3501–3521.
12 Tozzi’s perspective on regulation is available at www.thecre.com/regreview/index.html.
13 Agencies can request that the White House overrule OIRA.
14 This double standard is appropriate. When allowing the introduction of a potentially dangerous
product, an agency’s first obligation is to protect public safety.
15 Telephone interview with the clean air lobbyist for the Natural Resources Defense Council.
16 It is likely that the proposed rule that an agency such as the EPA or the FDA puts forth is as
strict as the agency believes it can achieve.
17 ISO is currently the most prominent organization addressing EMSs. Its membership is comprised
of national associations focused on standards. In the United States, the ANSI (American National
Standards Institute) is the most prominent body. ANSI was established in the early 1900s by
engineering societies that sought clearer standards that could enhance efficiencies. The ISO
14001 principles are now the most common standards for EMSs.
18 For an example of a waiver application to the SEC by JPMorganChase and the SEC response,
see www.nytimes.com/interactive/2012/02/03/business/art-of-filing-sec-waivers-document.
html?ref=business.
19 When campaigning in 1988, presidential candidate George H. W. Bush acknowledged that
global warming was a serious problem. On the campaign trail, Bush vowed to fight the
greenhouse effect with a ‘White House effect.’ Early in the Bush administration, the president
chose to emphasize the need for more study before taking major policy actions. By the end of
his administration, however, the administration initiated policies to reduce greenhouse gases.
The Obama administration made reducing greenhouse gases a major administrative goal and
encouraged the EPA and the Department of Energy to adopt the regulations necessary to
achieve this.
References
Anderson, R. (2002) ‘Incentive-Based Policies for Environmental Management
in Developing Countries’, Washington, DC: Resources for the Future Issue
Brief, August).
Bansal, P. and K. Roth (2000) ‘Why Do Companies Go Green: A Model of
Ecological Responsiveness’, The Academy of Management Journal 13(4): 717–36.
Baron, D. (2013) Business and Its Environment (9th edition), Upper Saddle River,
NJ: Prentice Hall.
Baumgartner, F., J. Berry, M. Hojnacki, D. Kimball, and B. Leech (2009) Lobbying
and Policy Change: Who Wins, Who Loses, and Why, Chicago, IL: University of
Chicago Press.
Birnbaum, J. and A. Murray (1988) Showdown at Gucci Gulch: Lawmakers, Lobbyists,
and the Unlikely Triumph of Tax Reform, New York: Vintage.
241
Broughel, J. (2017) Regulation and Economic Growth: Applying Economic Theory to

Public Policy, Harrisburg, VA: Mercatus Center at George Mason University.
Carr, D. and W. Thomas (1997) ‘Devising a Compliance Strategy under the ISO
14000 International Environmental Management Standards’, Pace Environmental
Law Review 15(1): 85–232.
Chubb, J. E. (1983) Interest Groups and the Bureaucracy: The Politics of Energy,
Stanford, CA: Stanford University Press.
Davenport, Coral (2017) ‘Counseled by Industry, Not Staff, EPA Chief is Off
to a Blazing Start’, New York Times, July 1, www.nytimes.com/2017/07/01/
us/politics/trump-epa-chief-pruitt-regulations-climate-change.html?mcubz=0.
Farber, D. A. (2007–2008) ‘Modeling Climate Change and Its Impacts: Law,
Policy, and Science’, Texas Law Review 86: 1655–87.
Fisher, E., P. Pascual, and W. Wagner (2010) Understanding Environmental
Models in Their Legal and Regulatory Context, Journal of Environmental Law
22(2): 250–80.
Godwin, K., S. Ainsworth, and E. Godwin (2013) Lobbying and Policymaking: The
Public Pursuit of Private Interests, Washington, DC: CQ Press.
Hahn, R. (1998) ‘State and Federal Regulatory Reform: A Comparative Analysis’,
Washington, DC: AEI-Brookings Joint Center for Regulatory Politics, http://
localgov.fsu.edu/papers/archive/Hahn_001.pdf.
International Organization for Standardization (2015) www.iso.org/
standard/60857.html.International Organization for Standardization (2016)
ISO/IEC Directives, Part 1, Consolidated ISO Supplement, Geneva.
Johnstone, N. and J. Labonne (2009) ‘Why Do Manufacturing Facilities
Introduce Environmental Management Systems? Improving and/or Signaling
Performance’, Ecological Economics 68(3): 719–30.
Jorgenson, D. and P. Wilcoxen (1990) ‘Environmental Regulation and Economic
Growth’, Rand Journal of Economics 21(2): 314–40.
Kerwin, C. (2003) Rulemaking: How Government Agencies Write Law and Make
Policy (third edition), Washington, DC: CQ Press.
King, A., M. Lenox, and A. Terlaak (2005) ‘The Strategic Use of Decentralized
Institutions: Exploring Certification with the ISO 14001 Management Standard’,
Academy of Management Journal 48(6): 1091–106.
Labott, E. and J. Diamond (2017) ‘Trump Administration Approves the Keystone
XL Pipeline’, www.cnn.com/2017/03/23/politics/keystone-xl-pipeline-
trump-approve/index.html.
McFarland, A. (2004) Neopluralism: The Evolution of Political Process Theory,
Lawrence, KS: University Press of Kansas.
McGuffie, K. and A. Henderson-Sellers (2005) A Climate Change Modelling Primer
(3rd edition) New York: John Wiley.
Melnyk, S., R. Sroufe, and R. Calantone (2003) ‘Assessing the Impact of
Environmental Management Systems on Corporate and Environmental
Performance’, Journal of Operations Management 21: 329–51.
242
NRDC (Natural Resources Defense Council) (2012) Climate Impacts of the

Keystone XL Tar Sands Pipeline, www.nrdc.org/energy/keystone-pipeline/tar-
sands-climate-impacts.asp.
NRDC (2011) Clean Water Action, and EarthJustice Meeting with OMB re: Perchlorate
Regulatory Determination. RIN: 2040-AF02 / 2040-AF08, January 13, www.
cleanwater.org/files/publications/national/OMB_perchlorate-regulatory-
determination_20110111.pdf.
Prakash, A. and M. Potoski (2006) ‘Racing to the Bottom? Trade, Environmental
Governance, and ISO 14001’, American Journal of Political Science 50(2): 350–64.
Randall, D. et al. (2007) ‘Climate Change Models and their Evaluation’, in the
Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 589–662.
Rondinelli, D. and G. Vastag (2000) ‘Panacea, Common Sense, or Just a Label?
The Value of ISO 14001 Environmental Management Systems’, European
Management Journal 18(5): 499–510.
Rossi, J. (1995) ‘Making Policy through the Waiver of Regulations at the Federal
Energy Regulatory Commission’, Administrative Law Review 47(2): 255–97.
Shapiro, S. and J. Morrall III (2012) ‘The Triumph of Regulatory Politics:
Benefit-cost Analysis and Political Salience’, Regulation and Government 6:
189–206.
Subramanian, V., E. Semenzin, D. Hristozov et al. (2015) ‘Review of Decision
Analytic Tools for Sustainable Nanotechnology’, Environmental Decision Systems
35: 29–41.
U.S. Environmental Protection Agency (2015) ‘Chemical Assessment Summary:
Perchlorate and Perchlorate Salts’, Washington, DC: Integrated Risk Information
System, http://cfpub.epa.gov/ncea/iris2/chemicalLanding.cfm?substance_
nmbr=1007.
VanWyhe, K. (2017) ‘Executive Orders and Presidential Memorandums
on Energy’, www.americanactionforum.org/research/executive-orders-
presidential-memorandums-energy/#ixzz4moNl4IaM.
Weiss, R. (2004) ‘“Data Quality” Law is Nemesis of Regulation’, Washington
Post, August 16, p A01.
Wyatt, E. (2012) ‘S.E.C. Is Avoiding Tough Sanctions for Large Banks’, New York
Times, February 3, www.nytimes.com/2012/02/03/business/sec-is-avoiding-
tough-sanctions-for-large-banks.html?_r=1&pagewanted=print.
243
THIRTEEN
Policy analysis and the nonprofit sector

Steven Rathgeb Smith1
The nonprofit sector in the United States is increasingly important and very
diverse; however, the largest component of the nonprofit sector in the United
States is public charities registered as 501(c)(3) tax-exempt organizations. Many
of these organizations provide public services to the citizenry with public funds,
advocate for important policy goals, and offer an opportunity for volunteer
and community engagement. In addition, private foundations and federated
fundraising organizations such as the United Way provide funding to countless
nonprofit organizations for their programs and services. A much smaller segment
of the nonprofit sector is 501(c)(4) organizations or ‘social welfare’ organizations
concerned with the broader public good. This chapter focuses on the use of
policy analysis by these two types of nonprofit organizations, offering services
to local communities and participating in the broader public arena, as well as
philanthropic institutions such as foundations and other philanthropic funding
organizations. In this chapter I will employ the broad definition of policy analysis
offered by Weimer (this volume): ‘professionally provided advice relevant to
public decisions and informed by social values.’ Using this definition, nonprofits
have used policy analysis since at least the late 19th century; however, the use of
policy analysis and related tools in the nonprofit sector has grown substantially
in recent years, including in the context of evaluation and strategic planning by
nonprofits. Moreover, the austere and turbulent fiscal environment for nonprofits
also encourages nonprofits to utilize policy analysis to advocate for their programs
and develop long-term plans for sustainability and improved performance.
Mapping the nonprofit sector

The nonprofit sector dates to the colonial period when many prominent
universities and hospitals were established. The sector grew in the 19th century
and early 20th century but many nonprofits were relatively small and primarily
dependent on private philanthropy, with some exceptions, especially in major
urban areas. The growth of the contemporary nonprofit sector dates to the 1960s
when the federal government launched multiple initiatives that relied upon the
nonprofit sector to provide local community services. Subsequently, the sector
has grown due to the continued expansion of government, growing demand
for nonprofit services such as community care, a shift in funding and service
responsibility from government to the nonprofit sector, and growing wealth
leading to many new foundations. Thus, the number of 501(c)(3) nonprofits
245
grew from approximately 150,000 in 1965 to over 500,000 organizations in 1995.

Since then, the number of nonprofit 501(c)(3) organizations registered with the
Internal Revenue Service (IRS) has more than doubled to 1.1 million (NCCS
2016). Private foundations (also 501(c)(3) organizations) also doubled between
1995 and 2016 (NCCS 2016). The increase in the number of nonprofits, though,
masks the relatively small size of many nonprofits. For example, more than half
of the nonprofits have less than $50,000 in expenses. And total expenses are
dominated by large institutions such as hospitals and universities. Many nonprofits
are completely volunteer such as a local chapter of a parent teacher association or
a neighborhood association. Local congregations are not included in these IRS
figures since they are not required to register or report to the IRS (although many
congregations do report because they have a significant social service mission).
Overall, the number of local congregations has remained relatively stable in
recent years at approximately 350,000 (Hartford Institute for Religion Research
2016). Like registered 501(c)(3) organizations, local congregations are exempt
from most taxes and also offer individuals a tax deduction for their philanthropic
contributions. In terms of the distribution of nonprofit 501(c)(3) organizations,
about 35% are human service agencies followed by education (17%), health
(12.9%), and public benefit organizations, which is a diverse category including
organizations with missions focusing on policy areas such as voter education, civil
rights, community improvement, and mutual aid (McKeever 2015). Also, many
of the large research organizations and think tanks such as the Urban Institute
and Mathematica and private nonprofit universities are registered as 501(c)(3)
organizations.
In contrast to the growing number of public charities, 501(c)(4) organizations
have been in an incremental decline for many years, dropping from approximately
120,000 in 1990 to about 85,000 in 2014 (Koulish 2016). Donations to these
organizations are not eligible for a tax deduction but IRS rules permit much
greater flexibility in terms of lobbying and advocacy as long as it is in support of the
public service mission of the organization. Recently, controversy has surrounded
these organizations because of the concern that the IRS was unfairly scrutinizing
new applications for 501(c)(4) status from conservative groups who had been
spurred to create new organizations in part by the 2010 ruling by the Supreme
Court in the Citizens United decision. The Court based its ruling on the First
Amendment right of freedom of association. Then, the Court also argued that
corporations are associations and hence entitled to spend unlimited amounts on
political ‘speech.’ The ripple effects on 501(c)(4) organizations has been profound
because these organizations can spend money on lobbying and electioneering
so long as it is not their principal activity. Further, these organizations can keep
their donors confidential. Thus, giving money to a 501(c)(4) organization is a
very attractive strategy to influence elections and public policy without the risk
of public scrutiny (also, see Sullivan 2013).
This controversy, though, has tended to emphasize a relatively small percentage
of the organizations in this tax category. For example, the most numerous type
246
of organization in this category is community service clubs such as the Lions

Club and the Rotary, with roots in the early 20th century (Koulish 2013). Yet,
the greater flexibility on advocacy and lobbying means that many large national
organizations such as the Sierra Club or the National Abortion Rights Action
League (NARAL) are under this tax category (or have a hybrid structure with
501(c)(3) and (c)(4) entities). These organizations also are heavily involved
in advocating for their positions in the policy process. In this sense, these
organizations could be regarded as interest groups that represent the goals and
priorities of their members in the policy process. And many of these organizations
employ policy analysts, produce significant reports based upon policy analysis,
and disseminate policy analysis to further their positions.
Historical perspective on the role of policy analysis in the

nonprofit sector
The nonprofit sector in the United States dates to the colonial period when
universities, hospitals, and service agencies were established as public charities. A
flowering of voluntary organizations occurred in the early to mid 19th century,
with many social reform associations established, especially in connection with
the abolitionist movement. However, the use of data, evidence, and policy analysis
in the nonprofit sector dates to the late 19th and early 20th centuries, with the
creation of large national foundations such as the Carnegie Corporation, the
Russell Sage Foundation, and the Rockefeller Foundation. During this period,
important national charities such as the Charity Organization Society (COS) were
founded and emerged as important sources of information on poverty and related
issues. Reformers such as Amos Warner tried to empirically assess the causes of
poverty and recommend concrete solutions to policymakers (Warner 1894; Leiby
1963). Progressive reformers also worked through local nonprofit Municipal
Research Bureaus to push for change in local government administration to
reduce corruption and favoritism. A multiplicity of other nonprofit organizations,
such as the National Committee on Child Labor, provided detailed analysis and
reports to Congress and state and local governments on important social problems.
Rockefeller, for example, supported extensive studies of public education,
especially in the South, in order to reform and improve public education (Walters
and Bowman 2010).
Many leading foundations of this period also provided support for these
nonprofit reform organizations, although the focus of their efforts shifted during
these early decades. Foundations established in the 19th century tended to focus
on direct aid, often to needy individuals, children, and families; it was mission-
based philanthropy based upon their social values. But the early 20th century
foundations, especially Carnegie, Rockefeller, Rosenwald, and Russell Sage,
introduced markedly different funding strategies and priorities. Direct donations
to individuals greatly decreased and efforts to leverage grants to improve future
conditions rose. Thus, foundations increasingly focused on project-related grants
247
and made fewer small grants. And many foundations had clear expectations
for outcomes including new organizations (such as libraries), expectations on
local matching support, and specific reports analyzing various social conditions.
Russell Sage, for example, supported a new social work journal which offered
information and data on social welfare conditions including the plight of poor
children. (They also had an explicit policy prohibiting individual relief.) As the role
of government expanded (primarily at the state and local levels), foundations also
strived to distinguish their efforts from government in order to avoid duplication
(Bielefeld and Chu 2010).
Foundation giving during these formative decades of the 20th century was also
mission driven so that the policy analysis that was undertaken was not dispassionate,
value-neutral evaluation and reports, but instead was underpinned by a specific
agenda of reform and change. Moreover, the theme of many foundation-supported
efforts was that government’s role needed to expand, especially in the fields of
social and health policy; thus, foundations supported analysis and evaluation
that would demonstrate the need for a more active government role. Through
their research and reports, the foundations hoped to influence policymakers and
opinion leaders to substantially change public policy.
Importantly, organized philanthropy was also changed in these early decades
by the establishment of chapters of the Community Chest, the forerunner of the
present-day United Way. The Community Chest movement started in Cleveland
and quickly spread across the country, thus hundreds of cities and towns around
the country had Community Chest chapters by the end of the 1920s. Each
chapter organized the important nonprofit agencies in their towns and conducted
a collective fundraising campaign every year; agencies were generally prohibited
from undertaking any additional fundraising outside the Community Chest
campaign. After the local fundraising campaign concluded every year, volunteer
community members, in concert with Community Chest staff, would review and
analyze the needs of the community and award allocations based upon need. Only
member agencies could receive funds and most of the money was raised through
workplace giving campaigns. Overall, then, the Community Chest represented
an effort to employ the latest needs assessment to local social problems and use
this information to guide funding decisions. In practice, local agencies tended to
get an allocation each year with the amount dependent upon the success of the
funding campaign. Agencies were still required to submit substantial amounts of
information to the local chapter review committees but this generally did not
influence funding decisions (see Hansan 2013; Brilliant 1990).
The Depression had a profound effect on the nonprofit and philanthropic
organizations. Many community agencies closed or drastically altered their
mission. Some agencies relied extensively on federal emergency relief funding to
support their operations. The very difficult economic circumstances of so many
individuals and families also meant that many nonprofit organizations focused
on responding to immediate needs rather than more deliberate analysis of local
social conditions. Many local nonprofits also resisted the permanent expansion
248
of government during the 1930s and tried to retain many traditional roles of
nonprofit community organizations such as relief (Hammack 2002). Throughout
the Depression, though, foundations and nonprofit service agencies continued
to be guided by mission-based programming and analysis, informed by their
own social values.
After World War II the economy recovered and the nonprofit and philanthropic
sector grew incrementally during the late 1940s and 1950s. Further, this period
was marked by growing government support for research in nonprofit institutions,
including universities and think tanks such as the Rand Corporation (Price 1965).
(See this volume’s chapter on think tanks by Rich.) Indeed, the development and
professionalization of policy analysis began in earnest during this period, largely
in nonprofit organizations.
The major break in the role of policy analysis within the nonprofit sector started
in the 1960s and thereafter. During the 1960s the federal government embarked
on a sharp expansion of the government’s role, including many new programs
in health care, education, economic development, and social services that relied
on nonprofit organizations for their implementation (Smith and Lipsky 1993).
Many foundations also supported pilot programs that eventually led to new
federal initiatives. For example, the War on Poverty, with its reliance on local
community agencies, was based, at least in part, on the Ford Foundation’s Grey
Areas program, designed to combat urban poverty and decline (Kohler 2007;
Mossberger 2010). In general, many foundations—both large national foundations
and smaller local community foundations—strived to support social reform in a
wide variety of policy areas including race relations, poverty, health care, public
education, and early childhood education.
Importantly, the use of policy analysis and evaluation within organized
philanthropy and local community organizations did not change significantly,
despite the advent of new federal funding and programs. Foundations continued
to provide mostly project-related funding that was also mission driven. Smaller
community foundations tended to provide short-term grants, especially for
capital projects. While foundations expected formal reports of their grantees,
including an identification of their accomplishments, these reports focused on
documentation of the activities rather than an analysis of long-term outcomes
and impacts. Community agencies, for their part, tended to be quite small and
lacked professional expertise to conduct rigorous policy analysis and evaluation.
And since public and private funders did not require this type of data gathering,
nonprofits also did not have a strong incentive to create these systems, especially
given the relative scarcity of funding.
A new emphasis on performance

The attention to outcomes within the nonprofit sector began to shift in the
1980s, driven in part by changes in public policy and the role of government
in supporting the sector. First, the Reagan administration reduced many federal
249
grant programs during the early years of their first administration, leading to
significant cutbacks for many community agencies (Bawden and Palmer 1984).
The resulting austerity at the local level encouraged nonprofit agencies to devote
greater attention to outcomes and program success, lest they lose more funding
in the future.
Second, the build-up of federal social programs reliant on local community
organizations and changes at the community level prompted scholars,
policymakers, and practitioners to devise new strategies of evaluation. This issue
was particularly important because the new community programs often relied
upon local nonprofit agencies to spark major community change in areas such
as poverty and economic development, often through the joint action of many
public and nonprofit agencies. The foremost new evaluation strategy was the
‘theory of change’ that was proposed by many evaluation scholars especially Carol
Weiss (1995). Working with the Roundtable on Community Change at the
Aspen Institute, Weiss hypothesized that a key reason that complex community
programs are so difficult to evaluate is that their underlying assumptions are
poorly articulated. Consequently, she argued, key stakeholders (such as nonprofit
leaders) are unclear of the actual change process and devote scant attention to
the short-term and medium-term changes necessary for lasting positive effects
on social problems (Center for Theory of Change 2016). Thus, she suggested
that policymakers and practitioners needed to map the assumptions and steps in
the implementation process for their community initiatives.
The work of Weiss and others has had a significant and enduring impact on
local community agencies and organized philanthropy. The theory of change and
associated concepts, including logic models and log-frames, have been broadly
adopted by public and private funders and their grantees (see also, Gugerty and
Karlan, forthcoming). For instance, most United Way chapters now require logic
models at the time of funding application and the expectation is that agencies
will adhere to their identified theory of change if they receive the grant. Many
federal, state, and local government agencies also require local nonprofit agencies
to employ a theory of change and logic models.
Classic policy analysis and logic models thus share similar assumptions and
techniques: the importance of data collection; the clear articulation of the
underlying programmatic assumptions; identification of key stakeholders; and
the importance of ongoing feedback and organizational learning as central to the
successful attainment of program outcomes. But logic models are also different
than policy analysis in important respects: the former are guided by the mission
of the organization and are in service of the organization’s priorities; the latter,
by contrast, is employed for a client such as a government agency where the
analyst evaluates a social intervention, its likelihood of success, and its efficiency
in reaching its goals. Logic models also do not necessarily weigh costs and benefits
like a typical policy analysis, although it is certainly possible to integrate these
considerations into a logic model frame.
250
Importantly, though, the expansion and professionalization of the policy analysis

profession has contributed to the broader adoption of the theory of change.
Increasingly, the graduates of public policy master’s programs are landing jobs in
the nonprofit sector and applying their analytic techniques and training learned
at graduate school to their own organizations. (See this volume’s chapters by
O’Hare and Rubaii on education in policy analysis.) Indeed, nonprofit agencies
have significantly ratcheted up their expectations on the evaluation and analytic
skills of their higher-level employees, partly due to the changing expectations
of funders. Many funders, including the Annie E. Casey Foundation (2004) and
the Kellogg Foundation (2006), have also broadly disseminated best practices on
logic models and the theory of change.
An important related development is the adoption of ‘evidence-based practices
and decision-making’ by many nonprofits (and other organizations), so services are
guided by best practices and informed by the latest research in a particular field
(Metz et al. 2007; Johnson and Austin 2008). Thus, the design and implementation
of a mental health counseling program would be guided by research on the
most effective intervention and counseling techniques. Evidence-based decision
making is often based on policy research on specific programs, such as job training
or substance abuse treatment. Thus, it does not fit the classic policy analysis
paradigm. However, the implementation of evidence-based decision making
is a better fit with the policy analysis model. For instance, proper and effective
implementation requires an initial assessment of the problem, an analysis of various
strategic options informed by the research, the creation of a logic model, ongoing
performance management throughout the implementation process, and ideally
a thorough outcome evaluation (Moore 2010). To be sure, many agencies are
not in a position to follow all of these steps but the shift toward evidence-based
decision making is pushing more agencies in this direction. This trend has also
been abetted by government, which has broadly disseminated evidence-based
practice guidelines in specific sub-fields such as substance abuse and child welfare.
Indeed, government has sometimes required adherence to evidence-based practice
standards by nonprofit (and for-profit) contract agencies as a condition of accepting
government funding. Important professions such as social work and psychology
have actively embraced evidence-based practice in their own research and training,
further encouraging the adoption of these standards. Leading foundations such
as the Arnold Foundation have also promulgated best practices and examples of
evidence-based policy (Coalition for Evidence-Based Policy 2015).
The broad support for evidence-based decision making and the theory of change
by local nonprofits is also due to the ripple effects of changes in government
funding and contracting procedures. Beginning in the 1990s, government agencies
began to more stringently regulate the programmatic content of nonprofits. This
trend was particularly apparent in social services and health care. To varying
degrees, depending upon the jurisdiction, many key social service contracts,
including welfare-to-work, mental health, workforce development, and child
welfare, are now performance based, wherein agencies are reimbursed for services
251
only if they meet specific performance targets (Desai et al. 2012; Smith 2016;
Fraser and Whitehill 2014; Freundlich and McCullough 2012). For instance, a
nonprofit job training agency with a state government contract is likely to be
subject to requirements that tie their reimbursement to the attainment of specific
job placement targets.
Performance-based contracts are now commonly included under the broader
category of ‘pay-for-success’ (PFS) (Corporation for National and Community
Service 2015; In the Public Interest 2015; Roman et al. 2014). In brief, PFS
seeks to link payment for services to the success of the intervention for clients
and the broader community. One variation of the traditional performance-based
contract are social impact bonds (SIBs)—a form of PFS that has achieved wide
attention in the United States and abroad as an innovative strategy to achieve
greater social impact. SIBs are complicated initiatives that depend upon private
investors assuming the risk of social programs, with the government paying off
those investments if and when the outcome goals are met. Private investors loan
money to an intermediary (usually a nonprofit), which then sub-contracts with
service providers (nonprofit and for-profit), who then deliver services with specific
performance targets. The project is evaluated by independent researchers and the
government sponsor repays the loan with interest to investors if the performance
targets are met. One of the most well-known SIBs was an ultimately unsuccessful
effort in New York City to reduce the recidivism rate among individuals leaving
Riker’s Island Correctional Facility. The city of New York partnered with
Goldman Sachs, the Bloomberg Foundation, MDRC (a nonprofit intermediary),
and the Vera Institute as the nonprofit service provider to offer an intensive service
to keep released prisoners (most of them on parole) in the community.
Despite widespread publicity, SIBs remain a very limited initiative, partly due
to the relatively high transaction costs. Also, SIBs require substantial scale in order
to justify the expense and conduct a meaningful evaluation with robust results.
Nonetheless, the SIBs movement reflects the ascendency among selected public
and private funders of a new orientation toward outcomes and performance. It
also reflects a rejection of the previous era of philanthropy embodied by funders
such as the United Way that financially supported the community organization
infrastructure.
The shift in funder orientation is also reflected in the enthusiasm for ‘collective
impact.’ As articulated by Kania and Kramer (2011), collective impact is a
performance management strategy that seeks to shift the focus of evaluation and
outcome measurement from specific programmatic effects (such as the impact
of one agency program on job placement) to community-wide impacts such as
the effect of multiple programs on poverty or joblessness in the community. The
idea of collective impact has been embraced in a wide variety of public–private
partnerships and collaborative initiatives. For instance, Kania and Kramer (2011)
cite the example of Shape Up Somerville (Massachusetts), which is a city-wide
public–private partnership focused on a comprehensive and successful effort to
reduce childhood obesity (also, Bielefeld 2014). Often these collective-impact
252
efforts are led by intermediary organizations created expressly to bring together

public and private agencies in local communities (Smith 2016; Nundy and
Chandler 2015).
The successful implementation of collective impact requires substantial analytic
capacity by funders and local nonprofits. Indeed, analysis is undertaken in the
formative stages of a new collective initiative in order to inform its programmatic
and organizational design. Then evaluation is ideally supposed to be an ongoing
effort. Like logic models, the underlying assumption is that outcomes are
measureable and the causal link between the collective efforts of multiple agencies
and local community outcomes can be identified. However, this linkage is often
difficult to establish, especially for complicated community endeavors.
One other performance-related development related to nonprofits is
noteworthy: Social Return on Investment (SROI). In essence, SROI was initially
conceptualized as an alternative to narrowly focused evaluation and policy analysis
(see Emerson et al. 2000; Javits 2008). For example, a classic policy analysis
of a nonprofit community re-entry program for ex-prisoners might assay the
immediate costs and weigh these costs against specific program targets such
as reduced recidivism. This focus on the costs (and efficiency) of a program,
though, tends to neglect longer-term benefits of successful re-entry programs.
An ex-prisoner who successfully gets a job becomes a contributing member of
society and pays taxes and purchases goods and services. SROI was designed to
capture these longer-term benefits by adapting a more conventional cost-benefit
analysis to fit a broader perspective on potential benefits. If successful, a nonprofit
program might then be a much better investment for public and private funders
than under a more conventional analysis.
SROI was also designed, like logic models, to link performance, evaluation, and
practice. As noted by an early proponent of SROI, Jed Emerson of the Roberts
Enterprise Development Foundation (now simply REDF), evaluation until the
1990s had largely been retrospective and externally focused on accountability to
public and private funders, especially on fulfilling the terms of grant and contract
expectations (Emerson et al. 2000). (Many performance contracts still have this
characteristic.) In the ideal SROI, nonprofits are regarded as investments, with
the public and private funders expecting a return; thus, they are entitled to a full
accounting of these costs and benefits. Further, nonprofits, unlike for-profits,
have a social mission that needs to be broadly assessed.
Despite extensive discussion by scholars, policymakers, and practitioners, SROI
has not been widely embraced by nonprofit agencies or policymakers. In part,
the complexity and cost of SROI—especially compared with logic models—has
limited its use. In addition, nonprofit leaders and scholars have correctly noted
that SROI tends to focus on the quantification of measureable costs and benefits;
thus, the more intangible benefits of nonprofits such as building community,
networking, and improved family relationships are often not directly considered
(also, Javits 2008).
253
Nonetheless, the thinking underpinning SROI has been influential. It has

encouraged funders to regard themselves as investors who are entitled to expect
good performance. SROI has also been part of the broader conversation on
‘strategic philanthropy’ that seeks to instill a culture of evaluation and outcome
orientation within nonprofit programs. For example, REDF has greatly expanded
in the last 10 years by offering an array of programs that are regarded as social
enterprises blending elements of the nonprofit and for-profit sectors, guided by
a rigorous approach to evaluation, including outcome transparency (see Rotz
et al. 2015).
Strategic philanthropy has been embraced by many foundations as a guide for
their grant-giving, data-gathering, and evaluation strategies (Kania et al. 2014).
The result has been significantly increased attention to evaluation and outcome
analysis among grantees, despite the continued debate within the sector on the
merits of strategic philanthropy and the extent to which it really is a new approach
to philanthropy (Prewitt 2014).
Strategic planning and policy analysis

A distinguishing characteristic of nonprofit organizations is the ‘non-
distribution constraint,’ which prevents nonprofits from distributing the ‘profits’
of the organization to the owners (typically the board) (see Hansmann 1980).
Consequently, nonprofits lack the profit measure to evaluate the success of their
operations, complicating the process of program design and implementation.
One common approach in nonprofits to this challenging dilemma is strategic
planning, which shares important similarities with policy analysis. Strategic
planning involves a sequence of steps and decisions: identification of organizational
mandates and values; clarification and discussion of agency mission; an assessment
of the strengths, weaknesses, opportunities, and threats (often called SWOT
analysis); identification of the strategic issues facing the organization; formulation
of plans and options to address the strategic issues; crafting an implementation
plan; identification of key benchmarks and performance targets; and provision of
opportunities for periodic assessment and organizational learning (Bryson 2016).
Typically, strategic planning entails the participation of the board of directors and
staff, as well as key external stakeholders.
An attraction of strategic planning for a nonprofit is that it does not necessarily
require a substantial amount of resources and it is a participatory activity that can
engage the staff, volunteers, and board members of the agency. Lacking a profit
measure, the mission of the organization is central to defining the programmatic
focus and values of the organization (Oster 1995). But, nonprofits are also
susceptible to shifts in mission or a lack of clarity in mission; thus, successful
and effective nonprofits need to periodically review their mission and values
in order to ensure that the organization is correctly positioned in terms of
mission, program priorities, and overall operations. Strategic planning can play
254
this important role, especially given the opportunity for broad participation and
input by agency personnel.
Strategic planning also has other virtues for program management. Many
nonprofits may start their organizational lives with a well-defined programmatic
focus and mission; however, the organization’s programs may grow quite diffuse
over time, reflecting funding opportunities and changes in staff and board
priorities. Strategic planning is an opportunity to refocus the agency and commit
itself to its most important and mission-related activities. Strategic planning
also offers an opportunity for the reallocation of resources between programs.
Through input from staff and volunteers, nonprofits may be able to create a climate
for resource allocation that would not otherwise exist. Further, the strategic
planning process can inject innovation into the program planning process since,
ideally, diverse input from the staff and stakeholders is solicited. And like policy
analysis, the planning and change process permits the explicit weighing of costs
and benefits of different programmatic options. The need to collect information
and performance data on an ongoing basis also means that the implementation
of the strategic plan can facilitate organizational learning (Bryson 2011, 2016).
Despite its potential benefits, strategic planning can go awry for many different
reasons. Classic strategic planning presumes that nonprofit agencies have a measure
of control over their resources and can influence their environment in ways that
can benefit the organization. Yet, many nonprofits are primarily dependent
on government funding or a major foundation grant; as a result, the staff and
volunteers may be subject to external forces and funds in a way that limits their
ability to choose particular programmatic options and even specific performance
measures (since these measures may be imposed by funders). Many nonprofits,
as noted, are small and thus may not have the staff resources to devote to a major
planning process. And, many nonprofits provide an array of services to the public;
thus, planning may require a diversion of staff time and resources away from
direct programming to planning. This redirection may be difficult for resource-
constrained nonprofits. Staff may also invest deeply in the planning process and
produce a well-defined and comprehensive plan; but the organization may lack
the resources or top-level support for successful implementation.
Despite these potential pitfalls, nonprofits are being pressured to undertake more
strategic planning due to the climate of enhanced accountability and transparency.
Indeed, many nonprofits now post their strategic plans on their website as a strategy
to attract donors and staff to their organization. Arguably, a well-crafted strategic
plan with specific performance targets and benchmarks is almost a necessity for
larger organizations interested in substantial fundraising. Even smaller foundations
and United Way chapters now expect a professional strategic plan. Funders,
donors, and community members cannot easily evaluate the quality of nonprofit
services, in part because the funders and potential community supporters are often
not the individuals actually using the services (many cultural institutions are the
exceptions) (see Hansmann 1980). Consequently, a funder or donor must trust that
255
the agency is providing quality services. A strategic plan can provide reassurance
and potentially enhance the trust of local funders and donors in the nonprofit.
The professionalization of nonprofit staff with more in-depth analytic skills,
including knowledge of policy analysis and sophisticated databases, is also
facilitating the increased utilization of strategic planning. To be sure, strategic
planning is a different process than classic policy analysis, especially since strategic
planning serves the organization rather than being directed to policymakers.
Nevertheless, the advance of strategic planning in nonprofits represents the
increased acceptance of policy analysis, broadly defined, within nonprofits and its
potential utility in guiding nonprofit program development and implementation.
Advocacy and lobbying

While nonprofits might use policy analysis for internal purposes, nonprofits
can also use it to influence public policy and represent citizen viewpoints. But
nonprofits also face a variety of constraints in their ability to undertake advocacy
and employ policy analysis in support of their organization’s goals and priorities.
As noted, the nonprofit sector is very diverse and is comprised of many different
types of tax-exempt organizations. Trade associations and think tanks (which are
usually nonprofit) are addressed elsewhere in this volume. Public charities (501(c)
(3) organizations) are tax exempt and allow donors to receive a tax deduction.
These organizations face serious limitations on their ability to undertake lobbying;
generally, federal IRS rules only allow these organizations to engage in lobbying
up to about 20% of their budget (with some exceptions). In addition, these
organizations are strictly forbidden from engaging in partisan political activity.
By contrast, 501(c)(4) organizations, defined in the IRS code as ‘social welfare
organizations,’ are permitted much greater flexibility on lobbying and advocacy,
although these organizations also cannot directly engage in partisan activity.
For 501(c)(3) organizations, 20% of the budget is quite a substantial sum for
most nonprofits whose principal mission is providing a service to the community,
rather than advocacy. Nonetheless, most 501(c)(3) organizations are not even
close to the 20% limit; indeed, even many sizable organizations do not undertake
any significant advocacy or lobbying. Responsibility for this limited activity
lies with many internal and external factors. Many nonprofits such as a parent
teacher association or a local girl scout chapter are primarily reliant on small
cash and in-kind donations and do not necessarily view their role as changing
public policy. In addition, many nonprofits tend to be relatively small and thus
lack the infrastructure and capacity to undertake advocacy. Moreover, the boards
of nonprofit 501(c)(3) organizations tend to be very cautious about advocacy
and lobbying due to a concern that active advocacy might negatively affect their
organization, especially given the legal restrictions on advocacy (Pekkanen et al.
2014; Berry 2004). Agencies with government funding tend to be more engaged
in advocacy than agencies strictly dependent upon private donations; however,
256
their advocacy tends to be focused on policy goals such as funding and regulations
that might benefit their organizations (Smith and Lipsky 1993; Mosley 2014).
501(c)(4) organizations are much more likely to be engaged in advocacy since
they do not face the legal restrictions of 501(c)(3) organizations, although they
are also prohibited from partisan political activity. Indeed, some 501(c)(3) public
charities have an affiliated (c)(4) arm that engages policy advocacy. (Some (c)(4)
organizations also create a charitable arm to raise private donations.)
Both types of nonprofits undertake policy analysis and policy research. First,
policy analysis of their own programs employing logic models and outcome
evaluation is often used by nonprofits to influence public policy regarding funding
and more generally to influence policy in a particular direction. For example, the
Nurse–Family Partnership (NPF) has published research, available on their website,
that demonstrates the positive impact of this early intervention program on the
reduction of maternal and child mortality (Nurse–Family Partnership 2014). NFP
has also been successful in receiving federal funding for the expansion of their
program. Many other nonprofits, including REDF, the Harlem Children’s Zone,
YouthBuild, and Teach for America, have undertaken similar types of evaluations
demonstrating their success. More generally, this type of program publicity can
also fit with the public education and engagement efforts of nonprofits. That is,
nonprofits may be reluctant to directly contact legislators on behalf of specific
legislation or funding priorities. However, nonprofits may nonetheless strive to
influence public policy by publicizing their own program success and relevant
studies of public problems. The importance of this public education role of
nonprofits is also underscored by the association of public education as part of
a nonprofit’s mission and its propensity to undertake advocacy (Pekkanen and
Smith 2014).
Importantly, national and local funders also support collaborative initiatives at the
community or regional level that bring together public and private organizations to
solve a community problem. One example is Cleveland’s Greater University Circle
Initiative (GUCI), financially supported by many local partners including the
Cleveland Foundation and Case Western Reserve University. These organizations
undertook joint problem solving, leading to multiple programs to support the
revitalization of the Greater University Circle area. The different partners used
local research to inform their strategic direction and program implementation
such as new business growth and retention incentives (see Cleveland Foundation
2013). Likewise, Thrive Washington, a collaboration of several public and private
funders in Washington state, focuses on early learning and education. Its activities
include grant making and policy analysis, as well as advocacy for increased funding
for early childhood education (Thrive Washington 2016).
Yet, changes in the political environment are changing the incentive structure
of nonprofits regarding advocacy and the use of policy analysis by nonprofits in
the policy process. Government austerity has led to reductions in public funding
for many nonprofits, leading them to redouble their efforts to influence public
policy through advocacy. These efforts can be direct advocacy by agency staff
257
or increasingly through intermediary associations such as coalitions of different

nonprofits whose mission is to represent nonprofit interests in the policy arena.
Moreover, the polarized political environment, including the political gridlock
at the federal level, means that simply producing solid, sound policy reports
on important issues such as poverty, job training, or early childhood education
may be ineffective in changing public policy. Thus, nonprofits need to employ
a political strategy to accompany their policy reports and analysis if they are
to be effective. For this reason, some funders have decided to support 501(c)
(4) organizations with a mission of advocacy for specific purposes, such as
influencing the implementation of the Affordable Care Act (Booth 2016). The
growth of hybrid 501(c)(3)and 501(c)(4) organizations also reflects this trend. In
a sense, nonprofits are in an unexpected and paradoxical situation: nonprofits are
conducting more policy analysis and research to influence public policy than ever
before; but their policy influence remains elusive given the political environment
and the limitations and constraints by nonprofits in undertaking and sustaining
advocacy and policy influence.
Conclusion
In the United States, the creation of a nonprofit entails relatively low transaction
costs; indeed, the process of incorporation and obtaining federal tax-exempt status
is relatively routine and low cost. The ease of formation has facilitated the sharp
growth of nonprofit 501(c)(3) organizations in response to rising demand for
community programs and widespread interest in social innovation and enhanced
effectiveness in public services. With more than double the number of nonprofits
during the last 20 years, nonprofits have provided an important vehicle for
more policy analysis, research, and evaluation. Public and private funders have
encouraged, and even in many cases mandated, more policy analysis by nonprofits.
Reform-minded foundations and policy entrepreneurs have used nonprofits to
promote their ideas and policy goals, including evidence-based decision making
and innovations in public services.
More nonprofits has in turn meant more consumption of policy analysis by
nonprofits themselves for internal planning purposes and to ensure the effective
implementation of their own programs. The demand for more policy analysis has
in turn spurred the professionalization of nonprofits in order to develop and sustain
evidence-based decision making and in general higher levels of accountability
and transparency within their own organizations. Yet the nonprofit sector is also
diverse, with many small nonprofits with little capacity to undertake policy analysis
and organizational missions focused on very local concerns. But the demand for
more rigorous policy analysis within nonprofits has encouraged a growing divide
between larger nonprofits with sizable staffs and diversified revenue sources and
smaller nonprofits. In an increasingly competitive environment for resources,
the larger nonprofits have a decided edge in the race for grants, contracts, and
private contributions.
258
Yet, new political developments in the United States may create serious
complications for all nonprofits interested in offering services to the citizenry. The
Trump administration has proposed quite drastic cuts in public funding of an array
of public services, including health and social services, the arts, and education.
Moreover, the administration has proposed reductions in funding for research and
statistical offices. Thus, nonprofit agencies are likely to find that it is much more
difficult to undertake policy analysis on important public policy issues. Moreover,
the presumption of many nonprofit programs today is that if the program can
be proven effective, by some measure, public and private funding will flow to
this organization. But the reductions in public funding and the relative scarcity
of private philanthropic funding will mean that even effective programs will
struggle to obtain funding. Consequently, a consolidation of nonprofit programs
is likely at the local level as funding to support community services declines. The
organizational consequences for the nonprofit sector and local communities could
be profound, as smaller nonprofits are absorbed by larger nonprofit organizations,
reducing the overall diversity and supply of community programs.
Given these challenges, sound and rigorous policy analysis may actually be
more vital than ever for nonprofit sustainability since nonprofits will be pressed
to make an effective case for their programs. In the current politically charged
and fiscally austere environment, nonprofits will need to think broadly about their
impact and their relationship to their local communities. Higher expectations
on accountability have tended to focus on specific programmatic interventions:
Did the mentoring program reduce high-school drop-outs? How many people
obtained permanent jobs through the job training program? How many previously
homeless were able to transition to permanent housing? These questions are of
course very important for policy and practice. However, it also means that the
more intangible but nonetheless valuable roles of nonprofits, such as community
building, citizen engagement in local affairs, and the representation of citizen
interests, may be downplayed or overlooked. Also, nonprofit programs with
elusive or difficult-to-measure outcomes, such as advocacy, the arts, or recreation,
may receive less funding and attention than programs that are perceived to offer
a higher potential long-run impact. Thus, a challenge for the practice of policy
analysis within nonprofits is to take a broad view of the potential outcomes
of nonprofits to ensure the impact of nonprofits and their programs on local
communities and the policy process are adequately assessed. This effort can also
help promote organizational learning, which can increase the effectiveness of
nonprofit governance, successful program implementation, and the contributions
of nonprofits to their communities and public policy.
Notes
1 The author is indebted to Putnam Barber, David Hammack, John Hird, Bob Ness, Beryl
Radin, and an anonymous reviewer for comments on earlier versions of this chapter. Meghan
McConaughey provided excellent research assistance.
259
References
Annie E. Casey Foundation (2004) Theory of Change: A Practical Tool for Action,
Results and Learning, www.aecf.org/m/resourcedoc/aecf-theoryofchange-2004.
pdf.
Bawden, D. Lee and Palmer, J. L. (1984) ‘Social Policy: Challenging the Welfare
State’, in J. L. Palmer and I. V. Sawhill (eds), The Reagan Record, Washington,
DC: Urban Institute, 177–215.
Berry, J. M. (2005) ‘Nonprofits and Civic Engagement’, Public Administration
Review, 65: 568–78. doi:10.1111/j.1540-6210.2005.00484.x
Bielefeld, W. (2014) Using Collective Impact to Improve Student Success, The Hubert
Project, Minneapolis, MN: University of Minnesota.
Bielefeld, W. and Chu, J. (2010) ‘Foundations and Social Welfare in the Twentieth
Century’, in H. K. Anheier and D. C. Hammack (eds), American Foundations:
Roles and Contributions, Washington, DC: Brookings 158–81.
Booth, M. (2016) ‘Could Foundations Have Mounted a Better Defense of the
ACA?’ Health Affairs Blog, http://healthaffairs.org/blog/2016/09/14/could-
foundations-have-mounted-a-better-defense-of-the-aca/.
Brilliant, E. L. (1990) The United Way: Dilemmas of Organized Charity, New York:
Columbia University Press.
Bryson, J. M. (2011) Strategic Planning for Public and Nonprofit Organizations (4th
edn), San Francisco, CA: Jossey-Bass.
Bryson, J. M. (2016) ‘Strategic Planning and the Strategic Change Cycle’, in D.
O. Renz and Associates (eds), The Jossey-Bass Handbook of Nonprofit Leadership
and Management (4th edn), Hoboken, NJ: Wiley 240–73.
Center for Theory of Change (2016) TOC Origins, www.theoryofchange.org/
what-is-theory-of-change/toc-background/toc-origins/.
Cleveland Foundation (2013) Cleveland’s Greater University Circle Initiative: A
Partnership Between Philanthropy, Anchor Institutions, and the Public Sector, Cleveland,
OH: Cleveland Foundation.
Coalition for Evidence-Based Policy (2015) Top Tier Evidence Initiative, http://
toptierevidence.org/.
Corporation for National and Community Service (CNCS), Office of Research
and Evaluation (2015) State of the Pay for Success Field: Opportunities, Trends, and
Recommendations, Washington, DC: CNCS.
Desai, S., Garabedian, L., and Snyder, K. (2012) Performance Based Contracts in
New York City: Lessons Learned from Welfare-to-Work, Albany, NY: Rockefeller
Institute of Government, www.rockinst.org/pdf/workforce_welfare_and_
social_services/2012-06-Performance-Based_Contracts.pdf.
Emerson, J., Wachowicz, J. and Chun, S. (2000) Social Return on Investment:
Exploring Aspects of Value Creation in the Nonprofit Sector, San Francisco, CA:
REDF, http://redf.org/current/web/app/uploads/2013/10/REDF-Box-Set-
Vol.-2-SROI-Paper-2000.pdf.
260
Fraser, J. and Whitehill, E. (2014) Introducing Performance-Based Contracts: A

Comparison of Implementation Models, Pittsburgh, PA: Allegheny County
Department of Human Services.
Freundlich, M. and McCullough, C. (2012) Privatization of Child Welfare Services:
A Guide for State Advocate, State Policy Advocacy and Reform Center and
First Focus, http://childwelfaresparc.org/wp-content/uploads/2014/07/17-
Privatization-of-Child-Welfare-Services-A-Guide-for-State-Advocates.pdf.
Gugerty, M. K. and Karlan, D. (forthcoming) The Goldilocks Challenge: Right-
Sized Monitoring and Impact for Social Sector Organizations, New York: Oxford
University Press.
Hammack, D. H. (2002) ‘Failure and Resistance: Pushing the Limits in Depression
and Wartime’, in L. Friedman and M. McGarvie (eds), Charity, Philanthropy, and
Civility in American History, New York: Cambridge University Press 263–80.
Hansmann, H. B. (1980) ‘The Role of Nonprofit Enterprise’, Yale Law Journal
89(5): 835–901.
Hansan, J. E. (2013) United Way of America, Social Welfare History Project,
Virginia Commonwealth University, http://socialwelfare.library.vcu.edu/eras/
civil-war-reconstruction/united-way-of-america/.
Hartford Institute for Religion Research (2016) Fast Facts about American Religion,
http://hirr.hartsem.edu/research/fastfacts/fast_facts.html#numcong.
In the Public Interest (2015) A Guide to Evaluating Pay for Success Programs and
Social Impact Bonds, Washington, DC: In the Public Interest, http://www.
inthepublicinterest.org/wp-content/uploads/ITPI-Pay-for-Success-Guide-
Dec-2015.pdf.
Javits, C. I. (2008) REDF’s Current Approach to SROI, http://redf.org/app/
uploads/2013/10/REDFs-Current-Approach-to-SROI-Opinion-Paper-2008.
pdf.
Johnson, M. and Austin, M. E. (2008) ‘Evidence-Based Practice in the Social
Services: Implications for Organizational Change’, Journal of Evidence Based
Social Work 5(1–2): 239–69.
Kania, J. and Kramer, M. (2011) ‘Collective Impact’, Stanford Social Innovation
Review (Winter), http://ssir.org/articles/entry/collective_impact.
Kania, J., Kramer, M. and Russell, P. (2014) ‘Strategic Philanthropy for a Complex
World’, Stanford Social Innovation Review (Summer): 26-33, https://ssir.org/
up_for_debate/article/strategic_philanthropy.
Kellogg Foundation (2006) W.K. Kellogg Foundation Logic Model Development Guide,
www.wkkf.org/resource-directory/resource/2006/02/wk-kellogg-foundation-
logic-model-development-guide.
Kohler, S. (2007) Case 33, Bedford Stuyvesant and the Rise of the Community
Development Corporation, Ford Foundation, Durham, NC: Sanford School of
Public Policy, Duke University, https://cspcs.sanford.duke.edu/sites/default/
files/descriptive/bedford-stuyvesant.pdf.
261
Koulish, J. (2013) There are a lot of 501 c 4 organizations. Most are not political,
Washington, DC: The Urban Institute, www.urban.org/urban-wire/there-are-
lot-501c4-nonprofit-organizations-most-are-not-political.
Koulish, J. (2016) From Camps to Campaign Funds: The History, Anatomy, and
Activities of 501(c)(4) Organizations, Washington, DC: The Urban Institute,
www.urban.org/sites/default/files/alfresco/publication-pdfs/2000594-From-
Camps-to-Campaign-Funds-The-History-Anatomy-and-Activities-of-501(c)
(4)-Organizations.pdf.
Leiby, J. (1963) ‘Amos Warner’s American Charities 1894–1930’, Social Service
Review 37(4): 441–55.
McKeever, B. S. (2015) The Nonprofit Sector in Brief 2015: Public Charities, Giving,
and Volunteering, Washington, DC: Urban Institute, www.urban.org/sites/
default/files/alfresco/publication-pdfs/2000497-The-Nonprofit-Sector-in-
Brief-2015-Public-Charities-Giving-and-Volunteering.pdf.
Metz, A. J. R., Blasé, K., and Bowie, L. (2007) Implementing Evidence-Based
Practices: Six Drivers of Success, Part 3 in a Series on Fostering the Adoption of
Evidence-Based Practices in Out-Of-School Time Programs, Child Trends,
www.childtrends.org/wp-content/uploads/2013/07/2007-29EVPSuccess.pdf.
Moore, K. A. (2010) Innovations in Evidence-Based Policymaking, Becoming Evidence-
Based: What Does It Take?, Bethesda, MD: Child Trends, http://www.childtrends.
org/wp-content/uploads/2013/12/2010-28InnovationsEBPPresent.pdf.
Mosley, J. (2014) ‘From Skid Row to the Statehouse: How Nonprofit Homeless
Service Providers Overcome Barriers to Policy Advocacy Involvement’, in R. K.
Pekkanen, S. R. Smith, and Y. Tsujinaka (eds), Nonprofits and Advocacy: Engaging
Communities and Governments in an Era of Retrenchment, Baltimore, MD: Johns
Hopkins University Press 107–34.
Mossberger, K. (2010) ‘From Gray Areas to New Communities: Lessons and
Issues from Comprehensive U.S. Neighborhood Initiatives’, https://greatcities.
uic.edu/wp-content/uploads/2013/08/Mossberger_Karen_5.09.pdf.
NCCS (2016) Quick Facts About Nonprofits, Washington, DC: Urban Institute.
http://nccs.urban.org/data-statistics/quick-facts-about-nonprofits.
Nundy, N. and Chandler, A. P. (2015) ‘Collective Impact: The Missing Link’,
Stanford Social Innovation Review, June 16, http://ssir.org/articles/entry/
collective_impact_the_missing_link.
Nurse–Family Partnership (2014) Nurse Family Partnership is the First Early
Intervention to Find Reductions in Maternal and Child Mortality, https://cchealth.
org/fmch/pdf/NFP-mortality-study.pdf.
Oster, S. K. (1995) Strategic Management in Nonprofit Organizations, New Haven,
CT: Yale University Press.
Pekkanen, R. J. and Smith, S R. (2014) ‘Nonprofit Advocacy in Seattle and
Washington, DC’, in R. K. Pekkanen, S.R. Smith, and Y. Tsujinaka (eds),
Nonprofits and Advocacy: Engaging Communities and Governments in an Era of
Retrenchment, Baltimore, MD: Johns Hopkins University Press 47–65.
262
Pekkanen, R.K., Smith, S.R., and Tsujinaka, Y. (2014) Nonprofits and Advocacy:
Engaging Communities and Governments in an Era of Retrenchment, Baltimore, MD:
Johns Hopkins University Press.
Prewitt, K. (2014) ‘Strategic Philanthropy for a Complex World’, Stanford Social
Innovation Review (Summer): 35–6.
Price, D. K. (1965) The Scientific Estate, Cambridge, MA: Harvard University Press.
Roman, J. K., Walsh, K. A., Bieler, S., and Taxy, S. (2014) Pay for Success and Social
Impact Bonds: Funding the Infrastructure for Evidence-Based Change, Washington,
DC: Urban Institute.
Rotz, D., Maxwell, N., and Dunn, A. (2015) Economic Self-sufficiency and Life
Stability One Year after Starting a Social Enterprise Job, San Francisco, CA: REDF.
Smith, S. R. and Lipsky, M. (1993) Nonprofits for Hire: The Welfare State in the
Age of Contracting, Cambridge, MA: Harvard University Press.
Sullivan, S. (2013) ‘What Is a 501(c)(4), Anyway?’ The Washington Post, May 13,
www.washingtonpost.com/news/the-fix/wp/2013/05/13/what-is-a-501c4-
anyway/.
Thrive Washington (2016) Policy and Advocacy to build a Strong, Responsive Early
Learning System, https://thrivewa.org/work/our-approach-to-policy-work/.
Walters, P. B. and Bowman, E. A. (2010) ‘Foundations and the Making of
Public Education in the United States 1867–1950’, in H. K. Anheier and D.
C. Hammack (eds), American Foundations: Roles and Contributions, Washington,
DC: Brookings 31–50.
Warner, Amos G. (1894) American Charities, New York, NY: Thomas Y. Crowell
Co.
Weiss, Carol H. (1995) ‘Nothing as Practical as Good Theory: Exploring Theory-
Based Evaluation for Comprehensive Community Initiatives for Children and
Families’, in J. P. Connell, A. C. Kubisch, L. B. Schorr, and C. H. Weiss (eds),
New Approaches to Evaluating Community Initiatives: Concepts, Methods and Contexts,
Washington, DC: Aspen Institute.
263
FOURTEEN
The media
Annelise Russell and Maxwell McCombs
The role of the mass media in political science is well known—from measures
of campaign communications, priming effects on mass publics, and its increasing
role in spurring social movements and cultural shifts. The media is an integral
part of political life, but as the role of media expands and changes we must
consider more than just political consequences, but also those of public policy
and governance. It is easy to make associations between media and politics when
Tina Fey is doing her famous Sarah Palin impression or candidates for office
flood the airwaves with unending commercials. But what is not so clear is how
the actions of media outlets and the actors within these institutions affect the
public policy process and the manner in which we govern (or not). In the last
20 years, scholars in public policy and political institutions have begun to better
understand the role of media in our governing systems and what that means for
how we attend to and make policy across numerous types of political systems
and institutional venues.
Some early studies of public policy considered the media to be just a ‘vehicle’
for information, serving as merely a conduit for an external process (Kingdon
1984), but more recent scholarship highlights the media’s ability to shape those
policies that do or do not receive attention, how those policies are received,
and how the public responds to elite action. The media is a dynamic political
institution (Cook 1998; Boydstun 2013) with norms and routines that affect not
only the production of news, but how policymakers and the public alike respond
to issues and events. The media can call attention to an issue—such as the firestorm
surrounding the events in Ferguson, Missouri after Michael Brown’s death—but
the media can also work within a political network to maintain issue frames, limit
outside attention, and perpetuate the status quo. The media contributes to both
positive and negative feedback cycles within the policy process that may speed
up or slow down the process (Wolfe 2012). And in addition to direct effects, the
media’s agenda-setting effect on the public’s issue awareness and priorities in turn
affects the actions of policymakers who are responsive to their constituents. The
system of influence via the media is complex and non-linear, similar to general
assumptions about the dynamism of public policy in general. The addition of
social media—such as Twitter, Facebook, and blogs—adds increased intricacy
to the role of media and policy. Elected officials are adapting new routines to
incorporate these new media platforms, and while many write off such activity
as pure constituent communication, these activities have an effect that has the
potential to alter policy agendas and individual policymaker priorities.
265
In this chapter, we argue that for too long public policy has lagged in its
understanding of the media as a political institution with real implications for
how we process and implement policy. We argue studies of public policy can
benefit from a broadened, integrative approach toward studying the media and
the policy process. In doing so, we will begin by providing an overview of the
growing literature on media and policy—with special consideration for media
effects on issue attention, agenda setting, and policy feedback cycles. Next, we
will discuss the role of new media and how research on social media has been
and can be applied to the policy process. Finally, we will consider two avenues
for future research on the media and public policy: (1) better integration of
media and policy studies with those of mass publics and (2) encouraging greater
communication and collaboration between media and policy scholarship. Both
proposals, though not easily achieved, would aid a much richer body of work
on media effects in the policy process.
The role of the media in politics

The role of the media in politics is most often considered according to its relation
to the public. The media connects constituents with the information they need
to be free and self-governing (Dahl 1971) but this information exchange does
not happen in a vacuum. Information is not a linear exchange, but rather part
of a complex system where the media operates not only as a transfer point, but
as an independent political institution. And that institution has an effect on both
elite and mass public priorities and preferences. The media matter—not only as
primers of the public agenda, but also as a resource and institution within elite
politics (Boydstun 2013). While it is argued that the media simply index elite
opinions (Bennett 1990), the media’s agenda-setting potential and its own voice
as a political actor in the process make it a formidable institution with its own
routines and norms (Cook 1998, Sparrow 2006). In newspapers or television
broadcasts, the priorities of the elite actors become integrated with the priorities
and norms of the news organization (Boydstun 2013; Sparrow 2006; Ku, Kaid
and Pfau 2003; Miller et al. 1998; Smith 1993).
One of the most critical roles that the media plays within our political system
is its role as an agenda setter for both elite and mass priorities. Policy and
communication researchers have often highlighted the media’s agenda-setting
effect on politics (Dearing and Rogers 1996), and many point to the media’s
ability to routinely influence political agendas and spur an increase in political
attention for certain issues (Sevenans et al. 2015; Edwards and Wood 1999; Van
Noije et al. 2008; Walgrave et al. 2008). The media can have a direct effect on
the policy process by influencing those issues on the institutional agenda, and it
can also have an indirect effect by setting the public agenda, which is then relayed
to elected officials via constituents. In their foundational work on media agenda
setting, Max McCombs and Donald Shaw (1972) find that prominent aspects
of the news become prominent aspects of public affairs among the public. This
266
The media
study has been replicated hundreds of times to show the link between media
and public salience (McCombs 2014). As Bernard Cohen pointed out in 1963,
the media may not be able to tell people what to think, but they are successful
in telling them what to think about. And that ability to prime the set of issues
up for consideration also applies to the priorities of elite publics in our policy-
making institutions. Agenda setting in the policy process stems from foundational
work by E. E. Schattschneider, who theorized the limited accessibility of political
agendas (1960). Scholars built on this work, including Cobb and Elder (1971),
who detailed the differences between public and institutional agendas and noted
a role for the media in the agenda-setting process. Cobb and Elder explored the
sources of the policy agenda, and the media is one contributing source to that
agenda. Policy agenda setting was furthered by John Kingdon’s multiple streams
theory and Jones and Baumgartner’s theory of punctuated equilibrium—each
addressing the formation of policy agendas and considering the media as one
factor in how issues become salient for policymakers.
Researchers of political science, public policy, and communication have long
studied the media—at least tangentially—by exploring patterns in attention, but
Wolfe, Baumgartner and Jones point out a distinct lack of connections between
studies of the media and studies of public policy processes (2013). While the arc of
media effects of salience is much longer for mass publics, more recent scholarship
within the last 25 years highlights the more nuanced integration of policy and
media scholarship (Wolfe et al. 2013).
As scholars explore the manner in which issues either rapidly or incrementally
appear on the institutional agenda, Amber Boydstun (2013) applies the notion of a
disjointed policy process to the media’s news-generation process and finds similar
results. She examines the political process by noting the policy issues that do and
do not make it onto the media agenda—linking policy attention to long-standing
patterns in media coverage. She finds that, on the whole, the majority of policy
issues receive scant attention on the front pages of newspapers, but a select few
issues receive explosive, or disproportionate levels of coverage. She attributes this
imbalance this to positive feedback effects propelled by media coverage norms,
where the presence of one story often leads to a follow-up, regardless of the
magnitude or scope of the issue. Boydstun’s work highlights the critical role of
the media as a feedback mechanism in the policy system because the episodic
nature of news coverage is similar to the patterns we see in the legislation that
is considered and passed.
An impressive example of the effects of explosive news coverage is the drug
problem in the mid 1980s (Shoemaker 1989). Public concern about drugs began
to build after the New York Times ‘discovered’ the drug problem in late 1985
and published the first of more than 100 stories. Following The Times’ lead on
this issue, the next year there was a Newsweek cover story, specials on two of the
national television networks, a surge in coverage of drugs in newspapers across
the United States, and, predictably, a sharp rise in public concern about the drug
problem. The culmination of this barrage of media coverage and public was the
267
Omnibus Drug Bill passed by Congress and signed by President Ronald Reagan
in late 1986. Another example is a seminal article on child abuse published in the
Journal of the American Medical Association, which stimulated considerable media
attention and subsequent actions by the Congress and many state legislatures
(Nelson 1984).
These powerful positive feedback mechanisms are further present when we
look at media coverage of large-scale media stories, what Boydstun, Hardy and
Walgrave term ‘media storms’ (2014). Media storms result from the positive
feedback forces that spur additive and successive attention. These storms reflect
the disproportionate nature of how political institutions process information,
where some information is rapidly propelled forward while other is consistently
ignored. Once the media fixates on an issue, instead of moving onto the next
scandal or story, the media maintains a narrow focus and sustains its relatively
excessive attention beyond an issue’s relative impact or influence. Boydstun, Hardy
and Walgrave identify four characteristics of a media storm—size, explosiveness,
duration, and reach across media institutions (Boydstun et al. 2014). They also
identify the two mechanisms that lead to media storms: (a) lower gatekeeping
thresholds: when media institutions alter their news generation process to lower
the threshold of what is considered ‘newsworthy’; and (b) mimicking: media
institutions’ tendencies to imitate each other’s news (Trumbo 1995; Boczkowski
2010; Boydstun et al. 2014). An example of a media storm was the continued,
high-volume coverage of events resulting from the devastation of Hurricane
Katrina in 2005. Media storms result in a strong skew toward some issues on
media and political agendas relative to many possible considerations.
A more routine opportunity for assessing the relationship between the media
and the president is the annual State of the Union address. The initial agenda-
setting examination of this question, which examined President Jimmy Carter’s
1978 State of the Union address, found that the coverage of Carter’s eight issues in
the New York Times and on the television networks during the month preceding
the State of the Union address had influenced the president’s agenda (Gilbert et
al. 1980). A replication based on the identical research design examined a very
different American president, Richard Nixon (McCombs et al). In this instance,
there was no evidence of any media influence on the president. However, the
agenda of 15 issues in President Nixon’s 1970 State of the Union address did
influence the subsequent month’s news coverage in the New York Times and the
Washington Post and on two of the three national television networks.
Enormous differences in presidential personalities are, of course, a major
factor to be considered in these kinds of historical analyses. However, even with
this factor in mind, there is also evidence of a shifting relationship between a
president and the news media over the years of their administration. Analyses of
President Franklin Roosevelt’s first seven State of the Union addresses, which
were delivered from 1934 to 1940, yielded highly mixed evidence about the
relationship between the news media and the president (Johnson et al. 1995).
Similar evidence of mixed effects is also found in analyses of President Reagan’s
268
The media
1982 and 1985 State of the Union addresses (Wanta et al. 1989). Sometimes the
president is able to direct the attention of the news media toward certain issues
and to set the agendas of the media and the public. At other times, he follows the
media and public opinion (Gonzenbach 1996; Wanta and Foote 1994).
But the media does not always propel issues forward. Sometimes, the media can
have a stalling effect on the policy process, as Michelle Wolfe finds in her 2012
study of media coverage of legislation in the U.S. Congress. Wolfe examines the
relationship between issue salience in the media and the rate of policy making
across issues. Wolfe adopts the notion of the media as a gatekeeper of issues
with the ability to influence the relative speed of the policy process. Her central
argument is that once a bill is introduced, the time it takes that bill to become a
law increases as the news coverage surrounding the debate on the bill reaches a
certain level. Her findings indicate that additional media attention can actually
slow down the speed of policy making, producing a negative feedback cycle that
is more likely to lead to minimal policy change because the media has expanded
the conflict and engaged additional interests.
So in addition to propelling positive feedback loops that contribute to rapid
change we see in an episodic policy process, the media also introduces friction into
the policy process that can hinder policy change. Wolfe’s research lends support
to the notion of the media as part of a complex information processing system
(Workman et al. 2009) that serves multiple functions within the policy feedback
process. Included among these functions is a phenomenon called agenda cutting,
the omission of certain topics or aspects of certain topics from the news agenda
(Colistra 2012; Wober 2002).
The relationship between the media and the policy process is not only about
the rapid or entrenched salience of issues, but also the manner in which they
are discussed. The media also has the ability to frame the scope of debate on
a particular issue. Framing theory centers on the notion that how an issue
is characterized will contribute to its general understanding (Scheufele and
Tewksbury 2007). Alternatively, Stuart Soroka and his colleges (2012) suggest
that it ‘refers to the selective exposure of information to an audience with the
intent of shaping their understanding of an issue’. McCombs (2004) describes
framing as the emphasis placed on the attributes of an issue in the news coverage
of the issue.
Interested individuals seek media attention for specific frames, and the portrayal
of an issue carries it through the policy process and potential policy change. For
example, Frank Baumgartner, Suzanna De Boef, and Amber Boydstun (2008)
connect the news media’s ability to frame issues from a particular perspective with
changes in death penalty policy. They identify how the media’s frames surrounding
the issue of capital punishment have a substantive impact on changes in how
policymakers address capital punishment policy over time (2008).
Similar framing effects are found in Rose and Baumgartner’s 2013 study that
assesses how perceptions of poverty have changed in the United States. They
identify the media’s narrative of poverty as one source of a cultural shift—a move
269
from assessing and analyzing the root causes of poverty or the economic and
cultural drawbacks of having large, poor populations, to rather focusing on new
images of the poor as cheaters, system schemers, and recipients of government
handouts that only perpetuate the cycles of poverty. This new image of poverty
in America influences the policies implemented by lawmakers and the authors
argue that this characterization of the unworthy or undeserving poor serves as
an impetus for alternative implementation of welfare public policy.
Media effects are felt across political institutions and governing systems—these
effects are not just unique to the American federal government. Much of the
growing research of media effects on government suggests we can look toward
alternative political systems to find similar patterns of influence. At the local level,
Simon J. Kiss (2013) highlights how media coverage and the shift of an issue from
research journals to the front pages of newspapers may, at least indirectly, lead state
legislatures to new policy proposals. In his study of the growing controversy around
the hazards of bisphenol A (BPA) exposure, Kiss hypothesizes that though there
was no consensus in the scientific community about these hazards, the growth of
the issue on local news shaped the legislative process (2013). He finds that media
coverage, particularly what he terms ‘high-impact newspaper coverage,’ shaped
the diffusion of legislative bans on products with BPA. His evidence of media
effects at the state level buoys existing research by Mead (1994) that found local
newspaper coverage could affect those reform issues on the municipal agenda.
Other examples of the impact of local news coverage on city government include
a new year’s community agenda of children’s issues on the editorial page of the
San Antonio Light—supported by subsequent news reporting during the year—
that resulted in vastly increased spending for children’s programmes by the city
government (Brewer and McCombs 1996), and two series of investigative reports
by a Chicago television station that led to policy changes in the city’s police and
fire departments (Protess et al. 1991). This research suggests a dynamic relationship
between policy development and the news generation process. Those issues that
the news decides to take up—often skewed as Boydstun (2013) points out—can
and do have an impact on those issues that lawmakers address. And this influence
goes beyond the United States—in fact, some of the most rigorous studies on
the relationship between media institutions and the policy process now occur in
comparative settings, although this was not always the case.
In 2006 Stefaan Walgrave and Peter Van Aelst assessed the scholarship to date
on the media’s precise role in political agenda setting and found that historically
there was relatively little attention and what research they did find suggested mixed
results as far as the media’s ability to influence policy was concerned. They found
that prior work concentrated on a select number of agendas within a narrow set
of domains and because of that, scholarship revealed little about the media’s true
effects (Walgrave and Van Aelst 2006). In their search, the authors found 15 of
18 published studies concerning the United States, but within those studies there
was no general consensus about the role of the media—some saw considerable
influence while others ignored or denied media impact altogether (2006). Since
270
The media
their assessment of the literature 10 years ago, policy, communication, and political
science scholars across the globe have wrestled with the question of just how
the media interacts with the policy process and how those patterns play out on
a global scale.
In a study of the Belgian Parliament through the 1990s, Walgrave, Soroka and
Nuytemans (2008) find that the media does, to some extent, determine the agenda
of Parliament. This influence is varied across types of media as newspapers are
more influential than television and that influence is stronger on certain issues,
such as law and order or the environment. The variability in media effects may
also be due to the type of measurement scholars employ. Van Aelst and Walgrave
(2011) forgo the common time series analysis for media influence and instead
conduct interviews with members of parliament in European democracies to
determine their impressions of media influence. The politicians report that the
mass media is one of the prime agenda setters, along with the prime minister
and political parties (2011). In the Israeli Parliament the agenda of questions
posed to government ministers by Knesset members reflects the agenda-setting
influence of newspapers. Historically, the number of questions grounded in news
reports steadily increased from 8% in the inaugural 1949–51 Knesset to 55% in
the 1969–73 seventh session (Caspi 1982). With the growth of international
governing bodies, such as the European Union and specialized United Nations
agencies, and broad international issues such as climate change, there are ample
opportunities for examining global agenda-setting effects on policy.
Social media and public policy

Media are increasingly being integrated into policy studies, but one avenue
that remains relatively untapped—outside of mass politics and campaign
communications—is the role of social media in governing institutions. Both
politicians and researchers have characterized social media as a mechanism for
constituent outreach due to its leverage for elite actors to bypass traditional media
partners and speak both directly and indirectly with members of the public.
Straus et al. (2013) compile data from the 111th Congress and find support for
the theory that members adopt Twitter to represent a broader constituency. For
example, with social media services like Twitter or Facebook, elite actors can
shape issue salience via appeals to both interested constituents that follow them
online and also journalists with constrained resources who follow elite activity
on social media and transmit that information to constituents. This two-prong
approach to constituent activity makes it very appealing for policymakers and
while they may be using it to connect with the public back home, much of the
information in those messages is policy relevant or issue relevant.
Schattschneider (1960, p 138) defines democracy as ‘a competitive political
system in which competing leaders and organizations define the alternatives
of public policy in such a way that the public can participate in the decision-
making process.’ Social media is the newest instrument for leaders to influence
271
the parameters of the decision-making process according to those issues at the top
of their personal agenda. Platforms like Twitter, at their more core function are
more than just a constituent communication platform—they are a mechanism for
conflict expansion that enables elected officials to communicate those issues by
which they seek to be judged by constituents. While most work up to this point
on the subject of social media has been primarily focused on mass politics, there
is some literature that gives us a sense of how policymakers are using the new tool
and what the policy implications might be given those communication patterns.
One of the first studies to explore policymaker use of Twitter looks at Congress
members’ Twitter patterns during two one-week periods in 2009—during the
initial development of the platform (Glassman et al. 2009). This report by the
Congressional Research Service divides member use into six categories and
suggests that the most frequent type of Twitter communications were press and
web link tweets. While the study did not consider policy messages specifically,
policy may be the underlying content in these tweets and opens a window into
the types of linked policy or press appeals that a policymaker employs. And those
linked policy appeals have the potential to affect relationships between elite and
mass publics. In a comparative use of Twitter within legislative bodies, Bang
(2013) conducted an extensive analysis of tweets among the members of the U.S.
House of Representative and among members of the Korean National Assembly.
A recent policy-specific analysis of a U.S. senator’s Twitter activity reveals that
these policymakers reference policy in about two thirds of all Twitter messages
they send, based on a two-month sample in 2013 (Russell 2015). These messages
are not just policy or constituent messages, but results show that members use a
multi-prong approach for policy, partisan, constituent, or media messages. For
example, a lawmaker may in one message deride the opposition party for backing
off a new oil pipeline that would create jobs for constituents back home. Policy
is not considered in a vacuum and neither are policymaker communications
within this political climate. Social media, unlike previous outlets, highlights
that endogeneity and provides a window into the complex agenda setting within
these new media platforms.
Additionally, those policies that members of Congress reference in their
social media communications are at least correlated with those policy issues the
public prioritizes. In a study of U.S. Senators in 2012, Russell (2014) finds that
policymaker priorities on social media and public priorities are significantly
correlated, although conditional on economic concerns.
Better integration of media and policy studies

Media studies are often driven by a choice of dependent variable: mass or elite
publics. In communication and media studies, that variable is most often the public
and the effects that the media has on the salience of issues or the effects of framing/
priming on attitudes and engagement. Policy studies and many in political science
use the media to explain elite behavior and how the media influences the policy
272
The media
process. But these two streams of scholarship are not mutually exclusive as both
policy and public reactions to media coverage can have both direct and indirect
effects in a more complex, or cyclical nature. Many researchers shy away from
this type of research due to the endogeneity of both public and elite effects and
problems of determining causation. While these are certainly concerns that each
researcher must contend with, they should not be barriers for further research on
the media’s simultaneous relationship with politicians and the public.
One example of this type of research is by Stuart Soroka in his 2003 study of
media coverage, public opinion, and foreign policy in both the United Sates and
the United Kingdom. The author highlights a correlation between the salience
in foreign topics in the media and the public. He extends this research via the
notion that governments react to issue salience and posits that issue salience can
drive issue spending by government (Soroka 2003). A similar study by Soroka
and Lim (2003) assesses whether changes in healthcare spending are tied to public
priorities on the issue in the United Sates and the United Kingdom. The authors’
time series analysis indicates greater responsiveness to public opinion in the United
Sates, and one source of that difference may be issue definition in each country
(2003). While the media does not play a large role in this second analysis, it is
the type of work that would be well suited for considering the media’s role in
this public–elite relationship.
One of the most comprehensive examinations of the media–public–policy
triad is Soroka’s (2002) analysis of a wide variety of issues in Canada from 1985
to 1995. Among the range of results particularly relevant to this chapter is a
finding regarding what Soroka terms sensational issues. Environment proved to
be a good example of such an issue, demonstrating the power of media to affect
both public and policy agendas (p 118). Arguably, the impact of news coverage
about the drug issue on the public and the federal government discussed earlier
is another example of a media-driven sensational issue.
Contributing to the divide between elite and public media effects is a tradition
of scholarship that considers two very different notions of agenda setting—
one a communications perspective and one a policy oriented framework. In
communication studies, the concept of ‘agenda setting’ is primarily a theory
about media effects on citizens, or rather how media coverage of select issues
influences the issue priorities of the public, and potentially their voting behaviors
(Van Aelst et al. 2014). Since the seminal study of McCombs and Shaw (1972)
the popularity of the media agenda-setting approach among communication and
journalism scholars has remained persistent and increasing as a common media
effects concept (Van Aelst et al. 2014; Bennett and Iyengar 2008; Dearing and
Rogers 1996).
For the analysis of media effects on policy makers as on the public, the dominant
focus of agenda setting, the theory offers three distinct perspectives (McCombs
2014). In line with the seminal Chapel Hill study and dozens of subsequent studies,
agendas can be defined in terms of public issues, public figures, or other objects
of attention. The transfer of object salience from one agenda to another is now
273
referred to as basic agenda-setting effects or first-level agenda-setting effects. In

other words, the prominence of the objects of attention on one agenda influences
the prominence of the objects of attention on another agenda.
In turn, for every object on a first-level agenda, there is an agenda of attributes,
those characteristics or aspects that describe the object. And the transfer of salience
from one attribute agenda to another is an attribute agenda-setting effect. The
prominence of particular attributes on one agenda influences the prominence
of those agendas on another agenda. This is referred to as the second level of
agenda setting.
Recently, a third level of agenda-setting effect has been introduced, network
agenda setting (Guo and McCombs 2016). Here the focus is on the elements of
the network—objects, attributes, or both—in tandem with the strength of their
association with each other. At the first and second levels of agenda setting the
focus is on the salience of discrete elements—objects or attributes. At the third
level the focus is on the linkages among these elements and a much stronger
hypothesis of effects is tested: The salience of the network of elements on one
agenda influences the network of elements on another agenda.
A phrase from the title of the opening chapter in Walter Lippmann’s Public
Opinion—‘The World Outside and the Pictures in Our Heads’—can be used to
distinguish these three levels of agenda-setting effects. At the first level: What are
these pictures about? At the second level: What are the dominant characteristics
of these pictures? At the third level: What are the pictures?
From the political science perspective, the political agenda-setting approach deals
mainly with the limited attention of political actors to a wide range of political
issues (Van Alest et al. 2014). Building on the insights of Schattschneider (1960),
Cobb and Elder (1971) were among the first who investigated why some issues
managed to get the attention of decision makers. The news media was considered
one of the possible factors that could influence policy salience, however it was not
a well-explored notion (Van Aelst et al. 2014). Gradually the role of the media
become a more central component of political agendas but it was seldom the main
focus of attention (Van Aelst et al. 2014; Baumgartner and Jones 2009; Kingdon
1984). It was this window of opportunity that spurred the research noted in the
previous section and enabled a much richer and more complex understanding of
media effects across political institutions and varying policy processes.
While these two literatures remain separate still, we echo the call by some to
bridge the divide. Wolfe, Jones and Baumgartner (2013) note this disconnect
between media and policy agendas and suggest that bridging the divide means
drawing on strong attributes of both disciplines. Communications offers long-
use media coverage as a measure of attention while at the same time studying
policy agendas as a dependent variable (Wolfe et al. 2013). We also suggest that
bridging the divide also requires an understanding of the complex nature of both
media and policy agenda setting, such that studies primarily looking at elite or
mass politics do indeed have the potential to have down-the-line effects beyond
your intended target. Scholars should take the next step and ask, when we see
274
The media
media effects on a public, what does that mean for the next step in the political
process? What does the disproportionate nature of media attention by elite actors
mean for how individuals receive information and vice versa?
Conclusion
The current state of the media literature and its potential highlights just how
complex our political institutions are and the role that the media plays in adding
to that complexity. The media is more than just a tool for information processing
and the gatekeeper between elite and mass publics. The news media—and the
news generation process—are both critical and effectual to policy making across a
number of political institutions at both the local and federal levels of government.
What is clear from the assessment above is that the media are not part of a linear
political process—where the media informs the public and the public then
uses the information to relay to political representatives their preferences. The
media does have the ability to affect the policy process indirectly through public
priorities, but it also has direct effects on political elites that have very little to
do with public support or attention.
The media is often an active participant in the policy process by choosing which
issues to highlight, selecting frames for those issues, and shaping the scope of
debate. The media gives political leaders the platform and potential to influence
the parameters of the decision-making process because a core function of the
media is its role as a mechanism for conflict expansion.
Postlogue—the Trump administration

The inauguration of President Donald Trump not only meant potentially sweeping
change for public policy, but also historic change in political communication.
President Trump’s shift away from traditional media and toward social media or
partisan blogs has implications not only for the role of the media in Washington
DC, but also for how policy information is communicated to constituents. The
relationship between the president and the press became increasingly aggressive
through the 1970s and beyond (Clayman et al. 2016). While journalistic norms
may have changed, so too have the president’s strategic communications. Trump
has fostered this adversarial relationship between himself and the press, from the
campaign to his election to office. President Trump routinely turns to Twitter to
chide the New York Times and CNN for their #FakeNews reports, but what is
often lost in the back-and-forth between these political institutions is the policy
at hand.
Presidents have traditionally used the media to trumpet their policy agenda and
used their role as the ‘agenda setter-in-chief ’ to foster media attention on their
preferred issues. Presidents use their publicity and media attention to ‘go public’
on issues where they want public support (Kernell 1997). But instead of holding
press conferences on healthcare proposals, President Trump has largely shied
275
away from traditional media, which he characterizes as biased and self-serving.

Going public for President Trump means turning to Twitter—where there are
no journalistic norms—to promote his policies and opinions. More often than
not, policy is not discussed when he turns to Twitter given the friction between
policy complexity and 140-character messages. When he does discuss policy, it is
often in broad terms and tinged with political rhetoric—reflecting the polarizing
politics that also appears across American political institutions. The disregard for
traditional press and the embrace of social media could mean either new sources
of information for public policy for constituents or a shift away from those
discussions altogether.
References
Bang, S. S. (2013) Over the Red and the Blue: A Comparative Study on the Political
Twitter Use by the U.S. House of Representatives and the Korean National Assembly,
unpublished doctoral dissertation at the University of Texas at Austin.
Baumgartner, F. R. and Jones, B. (2009) Agendas and Instability in American Politics,
Baumgartner, F. R., De Boef, S. and Boydstun, A. E. (2008) The Decline of the
Death Penalty and the Discovery of Innocence, Cambridge: Cambridge University
Press.
Bennett, L. and Iyengar, S. (2008) ‘A New Era of Minimal Effects? The Changing
Foundations of Political Communication’, Journal of Communication 58: 707–31.
Boczkowski, P. (2010) News at Work: Imitation in an Age of Information Abundance,
Boydstun, A. E. (2013) Making the News: Politics, the Media, and Agenda Setting,
Boydstun, A. E., Hardy, A. and Walgrave, S. (2014) ‘Two Faces of Media
Attention: Media Storm Versus Non-Storm Coverage’, Political Communication
31(4): 509–31.
Brewer, M. and McCombs, M. (1996) ‘Setting the Community Agenda’,
Journalism and Mass Communication Quarterly 73: 7–16.
Caspi D. (1982) ‘The Israeli Elections of 1981’, State, Government and International
Relations, 19-20: 5-17. (Hebrew.)
Cobb, R. and Elder, T. (1971) ‘The Politics of Agenda-building: An Alternative
Perspective for Modern Democratic Theory’, Journal of Politics 33: 892–915.
Colistra, R. (2012) ‘Shaping and Cutting the Media Agenda: Television
Reporters’ Perceptions of Agenda and Frame Building and Agenda Cutting
Influences’, Journalism and Communication Monographs 14(2): 85–146.
Cook, T. E. (1998) Governing with the News: The News Media as a Political Institution,
Dahl, R. A. (1971) Polyarchy, New Haven, CT: Yale University Press.
Dearing, J. W. and Rogers, E. M. (1996) Communication Concepts 6: Agenda-
setting, Thousand Oaks, CA: Sage.
276
The media
Edwards G. C. E and Wood, B. D. (1999) ‘Who Influences Whom? The President,

Congress, and the Media’, American Political Science Review 93(2): 327–44.
Gilbert, S., Eyal, C., McCombs, M., and Nicholas, D. (1980) ‘The State of the
Union Address and the Press Agenda’, Journalism Quarterly 57: 584–88.
Glassman, M. E., Straus, J. R., and Shogan, C. J. (2009) ‘Social Networking and
Constituent Communication: Member Use of Twitter During a Two-Week
Period in the 111th Congress’, CRS Report, Washington, DC: Congressional
Research Service.
Gonzenbach, W. (1996) The Media, the President and Public Opinion, Mahwah,
NJ: Lawrence Erlbaum.
Johnson, T. J., Wanta, W., Byrd, J. T., and Lee, C. (1995) ‘Exploring FDR’s
Relationship with the Press: A Historical Agenda-setting Study’, Political
Communication 12: 157–72.
Kernell, S. (1997) Going Public: New Strategies of Presidential Leadership, 3rd ed.
Washington, DC: CQ Press.
Kingdon, J. (1984) Agendas, Alternatives, and Public Policies, Boston, MA: Little
Brown.
Kiss, S. J. (2013) ‘Legislation by Agenda-setting: Assessing the Media’s Role in
the Regulation of Bisphenol A in the U.S. States’, Mass Communication and
Society 16: 687–712.
Ku, G., L.L. Kaid, and M. Pfau (2003) ‘The impact of Web site campaigning on
traditional news media and public information processing’, Journalism & Mass
Communication Quarterly, 80(3): 528–47.
Lei, G. and McCombs, M. (2016) ‘The Power of Information Networks: New
Directions for Agenda Setting’, Routledge Studies in Global Information, Politics
and Society, No. 8, New York, NY: Routledge: 175–89.
McCombs, M. E. (2004) Setting the agenda: The mass media and public opinion,
Malden, MA: Blackwell.
McCombs, M. E. (2014) Setting the Agenda: The Mass Media and Public Opinion
(2nd edn), Cambridge: Polity Press.
McCombs. M. E. and Shaw D. L. (1972) ‘The agenda-setting function of the
mass media’, Public Opinion Quarterly, 36: 176–87.
Mead, T. (1994) ‘The Daily Newspaper as Political Agenda Setter: The Charlotte
Observer and Metropolitan Reform’, State and Local Government Review 26(1):
27–37.
Miller, M., Andsager, J. and Riechert, B. (1998) ‘Framing the Candidates in
Presidential Primaries: Issues and Images in Press Releases and News Coverage’,
Journalism and Mass Communication Quarterly 75: 312–24.
Nelson, B. (1984) Making an Issue of Child Abuse: Political Agenda Setting for Social
Problems, Chicago, IL: University of Chicago Press.
Protess, D., Cook, F., Doppelt, J., Ettema, J., Gordon, M., Leff, D. and Miller,
R. (1991) The Journalism of Outrage: Investigative Reporting and Agenda Building
in America, New York: Guilford.
277
Rose, M. and Baumgartner, F. R. (2013) ‘Framing the Poor: Media Coverage and
US Poverty Policy, 1960–2008’, Policy Studies Journal 41(1): 22–53.Russell, A.
(2014) ‘#Corresponding Conversations: Public Issue Priorities and Elite Issue
Priorities on Twitter’, paper presented at the Comparative Agendas Project
Conference in Konstanz, Germany.
Russell, A. (2015) ‘Individual Agendas: Communicating Domestic Violence
Information in 140 Characters’, paper presented at the Comparative Agendas
Project Conference in Lisbon, Portugal.
in America, New York: Holt, Rinehart and Winston.
Scheufele, D. A. and Tewksbury, D. (2007) ‘Framing, Agenda Setting, and
Priming: The Evolution of Three Media Effect Models’, Journal of Communication
57(1): 9–20.
Sevenans, J., Walgrave, S., and Vos, D. (2015) ‘Political Elites’ Media Responsiveness
and Their Individual Political Goals: A Study of National Politicians in Belgium’,
Research and Politics, 2(3):1–7.
Shoemaker, P. (ed) (1989) Communication Campaigns about Drugs, Hillsdale, NJ:
Lawrence Erlbaum.
Smith, C. (1993) ‘News Sources and Power Elites in News Coverage of the Exxon
Valdez Oil Spill’, Journalism and Mass Communication Quarterly 70: 393–403.
Soroka, S. (2002) Agenda-Setting Dynamics in Canada, Vancouver: UBC Press.
Soroka, S. (2003) ‘Media, Public Opinion, and Foreign Policy’, The International
Journal of Press/Politics 8(1): 27–48.
Soroka, S. and Lim, E. (2003) ‘Issue Definition and the Opinion-Policy Link:
Public Preferences and Health Care Spending in the US and UK’, British Journal
of Politics and International Relations, 5(4): 576–93.
Sparrow, B. H. (2006) ‘A Research Agenda for an Institutional Media’, Political
Communication 23(2): 145–57.
Straus, J. R., Glassman, M., Shogan C. J., and Smelcer, S. N. (2013)
‘Communicating in 140 Characters or Less: Congressional Adoption of Twitter
in the 111th Congress’, PS: Political Science and Politics 46(1): 60–6.
Trumbo, C. (1995) ‘Longitudinal Modelling of Public Issues: An Application
of the Agenda-setting Process to the Issue of Global Warming’, Journalism
Monographs: 1–57.
Van Aelst, P., Thesen, G., Walgrave, S., and Vliegenthart, R. (2014) ‘Mediatization
and Political Agenda-Setting: Changing Issue Priorities?’, in F. Esser and J.
Stromback (eds), Mediatization of Politics, Basingstoke: Palgrave Macmillan
200–20.
Van Noije, L., Oegema, D., and Kleinnijenhuis, J. (2008) ‘Loss of Parliamentary
Control Due to Mediatization and Europeanization: A Longitudinal and
Cross-sectional Analysis of Agenda Building in the United Kingdom and the
Netherlands’, British Journal of Political Science 38(3): 455–78.
278
The media
Walgrave S. and Van Aelst, P. (2006) ‘The Contingency of the Mass Media’s
Political Agenda Setting Power: Toward a Preliminary Theory’, Journal of
Communication 56(1): 88–109.
Walgrave S, Soroka S., and Nuytemans, M. (2008) ‘The Mass Media’s Political
Agenda-setting Power: A Longitudinal Analysis of Media, Parliament, and
Government in Belgium (1993 to 2000)’, Comparative Political Studies 41(6):
814–36.
Wanta, W. and Foote, J. (1994) ‘The President–News Media Relationship: A
Time-series Analysis of Agenda Setting’, Journal of Broadcasting and Electronic
Media 38, 437–48.
Wildavsky, A. (1964) The Politics of the Budgetary Process, Boston, MA: Little Brown.
Wober, M. J. (2002) Agenda Cutting: The Hidden Twin of Agenda Setting’,
Media Tenor Quarterly 3: 15–19.
Wolfe, M., Jones, B., and Baumgartner, F.R. (2013) ‘A Failure to Communicate:
Agenda Setting in Media and Policy Studies’, Political Communication 30(2):
175–92.
Workman, S., Jones, B. D., and Jochim, A. (2009) ‘Information Processing and
Policy Dynamics’, Policy Studies Journal 37: 75–92.
279
FIFTEEN
Think tanks and policy analysis

Andrew Rich
At a moment when experts and analysts have become ubiquitous in American

politics, their numbers have become diluted and distorted amid the ranks of think
tanks. Both in Washington, DC and in state capitals, the number and variety of
sources of analysis have exploded during the last half century. Many are housed
in think tanks, public policy research organizations with origins that date back
to the early 1900s. This chapter builds on my research over the past two decades,
which examines the roles and impact of think tanks, and it identifies some of
the opportunities and challenges for policy analysts within them. As the ranks
of analysts have grown and the number of think tanks has proliferated, clarity
around the purposes and influence of think tanks has diminished.
For much of the 20th century, policy experts were understood as neutral,
credible, and above the rough and tumble of policy making. Progressive reformers
early in the 20th century turned to the burgeoning social sciences for salvation.
Reformers believed that the new ranks of policy experts trained in the new social
science disciplines at universities would be capable of usurping patronage politics;
experts would develop real solutions to the social and economic instabilities
that stemmed from the Industrial Revolution. Many would be housed at newly
formed think tanks, understood as independent, non-interest-based, nonprofit
political organizations that produce and principally rely on research and expertise
to obtain support and to influence the policy-making process (Rich 2004).
American politics and society would be better informed and much improved as
a result of think tank efforts.
During this period, the training of new policy experts became a central focus
of reformers, with the creation of schools for policy analysts at leading universities
and of agencies within government departments that produced research and
evaluation studies for decision makers. Twentieth-century scholars observed these
developments, and they contributed to the prevailing understanding of experts
in American policy making: as important background voices that bring rational,
reasoned analysis to long-term policy discourse based on the best evidence
available.
For much of the last century, this assessment was basically accurate; experts
and analysts fulfilled these mandates. By the beginning of the 21st century,
however, the ranks of real-life policy experts scarcely conformed to the promise
of making policy choices clearer and more rigorous and decisions necessarily
more rational. Many of these analysts were based at a growing number of think
tanks in Washington, DC and in state capitals, and it was as common for them
281
and their organizational homes to reflect clear ideologies and value systems as
seemingly unbiased commitments to objectivity or neutrality.
Think tanks have contributed to a transformation in the role of policy analysts
in American policy making. Many policy experts now behave like advocates.
They make themselves highly visible, and their work is highly contentious. They
more actively market their work than before; their work, in turn, often represents
preformed points of view rather than even attempts at neutral, rational analysis.
These developments apply particularly to a group of identifiably ideological,
particularly conservative think tanks founded since the 1970s.
Early national think tanks

The first think tanks embodied the promise of neutral expertise. They formed as
the social science disciplines of economics, sociology, and political science became
established fields of inquiry and as confidence grew in the uses of expertise as a
means for correcting social problems (Critchlow 1985). Through the first half
of the 20th century, think tanks largely sought to identify government solutions
to public problems through the detached analysis of experts. Think tank scholars
wrote on topics relevant to policy makers but typically maintained a distance from
the political bargaining that can take place in the final stages of the policy-making
process. This analytic detachment was a behavior to which analysts held fast and
which fostered an effective relationship between experts and policy makers. With
this approach, between 1910 and 1960, the analysts at think tanks often influenced
decision makers in government, who were receptive to their work (Smith 1991).
The Brookings Institution, formed in Washington, DC, in 1916, informed the
creation of the Bureau of the Budget early in the century (Smith 1989). The
RAND Corporation, formed in Santa Monica, California, in 1948, developed
systems analysis for the Department of Defense. The influence of these think
tanks was significant, and their research served important political purposes. But
the policy process did not compel experts to become directly involved in partisan
battles. Experts were mobilized by policy makers to prescribe possibilities for
change.
Through the first two thirds of the 20th century, while think tanks at times
produced politically contentious research, their input was sought and generally
respected by policy makers. Although think tanks were not the only source of
expertise, they were prominent, consistent, and visible providers.
Through the 1950s and 1960s, the ideas and expertise produced by think tanks
generally reflected a near consensus that developed among elites about the need
for government management of social and economic problems. Even when policy
entrepreneurs of a conservative bent established new think tanks, they usually
followed prevailing organizational norms, hiring academically trained staff and
avoiding any appearance of having links to a single point of view or political
party (Weaver 1989; Abelson 1995).
282
The ascendance of conservative ideology
By the end of the 1960s, as government grew larger, the desirability and possibility
of achieving social change through government programs began to be challenged.
Some of the problems themselves—notably civil rights for African Americans
and the Vietnam War—were highly divisive. Increasingly, critics described the
government as ineffective and overextended, both at home and abroad. Combined
inflation and unemployment in the 1970s— ‘stagflation’—contributed further
to the decline of confidence in ‘expertly devised’ government programs as well
as to doubts about Keynesian principles generally.
The growth of government fueled organization among those who disapproved
of it. Conservatives, by the 1970s, were united by strong opposition to communism
and a shared belief that government resources were better channeled toward the
nation’s defense and the fight against communism than to what they viewed as
bloated and ineffective domestic programs. A group of relatively small, politically
conservative foundations and wealthy individuals provided support for applying
these principles to public affairs. More than a dozen conservative foundations
and individuals formed a nucleus of support of conservative organizations that
emerged through the 1970s and 1980s. Think tanks were prominent among them,
and they collectively defined a movement of ideas (Rich 2005; Stefancic and
Delgado 1996; Stahl 2016). The explicit intent of their efforts was to discredit
the pro-government convictions that had dominated American politics since the
New Deal. Avowedly ideological, contentious, and politicized ideas and expertise
became tools in these endeavors.
The formation of the Heritage Foundation in 1973 was a turning point.
Heritage was the first of a new breed of think tanks that combined what Kent
Weaver described as ‘a strong policy, partisan, or ideological bent with aggressive
salesmanship and an effort to influence current policy debates’ (Weaver 1989).
The political entrepreneurs who started the Heritage Foundation were former
congressional staff and sought to create a highly responsive apparatus that could
react quickly to hostile proposals in Congress. By the late 1970s, as Lee Edwards
observed in his history of the Heritage Foundation for its 25th anniversary, ‘an
increasingly confident Heritage Foundation set an ambitious goal: to establish
itself as a significant force in the policy-making process and to help build a new
conservative coalition that would replace the New Deal coalition which had
dominated American politics and policy for half a century’ (Edwards 1997).
The proliferation of think tanks in the United States

To put numbers behind these trends, between 1970 and 2000, the number of
think tanks operating in the United States grew from fewer than 60 to more
than 300 (Rich 2004). These think tanks can be classified into three categories:
conservative, liberal, or what I call ‘centrist or of no identifiable ideology.’
Think tanks are grouped as conservative if they refer to a particular concern
283
for promoting the free market system, limited government, individual liberties,
religious expression, or traditional family values. Think tanks are grouped as
liberal (in the contemporary American sense) if they refer to a particular concern
for using government policies and programs to overcome economic, social, or
gender inequalities, poverty or wage stagnation, or if they call for progressive social
justice, a sustainable environment, or lower defense spending. Organizations not
classified into either broad ideological category reflect a ‘centrist or no identifiable
ideology’ (Rich 2004).
This last group, think tanks of centrist or no identifiable ideology, made up
the largest single category of think tanks in 2000 (141 think tanks or 45% of the
total). This finding is not surprising, given the long history of think tanks in the
United States of producing balanced or non-ideological research (Smith 1991).
These think tanks are mostly staffed by neutral analysts. What is remarkable,
however, is that a majority of think tanks since 2000 are identifiably ideological
in character, either conservative or liberal. In 2000, 165 of the 306 think tanks
in existence—54%—were avowedly conservative or liberal, broadly defined. By
contrast, in 1970 only 14 of the 59 think tanks in existence were identifiably
conservative or liberal; three quarters of the 59 pre-1970 organizations were
centrist or of no identifiable ideology.
Among the greatly expanded ranks of avowedly ideological think tanks,
conservative think tanks had come to substantially outnumber liberal
organizations. Of the 165 ideological think tanks, roughly two thirds (65%) were
avowedly conservative; only one third (35%) of them were identifiably liberal. The
differences in number between identifiably conservative versus liberal think tanks
were especially pronounced at the state and local level. By the mid-1990s, a full
one third of the think tanks operating in the United States—100 organizations—
were principally concerned with state and local issues, as opposed to national
matters. At the state level, conservative think tanks emerged at an overall rate of
3.5 each year between 1985 and 1995, more than three times the rate of liberal
organizations (0.9 each year). Among national think tanks, conservative think
tanks emerged at a rate (2.6 per year) that was twice that of liberal organizations
(1.3 each year) between 1985 and 1995. Nationally focused think tanks of centrist
or no identifiable ideology emerged at a rate of 2.7 each year through this period.
At the state level, this category of think tank emerged at a rate of 1.3 each year.
By the end of the 20th century, 47 identifiably conservative think tanks were
operating in 34 of the 50 states; by contrast only 22 liberal organizations were
visible in just 15 states. Almost half of the state and local think tanks (44%)
operating were conservative, compared to a bit less than one third of the nationally
focused organizations that were conservative. At both the state and national levels,
identifiably liberal think tanks made up only about one fifth of all organizations.
The asymmetries between conservative and liberal think tanks by 2000 went
beyond differences in numbers; conservative think tanks also tended to have
more resources and broader missions than their liberal counterparts. Whereas
conservative think tanks outnumbered liberal organizations by a ratio of roughly
284
2 to 1, conservative think tanks outspent liberal think tanks by more than 3 to 1

in the United States. In 1995, the total resources of conservative, state and locally
focused think tanks were roughly $28.4 million, compared to $8.8 million for
liberal organizations at the state and local level (Rich 2004).
In the first decades of the 21st century, the growth of ideological think tanks—in
number and size—has continued. Several more identifiably progressive think tanks
have emerged that begin to diminish the asymmetry between conservative and
liberal think tanks. Most notably, the Center for American Progress was created
in 2003 by former White House Chief of Staff John Podesta and a set of high
net worth progressive donors who believed that liberals should invest not just in
elections but also in expertise and ideas in order to challenge conservatives. By
2014, the Center had an annual budget of more than $40 million annually and
a wide-ranging, multi-issue focus on domestic, economic, and foreign policy.
All of these data combine to demonstrate a strong trend among the ranks of
think tanks in the United States. In the almost 50-year period since 1970, as
the number of think tanks in the United States has well more than quadrupled,
ideological think tanks have emerged in substantially greater numbers than think
tanks of no identifiable ideology. Through the 1970s and 1980s, advocacy-
oriented think tanks modeled after the Heritage Foundation proliferated, first on
the right but also on the left. Some of the older institutions, like the American
Enterprise Institute, adapted to become more advocacy oriented. The staffs of
these organizations tended to be ideologically homogeneous, and their leaders
used research and analysis as vehicles to advance their underlying ideologies.
This changed the environment for think tanks in Washington, DC, and in many
state capitals. These developments also altered the atmosphere for policy analysts.
The war of ideas

Think tanks have become a driving force in what is often called a ‘war of ideas’
in American politics, a war that conservatives have been winning. They are a key
organizational vehicle for promoting ideas. They offer a way for ideas to gain
adherents and to inform the substantive underpinnings of policy debates. And
think tanks are not just an engine for ideas. They are also a reflection of ideas.
Think tanks follow a variety of missions and pursue many strategies, some more
successful than others. As much as ideas—and differences in ideology—inform the
content and direction of American policy making, ideas also inform differences
in the missions and strategies of the organizations that aim to promote them.
Think tanks that may express loose commitments to the same goals—informing
policy making through research and ideas—understand these goals in different
ways and follow far from uniform strategies for achieving them. Knowingly and
not, differences in their missions and strategies are informed by aspects of the
very ideas that they seek to advance as organizations. In fact, there are important
differences in how conservative and liberal think tanks are organized, staffed, and
run, even when the leaders of both conservative and liberal think tanks seem to
285
have the same broad objectives. These differences have serious consequences not
just for think tanks themselves, but also for the power of their efforts—and of
analysts, experts, and ideas, generally—in American policy making.
A little more than a decade ago in July 2003, I administered a mail survey among
leaders of 115 of the 117 state-based think tanks nationally. The survey inquired
about the histories, missions, and strategies of these organizations, beginning
with questions about leader and staff backgrounds.1 The findings generally reveal
differences in the leadership of think tanks. Among conservative think tanks, a
significant plurality—almost 40%—of those who were the organizations’ first
leaders came from the private sector. By contrast, almost two thirds of those who
formed liberal think tanks came out of state government or from the nonprofit
advocacy community. These leadership differences seem to have a bearing on
decisions about how to organize operations and decision making.
The survey asked think tank leaders about the criteria they use when selecting
or promoting full-time staff. Out of nine response options, leaders of conservative
think tanks most often named political or ideological orientation as the most
important consideration when hiring staff; for liberals, ideology was far down
the list.2 Almost three quarters of the leaders of conservative think tanks named
political or ideological orientation as most or very important in making decisions
about whom to hire (73.6%). By contrast, less than half of the leaders of liberal
think tanks named ideology as most or very important (42.2%). Among the other
top priorities for the leaders of conservative think tanks were issue expertise
(61.8%), media and public affairs experience (35.3%), and a record of publication
(32.3%).
By contrast, in addition to being less concerned about political or ideological
orientation, the leaders of liberal think tanks expressed less concern with media
and public affairs experience (21.1%) and a record of publication (5.1%). Instead,
liberals place a premium on advanced degrees (either policy degrees, 42.1%,
or PhDs, 31.6%) and experience in government (36.9%), along with issue
expertise (57.9%). Leaders of conservative think tanks showed far less interest in
advanced degrees (23.5% for policy degrees and 8.8% for PhDs) and experience
in government (20.5%).
These results about the hiring preferences of think tank leaders are consistent
with whom think tank leaders report that they actually employ. Almost three
quarters of conservative think tank leaders indicated that all or some staff came
from the business community or private sector. By contrast, liberal think tank staff
came from the nonprofit advocacy community in almost the same proportion.3
One more difference between survey responses from conservative and liberal
think tanks is worth noting: how they rank the significance of different kinds
of staff activities to their organizations. Respondents were asked, ‘How do
you rate the importance of the following activities in relation to fulfilling your
organization’s mission?’ They were provided 10 choices.4 Leaders of both
conservative and liberal think tanks most often named advising policymakers and
286
the news media about their research products as either most or very important to
fulfilling their mission.5 But from there, differences quickly emerged.
The leaders of conservative think tanks were significantly more likely to name
‘advising legislators on immediately pending policy issues’ and ‘shaping public
opinion on policy issues’ as high priorities compared with the leaders of liberal
think tanks. Three quarters of the leaders of conservative think tanks named
advising legislators as most or very important (76.5%), whereas just more than
half of liberal think tanks named that as important (57.9%). Likewise, three
quarters of the leaders of conservative think tanks named shaping public opinion
as important (73.5%), while only half of the leaders of liberal think tanks report
that as important (52.6%).
The leaders of liberal think tanks, by contrast, named informing nonprofit
advocacy groups about their research as important at much higher rates than
those at conservative think tanks. More than three quarters of the leaders of
liberal think tanks named the nonprofit advocacy community as very or most
important (78.9%), whereas only one fifth of the leaders of conservative groups
named it as a priority (20.5%).
Differences in the strategic priorities of think tanks

Overall, these findings begin to illustrate important ways that ideology affects
the organization of think tanks and the environment for policy analysts. The
differences in their priorities with respect to staffing decisions offer perhaps the
starkest contrast between conservative and liberal think tanks. Consistent with
a view that ideas matter—and that differences in ideology are important—the
leaders of conservative think tanks place substantial importance on the ideological
and political predilections of those they hire. Conservative think tanks are
interested in hiring politically conservative people above all else.
Next in importance for conservative think tanks is that those that they hire be
prepared to make a contribution to the war of ideas. Conservative think tank staff
need to have an issue expertise; they need to have experience in media and public
affairs. And staff should have a record of publication. The leaders of conservative
think tanks were much more likely than their liberal counterparts to express a
preference for staff who are ready to hit the ground running in the public battles
to shape the terms of American policy debate. Responses to the question about
staff qualifications were wholly consistent with the view that conservatives see
think tanks as idea promoters. This finding is supported by in-depth interviews
with many national think tank presidents.
The legitimacy of this understanding of the role of think tanks among
conservatives was also confirmed in a final survey question. Think tank leaders
were asked to choose from among three descriptions of think tanks: as places for
(1) public intellectuals, (2) policy researchers, or (3) issue activists. The majority of
conservative think tank leaders (56.0%) selected the response that described think
tanks as a place for ‘public intellectuals—for those with well-formed ideas about
287
the role for government and talents in producing and organizing policy research
about these ideas in ways that might inform policymaking.’ For conservatives,
think tanks are important as promoters of ideas; the research that takes place at
think tanks is in the service of a broader ideological agenda.
This is how we understand the power of the best-known national conservative
think tanks—the Heritage Foundation, Cato Institute, American Enterprise
Institute. Conservative think tanks have used this power to advance successfully
a range of conservative policies, including deregulation of a number of industries
in the 1970s and 1980s, and welfare reform in the 1990s. In the early 2000s, the
neoconservative scholars based at conservative think tanks played a central role
in making the case for the war in Iraq—a successful story of influence on their
part that has informed what by most accounts was a poorly conceived and badly
managed war. This same sort of power—as promoters of ideas—is evident among
the ranks of state-focused conservative think tanks.
By contrast, the leaders of liberal think tanks selected the description of think
tanks as places for ‘public intellectuals’—as places for those with well-formed
ideas—least often among the three choices offered. Instead, they were split
between those who described think tanks as ‘for policy researchers—for those
with interest in the researchable dimensions of particular issue areas and talents
in producing applied policy research that might inform policymaking’ (31.6%),
and those who saw think tanks as ‘for issue activists—for those with concerns
about specific policies and populations and talents in producing research and
organizing citizens in ways that might inform and affect policymaking’ (36.8%).
The leaders of liberal think tanks view their organizations first and foremost as
research organizations, not as idea promoters. Yet these are the think tanks that
one might assume are poised to do battle for the left in the war of ideas. In fact,
this is the stated reason why some of them were formed. Instead, the results on
this and other questions in the survey reveal identifiably liberal think tanks as
virtually indistinguishable in many ways from think tanks coded and confirmed
as of no identifiable ideology.
The leaders of liberal think tanks are most concerned with hiring staff with
issue expertise and with research/academic credentials, rather than staff with
media experience or with records of popular publication. For the leaders of
liberal think tanks, it is most important that the organization be able to produce
credible, rigorous research rather than promote that research or fit it into a broader
ideological project. Research is the product of think tanks, and its completion is
the core purpose of the organization.
When ideology affects organization

The findings from the survey suggest that many of the differences in how think
tanks approach their missions are closely related to differences that come out of
the ideologies they seek to promote. As a practical matter, this conclusion suggests
a far bigger problem for liberals than for conservatives if engaging in the war of
288
ideas in American politics is the goal. To the extent that this war is ongoing and
think tanks are important to it, conservatives have the advantage.
Conservatives have more think tanks that are better funded. The dramatic
growth of conservative think tanks was made possible principally with support
from a small corps of newer conservative foundations, such as the Bradley,
Smith Richardson, and the Sarah Scaife foundations. Before the 1970s, many
conservative foundations and their patrons reviled government so much that
they refused to support efforts related to what was going on in Washington. But
with the advent of increased government regulation in the late 1960s, the leaders
of these foundations wanted to stop the tide of government activism. Funding
organizations to fight the war of ideas became their way of doing it (Rich 2005).
In recent years, the Koch brothers have become even more generous donors to
conservatives—and especially libertarian—think tanks (Drezner 2017).
For liberal think tanks, the problem has been that the mainline/progressive/
liberal foundations are often not organized to provide support to progressive
think tanks or other organizations in the broad-based war of ideas—or even to
see that as their role. These foundations tend to be organized by issue area; as a
result, think tanks on the left tend to be organized by issue area as well—around
women’s issues, poverty, or the environment—rather than taking on the broad
range of issues with which Congress and the president deal. The consequence is
that these liberal organizations are not organized to do battle in the same ways
as their conservative counterparts, across a broad range of topics (Rich 2005).
Even when liberal think tanks have resources, they are typically not structured
or organized to be effective counterweights to conservative organizations in the
war of ideas. Donors have an effect on them. In addition, liberals approach think
tanks from a different point of reference—a century-long tradition of investing
in the production of objective policy research and analysis. Since the creation
of the social science disciplines during the Progressive Era, liberals have been
committed to the view that research is essential to an informed policy-making
process. Research may lead to ideas—but ideas that are pragmatic and well
reasoned, not value laden. The tradition of research among liberals is one that
‘tends to minimize disagreement over political values, and at times seems to ignore
underlying values if not wish them away altogether’ (Smith 1989).
In this context, liberals are inclined to approach a war of ideas in American
politics by, in some sense, denying its very legitimacy. Ideological battle is political
nonsense; the results of rigorous, objective research can—and should—best inform
the appropriate possibilities for government and society. Politics should not be
about winners and losers so much as it should be about building consensus, and
research can point the way toward that consensus. To view the role of research—
and research organizations—in any other way would be inappropriate. It is the
obligation of the disinterested expert to develop optimal policy-based solutions.
The findings here lend support to this conclusion. These are attitudes pervasive
not just among think tank leaders at the state level, but among those who run
national organizations as well.
289
These findings suggest that the puzzle for the ranks of policy analysts—at think
tanks and elsewhere—is complex. The neutral analyst is in part a remnant of an
earlier era, a period in the 1950s, 60s, and 70s when there was far greater consensus
among policymakers and policy elites in the United States about the desirability
of government programs and government intervention in the economy. During
that era, the liberal think tanks of today did not need to behave in an ideological
way when promoting government progress because there was a ‘liberal consensus’
behind which to hide. These think tanks and the analysts found in their ranks have
not done so well at making the transition to the more contentious environment
of the past few decades.
Conservatives begin their thinking on these issues in a very different place.
They begin from the perspective that ideas and values motivate—rather than result
from—research. In their view, all research is ideological insofar as ideas or ideology
at least inform the questions that so-called ‘neutral’ researchers ask. There is no
such thing as disinterested expertise or the disinterested expert. Instead, there
are ‘permanent truths, transcending human experience, [that] must guide our
political life’ (Smith 1989, p 192). These truths motivate research, and research
is a means to a more important end: realizing the ideas that are a reflection of
this core truth (Medvetz 2012, Drezner 2017).
Conservatives believe at a fundamental level that ideas have power (Weaver
2013). Ideas inform preferences and behavior far more than research. And ideas
not only are—but should be—more powerful than expertise. One engages in (or
supports) policy research for the same reasons one supports political advocacy:
because both contribute to the larger causes of shifting the terms of debate in
American policy making and to amplifying the power of conservative ideas.
For conservatives, the war of ideas provides the rationale for creating think
tanks. Think tanks are the engine for conservative ideas. And conservatives apply
an entrepreneurial spirit to their organization with the view that in a war of ideas,
conservative ideas need a machinery—an artillery—to promote and disseminate
them from every angle possible. Conservative think tanks should operate across
the full range of issue domains, and they should be poised to interject ideas into
any issue debate that captures the attention of policymakers or the public. This
has been the guiding philosophy of everything from the Heritage Foundation
and Cato Institute at the national level to the Mackinaw Institute for Public
Policy in Michigan and the Manhattan Institute in New York City at the state
and local levels.
Thanks to their alternative views about research, liberals have a much more
difficult time reconciling the formation of research organizations with the
promotion of ideas in American policy making. Rather than approaching think
tanks with a commitment to advancing a particular worldview as their main
priority, liberals view think tanks with a pragmatic eye, relying on them to produce
research that might speak to the policy needs of different issue domains. As the
president of one liberal think tank put it to me:
290
‘The important thing for us, and it’s not true—and I don’t say this
purely out of a spirit of rivalry and competitiveness—but it’s not true,
for example, for the Heritage Foundation. They don’t really care
whether their numbers meet academic standards. For us, it’s a question
of survival. We know that we can’t make it unless we continue to be
credible to places with our numbers. So we try to be bold politically
but we spend a lot of energy making sure our numbers are right’
(Author interview 2005).
This is a challenge for all policy analysts across the United States and for all who
care about the role and impact of analysis in U.S. policy making.
The influence of think tanks

By 2012 more than 320 think tanks were active in national and state policy
making. These are institutions that matter. Yet, despite their greater numbers,
the nature and extent of their influence was questionable. As the ranks of think
tanks have grown and diversified, beyond complicating an explanation of what
think tanks are, they have contributed to confusion about how analysis should
be received and understood among journalists as well as policymakers and policy
staff in many corners of government. There is research to support every point
of view imaginable—not just ideological work but also analysis that reflects the
preferences of corporations, foreign governments, and lobbyists, all of which have
now taken to funding think tanks (Lipton et al. 2016). In the face of all of this,
the consumers of think tank work are often attracted to that which supports their
existing point of view rather than that which might reflect the best data and the
most rigorous methodologies. The distinctions between quality commentary and
careful analysis are all too often difficult for the intended audiences of expertise
to discern. That can lead to all of it being dismissed.
Perhaps the good news from this, at least for those who care about the influence
of analysis in American policy making, is that the cumulative effect for think tanks
of the organizational developments that I describe in this chapter is that think
tanks often focus more of their efforts on producing commentary about urgent
and immediate policy decisions rather than intervening at the earlier stages of
policy making. This might be seen as good news because at these latter stages
of the policy process, the commentary produced by think tanks often serves as
little more than ammunition for policy makers who need public justification for
their preferred policy choices.
The more important lessons for analysts, however, are three fold. First, given
the more complex organizational environment among think tanks in the United
States (not to mention among interest groups, trade associations, and lobbyists,
handled elsewhere in this volume), policy analysts must be clear-minded about
the organizations with which they affiliate. They must be sure that their personal
and professional goals conform with the purposes of the organizations with which
291
they affiliate and the role of analysis in the missions of these organizations. Second,
in this more complex and more densely populated organizational environment,
analysts must be more intentional about who constitute their intended audiences
and how best to reach them in order to have impact. The policy and media
audiences for analysis are deluged with information; they cope by often focusing
on well-known and trusted sources. As a result, analysts must select their audiences
carefully and work to establish trust with them. Lastly, the lesson for analysts—at
think tanks and elsewhere—is that they must be dogged in the promotion of
their work among relevant audiences. The times when good analysis spoke for
itself are in the past, if they ever existed at all. Policy analysts must combine the
pursuit of rigorous findings with relentless marketing of themselves and their
work in order to break through and to make a difference.
Notes
1 I received 78 responses, a 67.8% response rate, from think tanks that were broadly representative
of the larger population of state think tanks with respect to ideology, along with geography
and size. I received responses from 34 conservative think tanks, 19 liberal think tanks, and 25
think tanks of no identifiable ideology.
2 The response choices for this question were: (1) specific issue expertise, (2) media/public affairs
experience, (3) coherent/appropriate political or ideological orientation, (4) record of previous
publication, (5) advanced policy degree (MA, MPA, MPP), (6) advanced research degree (PhD),
(7) experience working in politics, (8) experience working in/around government, (9) academic
experience.
3 State government was the second most frequently named background for think tank staff by both
conservative and liberal think tanks. And government was the most often named background
characteristic raised by leaders of think tanks of no identifiable ideology.
4 The response options were (1) advising legislators on immediate pending policy issues, (2)
advising legislative staff on immediately pending policy issues, (3) advising Executive Branch
officials on immediately pending policy issues, (4) advising the news media about immediately
pending policy issues, (5) informing the news media about research products, (6) informing
nonprofit advocacy groups about research products, (7) informing lobbyists and/or trade
groups about research products, (8) informing policymakers (legislators and executive branch)
about research products, (9) informing the policy research community (e.g., other think tanks,
academics) about research products, (10) shaping public opinion on policy issues.
5 85.3% of leaders of conservative think tanks named this response as most or very important,
and 79.0% of liberal think tanks did the same.
References
Abelson, D. E. (1995) ‘From Policy Research to Political Advocacy: The Changing
Role of Think Tanks in American Politics’, Canadian Review of American Studies
25: 93–126.
Critchlow, D. T. (1985) The Brookings Institution 1916–1952: Expertise and the Public
Interest in a Democratic Society, DeKalb, IL: Northern Illinois University Press.
Drezner, D. W. (2017) The Ideas Industry, New York: Oxford University Press.
Edwards, L. (1997) The Power of Ideas, Ottawa, IL: Jameson Books, Inc.
Lipton, E., Confessore, N. and Williams, B. (2016) ‘Think Tank Scholar or
Corporate Consultant? It Depends on the Day’, The New York Times, August 8.
292
Medvetz, T. (2012) Think Tanks in America, Chicago, IL: University of Chicago

Press.
Rich, A. (2004) Think Tanks, Public Policy, and the Politics of Expertise, New York:
Cambridge University Press.
Rich, A. (2005) ‘The War of Ideas’, Stanford Social Innovation Review, Spring:
pp 18–25.
Smith, J. A. (1989) Brookings at Seventy-Five, Washington, DC: The Brookings
Institution.
Smith, J. A. (1991) The Idea Brokers: Think Tanks and the Rise of the New Policy
Elite, New York: The Free Press.
Stahl, J. (2016) Right Moves: The Conservative Think Tank in American Political
Culture since 1945, Chapel Hill, NC: University of North Carolina Press.
Stefancic, J. and Delgado, R. (1996) No Mercy: How Conservative Think Tanks
and Foundations Changed America’s Social Agenda, Philadelphia, PA: Temple
University Press.
Weaver, R. K. (1989) ‘The Changing World of Think Tanks’, P.S. Political Science
and Politics, September 22: 563–79.
Weaver, R. K. (2013) Ideas Have Consequences, Chicago, IL: University of Chicago
Press.
293
Part Four
Policy analysis education and
impact internationally
295
SIXTEEN
Public policy education in

the United States
Michael O’Hare
Introduction
Policy analysis (i) predicts futures conditional on alternative policies’ adoption, (ii)
estimates outcomes of interest for each policy, (iii) combines these estimates into a
scalar (or at least dimensionally reduced) measure of social welfare, to (iv) inform
political choice of a ‘best’ alternative. In that abstract form, it’s what governments
and their advisors have undertaken for all of human history, but as a formalized
enterprise with specialized training and broadly accepted standard practices, it
is the scope of professional education in graduate schools offering the Masters
in Public Policy (MPP) degree, and has spread through older Masters of Public
Affairs (MPA) programs and even more widely.
The present chapter reviews the practice of policy analysis education in the
United States today. Because a national conference of the association of policy
schools in 2006 commissioned a series of 10 papers and responses about this
enterprise, in more detail and with much more attention and research than is
possible here, the reader is referred to that set of works, published in the four
2008 issues (vol. 27) of the Journal of Policy Analysis and Management (JPAM)
Curriculum and Case Notes section, for a deep dive into the current state of affairs,
and discussion here will attend more to unsettled issues and developing trends.
History
The MPP schools have their roots in two main traditions. The first is the
progressive-era public administration MPA programs, mostly situated in state
universities, designed to provide competent, skilled, apolitical managers—mostly
state and local—that (under the ‘political-administrative dichotomy’ model) would
be designed and enacted by political processes. Most of these MPA programs
endure today, but especially at the élite state schools, increasingly resemble
MPP programs. The professional associations of these schools, the Association
for Public Policy Analysis and Management (APPAM) and the one now called
Network of Schools of Public Policy [sic], Affairs, and Administration (which
still uses its former acronym NASPAA) have greatly overlapping individual and
297
institutional membership. (See the chapter by Rubaii in this volume for more
about NASPAA and APPAM.)
In the 1960s and 1970s, widespread discontent with government policies—
including especially the Vietnam War, slow progress on civil rights, enduring
poverty, and the failure of the Great Society initiatives to deliver on many of
their promises—grew along with conservative dismay at some of the expressions
of that discontent (urban riots, the Free Speech Movement at Berkeley, and the
1968–69 student demonstrations and building takeovers). It appeared to groups
of academics in different fields, at several élite universities, that if government
policy were formulated with more respect for the principles and knowledge
that characterized their respective disciplines, government would be more likely
to perform better. Operations research methods developed during and after
World War II, and the planning/programming/budgeting family of government
allocation rationalizations promised real gains in government performance from
better, more technical, evaluation methodology.
Founding groups at different schools had varied composition and backgrounds
but members were commonly stars of their respective disciplines who were
also somewhat ‘off-center’ from the center of gravity of their peers, whether
intellectually or ideologically. An engaging personal history of the early years of
the Kennedy School by its founding dean is in that program’s oral history archive
and rewards viewing (Allison 2011).The Ford Foundation provided important
impetus for academically based, multidisciplinary, public policy analysis, with
large grants to support MPP programs at eight élite, mostly private, schools,
significantly bypassing the state university MPA programs long in existence. An
unstated assumption underlay the enterprise from the start: MPP super-analysts
would determine what programs should be enacted, while MPA bureaucrats would
make them happen, implicitly privileging academic knowledge and intellectual
tools over hands-on experience and craft skills. For a thoughtful review of the
MPP trajectory and its relationship to the MPA, see Ellwood (2008).
Starting with small entering classes of especially distinguished applicants, the
programs began generating alumni who went into government as policy analysts,
mostly in the Executive Branch of the federal government but also agencies like
the Congressional Budget Office. The political flavor of these programs was, at
least in some schools such as Berkeley and Harvard, distinctly conservative, suffused
with a deep skepticism about government capacity to achieve ambitious purposes
in the face of political power bases invested in the status quo and the standardized
routines of existing agencies. The founding dean of the Berkeley program, Aaron
Wildavsky, for example, was the author (with Jeffrey L. Pressman) of an influential
1973 book whose subtitle was How Great Expectations in Washington Are Dashed
in Oakland; Or, Why It’s Amazing that Federal Programs Work at All (Pressman and
Wildavsky 1984). In the 1980s, a common jape at the Kennedy School was that
it was better called the ‘Kennedy School of Less Government’.
Perhaps because the analytic justification for the Vietnam War provided by
academics had been widely savaged—ridiculed in Halberstam’s sarcastic book title
298
Public policy education in the United States
The Best and the Brightest—much intellectual attention was given to anticipating
modes of failure and analytic deficiencies in proposed and new programs, giving
the enterprise of policy analysis an early flavor of being a tool with which to
‘shoot the dogs’ at least as much as to justify new initiatives.
As the government careers of the early graduates of these programs unfolded,
a phenomenon impossible in theory was observed: an analysis meeting all the
standards of the policy analysis paradigm was prepared and released, and yet
the bill did not pass, or the failing program was not terminated. Observing
a phenomenon impossible in theory always triggers two kinds of responses
in academics, one constructive and one not. In this case, it initiated a sort of
‘Thermidorian reaction’ in the field, causing a rethinking of the academic/craft
skill ranking. MPP faculty realized that if policy analysts were going to have real
influence, they would need to learn to work government—importantly, not just how
to work government, much less how government works, because motivation and
action are as important as understanding. Accordingly, public management and
implementation courses were added to the original core offerings, along with
research projects that supported the international ‘New Public Management’
movement and peripheral activities like the Innovations in Government Awards
competition (1985).
Possibly supportive of this expansion in the MPP world were the failed efforts of
at least three business schools whose public management tracks could not attract
faculty or student engagement and support, in contrast (for example) to European
business schools like Bocconi that have strong and enduring public management
programs. Admitting ‘MPA-like’ management training to the MPP core was the
most recent fundamental change in the MPP enterprise; developments since that
time have been evolutionary and adaptive.
The ‘standard’ MPP curriculum

The American MPP has standardized, with variations, on a two-year model
comprising a required core (about half the program) plus electives, one or more
capstone/integrative courses, and a summer internship. The core courses are
usually defined by the academic disciplines of economics, political science, and
statistics/econometrics, plus management and implementation. The framework
was established very early in the programs’ history (Dunn 1975) and has changed
relatively little since then.
In the summer of 2006, APPAM organized a conference in Park City, Utah
on the 20th anniversary of a similar conference in Hilton Head Island, to review
the state of policy analysis and public administration education. Commissioned
papers reviewed the elements of the standard MPP curriculum and were discussed
at the meeting in parallel sessions, and commentators wrote up reflections on
those discussions. The papers, published in the Curriculum and Case Notes
section of the four numbers of volume 23 of JPAM, along with the rapporteurs’
discussions that follow each one, all deserve attention and greatly enrich the
299
attenuated discussion possible in the present chapter. Several include surveys and
analyses of syllabuses at different schools.
The following paragraphs characterize a typical program. Different schools alter
the recipe in one way or another, a flexibility retained as the APPAM schools
occasionally consider accreditation formalities for the MPP degree and decide
against them.
Economics
The economics core course is a semester (sometimes two) of neoclassical

microeconomic theory with applications to public policy and emphasis on market
failure. Macroeconomics is rarely included in the core. Like the field itself, this
course’s content is being challenged in some places to accommodate behavioral
economics, and it confronts affective resistance: students often fear that attention
to efficiency is in some deep way in competition with concern for distributional
equity, or that models that treat environmental pollution followed by remediation
as a GNP gain, or that can be used to argue with a straight face that people should
always give money rather than purchased (or home-made) Christmas presents
(Waldfogel 2010), are missing something important. As commonly taught, the
graphs and equations in which economic insights are typically framed for teaching
seem to turn off as many students as they light up. Nevertheless, this content is
universally recognized as essential to any responsible policy analysis.
The Park City economics paper (not published in JPAM) was the first two
chapters of Robert Frank’s book, The Economic Naturalist (Frank 2008), which
argues that introductory economics courses generally, and MPP courses also,
cover too much material and do not let the students use the core ideas of the
discipline enough to really understand them. The discussion report from Park
City is in Mendeloff (2008).
Politics
Almost all programs require some sort of politics course, with emphasis ranging
from applications and cases to political science theory, and usually including
attention to the policy-making process in the U.S. context. One way it can
be framed is as ‘management up and out’, getting resources and authorization
for policy initiatives and agency operations, where another course attends to
‘management down and in’—organizing work using tools like budgeting,
accounting, human resource procedures—and the like. There is, however, more
difference across MPP politics courses than economics and statistics offerings,
and the Park City conferees, at least, felt that this curriculum element (especially
insofar as it might improve graduates’ practical political skills) needed updating
and revision in the light of recent history. The Park City paper by Straussman
with discussion report by Radin discusses a range of issues confronting MPP
programs in defining and delivering this element of the core, and also reviews
300
the management content both adopted from MPA programs and developed
specifically for the MPP (Straussman 2008). The 2016 U.S. elections and the
increasingly bizarre political history that has followed them can be expected to
trigger a new round of revision of these offerings.
Statistics
The core statistics course is one (sometimes two) semesters of standard inferential
statistics, beginning with probability and ending with multivariate regression
models. This course is frequently problematic for students, possibly because its
instruction relies on proofs that add little or nothing either to the credibility of its
propositions or to a deep understanding thereof, possibly because of insufficient
practice in actually using the material to answer questions students want to ask.
This course almost never undertakes to replicate the training required by a statistics
scholar who develops new methodology, but it does present a constant tension
between the expectation that its graduates will actually do statistical analysis
of data, and the more realistic (for MPPs) likelihood that they will mainly be
consumers of such studies done by others.
For the Park City conference, Desai reviewed syllabuses for this course in most
of the APPAM schools and came away with the impression that ‘our students
are comfortable using statistical techniques … [but he is] not entirely sure they
know when to do so … we are continuing to choose to teach techniques over
methodological reasoning’ (Desai 2008). He also highlights the tension between
doing statistical analysis as professors do in their research, and using such studies
performed by others, as is more likely for analysts in government. A striking
feature of Desai’s syllabus topic survey is the complete absence of any Bayesian
methods. Especially in view of now readily available numerical methods that
obviate the need for conjugate priors, and instructional materials like those of
Kruschke (2014), this important toolkit should be more commonly offered.
MPP alumni repeatedly confront decisions about which they have considerable
but not dispositive knowledge, and which must be made (i) on an exogenous
schedule, (ii) without traditionally ‘certain’ evidence (p < 0.1), and (iii) in view
of differential costs of error.
This course almost always requires that students use software such as Stata or
R; a debate continues about which software is best suited to its instructional
purposes … and to the requirements of employers, for whom this turns out to
be Excel (Adams et al. 2013).
Management/implementation
Management is usually taught with cases and widely varying emphasis on

different components such as leadership skills, quantitative management science,
and standard management tools such as accounting, human resources, and
performance evaluation. Most management courses are oriented to leadership
301
techniques useful to managers with positional authority, though courses in MPA

programs tend to cover management techniques such as organizational design,
performance evaluation, accounting, and the like. The Straussman Park City
paper (Straussman 2008) and its commentary by Radin also discuss management
courses, and reflect on why this field has not produced the hoped-for body of
public management research.
Available teaching cases, which naturally affect course content and design,
continue to multiply. They are distributed almost entirely through the Kennedy
School’s case program (Harvard Kennedy School n.d.) and the Electronic Hallway
at the University of Washington (Evans School of Public Policy and Governance
n.d.), and also the more business-oriented UK Case Centre (Case Centre n.d.)
and the Harvard Business School (Harvard Business Publishing n.d.). While the
traditional case is a text narrative, more and more are being developed as multi-
media and interactive packages. Choosing cases to teach requires navigating several
conflicting purposes. Many professors have ‘oldie but goldie’ classic favorites with
which they have developed real expertise, but these often become musty, especially
for students as their technology content obsolesces and their context recedes into
history. The lessons they offer may be quite timeless, but if the principals are using
carbon paper or even floppy disks, that may be hard to see. Students, especially but
not only foreign students, constantly ask for more ‘international cases’, and cases
in non-U.S. contexts exist—but is a case set in Peru any more ‘international’, or
more engageable, for a Pakistani student than one in Chicago? Or more relevant
to an American whose next job will be in Australia? The problem here is that
while ‘American’ is a nationality, ‘foreign’ is not.
Another challenge to the case inventory is a shortage of women and minorities
as principals, which mirrors the reality of most governments; a more general
challenge to a case convention in which ‘lone heroes get by with little help or
input from politicians, the public, or organizational subordinates’ was offered in
2001 by two policy professors and still has relevance (Chetkovich and Kirp 2001).
Policy analysis capstone/integration
Almost all cores include an overview of policy analysis, using one of several
textbooks such as Bardach and Patashnik (Bardach and Patashnik 2015) or Weimer
and Vining (Weimer and Vining 2010) as a framework. The MPP usually ends
with a capstone course, built on student projects that invoke the various skills
and methods of the core in service to a public sector or nonprofit client. These
can be seminars with class meetings, or independent study under one to three
faculty advisors.
Electives
About a third of MPP coursework is in electives in the public policy school or

across its host university. Some programs require students to specify a concentration
302
with specified courses. The electives are usually field defined (for example, ‘Energy
Policy’) rather than methodological.
Variations
Law
A few MPP programs include a law course in the core but most do not. This
rather puzzling lacuna seems to be a matter of historical accident; the Kennedy
School MPP originally did, for example, but the founding professor who taught
it returned to the law school full time and the course was not maintained. One
reason this course has failed to hold a place in the MPP curriculum seems to be
availability of faculty qualified to teach it, another may be resistance from foreign
students who do not expect a common law framework to be useful in their
home contexts, and a third may be the perception that joint MPP-JD programs
(see below) satisfy the demand adequately. None of these seems to constitute a
reasoned justification for omitting content so central to government actions and
capacities, and to professionals who will be constantly engaging with lawyers in
their work (Jensen 2008).
Operations research methods
In the early years of MPP programs, the quantitative tools included in the core
often included decision theory (including Bayesian statistics), queuing theory,
linear programming and optimization models, and other tools more useful in
designing and operating programs and delivering services than writing the
kind of data-based academic articles currently expected from MPP faculty. The
Kennedy School and the Goldman School (Berkeley), for example, divided a
year of quantitative methods (a quarter of a student’s first-year class time) into
about half inferential classical statistics, and half these ‘operations research’ topics;
indeed, the textbook written for the latter course at Harvard was called A Primer
for Policy Analysis (Stokey and Zeckhauser 1978). Most of this material has
been displaced from the core by frequentist classical statistics and econometrics,
especially including advanced regression methods, though it may be available
as electives (for example, at Berkeley). Whether this displacement has occurred
as a result of faculty judgment that the material is not—or no longer—useful
to MPP alumni, or because the younger faculty assigned to teach ‘quantitative
methods’ usually have had no experience with these tools and don’t use them in
their scholarly work, is not clear.
Contested territories
Three curriculum areas—ethics, policy design, and implementation/

management—have never fit comfortably in the MPP core despite their apparent
303
relevance to policy-making practice, for different reasons. The problem with

ethics is that no one can agree on what it means as a curriculum element. Do
we expect to improve the moral behavior of our alums by making them more
likely to do the right thing? To deliver a set of ethical precepts, applicable in
practice like economic theorems about deadweight loss, and censored data
correction techniques? Perhaps an ethics course should introduce students to
the elements of philosophical ethics, in the expectation that being literate in that
field—knowing ‘what Kant said about that’, or the difference between utilitarian
and deontological prescription—will improve policy analysis in some unspecified
way? It is known that having knowledge of statistical techniques and principles
will improve data analysis practice but there is no evidence that being expert in
philosophical ethics, or theology, or even being a nice person (nothing wrong
with any of those), causes a policy analyst to create more public value on the job.
Formally, ethics appears (when it does) in different ways. One approach is to
have an ethics course in the core. Such a course may take many forms, from
‘preaching’ in a good sense, to discussion of action-forcing cases with the real
ethical ambiguity that arises from a collision of legitimate values, to a ‘great books’
tour. If such a course is offered, what are the qualifications for its professor?
Another framework is to expect that ethical challenges and puzzles raised
throughout the core will be highlighted and discussed in many courses when the
content (otherwise selected) makes them salient. However, there’s no assurance
that the professor expert in this or that technical material will not just make a
hash of the ethical dimensions of a case or example.
Finally, important uncertainty surrounds the substrate of ethical debate. Is ‘ethics
in policy’ about practitioners’ own ethics— how they should treat subordinates,
or when to leak to reporters, or resign in protest? Or is it about the intrinsic
ethics of policies, including (for example) their distributional consequences, how
clients are treated, or balancing (say) rehabilitation versus retribution in criminal
justice policy? Or is it about the professional ethics of policy analysis, like proper
citation practices, not reporting results with spurious precision, and handling
confidential data responsibly? For an extended discussion of the unstable and
unsettled state of ethics teaching in MPP programs, see Jensen (2008) and its
included discussion by Cole.
Dissatisfaction continues since the Park City conference. In thoughtful
remarks for a panel discussion of James Q. Wilson’s The Moral Sense, the dean
of the Berkeley MPP program says ‘we should have more modules on human
nature, not just one module that gets students to think of people as maximizing
their utility with respect to budget constraints and that depicts them as sets of
preferences represented by indifference curves. We have got to get beyond that
… and talk about empathy and fairness, and duty and self-control’ (Brady 2016).
This insight presumably has even more weight in view of the still-growing ethics
conflagration following the 2016 elections, and campus conflict over controversial
speakers, but it remains to be seen how the academic policy analysis community
should respond to it.
304
Policy design is, if anything, an even more daunting challenge to the MPP
teaching enterprise. The Bardach ‘eightfold path’ framework (Bardach and
Patashnik 2015) begins with ‘Alternatives’ to be compared, policies more or
less well described, like ‘tax CO2 emissions and spend the revenue on carbon
sequestration’. But where do these alternatives come from? In most cases, they
are in place or already sloshing around the ‘garbage can’ of Cohen et al. (1972),
but policies are also deliberately designed, and this can be done well or poorly. It
is perfectly possible to ‘teach design’ in a way analogous to art history or literary
criticism, examining policies and identifying better and worse design features.
But teaching about design, or ‘design appreciation’, or describing and cataloging
useful features of good design as in Weimer (1992) is very different from teaching
to design, an element almost completely absent in current policy analysis programs.
The latter praxis has been standardized over centuries with a pedagogy universally
adopted in every context where the task is to build a skill rather than acquiring
content embodied in propositions (for example: welding, sailing, playing the
piano, …). Architecture students spend at least 10 times as much class time in a
studio, actually designing buildings, as they do in architectural history class with
great buildings designed by others.
This pedagogy, a pure form of ‘theory C [for coaching]’ teaching as distinct
from ‘theory T [for telling]’, has three basic steps. First, the instructor assigns
the students a task somewhat beyond their current abilities; in architecture, for
example, the assignment on the very first day might be to design a house for a
small family on a suburban lot; in an opera staging class I visited many years ago,
four students prepared the café scene from La Bohème and performed it for the
class to critique. Second, the students undertake the task, consulting and advising
each other and—after each has made some progress and there’s something to
talk about—with comments and questions from the professor: “Did you try
putting the kitchen on the east side where it would get morning sun?” A life
drawing class is conducted in a studio, where every students’ work is visible to
all the others, who regularly circulate and kibitz each other. Finally, everyone
talks about what they did, and what about each effort was successful and how it
could have been even more so.
The cycle repeats with a more challenging assignment, and peripheral content is
introduced, where possible, as solutions to problems the students have discovered
they have, rather than new problems they didn’t ask for. The lecture portion of the
opera staging class was, and I quote verbatim, “If you’re going to move to your
right, start with your right foot.” Schön usefully examines this pedagogy, and its
potential extensions, in his classic Educating the Reflective Practitioner (Schön 1990).
If better-designed policies have social value, and design is not optimized only
by choosing among existing alternatives, real design training would seem to be
a worthwhile component of the MPP curriculum, even though it would be a
stretch for current faculty to learn how to provide it.
Management presents many of the same challenges in the existing MPP culture
as design, though it is included in the core of most programs, and entails the
305
same displacement of didaction with experiential learning. The purposes of such

training, presumably, are (i) to make MPP alumni more effective managers as their
careers progress and their seniority increases (whether of policy shops or executive
agencies), and (ii) to make their policy arguments lead to programs that are more
manageable on the ground. Here I think the biggest challenge is epistemological
and cultural: management knowledge and skills (with the exception of technical
material such as administrative science tools and accounting) are not like the
knowledge that characterizes disciplinary scholarship and therefore are (i) not
generally transmissible by didactic means (ii) distasteful to and unrespected by
colleagues immersed in a ‘scientific’ positivist world view. When embodied in
‘if-then’ propositions, this knowledge appears either banal and bromidic, or
impossibly contingent on a situation that will never exactly recur, and it is almost
never possible to apply Popperian falsifiability tests to data to show what is ‘true’
in a way an economist would find persuasive. What good managers ‘know’, in
contrast, is embodied for the most part in stories, much as what great pianists
know is embodied in performances (their own and others’), even in the presence
of propositions like ‘to properly transition a Schubert modulation, be sure to
play the inner voices clearly’. And there is no prospect of assembling a database
of comparable managerial actions that retains degrees of freedom in the face of
all the confounding variables and correlates, to prove that this or that managerial
strategy or tactic yields better outcomes than another.
Managers typically make decisions with incomplete and inconsistent
information, and without sufficient time for thorough analysis. Skill at this kind of
task is accumulated through reflective practice and accumulation of examples that
will come to mind usefully (even if not decisively) when needed. Good managers
also typically draw on more of Gardner’s eight kinds of intelligence (Gardner 2006)
than the two or three that dominate formal education. The academic literature of
public management is overwhelmingly a series of narratives, many quite gripping,
almost entirely devoid of statistical inferences from data, analyzed and adumbrated
with inferred principles that are more useful as mnemonic hooks than rules of
action. A management course, taught in the usual fashion as a series of decision-
forcing cases analyzed in plenary class session, is a way to accumulate the kind
of experience (and, to some degree, a useful ‘naming-of-parts’ vocabulary) that
good managers draw on, often unconsciously, making consequential decisions.
Indeed, the cases are sometimes fictional; Ibsen’s Enemy of the People and Chapter
XII of the Book of Exodus have been used to teach leadership; with this kind of
content, the alienation of management from positivist pedagogy and intellectual
culture is extreme if not complete.
Perhaps it will be easier for management education (and research) in policy
schools to hold a central position as the current administration’s helpless flailing
on this particular dimension continues to dominate the news and cripple its ability
to perform. Unfortunately, as ‘hard cases make bad law’, the bonfire of 2017
national governance actually offers little in the way of useful evidence beyond
306
President Obama’s caution, “don’t do dumb stuff”, so the influence of events, if

it has any, is likely to lie entirely on the motivation side.
Faculty
Core faculty
MPP core faculty, with the frequent exception of those teaching management
and ‘skills’ courses like negotiations, are almost exclusively PhDs in academic
disciplines. Their research is policy related, sometimes in the form of program
evaluation, but usually not framed as client-focused policy analysis of the type
that is claimed to inform the MPP curriculum. Few MPP faculty have PhDs
in public policy. Joint appointments with disciplinary departments are fairly
common, in fact the Woodrow Wilson School faculty must all have positions
with other departments.
The implicit requirement that (for example) an economist hired in a policy
school should be appointable in its university’s economics department is expressed
both informally in the hiring process and practically when it requires approval
of a university-wide committee, almost all of whose members are scholars ‘of
the usual sort’ with the usual disciplinary breadth (not much) and respect for
academic publication.
The lack of a clearly defined discipline of public policy analysis, in the sense
that there is a discipline of, say, anthropology, defined by its methodology and
assessable in a qualifying examination, is not only a strength but also a liability for
MPP programs. At present, we note that an interdisciplinary academic program
will almost never have individuals who are themselves interdisciplinary, and faculty
members willy-nilly benchmark themselves against peers outside the policy
school, in conventionally defined departments, and publishing in disciplinary
journals. This effect is most pronounced, at least recently, in economics and
political science; it is much easier to identify a national star in those fields than
a ‘nationally distinguished policy professor.’
The academicization of MPP programs, in particular what has occurred
since their original launch by already established academics, obstructs bringing
real government experience to the classroom, to curriculum design, and to
faculty self-replication. Some junior faculty take up their appointments having
had spells in government on committee staff or bodies like the Council of
Economic Advisors, but even this ‘time out’ weakens the purely academic case
for appointment to the degree that committee members inside and outside the
school score ‘papers published since the PhD’, and it is very rare that a new
hire will have had experience—especially responsible senior experience—in an
executive agency that actually delivers services to citizens, or for that matter in
winning elections. This kind of experience might be gained after a couple of
years teaching, at least at a junior staff level, on academic leave, but if the school
is not in Washington, DC or a state capital, it is disruptive to families. Even if
307
a tenure clock is formally stopped for public service (which the University of
California, for example, forbids), it requires some real courage on the part of a
junior faculty member conscious of an upcoming tenure decision.
Policy schools, despite genuine faculty desire to hire and promote women and
minorities, have found it difficult to achieve real faculty diversity on dimensions
of gender, race, and ethnicity in their faculties, for all the usual refractory
reasons, including not only discrimination and unequal opportunity far back in
the pipeline, but also the unconscious prejudice on the part of well-meaning
colleagues that broadly afflicts academia. The Goldman School at Berkeley, no
hotbed of reaction, misogyny, or conservative politics, was astonished to realize in
2011 that out of its 15 FTE faculty, 1.25 were female, and made a concentrated
effort that has raised the number of ladder-rank women … to seven. Pipeline issues
include the effects of poverty and residential segregation on K-12 education (for
minorities), as well as a pervasive sexism best illustrated by the different reactions
a grade school teacher could expect if they began the day with ‘good morning,
black children and white children’ instead of ‘good morning, boys and girls’.
For the most part, these are beyond the reach of graduate schools, but 41 now
belong to a program that, since 1984, has run a summer program and fellowships
to prepare and encourage undergraduates in underrepresented groups to enter
public policy graduate programs (women are the majority of MPP students);
this may have had some indirect though very diffuse, delayed, and imperfectly
targeted effect on faculty hiring (PPIA Program 2016).
Other faculty
Policy schools have responded to the ‘government experience’ problem in part

by non-ladder faculty appointments. A significant part of the MPP curriculum in
most schools is presented by adjunct faculty, lecturers, short-term appointees, and
visitors who are qualified more by experience than their scholarly records. These
non-core faculty teach skills courses like accounting and negotiations, and area
electives like housing policy. Though some are ‘in-and-outers’ who have moved
between government or nonprofits and academia, and publish scholarly or quasi-
scholarly work, most are hired to teach, and write mostly as public intellectuals
in the popular press if at all (Garris et al. 2008). Many retain consulting practices
with agencies, and in any case have real value to offer to students, but a status
divide between them and the core, academically certified, faculty affects their
weight in school governance, especially regarding curriculum.
Pedagogy
The value of professional education lies in the use students make of it. Supply-side
analysis offers useful evaluation shortcuts (‘is the normal distribution somewhere
in your syllabus?’) but none of that matters unless it changes how students live
their lives and do their work. For this to happen, students must ‘get it and do
308
it’: understand the content, retain it, and not only know how to use it but be
motivated to do so. These consequences are greatly affected by more and less
effective teaching, so how policy analysis is taught is as important as what the
curriculum comprises.
It was not possible to do a real survey of pedagogical practice for this study, but
conversation with colleagues and review of posted syllabuses indicates that for
the most part, instruction in policy schools follows a traditional arts-and-sciences
model with local variations. The ‘course’ is the basic unit, defined (except in
the case of area courses like ‘Environmental Policy’) by a discipline, and meets
for a semester or a quarter, as a quarter of a student’s course load. Each week
comprises one or two plenary sessions, totaling about three hours, and smaller
section meetings, led by graduate student instructors, of an hour each. A syllabus
schedule assigned readings from the scholarly literature or a textbook, and students
are given tasks like ‘problem sets’ and short essays. Their work is assessed, with
various weightings, on the basis of one or two invigilated midterm exams, a
final exam or a term paper, homework assignments, and class participation. A
few special courses, like a capstone thesis seminar and some specialized electives,
have different frameworks, but an MPP student spends most of their class hours
in the structure described above.
This formal scheduling structure is not the only way to organize professional
education within a single school, especially an avowedly interdisciplinary program,
and especially when the first year is a ‘lockstep’ curriculum with little or no
enrollment in courses outside the program. For example, it’s not obvious why
the required statistics, politics, and economics courses are optimally assigned the
same amount of student attention and have to start and stop together at semester
breaks, nor why instruction might not be better packaged around projects that
require several disciplinary approaches at once, like the professional practice of
governing. Administrative convenience and standard faculty workload calculations
may have more influence than they should over experimentation in course design
and scheduling, and political maneuvering among faculty members who identify
themselves by their original disciplines may have more to do with the overall
allocation of student time across subjects than the interests of the students.
I use the term plenary session even though the common name is lecture, to
emphasize that what actually happens in these meetings may be either of two
very different kinds of activity, and conventional practice may be undergoing
monotonic secular change, in which lecturing is displaced by what is variously
called ‘active learning’, ‘discussion teaching’, or a ‘flipped classroom’. The key
difference between these pedagogies’ is how much of the time the professor is
telling the students content, and how much of the time they are using, examining,
testing, and examining content obtained elsewhere, for example from readings, the
web, online videos, and a textbook. The underlying motivation for this kind of
teaching is a judgment that books and the internet, whose pace and engagement
order students can control, are better ways to acquire content than being told
it in a lecture. As research undermining implicit faith in lecturing as a content
309
delivery channel accumulates, and diffuses through the academic community, the
active learning model is gradually gaining ground.
What happens in an active learning classroom generally has two forms. The
first is discussion, led by the professor, focused on a controversy, or a decision
faced by the principal in a teaching case that the students have read and reflected
on in advance. This last is the version of ‘case-method teaching’ developed at
the Harvard Business School (Barnes et al. 1994) and widely used elsewhere,
especially in management courses, in MPP programs. (It is not to be confused
with the ‘case-method’ instruction used in law schools in common law regimes,
which is a fundamentally different process.) A teaching case is classically a story
without an ending, featuring a protagonist facing an important decision, that
provides relevant background (financial reports, press clippings, etc.) of the type
the real protagonist had available at the time, and the learning process is exercising
various inferential methods to predict the outcomes of different courses of action
and identify the best. For an extended discussion of the Theory C/Theory T
contrast mentioned above, see O’Hare (2008).
Another typical activity is small-group discussion of an assigned problem or
task, possibly homework begun before class, possibly assigned by the professor
on the spot in response to a difference of views revealed in discussion. Classroom
response systems (CRSs, or ‘clickers’), with which every student can register
an anonymous response, fit well with this pedagogy. A typical use is to pose an
interesting question—The tract should be used for (a) the youth center (b) a supermarket
(c) a park (d) a parking lot (e) the new corporation yard—and responses displayed in a
histogram. “I see people disagree; OK, break up into your small groups for 10
minutes and discuss this, and let’s see if people have changed their minds and why.”
Often the professor will circulate around the room kibitzing the small groups
sequentially. The literature on this pedagogy is too large to summarize here, but
a good introduction, with links and references, may be found in Brame (n.d.).
It bears notice that active learning of either kind (plenary discussion or small-
group breakouts) requires specific capital resources. It is not possible to have
a discussion course in a lecture room in which everyone is forced to face the
‘sage on a stage,’ and small-group teaching requires a room with a flat floor and
movable tables and chairs (O’Hare 1998). Fortunately, as a fraction of the total
cost of education, properly designed classrooms are cheap.
The ‘science teaching problem’ in the MPP
In 1987 the Harvard-Smithsonian Center for Astrophysics released the now-classic

film, A Private Universe (Schneps 1987), opening with a Harvard Commencement
at which only two out of 23 graduates, alumni, and faculty could explain ‘why
the summer is warmer than the winter’ even though all had assuredly been taught
about the earth’s ‘tilt’ and if they thought about it, would also realize that only half
of the earth is warmer in the (northern) summer (the modal wrong explanation
was that the earth is closer to the sun in the summer). This film set off a wave
310
of reflection and reformation in science teaching, echoing the Physical Science

Studies Committee ‘reforms’ that had flowered, and faded, 30 years before.
The lesson of A Private Universe, and the later PhET project established by Carl
Wieman (PhET n.d.), is that permanent learning cannot occur until students
reveal their beliefs about a phenomenon or subject and those (often wrong)
beliefs are engaged. Otherwise what appears to be learning is actually memorized
recipes and ‘official truths’ for repetition on an exam, content troweled across
their prejudices, which slough off quickly leaving the original misapprehensions
in place. The new material never becomes part of what students ‘really know’
about the real world.
Economics and statistics in professional programs experience the same
challenge: students pass final examinations but to the dismay of capstone and
follow-on course professors, cannot bring the material to bear on real problems
without having their hands guided, because the content is not part of their deep
understanding of reality. After a year of probability and statistics, smart students
still need extensive prompting to characterize an uncertain quantity as a random
variable, or their knowledge of it with a probability density function. It is not
unusual for second-year MPPs to say things like “used paper prices are low now,
but an aggressive recycling program will increase volume so the price will go up”
even when they ‘know’ what happens when a supply curve moves to the right.
Frank’s discussion of standard economics teaching relates an infamous experiment
in which economists [sic] at an American Economics Association meeting could
not identify opportunity cost correctly in a simple example.
A classic example of official ‘knowledge’ completely disconnected from real-
world experience is that nearly all adults will attempt to explain how an airplane
flies with a wing that is curved on top and flat on the bottom, and an invocation
of Bernoulli’s theorem relating velocity to pressure. They were told this incorrect
piece of pseudoscience in school, and repeat it on cue even though as they have
all seen an airplane flying upside down, and made and flown paper airplanes, it
cannot be the real explanation.
It is mystifying that such wonderful, eye-opening material as statistics and
economics, which so illuminate a complicated reality, can fail to engage students
in the kind of visceral way we learn a song or to play a sport, but as Frank
points out, the same phenomenon—acquiring a set of official answers to pass
a written test displacing real learning—also occurs in language courses. The
general principle, attested by the PhET research and more, is that people learn
to use tools by using them, not by being told how to use them, no matter how
spectacular the ‘pseudo-teaching’ classroom performance is (Noschese 2011).
The great pedagogical challenge of policy education is to transform conventional
economics and quantitative methods instruction from the didactic routines used
on beginning PhD students (whether or not they are appropriate even there) into
a pedagogy that will look to professional students like solutions to problems they
know they have, rather than new problems they don’t really want.
311
Quality assurance
With a variety of pedagogical approaches available, one may reasonably ask,

“how do faculty members improve their practice and learn new techniques?”
This question may be usefully rephrased as “what quality assurance practices do
policy schools use for teaching?” because quality assurance is a well-developed
set of principles and practices in industry (including service industries) and
government. The quality assurance machinery for teaching in policy programs,
unfortunately, is (with significant exceptions) not much better than the wretched
resources provided elsewhere in higher education.
Resources shared across schools include the spring conferences of APPAM,
whose much larger fall conference is devoted to research with occasional
sessions devoted to teaching, and NASPAA meetings which usually have more
time committed to teaching issues. NASPAA publishes a Journal of Public Affairs
Education but nearly all its content (inferring from a review of the last three years’
issues), like the presentations at the APPAM meetings, is about curriculum, with
no more than a half-dozen first-person articles describing a pedagogical innovation
or experiment in that period. None is from a professor in a top-ranked school,
or an MPP program. The JPAM used to have a regular ‘Curriculum and Case
Notes’ section but it received few submissions and was in any case shut down a
few years ago by an editor much concerned with the journal’s impact score. There
appears to be little professional reward for policy faculty to actively experiment
with pedagogy and share results formally.
Most universities have a teaching resource center of some sort whose services
are available to the policy school (consulting, videotaping classes and discussion
with a coach, a website with readings and resources, occasional seminars and
meetings), but the reach of these services, as a fraction of faculty who actually
engage with them, is very small. Some MPP programs, including, for example,
the Kennedy School, maintain a well-staffed internal teaching help center of this
kind, but they are rare. Many campuses award an annual teaching prize, though
the theory by which this practice increases learning is unclear; in any case it flatly
violates the prescription of W. Edwards Deming (Deming 1994). Are all teachers
expected to try harder in order to increase their tiny odds of winning it? Are the
practices of the winners shared with everyone to copy?
Teaching in every school known to me is evaluated summatively for purposes
of promotion and tenure at least in part on the basis of student surveys (Student
Evaluations of Teaching or SETs) administered at the end of each course; peer
visit reports, syllabuses and assignments, and critiques of student work may or
may not be considered. Unfortunately, this assessment is now known to be either
uncorrelated, or negatively correlated, with student learning (Stark and Freishtat
2014) and there is evidence that it differentially disadvantages women faculty
(MacNelll 2015) (the experiment with ethnic minorities has not been done
yet), so using this measure for this purpose (however useful and informative the
312
surveys may be otherwise) is counter to increased learning, which is presumably

the goal of teaching, and may be illegal.
Teaching, especially in higher education, is widely recognized as the most
isolating profession. We never watch each other work or talk about it; the contrast
with quality assurance conventions for research, which is built on collaboration,
peer review, conferences, and standing on each other’s shoulders in order to see
farther, is remarkable. A conversation about teaching among professors always
turns immediately to curriculum; if forced to attend to pedagogy, it is always
about what the professor does and never about what they have seen colleagues
do, nor importantly what the students do. It is worth noting how far industrial
quality assurance practice, and quality assurance practices for research, are from the
standard for professors as teachers. The first two have a few basic principles,
including (my rephrasing):
• watch each other work;

• talk together in vertically integrated groups about what we see;
• measure everything to look for instructive excursions (either way),
• do not allow scientific/statistical evidence and measurements to displace
intuition, craft skills, and experiential expertise;
• do not reward or punish individuals for successes and failures;
• recognize that quality ‘starts at the top’ and is the responsibility of management
and the systems it establishes and maintains;
• suppress defect finding and do not try to inspect your way to quality: manage
on the derivative of quality for continuous improvement with no ceiling;
• good work is the result of organizational and individual learning, not individual
traits.
The most recent survey of quality assurance practices for teaching in MPP
programs, now two decades old, found an industry not ‘in the grip of quality fever
… a quality assurance program that would command the respect of a successful
private sector firm is nowhere to be seen here’(O’Hare 1996). I have found little
evidence that much has changed, though a systematic survey is long overdue;
adoption of real quality assurance practices here is probably the single most fruitful
opportunity to improve policy analysis, and in turn, policy.
Border areas
Undergraduate and doctoral policy education
Above and below the MPP and MPA in a hierarchy of seniority are the PhD
and undergraduate education. Many policy schools maintain offerings at these
levels; the undergraduate policy program can be either a major, like the large
program at Duke, or a minor. The PhD in public policy remains an awkward
animal, partly because the market for professors with this degree is limited by
313
policy schools’ preference for appointing faculty with disciplinary degrees, and
partly because the PhD itself traditionally comprises training to do research in
an established discipline. It is hard to imagine how general examinations and a
dissertation could embrace the breadth of content represented by policy analysis
and also reach the depth expected of a PhD preparation. However, a policy PhD is
quite marketable to ‘think tanks’ and government agencies like the Congressional
Budget Office, so the overproduction recently afflicting, for example, law schools
and humanities graduate programs has not been in evidence.
Undergraduate policy programs are generally popular with students who
appreciate the variety of issues engaged in courses and the ‘real-world’ relevance
of the content. They fit well with the general education model of American
higher education, and usually (unlike undergraduate business majors) do not
simply repackage the MPP curriculum. The courses probably include an overview
‘Introduction to Policy Analysis’ but are mostly area subjects like ‘Health Policy’
and each presents a ‘policy analytic perspective’ on its topic. Policy bachelor’s
degrees are accepted preparation for a variety of graduate degrees.
In fact, because most policy majors and minors do not go on to MPPs, but
pursue other career paths, the undergraduate public policy programs probably
create a great deal of value by transmitting policy-analytic thinking into places it
would otherwise not reach, like corporate executive suites, law firms, health care
administration, and the like. Undergraduates are also fun to teach, and faculty
typically have extra freedom in designing these courses.
Another advantage of such programs is that while few policy school alums
accumulate significant wealth, undergraduates often do and their appreciation
of their policy courses can make them (if the school manages the relationship)
important to the development office. The policy PhD and undergraduate policy
programs are discussed with the extended attention they deserve in (Cordes al
2008) and the included discussion report by Cordes and Conger.
Joint degrees
Policy analysis training meshes gracefully with several other professional

formations, and many policy schools have established joint degree programs in
which a student can obtain both degrees in one year less than the two would take
if pursued seriatim. Occasionally, these are custom made; I once supervised an
MPP/DDS. The usual ‘recipe’ is to cross-credit some courses from each degree
against the other; for example, if a Master of Public Health and an MPP program
both require a statistics course, the student can take one or the other.
The most common combinations are with law (MPP/JD) a strong preparation
for a government career; public health (MPP/MPH); and social work (MPP/
MSW). Less common are joint degrees with engineering (MPP/MS or MEng)
and city planning (MPP/MCP).
314
Looking forward
Despite the challenges discussed above, especially (i) the need to institute quality
assurance practices for pedagogy that maximize learning per unit of resources
consumed and (ii) the continuing tension between academicized Popperian
epistemological conventions and the need for implementation and management
craft skill building, the policy analysis educational enterprise can be considered a
success and a stable, established enterprise. MPPs are increasingly encountering
more senior versions of themselves in government and nonprofit workplaces,
policy school faculty maintain a respectable rate of research productivity, and
the employment market can be said to have a set of known expectations for the
competences an MPP will provide. Policy schools can be said to have claimed
a secure niche in the higher education ‘industry’ and there is no reason their
contribution to governance shouldn’t increase if the innovative spirit of the first
decades of the experiment can be maintained.
References
Adams, W. C., Lind Infeld, D. and Wulff, C. M. (2013) ‘Statistical Software for
Curriculum and Careers’, Journal of Public Affairs Education 19(1): 173–88.
Allison, G. (2011) Interview, Harvard Kennedy School Oral History Project, Harvard
Kennedy School.
Bardach, E. and Patashnik, E. (2015) A Practical Guide for Policy Analysis,
Washington, DC: CQ Press.
Barnes, L. B., Christensen, C.R., and Hansen, A. (1994) Teaching and the Case
Method, Cambridge, MA: Harvard Business Review Press.
Brady, H. (2016) Encountering the Moral Sense in the Public Affairs Classroom,
Goldman School of Public Policy, University of California, Berkeley.
Brame, C. J. (n.d.) ‘Flipping the Classroom’, https://cft.vanderbilt.edu/guides-
sub-pages/flipping-the-classroom/.
Case Centre (n.d.) ‘Case Centre’, https://www.thecasecentre.org/main/.
Chetkovich, C. and Kirp, D. (2001) ‘Cases and Controversies: How Novitiates Are
Trained to Be Masters of the Public Policy Universe’, Journal of Policy Analysis
and Management 20(2): 283–314.
Cohen, M. D., March, J. G., and Olson, J. P. (1972) ‘A Garbage Can Model of
Organizational Choice’, Administrative Science Quarterly 17(1): 1–25.
Cordes, J., Conger, D., Ladd, H., and Luger, M. (2008) ‘Undergraduate and
Doctoral Education in Public Policy: What? Why? Why not? Whereto?’ Journal
Deming, W. E. (1994) The New Economics: For Industry, Government, Education,
Cambridge, MA: MIT Press.
Desai, A. (2008) ‘Quantitative Methods, Economics, and or Models’, Journal of
Dunn, W. L. (1975) ‘A Comparison of Eight Schools of Public Policy’, Policy
Studies Journal 4(1): 68–73.
315
Ellwood, J. W. (2008) ‘Challenges to Public Policy and Public Management

Education’, Journal of Policy Analysis and Management 27(1): 172–87.
Evans School of Public Policy and Governance (n.d.) ‘Electronic Hallway’, http://
hallway.evans.washington.edu/.
Frank, R. H. (2008) The Economic Naturalist: In Search of Explanations for Everyday
Enigmas, New York: Basic Books.
Gardner, H. (2006) Multiple Intelligences: New Horizons in Theory and Practice, New
York: Basic Books.
Garris, R., Madden, J., and Rodgers III, W. M. (2008) ‘Practitioners’ Roles,
Internships, and Practicum Courses in Public Policy and Management
Harvard Business Publishing (n.d.) ‘Cases’, https://cb.hbsp.harvard.edu/cbmp/
pages/content/cases.
Harvard Kennedy School (n.d.) ‘Case program’, http://case.hks.harvard.edu/.
Jensen, E. R. (2008) ‘Law, Ethics, and Craft Skills’, Journal of Policy Analysis and
Management 27(4): 977–82.
Kruschke, J. (2014) Doing Bayesian Data Analysis, Cambridge, MA: Academic
Press/Elsevier.
MacNell, L., Driscoll, A., and Hunt, A. N. (2015) ‘What’s in a Name: Exposing
Gender Bias in Student Ratings of Teaching’, Innovations in Higher Education
40(4): 291–303.
Mendeloff, J. (2008) ‘Discussion Report: Do We Try to Teach our Students Too
Much?’ Journal of Policy Analysis and Management 27(3): 617–24.
Noschese, F. (2011) ‘Pseudoteaching: MIT Physics’, Action-Reaction, https://
fnoschese.wordpress.com/2011/02/21/pt-pseudoteaching-mit-physics/.
O’Hare, M. (1996) ‘Quality Assurance for Teaching in APPAM Schools’, Journal
O’Hare, M. (1998) ‘Classroom Design for Discussion-based Teaching’, Journal
O’Hare, M. (2008) ‘Pedagogy for Policy Analysis and Management’, Journal of
PhET (n.d.) ‘Research’, https://phet.colorado.edu/en/research.
PPIA Program (2016) ‘Home page’, www.ppiaprogram.org/ppia/.
Pressman, J. L. and Wildavsky, A. (1984) Implementation: How Great Expectations
in Washington Are Dashed in Oakland; Or, Why It’s Amazing that Federal Programs
Work at All, This Being a Saga … Morals on a Foundation (Oakland Project) Berkeley,
CA: Berkeley, University of California Press.
Schneps, M. (1987) A Private Universe, Cambridge, MA: Annenberg Learner.
Schön, D. A. (1990) Educating the Reflective Practitioner, New York, NY: Jossey-Bass.
Stark, P. B. and Freishtat, R. (2014) ‘An Evaluation of Course Evaluations’,
ScienceOpen, https://www.scienceopen.com/document/vid/42e6aae5--246b-
-4900--8015-dc99b467b6e470, DOI: 10.14293/S2199-1006.1.SOR-EDU.
AOFRQA.v1
316
Stokey, E. and Zeckhauser, R. (1978) A Primer for Policy Analysis, New York:
W.W. Norton.
Straussman, J. D. (2008) ‘Public Management, Politics, and the Policy Process in
the Public Affairs Curriculum’, Journal of Policy Analysis and Management 27(3):
624–35.
Waldfogel, J. (2010) Scroogenomics: Why You Shouldn’t Buy Presents for the Holidays,
Princeton, N.J.: Princeton University Press.
Weimer, D. and Vining, A. (2010) Policy Analysis: Concepts and Practice, Abingdon:
Routledge.
Weimer, D. L. (1992) ‘Claiming Races, Broiler Contracts, Heresthetics, and
Habits: Ten Concepts for Policy Design’, Policy Sciences 25(2): 135–59.
317
SEVENTEEN
The status of the profession: the role

of PhD and masters programs in public
policy education
Nadia Rubaii
Introduction
The United States is at the center of public policy analysis worldwide and thus it
should come as no surprise that U.S. higher education programs for public policy
and policy analysis are standard bearers for programs elsewhere (Geva-May and
Maslove 2006. See also Peters in this volume). U.S. programs for public affairs
education were established and crystallized before their counterparts elsewhere,
but beyond the mere temporal sequencing, there is considerable evidence that
U.S. programs served as a model for shaping the curriculum content and delivery
methods established elsewhere, as well as influencing ideas about how to assess
quality and on what criteria to accredit such programs (Geva-May et al. 2008). In
some respects, the widespread influence of U.S. policy education is unremarkable,
however when we consider the extent to which program design and pedagogy
reflect the unique historical, economic, political, and cultural contexts of any
society (Geva-May et al. 2008), it becomes evident that an understanding of how
and why public policy education developed in the United States is important.
This chapter traces the evolution of graduate-level public affairs education in
the United States in terms of focus, mission, curriculum, institutional locus,
and enrollments, and highlights the role of two key professional associations in
the evolution of the field. It also highlights some persistent challenges regarding
how broadly or narrowly to define the field, how clearly to differentiate among
the related fields of study, and how to define and ensure quality.
Terminology
There is a striking absence of precision in the terminology surrounding
public policy education in the United States. The terms public policy, public
administration, public affairs, public management, policy analysis, policy studies,
and public service are alternatively used interchangeably and presented in contrast
to one another. This chapter applies a broad definition of public policy education,
encompassing not only degrees titled Master of Public Policy or Ph.D. in Public
Policy, but also those in Public Administration, Public Affairs, or related fields
319
with a policy component. In doing so, the chapter perpetuates the widespread
practice of using terms interchangeably. Despite strong assertions about the
importance of clear distinctions in program content and delivery across these
fields, the reality is that ‘blurry borders’ (Geva-May and Maslove 2006, p 415)
and lack of precision of degree titles remains a characteristic of U.S. public affairs
education. Where possible, the emphasis is on masters and doctoral programs
with a specific emphasis in public policy, policy analysis, policy studies, or public
management, within the framework of and in comparison to the other elements
of the broader field of study.
Evolution
The field of public policy has evolved as a profession, a research focus, and an
area of study in interrelated ways. Within the realm of educating for public policy
analysis in the United States, it is important to look at the programs themselves as
well as two organizations that have played a critical role in guiding, supporting,
and coordinating the activities of those programs. The development of these
organizations, generally referred to only by their acronyms NASPAA (Network of
Schools of Public Policy, Affairs, and Administration) and APPAM (Association for
Public Policy and Management), provides insights into the evolution of the field
as a whole and ongoing struggles of programs which offer masters and doctoral
degrees. Both associations are international in scope and membership, but each
is physically based in Washington, DC, and each has membership predominately
from the United States.
Key professional associations

NASPAA
In 1970, the National Association of Schools of Public Affairs and Administration

(NASPAA) was established as a successor to the Council on Graduate Education
for Public Administration (CGEPA), which had been founded in 1950
(Henry 1995). NASPAA, like the field itself, was developed in response to a
sociopolitical environment of change and the need for more professionalized
public administrators to address political crises, global conflicts, social unrest, and
economic hardships (Henry 1995). NASPAA has grown rapidly from roughly
65 institutional members at the time of transition from CGEPA to NASPAA in
1970, to more than 300 as of 2016. NASPAA’s two-fold mission is ‘to ensure
excellence in education and training for public service and to promote the ideal
of public service’ (www.naspaa.org).
One of the most important roles that NASPAA plays is in setting standards for
accreditation of professional master’s degrees. Public administration, public policy,
and related fields lack the professional certification or licensure requirements of
other professions such as law, medicine, or social work, so accreditation is an
320
The status of the profession
important means by which to ensure some degree of quality and to clarify what
unites the profession. Accreditation is one of the most notable examples of how
public affairs education in the United States has addressed and resolved an issue
to which many other countries have not yet responded (Geva-May and Maslove
2006). The evolution of NASPAA accreditation standards both guided and
reflected changes in the profession and the discipline. Discussions and decisions
about changing standards provides insight into the debates and demands of the
profession.
NASPAA standards are characterized by three distinct phases or ‘generations’
(Rubaii and Calarusse 2012). In the initial period (1986–1992), the emphasis was
on resources or inputs in the form of sufficient numbers of faculty, infrastructure,
library holdings, and demonstrated curriculum content in specified areas. The
standards articulated common curriculum requirements in three broad areas: (1)
Management of Public Service Organizations, (2) Application of Quantitative
and Qualitative Techniques of Analysis, and (3) Understanding of the Public
Policy and Organizational Environment. Across these three areas, more specific
components were identified consisting of: human resources; budgeting and
financial processes; policy and program formulation, implementation, and
evaluation; decision making and problem solving; political and legal institutions
and processes; economic and social institutions and processes; and organization
and management concepts and behavior (NASPAA 2008). Although NASPAA’s
Commission on Peer Review and Accreditation (COPRA) repeatedly assured
programs that the list of curriculum requirements did not represent specific class
requirements, nor did they need to be provided equal attention, public policy
programs struggled and resisted some aspects of the standards (such as the need
to teach about human resource management) as reflecting too much of a public
administration emphasis.
The second generation of standards (1992–2009) allowed for greater flexibility
in accordance with a program’s stated mission and began the process of requiring
rudimentary program evaluation. Added to the common curriculum requirements
in this stage was information management, technology applications, and policy,
and a requirement that programs demonstrate diversity across the entire curriculum
(Rubaii and Calarusse 2012; NASPAA 2008). In this phase programs were allowed
to request deviations (exemptions) from the standards, based on their particular
mission.
The third generation (2009–present) strengthened the mission-based philosophy
and brought in an emphasis on demonstrated evidence of student learning across
five broad competency domains. Programs seeking NASPAA accreditation must
demonstrate that their graduates are prepared to: (1) lead and manage in public
governance, (2) participate in and contribute to the policy process, (3) analyze,
synthesize, think critically, solve problems and make decisions; (4) articulate and
apply a public service perspective; and (5) communicate and interact productively
with a diverse and changing workforce and citizenry (NASPAA 2014). Programs
define each of these universal competency areas within the context of their
321
missions, with the expectation that a public policy and a public administration
program will do so quite differently.
In 2012, in my last act as NASPAA President, I oversaw the vote in which
NASPAA members approved making two substantive changes to the name of
the organization while maintaining its acronym. The first initial ‘N’ which had
referred to ‘National’ for the first 40 years of the Association’s existence, was
replaced with ‘Network’ to better reflect the growing international scope and
membership of the Association. Additionally, the ‘P’ which had simply referred
to ‘Public’ as part of reference to ‘Public Affairs and Administration’ (PAA)
was expanded to explicitly reference Public Policy. This change was made in
response to concerns expressed by some NASPAA members from policy schools
that the prior name left them feeling somewhat excluded or ignored. Thus, in
2012, NASPAA transitioned from the National Association of Schools of Public
Affairs and Administration to the Network of Schools of Public Policy, Affairs, and
Administration. Like many professional associations whose names have evolved
over time, the association currently relies only on its acronym and tagline –
‘NASPAA: The Global Standard in Public Service Education’ – and reserves the
use of the full name for legal purposes. Despite the recent name changes, the
relationship between NASPAA and public policy programs reflects an historical
‘period of wavering’ after which many ‘decided to stay with NASPAA, although
with allegiance divided between it and the new APPAM that offered individual
memberships to their entire faculties’ (Henry 1995, para. 24).
APPAM
The Association for Public Policy and Management (APPAM) was founded
in 1979 to serve as a new professional association of graduate schools of public
policy and management. The mission of APPAM is to improve ‘public policy
and management by fostering excellence in research, analysis, and education’ and
it does so through conferences, a research journal, linking policymakers with
scholars, and fostering participation by students of public policy and management.1
APPAM welcomes both institutional and individual members and as of 2016
consisted of 100 of the former and approximately 1,500 of the latter, many of
whom hail from the member institutions. The institutional members include both
universities and think tanks and, while international in scope, the membership
list is predominated by U.S. schools and U.S.-based research organizations. Of
the 98 institutional members as of August 2015 listed on the APPAM website,
81 are universities or university-based research centers, 74 of which are within
the United States, representing 69 distinct institutions.
Distinctions and overlap
While there are overlaps in the missions and membership rosters of NASPAA
and APPAM, there are also notable distinctions. NASPAA explicitly focuses
322
on increasing the quality of education for public service, in part through its
accreditation processes, and it focuses on the concerns of program deans and
directors. NASPAA’s decision to limit membership to institutions rather than
individuals maintains this focus as does the scope of its journal, the Journal of Public
Affairs Education (JPAE), which largely addresses issues of program management
and effective teaching. APPAM, in contrast, has a smaller but more focused
institutional membership and a stronger research focus. The engagement of
individual policy scholars is evident via APPAM’s research conference and its
research-focused journal, the Journal of Policy Analysis and Management (JPAM).
Among the U.S. schools, 40 are members of both associations as of 2016. The
vast majority of NASPAA member schools have not sought APPAM membership,
and some of the leading public policy schools which are active in APPAM are
not members of NASPAA. The two associations share some common strategies
and have some unique methods of advancing the interests of the profession; they
have on occasion collaborated, and at other times seemed at odds.
NASPAA and APPAM are the most important professional associations when
considering graduate-level education for public policy in the United States, but
they are by no means the only ones. Other professional associations, including
the American Society for Public Administration (ASPA), the National Academy
of Public Administration (NAPA), the International City/County Management
Association (ICMA), and more recently the International Comparative Policy
Analysis (ICPA) Forum, have also helped shape the profession and the scholarship
informing the profession, but are less explicitly focused on education.
The evolution of U.S. public policy education

In Chapter 18 of this volume, B. Guy Peters notes that the integration and
institutionalization of policy analysis schools and curricula in the United States
predates and has informed developments in other countries, and that U.S. pre-
eminence in policy education continues. Throughout its evolution, U.S. policy
education has focused on the graduate level, emphasizing either professional
master’s degrees to prepare individuals to begin or advance in their public service
careers, or doctoral programs intended to advance the scholarship and teaching.
The two most well-established degrees at the master’s level are the Master of Public
Administration (MPA) and Master of Public Policy (MPP). One might expect the
degree name to signify important distinctions and to serve as the basis for attracting
particular types of students, reflecting the educational emphasis and placement
opportunities, and conveying a message about what prospective employers can
expect from graduates. A closer look suggests that the lines between the MPA
and MPP are not so clear. Both NASPAA and APPAM acknowledge that in how
they discuss degree options; they speak of the two degrees together in identifying
the key characteristics of public service education, rather than distinguishing
between them. In describing Ph.D. programs, APPAM also refers to programs
in public administration/public policy rather than differentiating between them.
323
The evolution of programs in the field of public administration and public policy
reflects societal changes. Dynamic changes in American society from post-World
War II through Watergate resulted in:
… explosive economic and population growth; suburbanization and

urban crises; demands for civil rights and equal opportunity; marches
in the streets and demonstration on the campuses; energy crises and
episodes of recession and inflation. The public environment was, to
put it mildly, superheated, with profound effects on the attitudes of
university leaders, faculty members, and students; not least of these
was a heightened appreciation of the importance of politics and public
service (Henry 1995, n.p.).
Using the umbrella term of ‘public service,’ Donald Stokes (1996), who served as
APPAM president in 1984–85, documents how the changing needs of government
and society are reflected in developments in the curricular and research foci of
university-based programs through four successive waves. The first wave, what
Stokes labels the Public Administration movement, was an outgrowth of the
Progressive Era’s emphasis on isolating administration from the corrupt influences
of politics and instilling a sense of professionalism. Beginning with ideas expressed
by Woodrow Wilson (1887) in his iconic paper, efforts to separate politics
from administration ultimately led to the creation of the nation’s first public
administration training program in 1911, based not in a university but rather as
a training school within the New York Bureau of Municipal Research (see the
chapters by Weimer and Mossberger et al. in this volume for more on municipal
research bureaus). This program was established after some of the nation’s top
universities, including Yale and Columbia, opted not to establish public service
programs, having judged them to be too practical and not sufficiently academic
(Henry 1995). The Bureau was later moved to Syracuse University, where it
became the foundation of the Maxwell School of Citizenship and Public Affairs.
The public administration curriculum of this era was dominated by teaching
Luther Gulick’s (1936) famous PODSCORB skills of Planning, Organizing,
Directing, Staffing, Coordinating, Reporting, and Budgeting, as well as advocating
arrangements such as the council-manager form of government at the local level
to remove politics from the exercise of administration.
The second wave – what Stokes refers to as the Public Affairs movement –
tempered the earlier emphasis on a strict wall between politics and administration
with a recognition that addressing the nation’s problems required administrators
who were willing to get involved in policy content and implementation. During
this era, public service programs educated individuals to simultaneously be
professional administrators and policy leaders. Beyond the PODSCORB skills,
public affairs programs instilled in their students the ability to identify social needs,
articulate policy options, and build coalitions to achieve policy goals (Stokes 1996,
p 160). Master of Public Affairs (MPAff) programs established at this time sought
324
to be comprehensive, international, and comparative, to encompass urban, social,

and economic development, and to blend public policy and public management
(Henry 1995).
In these early periods some public administration curriculum appeared within
political science programs, but since the early times and continuing to the
present, the relationship between public administration and political science has
been complicated. As political science departments placed greater emphasis on
positivism, empiricism, and quantification, public administration education was
often disparagingly characterized as mere training (Henry 1995). As political
science loosened its grip, public administration became more multidisciplinary,
drawing upon sociology, social psychology, economics, law, and management,
among other fields, leading to confusion about what, if anything, formed the
intellectual core of public service education (Elmore 1986). Despite these
criticisms, the demand for and supply of MPA programs grew steadily.
The 1950s and 1960s ushered in the third wave – the Public Policy movement
– based on a recognition that more systematic and sophisticated analysis was
required to inform policy decisions (Stokes 1996). Developments within the
practice of policy analysis during the 1960s and 1970s, including greater emphasis
on efficiency, rationality, systematic policy analysis, and ‘speaking truth to power’
(Wildavsky 1987), led to the proliferation of public policy programs (Geva-May
and Maslove 2006). These changes manifested in educational programs specifically
focused on public policy beginning in the late 1960s and early 1970s. Much in
the same way that Beryl Radin describes in Chapter 2 how the policy analysis
profession itself developed in response to the emergence of new analytical needs
within the U.S. Department of Defense in the 1960s, so too did the educational
programs. There was an emergence of new programs and schools of public
policy during this period, with many universities careful to avoid the term public
administration lest it conjure up the negative stereotypes previously assigned to
it by political science (Henry 1995).
Public policy programs of this era were modeled after ‘the Rand experience’
of applying ‘economic theory and systems analysis to government problems’ as
a means of making ‘public decisions rigorously and analytically on the basis of
systematic quantitative evidence’ (Yates 1977, p 364). Public policy education has
three principle assets relative to public administration: ‘a well-specific body of
theory (primarily price theory and welfare economics), a simple model of choice
(problem, criteria, alternatives, outcomes, decision), and a clear set of analytic
techniques by which the first could be connected to the second (statistics, benefit-
cost, optimization methods)’ (Elmore 1986, pp 71–2). MPP programs place greater
emphasis on tools such as cost/benefit analysis, micro-economic theory, decision
analysis, statistics, and political and organizational analysis (Yates 1977) and leave
out courses on human resources, information technology, and organization theory
(Henry 1995). MPP programs also offer electives in substantive areas of public
policy. But other than that, the MPP curricula at top schools are much like their
MPA counterparts (Henry 1995).
325
Whereas earlier public administration programs were designed to prepare

government professionals at the national, state, and local levels, programs in public
policy and policy analysis were targeted primarily to the federal government
based on a sense that state and local governments were not sufficiently organized,
large, or well resourced to hire and pay adequately analytical staff (Yates 1977).
Despite this perception, there is a rich history of state-level policy analysis on a
range of issues and across an array of states (as documented by VanLandingham
in Chapter 6). Public policy programs at their inception largely focused on the
national level of government, with the notable exception of the University of
Texas Austin, where a large percentage of public policy graduates enter state
government in Texas (Yates 1977).
In the mid 1980s, while analysis and economic theory remained hallmarks of
public policy programs, there was some shift toward preparation of students for
managerial roles through the development of a public management approach.
Stokes (1996) labels the fourth wave a Public Management movement, in which
public service education was positioned within business schools, generic schools
of management, or as part of public policy schools. In this period we see Masters
of Management degrees premised on a resurgence of the idea that management is
generic (Henry 1995). Public Management has a ‘deep skepticism about public
intervention and an active decision-forcing attitude toward the practice and
teaching of management’ (Elmore 1986, p 69). Public management broadens
public policy by moving the focus from staff to management levels, and shifts
public administration by replacing neutral competence with goal-setting and
shaping mandates. Public management has been characterized as a merger of the
best of public policy and private sector management. In contrast to traditional
public administration, public management is more skeptical of government and
tends to look more favorably on the private market, treats managers as decision
makers who should rely on formal analysis, and rejects the politics–administration
dichotomy (Elmore 1986).
Writing two decades ago, Stokes (1996) envisioned a fifth wave that would
emphasize effective governance in ways that were distinct from the earlier waves
in at least three respects: a focus on outcomes rather than inputs, extending
the scope from the federal government to state and local governments, and
transcending government to more broadly encompass governance by public
service organizations spanning all sectors. He suggests that this approach requires
a rethinking of program missions and curricula, as well as research agenda of
faculty. Among the skills he suggests would be most important are financial
analysis, negotiation, political analysis, legal analysis, and ethical analysis. He also
stressed the increasing importance of values and commitment to general or public
interest as key considerations moving forward. Despite the changes in emphasis
over time, arguably all variations of public service education programs ultimately
‘derive from the progressive era’ (Ellwood 2008, p 174). In what might be labeled a
Governance movement, public affairs education has expanded beyond government
326
to include nonprofit and private sector actors, and increased emphasis on global
interdependencies in responding to social problems.
Masters programs
Among NASPAA-member schools and more generally within the United States,
the MPA degree has been and continues to be the most common on an order
of magnitude relative to the MPP, MPAff, and other degrees. As of 2013–14,
more than 21,000 students were enrolled in NASPAA-member MPA programs,
compared to less than 4,000 in MPP and fewer than 2,000 in MPAff programs
(NASPAA, n.d.). As described above, the different degree titles have origins in
specific time periods and reflect different educational and public service priorities.
The field has shown growth over an extended period of time in terms of both
numbers of masters programs and enrolled students. A survey conducted by
NASPAA’s predecessor, CGEPA, in 1959–60 identified 100 institutions offering
some form of master’s degree in public administration, with the majority housed
in small programs within political science departments. Most of the 3,000 students
enrolled at that time were concentrated in fewer than a dozen large programs
which were organized as separate departments or schools. Observing patterns in
the 1970s, Fritschler et al. (1977) characterized graduate education in public affairs
and public administration as ‘one of the few growth areas in higher education’
of the time (p 488). By 2013–14, the NASPAA survey generated responses from
215 masters programs, ranging in size from 12 to over 1,200 students, which
collectively enrolled more than 25,000 students at the master’s level.
Consistently, over time and across programs, a majority (55–57% over the
years) of students in public affairs programs are women. An important caveat to
the discussion of U.S.-based public policy education is that these programs are
not simply educating U.S. students. These programs are experiencing a growing
number and proportion of students from foreign countries, representing one third
of incoming students in any given year across all programs and as many as one half
of students in some programs (Fritzen 2008). Despite the internationalization of
the student body, the education provided is still largely U.S. centric, with examples
and case studies drawn predominantly if not exclusively from the U.S. context
(Fritzen 2008). International students are represented more heavily within MPP
programs relative to MPA programs (20% compared to 10%). MPP programs also
enroll a larger percentage of out-of-state students (19%) than MPA programs (8%),
but demonstrated less diversity among students (25% vs. 33%) and fewer part-
time students (20% vs. 48%) (NASPAA n.d.). An increasing number of programs
of both types are offered in a hybrid format combining tradition in-person
education with online coursework. The most common length of study period
is four semesters with 107 of 225 reporting this, and 66 reporting 5 semesters.
The institutional locus of NASPAA-member programs has shifted over time.
While nearly half (46%) were based in a department or program within a school
of arts and sciences (most often in a political science department) in 2000, by 2013
327
this arrangement represented only one quarter of all programs and was surpassed
by the programs in a separate professional school or college representing nearly
one third (31%). In contrast to public administration programs which are well
established across both public and private institutions, schools of public policy
in the United States are disproportionately housed in private universities, and
they often rely on faculty who have held senior government positions (both
appointed and elected) prior to transitioning to academia (Geva-May and Maslove
2006). Compared to their counterparts in other countries, public policy schools
in the United States place considerable emphasis on analytical techniques and
management, relative to theory or abstract concepts (Geva-May and Maslove
2006).
Employment patterns of graduates have also demonstrated some transition over
time. A 1975 survey showed that the plurality of graduates (35%) were employed
in local government, followed by state government (21%), and federal government
(18%). Significantly smaller proportions of students pursued additional studies
(8%), were employed in the private sector (8%) or quasi-government agencies
(7%), or teaching (3%) (Fritschler et al. 1977, Table 5, p 492). By 2014, graduates
are most likely to be employed in either the nonprofit (26%) or private sectors
(21%), with only 10 to 15% in local, state, or national governments (NASPAA
n.d.). Not well illustrated in the aggregate-level data from NASPAA is the
prevalence of federal government placements among MPP students as documented
by Rebecca Maynard in Chapter 5 of this volume.
Numerous scholars have called for a clearer distinction between the MPA and
MPP in keeping with their original design, and numerous studies have illustrated
the absence of clarity. In a 1991 study that analyzed curricula and syllabi from
a random sample of 60 NASPAA schools offering the MPA, and 21 APPAM
schools offering the MPP, Averch and Dluhy (1992) found minimal differentiation
between NASPAA and APPAM schools. The biggest distinction they found was
the prominence of public personnel in NASPAA programs and economics in
APPAM programs. More than 15 years later, Ellwood (2008) found a general
but not complete convergence of MPA and MPP programs. A report from an
ASPA Task Force on Educating for Excellence in the MPA Degree, is critical of
developments in public administration education, particularly the proliferation of
degrees and curricula which have blurred the core mission of public administration
education, the MPA, and the values on which it is founded (Henry et al. 2009).
The criticisms about convergence can themselves be criticized for possibly
being overstated and unduly negative in tone. The areas of emphasis with the
two types of programs suggest important distinctions. According to data collected
by NASPAA for the 2013–14 academic year, the most frequently offered
concentrations within MPA programs were: (1) nonprofit management; (2)
general public management; (3) city/local government administration; (4) health
administration; and (5) public policy analysis. In contrast, MPP programs from the
same time period identify their most common specializations in the areas of: (1)
social policy; (2) environmental policy; (3) public policy analysis; (4) health policy;
328
and (5) education policy. There is overlap but also some notable differentiation.
A more detailed analysis of policy school realignments and typical curricula of
policy programs is provided by Michael O’Hare in Chapter 16 of this volume.
While the lack of a clear dividing line between MPA and MPP programs
is frequently characterized as a shortcoming of the field, convergence is not
necessarily a negative characteristic. It may simply reflect ‘a more holistic approach
that recognizes the symbiosis between the fields of public management, public
administration, and policy analysis’ in which ‘over the years the MPA and MPP
degrees in the United States have moved closer together’ (Geva-May and Maslove
2006, p 418) and an example of the field becoming more pluralistic and open,
and thus more quintessentially American (Radin, Chapter 2).
Depending on one’s perspective, a single ranking system for public affairs
programs also contributes to either an unfortunate blurring of the lines between
MPA, MPP, and MPAff programs, or a positive force means of illustrating the
public service contributions they collectively provide. The surveys identify top
schools based on a list of programs provided jointly by NASPAA and APPAM.
Among the top 10 programs according to U.S. News and World Report
(USNWR) rankings based on peer assessment surveys administered in fall 2015
are an interesting mix of programs from various eras and various philosophical
perspectives (see Table 17.1).
Although ostensibly a ranking of master’s programs, the USNWR reputational
ranking process makes it difficult for evaluators to disregard their perceptions of
quality of doctoral programs in the process. The top-rated programs consistently
offer doctoral as well as master’s programs, and it is impossible to ascertain the
extent to which the relationship is merely correlational or has some causal
elements as well.
Doctoral programs
The case for doctoral-level programs in public policy was first articulated by Albert
Lepawsky in an article published in a 1970 issue of Policy Sciences in which he
asserted the need for an interdisciplinary program which would both ‘aggregate
the most relevant materials and usable generalizations of the sociopolitical sciences’
and ‘strike a balance between viable theory and tested experience’ (Lepawsky
1970, p 443). In making this call he acknowledged the high levels of skepticism
among American intellectuals and the general public about the social sciences.
He saw the potential to establish greater credibility and to improve the ‘training
of high-priority policymakers, policy administrators, and policy scientists’ (p
444). He presented a detailed curriculum proposal grounded in political science,
public administration, sociology, psychology, and economics, with coursework in
the policy process, and fundamental concepts of public interest, decision-making
theory, and substantive broad policy issues areas such as economic policy, social
policy, environmental policy, community and regional development, etc. The
doctoral programs established in response largely began within freestanding schools
329
or colleges (Overman, Perry and Radin 1993) and later shifted to be more evenly
distributed between that structure, and within departments of either political
science or public administration/urban policy (Brewer et al. 1998).
Table 17.1: Top 10 masters in public affairs according to USNWR, 2016

University/school Masters NASPAA NASPAA APPAM Doctoral degrees
offered member accredited member offered
1* School of Public and MPAff    PhD in Public Affairs
Environmental Affairs PhD in Environmental
(SPEA), Indiana University – Sciences
Bloomington Joint PhD in Public
Policy/Political Science
Maxwell School, Syracuse MPA    PhD in Public
University Administration
3 Kennedy School at Harvard MPA   PhD in Public Policy
University MPP PhD in Health Policy
MPA/ PhD in Social Policy
International PhD in Political
Development Economy & Government
4* Woodrow Wilson School MPA  PhD in Public Affairs
Princeton University MPP with two academic
clusters (Security Studies,
or Science, Technology,
and Environment Policy)
University of Georgia MPA   PhD in Public
Administration
Price School, University of MPA    PhD in Public Policy &
Southern California Management
PhD in Urban Planning &
Development
Evans School, University of MPA    PhD in Public Policy &
Washington Management
8* Goldman School, University MPP  PhD in Public Policy (no
of California – Berkeley MPAff core curriculum, must
have MPP)
Ford School, University of MPA   Joint Doctoral Program
Michigan – Ann Arbor MPP in Public Policy and
Social Science
Humphrey School, University MPP    PhD in Public Affairs
of Minnesota – Twin Cities
MPA = 7 8 6 9 Public Admin. = 2
SUMMARY MPP = 5 Public Policy (includes
MPAff = 2 joint and Public Policy
Other=1 & Management) = 6
Public Affairs = 3
Other = 5
330
By 1995, a survey of the 56 NASPAA-affiliated doctoral programs showed

2,474 students enrolled with an average size of 26 full-time students (ranging
from 0 to 258) and 20 part-time students (ranging from 0 to 9) (Brewer et al.
1998). Programs reported that full-time students generally took three years of
coursework beyond master’s degree, while part-time students required 4.8 years.
The dissertation process consumed another 1.7 and 2.7 years, respectively. Across
all programs, a majority of students (61%) had some full-time public sector
work experience. In contrast to the dramatic growth in the number of master’s
level programs and enrollments over time, the number of doctoral programs has
remained relatively steady and the number of students has increased slowly. As
of 2009–10, NASPAA data documented 49 programs (46 PhD and 3 doctorate
of public administration (DPA)), enrolling 2,012 students. Although the field
has experienced growth in the number of doctoral programs and students, it still
remains well below political science, sociology, or economics on those measures
(Cordes et al. 2008). As is the case with master’s degree programs, the titles
of doctoral programs vary considerably (Cordes et al. 2008). Despite relative
consistency in the overall number of doctoral programs, there has been a notable
movement away from the more applied DPA to the more research-oriented PhD.
As Brewer et al. (1998) emphasize, ‘doctoral programs are the flagships of
public administration education, and they are important incubators of public
administration inquiry. What is more, the state of doctoral education elucidates
the state of public administration’ (p 132). As such, since the earliest years and
continuing to the present, there has been a steady stream of studies designed to
describe doctoral education in the field and assess its quality. One of the earliest
reflections on the status of doctoral education in public policy took place at
a meeting organized by the Ford Foundation in 1975. This meeting brought
together 10 institutions2 with the express purpose of assessing what was happening
to the public policy movement, what problems had developed, what contributions
had been made, and what careers graduates were entering (Yates 1977, p 363).
Among the group that gathered, there was a general consensus that policy analysis
as a discipline had overcome early suspicion from academics that it was too applied
and unsophisticated, and agreement that their students were better than those in
public administration. They also identified notable differences across and among
their programs. Some emphasized training policy analysts or systems analysts,
while others prepared generalists who would be knowledgeable consumers of
analysis. There were disagreements on whether the emphasis should be on the
acquisition of analytical tools or an appreciation of political and organizational
contexts. They differed on whether their public policy programs were governed
by stand-alone faculty or by faculty with joint appointments in other academic
units, and whether the curriculum should be controlled centrally, as was the case
at Berkeley under Aaron Wildavsky, or through numerous intellectual clusters of
faculty, as Graham Allison explained the method used by Harvard (Yates 1977).
Subsequent studies generally evaluate doctoral programs spanning the broad
field and are not limited exclusively to public policy. Collectively, these studies
331
provide a sense of the content and quality of public affairs doctoral education in
the United States. In 1995 the most common substantive fields of study offered
within these doctoral programs were public policy (offered by 85% of programs),
general public administration (62%), public finance, organization behavior/
theory, management, and research methods (50–55% each), with fewer programs
offering comparative public administration or comparative public policy (40%),
personnel (30%), and public/administrative law or information systems (fewer
than 25% each) (Brewer et al. 1998). Regardless of specialty area, the doctoral
program curriculum was generally multidisciplinary, drawing most frequently
from economics (more than 50%), business/management (42%), sociology
(38%), political science (25%), and education, statistics, geography, planning and
psychology (each between 10% and 20%) (Brewer et al. 1998).
Two studies from 2008 provide insight into doctoral program content using
similar methodologies applied to different sampling frames. Holzer, Xu and Wang
(2008) use data collected through a 2003 questionnaire sent to 64 NASPAA
members with a PhD or DPA (from which they obtained 37 responses) combined
with data collected from program websites in 2005–06. The study by Cordes et
al. (2008) uses data from an online survey of public affairs programs administered
by Hank Jenkins-Smith to APPAM-member institutions, combined with data
collected from the websites of 40 APPAM-member universities with doctoral
programs, as well as data from the Survey of Earned Doctorates on graduating
PhDs and titles of doctoral dissertations listed in the summer 2005 issue of JPAM.
These studies provide insights into admission requirements, curriculum content,
and career paths.
Fully 60% of respondents in the NASPAA survey indicated that a master’s
degree was required for admission, but only 8% explicitly required an MPA or
MPP. Standardized tests (GRE, GMAT, or LSAT) are required by nearly all (94%)
programs. The GRE is the most common test, but only half of programs consider
these scores extremely important in admissions decisions (Holzer et al. 2008). On
average, 30% of coursework completed by doctoral students was at the master’s
level, but in some programs this reached 80% (Holzer et al. 2008). Roughly
one third of the NASPAA doctoral programs offered only daytime classes, while
nearly half provided only evening classes, and less than 20% used a combination
of day and night classes; the scheduling of classes was correlated closely with the
status of students as full- or part-time (Holzer et al. 2008).
Nearly all programs (95%) require a comprehensive or qualifying exam in the
form of supervised exams (46%), take-home exams (32%), or a combination
of methods (14%). Programs reported admitting, on average, nine students per
year, and experiencing dropout rates of 10% before the second year, 14% before
comprehensive exams, and 20% after the comprehensive exam (Holzer et al.
2008). In terms of research and teaching requirements, programs in the NASPAA
study were roughly evenly split on whether publishing a peer-reviewed article
was required or not, and whether teaching experience or teacher training was
required (Holzer et al. 2008).
332
Despite the movement from practitioner-oriented DPAs to research-oriented

PhDs noted earlier, programs report continuing placement outside of academia.
Faculty in a NASPAA survey reported that they expected half of their doctoral
students to pursue academic teaching/research careers, and the other half to
remain full-time practitioners (Holzer et al. 2008). Among the APPAM schools,
Cordes et al. (2008) found that, compared to their counterparts in political science,
economics, and sociology, the public policy and public affairs doctoral graduates
were less likely to seek academic employment and more likely to be interested
in employment in the US federal government.
The different sampling frames of the two studies understandably produced some
different findings in terms of curriculum. In the program focused exclusively on
APPAM-member universities, the only courses which were mandatory across
almost all programs were microeconomics, research methods, introduction
to public policy and policy analysis, basic statistics, and multiple regression
(Cordes et al. 2008). The study based on the wider array of programs within
the NASPAA-member schools found the most common core courses to be in
the areas of research methods I and II, public policy, organization theory, public
administration, research design, public finance, and human resources (Holzer et
al. 2008).
Stepping back from the micro level of analysis of course requirements, Cordes
et al. (2008) look at the disciplinary or theoretical grounding of programs. They
report that among the doctoral programs housed within APPAM-member
schools, some of the leading programs require students to ground themselves in
a particular social science discipline, including the programs at Harvard, Indiana
University, RAND, the University of Michigan, the University of Pennsylvania,
and the University of Texas-Dallas. Many schools allow students to explore a
range of substantive policy areas, while others are more focused. For example,
Brandeis requires students to pursue social policy, and the University of Pittsburgh
focuses on international affairs. The titles of dissertations provide an indication
of interests; the most popular dissertation topic areas in 2005 were management,
health policy, technology, social policy, education, and economic policy (Cordes
et al. 2008).
Dissertation titles are an indicator of topical interests, but a closer look at
the theories, methodologies, and findings of doctoral dissertations is required
to evaluate research quality and contribution to the knowledge base. Many of
the studies of program quality rely on analysis of dissertations abstracts and they
identify as key concerns whether the research reflected in doctoral dissertations
addresses significant issues (McCurdy and Cleary 1984; Stallings 1986), apply
rigorous methods (Cleary 1992; McCurdy and Cleary 1984), and contribute
to the knowledge base of the field (White 1986). A common critique is that
the dissertations are largely in the form of practitioner projects which provide
a case study in the form of a narrative descriptions of the author’s organization
(Cleary 1992) and are in essence advanced MPA or MPP projects (Felbinger et al.
1999). Some scholars even challenge the notion that the public affairs doctorate
333
is a research degree given so many studies that have ‘shown, alarmingly, that
very few doctoral graduates have contributed to a common body of knowledge
about the field’ (Felbinger et al. 1999, p 460). These shortcomings have been
attributed to the high proportion of part-time students (Felbinger et al. 1999);
the practitioner/academic orientation of programs (Brewer et al. 1998); the
number of students who do not want a research degree, but rather seek to
gain a promotion, to get consulting grants, or simply to be able to claim they
are a doctoral candidate (Felbinger et al. 1999); and the reliance on relatively
unsophisticated quantitative analysis in the form of regression-based analysis,
correlation analysis, or cross-sectional analysis, rather than longitudinal analysis,
structural equations, simulations, or controlled experiments (Brewer et al. 1998).
Despite the widespread criticisms, or perhaps in response to them, there have
been measurable improvements in research quality over time. In comparing
dissertations from 1981, 1990, and 1998 (n=142, 165, and 168, respectively)
Cleary (2000) finds improvements in research purpose, validity, and causal
relationships, although little change and continued weaknesses in theory testing
and selection of important topics. Moving forward, doctoral program quality can
be measured not only by contributions to the academic knowledge base, but also
in terms of how research informs policy analysis in government and think tanks.
Another indicator of improved quality is recognition by the National Research
Council (NRC). Doctoral programs in ‘Public Affairs, Public Policy, and Public
Administration’ were evaluated and ranked for the first time in 2007 alongside
other PhD programs. Initially, NRC had not planned to include this category in
its survey of doctoral program quality in 2007 because of concerns about the low
number of PhDs awarded, the number of graduates entering the professoriate,
the amount of external research, and the quality of published research. Joint
presentations by NASPAA and APPAM convinced the NRC to change its position
and to include the field in its study (Cordes et al. 2008).
Trump-era implications for NASPAA, APPAM, and the U.S.-based

policy schools
If we revisit Donald Stokes’ waves of development, we might ask if the Trump
era will represent a sixth wave and, if so, what label and characteristics it might
engender. In a time when U.S. government policies and rhetoric are emphasizing
isolationism, supremacy, and ideology, public policy education must redouble its
commitment to rational analysis and evidence-based decision making but in a
practical and strategic way.
When top national leaders in the United States and the public which supports
them reject the relevance of evidence and analysis, professional policy associations
and policy schools face unique challenges. The political environment of the Trump
administration requires that NASPAA, APPAM, and programs offering MPP and
policy-focused doctoral degrees consider their curriculum, their recruitment
strategies and enrollment prospects, and the employment opportunities that will
334
be available to their graduates. Several responses are beginning to emerge in

response to this new environment.
While none of the organizations or programs are abandoning their commitment
to policy analysis within the national government, they are increasingly shifting
their attentions down to the state and local levels, and up and out to international
levels. These shifts are a matter of emphasis, not wholesale change, and can take
several forms. For example, in the case of NASPAA, there is increasing attention
to the United Nations’ Sustainable Development Goals (SDGs) more so than U.S.
domestic policy goals as a framework for evaluating public policy implications.
MPP programs, which have historically emphasized placement of their graduates
as policy analysts in national government agencies, are focusing more attention
on prospects within the states, and this has implications for the policy issue areas
they study within their curricula.
Trump-era policies are also jeopardizing some of the progress that had been
observed in the internationalization and diversification of program enrollments.
As national policies have made it more difficult for students to obtain visas and,
perhaps equally as importantly, the political and social climate is perceived to be
more hostile to international students, universities and programs have struggled
to maintain prior gains in international student enrollments.
The most significant implications for U.S.-based public policy education
will be in how programs prepare their students to be effective within a highly
politicized and polarized environment without sacrificing their commitment to
rational analysis. The skills and talents students will need in this environment
extend beyond their capacity to engage in sophisticated analysis of data to also
‘package’ and ‘market’ the results of their analysis to audiences and decision
makers who prioritize ideology and image more than evidence and analysis.
This pressure may contribute to a further blurring of the lines between policy
and administration programs.
Future prospects
In the United States, education for public policy, public affairs, and public
administration has played and continues to play a leading role in defining the field.
Public policy and public administration education in the United States reflects
transitions in the needs of American society over time as well as lingering tensions
or challenges. While there have been noticeable new developments over the years,
each successive wave has added to, and not replaced, earlier models of education
for public service. We have now a mix of programs focused on politically neutral
administration and active policy-advisory roles, a focus on both management and
analytical skills, and an emphasis on traditional government service and broader
notions of governance involving actors in think tanks, private corporations,
nonprofit organizations, and interest groups.
Despite repeated calls for more differentiation among degree titles as distinct
disciplines, a variety of forces continue to promote the notion that these are
335
variations within a single field which Stokes (1996) collectively referred to as

public service. Although NASPAA and APPAM have different emphases and
memberships, both speak of MPA and MPP degrees together. It is unlikely that
the NRC would have included any of the component areas on their own and
instead opted to combine them into a single category. Studies of curriculum
content at both the masters and doctoral levels reveal little systematic variation
association with degree titles. Even so, the debate is likely to continue.
What does the future hold for public policy education in the United States?
A decade ago it seemed apparent that given the growing diversity of students,
including an increasing number of international students bringing a wider array
of languages and relevant country contexts, U.S.-based programs may need to
incorporate more ‘comparative analysis and lesson drawing across policy contexts
and time’ (Fritzen 2008, p 213). More recent drops in international applications
and enrollments now call into question the rationale for comparative analysis,
but they do not reduce the importance of the change.
Additionally, given Steven Rathgeb Smith’s explanation in Chapter 13 of the
growing importance of the voluntary sector to public policy and policy analysis in
the United States and the employment trends among master’s program graduates,
we may expect to see MPP and doctoral programs respond with more nonprofit-
focused course offerings and research projects as many MPA programs have
done. Regardless of the specific changes, we can expect that U.S. public policy
education will remain a role model for programs in other countries as it has been
since the earliest days. Ironically, to maintain this position of leadership in public
policy education, it will be increasingly important for U.S-based public policy
programs and professional associations to distance themselves from the policies
and policy decision processes of the U.S. government itself.
Notes
1 www.appam.org/about-appam.
2 Included were representatives of Harvard University (John F. Kennedy School of Government),
University of Texas, Austin (Lyndon B. Johnson School of Public Affairs), University of Michigan
(Institute of Public Policy Studies), Carnegie-Mellon University (School of Urban and Public
Affairs), Duke University (Institute of Policy Sciences and Public Affairs), Stanford University
(Graduate School of Business), Princeton University (Woodrow Wilson School of Public and
International Affairs), University of California Berkeley (Graduate School of Public Policy),
The Rand Corporation (Graduate Institute for Policy Studies) and Yale University (School of
Organization and Management).
References
Averch, H. and Dluhy, M. (1992) ‘Teaching Public Administration, Public
Management, and Policy Analysis: Convergence or Divergence in the Masters
Core’, Journal of Policy Analysis and Management 11(3): 541–51.
Brewer, G. A., Facer II, R. L., O’Toole Jr., L. J., and Douglas, J. W. (1998) ‘The
State of Doctoral Education in Public Administration: Developments in the
Field’s Research Preparation’, Journal of Public Affairs Education 4(2): 123–35.
336
Cleary, R. E. (1992) ‘Revisiting the Doctoral Dissertation in Public Administration

– An Examination of Dissertations of 1990’, Public Administration Review 52(1):
55–61.
Cleary, R. E. (2000) ‘The Public Administration Doctoral Dissertation
Reexamined: An Evaluation of the Dissertations of 1998’, Public Administration
Review 60(5): 446–55.
Cordes, J., Conger, D., Ladd, H., and Luger, M. (2008) ‘Undergraduate and
Doctoral Education in Public Policy: What? Why? Why Not? Whereto?’ Journal
of Policy Analysis and Management 27(4):1009–26.
Ellwood, J. W. (2008) ‘Challenges to Public Policy and Public Management
Felbinger, C. L., Holzer, M., and White, J. D. (1999) ‘The Doctorate in Public
Administration: Some Unresolved Questions and Recommendations’, Public
Administration Review 59(5): 459–64.
Fritschler, A. L., Mackelprang, A. J., Blaustein, E., and Brozen, R. (1977)
‘Graduate Education in Public Affairs/Public Administration: Results of the
1973 Survey’, Public Administration Review 37(5): 488–94.
Fritzen, S. A. (2008) ‘Public Policy Education Goes Global: A Multi-dimensional
Challenge’, Journal of Policy Analysis and Management 27(1): 205–14.
Geva-May, I. and Maslove, A. (2006) ‘Canadian Public Policy Analysis and Public
Policy Programs: A Comparative Perspective’, Journal of Public Affairs Education
12(4): 413–38.
Geva-May, I., Nasi, G., Turrini, A., and Scott, C. (2008) ‘MPP Programs
Emerging Around the World’, Journal of Policy Analysis and Management 27(1):
187–204.
Gulick, L. H. (1936) ‘Notes on the Theory of Organization’, in L. Gulick and
L. Urwick (eds), Papers on the Science of Administration, New York: Institute of
Public Administration, pp 3–35.
Henry, L. (1995) ‘Early NASPAA History’, www.naspaa.org/about_naspaa/
about/history.asp.
Henry, N., Goodsell, C. T., Lynn Jr., L. E., Stivers, C., and Wamsley, G. L. (2009)
‘Understanding Excellence in Public Administration: The Report of the Task
Force on Educating for Excellence in the Master of Public Administration
Degree of the American Society for Public Administration’, Journal of Public
Affairs Education 15(2): 117–33.
Holzer, M., Xu, H., and Wang, T. (2008) ‘The Status of Doctoral Programs in
Public Affairs and Administration’, Journal of Public Affairs Education 13(3/4):
631–47.
Lepawsky, A. (1970) ‘Graduate Education in Public Policy’, Policy Sciences 1(4):
443–57.
McCurdy, H.E. and Cleary, R.E. (1984) ‘Why Can’t We Resolve the Research
Issue in Public Administration?’, Public Administration Review, 44(1): 49–55.
NASPAA (n.d.) ‘NASPAA Data Center’, www.naspaa.org/DataCenter/index.asp.
337
NASPAA (2008) General Information and Standards for Professional Masters Degree
Programs (January 2008 edition), https://naspaaaccreditation.files.wordpress.
com/2014/04/old-accreditation-standards.pdf.
NASPAA (2014) Accreditation Standards for Master’s Degree Programs (adopted
October 16 2009; amended November 6 2014), https://naspaaaccreditation.
files.wordpress.com/2015/02/naspaa-accreditation-standards.pdf.
Overman, S., Perry, J., and Radin, B. (1993) ‘Doctoral Education in Public
Affairs and Administration: Issues for the 1990s’, International Journal of Public
Administration, 16(3): 357–80.
Rubaii, N. and Calarusse, C. (2012) ‘Cultural Competency as a Standard for
Accreditation’, in Norman-Major, K. A. and Gooden, S. T. (eds), Cultural
Competency for Public Administrators, Armonk, NY: M. E. Sharpe.
Stallings, R. A. (1986) ‘Doctoral Programs in Public Administration: An Outsider’s
Perspective’, Public Administration Review 46(3): 235–40.
Stokes, D. E. (1996) ‘“Presidential” Address: The Changing Environment of
Education for Public Service’, Journal of Policy Analysis and Management 15(2):
158–70.
White, J.D. (1986) ‘Dissertations and Publications in Public Administration’,
Public Administration Review, 46(3): 227–34.
Wildavsky, A. (1987) Speaking Truth to Power: The Art and Craft of Policy Analysis
(2nd edn), Piscataway, NJ: Transaction Publishers.
Wilson, W. (1887) ‘The Study of Administration’, Political Science Quarterly 2(2):
197–222.
Yates, D. (1977) ‘The Mission of Public Policy Programs: A Report on Recent
Experience’, Policy Sciences 8(3): 363–73.
338
EIGHTEEN
The influence of policy analysis in

the United States on the
international experience
B. Guy Peters
Although there could well be quibbles about the extent to which this
characterization is valid, it is difficult to argue that the modern science, or art, of
policy analysis did not begin in the United States. Policymakers had always had
advisors who gave them ideas about policy1 and attempted to prevent them from
making egregious errors, but that advice was generally far from systematic. With
the work of Harold Lasswell and his colleagues (Lasswell 1951) a more systematic
conception of public policy and policy analysis was developed and began to be
diffused within the United States and later to the rest of the world.
To argue that the United States has played a central role in the development
of policy analysis is not meant to chauvinistic, nor does it deny the important
contributions of scholars and practitioners in many other countries (see Radin
2013). Rather, it is intended to recognize that despite the general disdain that
the population has for government, or perhaps because of it, government in the
United States has been the source, and target, of many innovations in public
policy analysis, and at least some elements of academia in the United States have
had a greater commitment to involvement with policy than has been true for
the academy in other societies.
To understand the influence of policy analysis in the United States on the rest of
the world also requires more than a little attention to the historical development
of policy analysis and the tools associated with that mode of policy advice within
government. Some of the specific mechanisms for policy analysis discussed
below have long fallen out of favor with policymakers and with their academic
counterparts, but they remain significant components of the foundation on which
contemporary policy analysis is constructed. As such, these ideas concerning public
policy condition a good deal of contemporary policy analytic work. In addition,
processes of diffusion in which the policy analysis community in the United
States has played a significant part in the development of policy analysis around
the world, and to understand that diffusion, requires an initial understanding of
the developments within the Unite States.
The above having been said, European scholars in particular have been doing
significant amounts of high-quality public policy work, albeit often under different
guises (see Sabatti 1973 for an early statement). Somewhat like Monsieur Joudain,
they had been speaking policy all their life without knowing it. Some of what
339
would be considered policy studies in the United States went under the label of
political economy in Europe, long before that term became more popularized.
Public administration scholars in Europe also did a great deal of research that
might have been labeled public policy in North America, whereas those two
streams diverged more in the United States, and finally there was also a strand
of political sociology in Europe that had a strong policy concern and made
significant contributions to thinking about public policy. So the very American
story being told here needs to be tempered somewhat with an understanding of
these alternative academic traditions.2
The dimensions of influence

To understand how the development of policy analysis in the United States has
developed, and in turn influenced the development of this field internationally, I
will examine five dimensions of the field of policy analysis. One might be able to
divide up the field of policy analysis in different ways, but these five dimensions
do capture, I believe, the more important aspects of public policy as a field of
study and of practice, and also help to demonstrate how developments in the
United States have had a much wider international influence.
Policy analysis: content
Perhaps most importantly, policy analysis in the United States has been influential
because much of the intellectual development in this field, and especially
development early in the existence of the field, was in the United States. This
centrality to policy analysis is hardly an accident, given the importance of the
pragmatism of John Dewey and others with the same general approach to
knowledge in American philosophy and social thought. The assumption of this
pattern of thought, and of a significant component of American education, is that
knowledge is important primarily because of its capacity to improve the society
within which it is developed. Charles Merriam (1945, p 7) called this tradition
‘democratic engineering’ and Lasswell (1951, p 3) referred to it as being a ‘social
therapist.’ Policy analysis is a natural extension of that pragmatic logic, applying
the methods and the theories from the social sciences in order to ameliorate, if
not solve, public problems.
Cost-benefit analysis is one of the principal strands of policy analysis.
Although its intellectual roots may be more European, especially in the work
of the economist Alfred Marshall, it first became institutionalized in the public
sector in the United States, especially in the Army Corps of Engineers and its
implementation of flood control projects. This analytic approach was then diffused
to other governments, as well as to international organizations and the private
sector (Hira and Parfitt 2004). Although critics have decried the ‘economism’
of this approach to public policy (Self 1977), it continues to be widely used in
340
The influence of policy analysis in the United States on the international experience
a range of policy areas, and with extensions such as risk-benefit analysis can be
considered the foundation of much of contemporary policy analysis.
In addition to the ‘hard’ economic analysis of policy, the American policy
analytic community has also been at the center of several ‘softer’ versions of policy
analysis. Beginning with Lasswell (1956), there has been a significant interest in
the policy process, and the impact of process on final policy outcomes. While
the policy process literature has been dominated by political scientists interested
perhaps more in the nature of the political aspects of the process than in the
policy outcomes (see Jones 1984; Sabatier and Weible 2014), understanding the
process can still provide insights for policymakers attempting to understand how
best to advance their programs within the political system.
In addition to the relatively simple linear and political models of the policy
process that have been central to the study of the policy process, more analytic
models have also been developed in order to analyze how policy is made and to
identify the effects of process on policy outcomes. For example, William Dunn
(2010) developed a model of the policy process that, while sharing some of the
same linearity, identifies the stages in a more analytic manner and links policy
analysis more directly to the process of policy making.
In addition to the general models of the policy process, the principal approaches
to policy change, all of which have a clear process focus, have definite American
roots (see Real-Dato 2009). If we consider the three major models of policy
change—the Advocacy Coalition Framework (ACF), Punctuated Equilibrium
Theory (PET), and Multiple Streams—then they all have been developed primarily
by American scholars. The ACF is of course associated principally with the late
Paul Sabatier, but has been developed by other scholars such as Chris Weible and
Hank Jenkins-Smith (Sabatier and Jenkins-Smith 1993) and the PET approach
is increasingly diffused among political scientists around the world interested in
agenda setting and policy change (Princen 2013) The multiple streams approach
is generally associated with John Kingdon (2003) though the garbage can version
of this approach has involved one European (Cohen et al. 1971).
A concern with participation and democracy is another strand in the American
policy analysis literature that has had some significant impact outside the United
States (Schneider and Ingram 1997; Ingram et al. 2016). For example, the
development of participatory modes of policy analysis was important in shaping
how policy was to be conducted (Durning 1993). This version of policy analysis
has not been as easy to diffuse as are the ‘harder’ economic models, given that it
depends on social and political values that may not be shared (even within some
quarters of the American policy community). But at the same time, attempting
to infuse the analysis of policy with concern for democratic values does go back
to the roots of policy analysis with Merriam and Lasswell.
While much of the diffusion of ideas about policy has been to Europe, there
has also been a significant influence on policy analysis in other parts of the world,
perhaps notably Latin America. Scholars such as Guillaume Fontaine in Ecuador
and Luis Aguilar and Jose Luis Mendez in Mexico have written major works of
341
policy analysis that incorporate some of the American perspectives, while adding
their own distinctive conceptions of policy. The ideas of policy analysis have
also reached less democratic regimes, with a number of think tanks now being
formed in China and Russia, all using some of the basic ideas coming from the
largely American tradition.
Policy analysis: structures
As well as pioneering many aspects of policy analysis, the structures for performing
and promoting that analysis also have some distinctive American roots, and
perhaps the most important of these is the ‘think tank’ (see the chapter by Rich
in this volume.) Think tanks are perhaps the most significant structure for policy
analysis, and they have rather distinctive American roots in organizations such as
the Brookings Institution. Washington is well populated with organizations of all
political allegiances that attempt to influence government. These organizations
have created a vibrant and creative market for policy ideas in government.
Of course, there has been a major diffusion of the idea of the think tank around
the world, and there have been significant developments of organizations of this
type in all democratic systems, and even in some less democratic ones (Stone and
Denham 2004). This diffusion represents an important success for making policy
analysis more available to decision makers in any number of countries, but in
most cases there is not the wide range of organizations and ideas that continue
to characterize these organizations in Washington.
In addition to think tanks, one of the roots of policy analysis for state and local
government is the tradition of ‘bureaus of government’ or other organizations of
that ilk in public universities—especially the land grant universities. These public
universities, with a clear mandate to improve the quality of life within their states,
have developed a relatively common model of having a research unit that provides
advice to the state itself, but especially to local governments within the state.
As will be noted below, the tradition of applied education going hand in hand
with more theoretical education has been institutionalized in state universities
for well over a century.
The transformation of government accounting from ‘green eyeshade’ financial
accounting toward more policy-focused versions of accounting also began with
the transformation of the General Accounting Office, now the Government
Accountability Office (GAO), in the United States (Mosher 1979). Like most
government accounting offices the GAO is responsible to the legislature, and the
transformation from financial accounting to performance accounting represented
a major injection of policy analysis into the process. That style of accounting has
since then been adopted by the Bundesrechnungshof, the Riksrevisionsverket, and
other audit organizations around the world.
In addition to policy analysis through legislative auditing organizations, the
Congress of the United States developed other policy analytic structures that
have been used as models for other legislatures. The Congressional Reference
342
Service has numerous analogies in legislative organizations responsible for

assisting legislators in drafting legislation. What is more distinctive in the United
States has been the policy analytic capacity of committee staffs. Whereas most
legislatures have small staffs with limited capacities (see Peters 2016) the U.S.
Congress has literally thousands of staffers involved in policy advice. In addition,
the Congressional Budget Office (CBO) (see Joyce 2011) provides Congress with
high-quality information on public expenditures and the policy consequences of
their budgetary choices, and has been copied by other legislatures, for example
in the form of the Parliamentary Budget Office in Canada (Levy 2008). (See the
chapter by Joyce in this volume for more on the GAO and CBO.)
Policy analysis: evaluation research
Evaluation research has been another major American contribution to policy

analysis. Governments have always attempted to understand the effects of their
policies and to develop information that can be used to improve those policies,
or to repeal poorly performing ones. But the systematic evaluation of policies,
whether in the public sector itself or through private sector consultants, was largely
an innovation in the United States. In historical terms, this is often associated
with the Great Society of President Lyndon Johnson, which allocated large sums
of money to evaluation (Aaron 1973; Shadish et al. 1991). But the intellectual
roots for evaluation research are somewhat older and deeper, going back to the
fundamental pragmatist notions of usable knowledge.
That pragmatic interest in usable knowledge has manifested itself in the work of
a number of scholars interested in evaluation and in the use of the social sciences
as mechanisms for improving society. For example, Carol Weiss has been a leading
scholar in evaluation research, and has emphasized the development of knowledge
that could be used to improve policy, especially educational policy (1977), and
Charles E. Lindblom and his colleague David Cohen (1979) argued for making
the social sciences more concerned with usable knowledge, rather than simply
pursuing their usual theoretical arguments and discussions.3
The above having been said, evaluation research has suffered something of a
decline, beginning with the neoliberal governments of Reagan and Thatcher, inter
alia. Not only did these governments not care particularly about the effectiveness
of policies, rather only about efficiency, but they also thought they knew the
answer—the market—to the policy problems confronting society. To a great
extent more in-depth evaluation was replaced by ‘performance measurement’
that would focus on short-term achievement of goals rather than asking whether
those goals were the proper ones, or whether there were better programs possible
to achieve them.
Evert Vedung (2010) has, however, argued that the more recent emphasis on
evidence-based policy is turning the tide back to thinking more comprehensively
about policy. The evidence-based policy movement has been, however, more
European and Antipodean than American. In particular the British government
343
and several organizations in health care in Europe began to press for clear evidence
of the effectiveness of programs, and to some extent also began to ask more
fundamental questions about policy goals and program design.
Policy analysis: the budget
One of the clearest cases of ideas in policy analysis in the United States being
diffused to the rest of the world can be found in public budgeting. In the first
instance the incremental model of budgeting—what was for several decades the
standard model of budgeting—was a product of Aaron Wildavsky’s research.
A series of reforms in the budgetary process in the United States injected policy
analysis directly into the consideration of public expenditure. These reforms
of the budgetary system are clearly extensions of the traditions of rational and
systematic analysis of public policy, and although they were less than successful
in practice they did represent significant analytic advances.
The most important of these advances was Planning-Programming-Budgeting
Systems (PPBS, or program budgeting). Whereas conventional budgeting was,
and continues to be, focused on individual organizations, the assumption of PPBS
was that there were more fundamental programs in government that cut across
organizations, even within a single ministry or department. The development of
this methodology for budget reform involved extensive policy analysis because
a major component of the logic was to be able to justify the best mixture of
expenditures, generally through utilitarian methods such as cost-benefit analysis
(see above).
The basic concept of PPBS was exported to a number of other countries.
For example, in France it was called RCB (rationalisation des choix budgetaires;
Lord 1973) and very similar ideas were adopted in Sweden (Jacobssen, Pierre
and Sundeström 2015, pp 108–11). Various donor organizations mandated this
style of budgeting and policy analysis for poor countries, often with disastrous
results. The informational demands and the need for analytic capacity tended to
overwhelm many less-developed systems, and to some extent the governments
of more developed governments, rather than help them develop more rational
public budgets.
Although PPBS is the clearest example of a policy analytic approach to budget
reform, several other attempts at budget reform also have attempted to introduce
greater analytic rigor into the budgetary process. Although largely dismissed
by Aaron Wildavsky (1978), the doyen of students of budgeting, Zero-based
Budgeting (ZBB) did require greater analytic rigor than traditional forms of
budgeting. In particular, ZBB required organizations to identify the bundles of
services they currently delivered and what additional bundles of services could
be delivered. This required the organizations in government to get away from
the usual incremental budgeting to think about linking budgetary politics with
policy analysis.
344
The irony of this history of innovation in public budgeting is that despite the
innovations made in the past, public budgeting in the US—most clearly at the
federal level—remains one of the most antiquated systems among the developed
democracies.4 This failure to sustain innovation, and the failure to adopt other
innovations in public budgeting such as accrual budgeting, reflects the increasing
domination of politics over analysis in making policy. This failure helps to make
the general point that no matter how good the policy analysis, that analysis must
be utilized in a politicized environment.
Policy analysis: education
Finally, as mentioned briefly above, policy analysis was institutionalized in

American universities earlier than in other countries and the ‘School of Public
Policy’ has become an integral part of American higher education. Many of
the leading universities have schools of public policy and provide masters and
doctoral-level programs (and some undergraduate programs) in policy studies5
(see the chapter by Rubaii on graduate education in this volume). This model of
policy education is increasingly being copied in other countries but the heartland
of systematic policy education remains in the United States.
Again, there are some historical roots to this somewhat applied dimension of
higher education in the United States. The Morrill Act 1862, creating the system
of land grant colleges, was one of the greatest policy interventions in higher
education in American history.6 The land grant colleges, now universities, were
for scientific instruction in agriculture and the ‘mechanic arts.’ In addition to their
original purposes, these universities have maintained a commitment to public
service and applied education, even in the context of more general university
structures.7 The ‘bureaus of government’ mentioned above are particularly
prevalent in these universities, as well as having been diffused to other public
universities.
There are, of course, good examples of this more applied perspective on higher
education coming from Europe, and to some extent also education on policy
analysis per se. The Fachhochscule in Germany, as a school of applied sciences,
was copied in Austria and Switzerland, with analogues in the Scandinavian
countries. While these schools did not initially offer programs in subjects such
as public policy or public administration, and many still do not, they did have
an orientation toward more applied learning that is important as a potential
foundation for policy education.
Barriers to diffusion
Although the tradition of policy analysis in the United States has been widely
diffused around the world, there are also some important barriers to that diffusion.
Some of these are institutional and structural, some are more cultural, and some
are political. Thus, although some aspects of the American style of policy analysis
345
can be found throughout the world, it may not be as pervasive as assumed by

some, and in some cases it may be misunderstood. That misunderstanding is not
unusual when ideas and practices are spread to different cultures, but it is important
to separate the real adoptions from the more apparent adoptions.As noted above,
policy analysis has been institutionalized in, among other organizations, a large
number of think tanks in Washington, DC and in state capitols, and even in
some cities in the United States, and these organizations have become a standard
part of the policy process in the United States. On any major policy issue, and
most minor ones, one or more think tanks will be weighing in with analysis
and opinion, so that the task of policymakers becomes attempting to assess the
quality of the evidence.
The idea of the think tank has now been diffused widely to a number of other
countries, albeit often without the same independence that has characterized
many of the think tanks in the United States (Stone and Denham 2004). Many
think tanks in Europe, for example, are organized by political parties to assist their
members of parliament in making decisions, or by interest groups to advocate on
particular policy issues. In these European cases, the career civil service has been
central to policy advice and remains so, if in a somewhat diminished state, so that
there has been somewhat less space for the think tanks to function. Likewise,
think tanks in Europe tend to be more closely allied with political parties and
hence may be serving one client more than attempting to be more general policy
entrepreneurs.
The openness and multiple points of access of the policy process in the United
States also provide more of a market for policy ideas. With the 50 states and
multiple cities having capacity for autonomous policy development, and the
executive and legislative branches vying for policy dominance in Washington
and in state capitols, there is a broad market for policy ideas. (See the chapters
by Van Landingham on state policy analysis and Mossberger et al. on local policy
analysis in this volume.) Political systems with fewer opportunities, and a more
dominant civil service system, may provide fewer alternatives for the influence
of think tanks or other policy advocates from outside the system.
The United States as follower

Although the United States has been a leader in many, if not most, areas of policy
analysis, there are several areas in which it has been more of a follower. These areas
of policy analysis to some extent reflect different intellectual traditions, as well
as different ways in which public policies may be considered within government
itself. Further, there are some aspects of policy analysis that may have begun
in the United States and are now not as popular either in the academic or the
government communities but which are alive and well in Europe.
Although there have been several leading American scholars in this field,
notably Deborah Stone (2012) and Frank Fischer (this volume, and 2016),
interpretative forms of policy analysis have been less central to the field within the
346
more quantitative and positivist traditions of the United States. There was some
discussion of post-positivist forms of analysis in American policy analysis, but
the dominant approach has been and remains more quantitative (Hawkesworth
1989). Likewise discursive approaches to public policy, and to the social sciences
more generally, have been more central to the field in Europe than in the United
States (see Hajer and Wagenaar 2003; Torfing 2010).
Network governance and policy analysis is a second and related area in which
non-American scholars have been more central to the research tradition. Again
there are notable American scholars working in this area (see Laumann and Knoke
1987) but the heartland of this approach has been in the Low Countries and
Scandinavia (Koppenjan and Klijn 2004; Sørensen and Torfing 2007). Perhaps
because of the corporatist tradition in governance and the greater legitimacy of
interest groups’ involvement in these governments, the development of network
analysis and the relationship of these structures has been more important than in
the pluralist United States.
Critical theory and Marxist approaches to policy are also much more European
than American. Those intellectual traditions have been weak in the United States
and have had relatively little influence on policy analysis, or on the social sciences
more generally. In Europe, however, political economy approaches have included
some Marxist versions within that general approach, and discourse models have
included aspects of critical theory (see the discussion in Knoepfel et al. 2011).8
Finally, the substantive interest in the welfare state has been much greater in
Europe than in the United States, perhaps for obvious reasons. For example the
succession of British scholars such as Richard Titmuss (1969), Brian Abel-Smith,
Rudolf Klein, and Julian Le Grand have created a lively and persistent tradition of
social policy research at the London School of Economics and Political Science
and other British universities. In Scandinavia, Gunnar (1960) and Alva Mydal
(1960) created an analogous tradition in studying the welfare state, as well as
development policy, within Sweden. More recently a series of German scholars
such as Stephan Leibfried (2001), as well as Bruno Palier (2010) and others
in France, have made major contributions to understanding social policy, and
especially the politics of social policy, as pressures for downsizing the welfare state
become more prominent. The United States certainly has not been totally absent
from the field of social policy—scholars such as Theodore Marmor (Marmor et al.
1994) and Paul Pierson (2001) come readily to mind but that said, the distinctive
pattern of study for social policy is not so well established as in Europe.
United States as leader, but in the opposite direction

Having sung hymns of praise to the development of policy analysis in the United
States, it is also important to note the current retrograde movements. The United
States unfortunately has become the leader in a movement away from careful
use of evidence and analysis in policy making toward a more politicized style of
policy. While all political actors have had their own opinions as long as there has
347
been politics, now everyone has their own facts. The failure to accept evidence,
even coming from reputable scientific sources (climate change for example) makes
effective policy making all the more difficult.
For policy analysis in the United States this is now the best of times and the
worst of times. Never has so much data been available, and never have there been
so many people trained in the use of policy analysis and other forms of analysis
who could be contributing to better policies. But these efforts are stymied by
relentless partisanship and distrust, and a failure to agree on the most fundamental
aspects of public policy.
Summary and conclusions

There appears to be little doubt that much of the development of public policy
studies has been centered in the United States. This central role for American
approaches might be expected, given the size of the academic community in
the United States, as well as the dominance of the English language. But even
given those numerical and linguistic advantages, the dominance of the American
approaches to policy analysis have been greater than might have been expected.
Although there may be a variety of explanations, the pragmatic streak in American
political philosophy and education may account for at least some of the emphasis
on policy analysis.
Although we can identify the consequences of the several strands of diffusion
of ideas and approaches about policy analysis, I have not been able to identify the
‘smoking gun’ of the mechanisms through which the influence of the patterns
and styles of policy analysis in the United States has been exerted in other
governments and other academic communities. However, if we utilize the logic
developed to explain institutional isomorphism (Dimaggio and Powell 1983) we
can consider at least three mechanisms through which policy analytic styles may
have been transferred.
The simplest is perhaps mimesis, or simple copying. The logic is that a prospective
adopter of the institutional structure, or the policy analytic perspective in this case,
identifies that there is a good option available and copies that.9 In some cases the
adoption may occur simply because the program is considered to be ‘modern’ or
scientific, or perhaps simply because other countries are already implementing the
program. The spread of ideas through international organizations may facilitate
this rather simple form of diffusion.
Although similar to mimetic forms of isomorphism (see Mizruchi and Fein
1999), normative isomorphism depends on the appeal of the ideas associated with a
particular program or institution based on how well it conforms to fundamental
social and political values. In other words, policy analysis instruments and
institutions are adopted because it is the right thing to do. In this case the ‘right’
may mean that the pursuit of efficiency and effectiveness through analysis is an
important normative value. Or in other cases a more democratic or participatory
form of policy making may have substantial normative appeal.
348
As noted above concerning some of the budgetary reforms, coercion may also
lead to the diffusion of ideas just as it has for institutional structures. The most
obvious example of this form of diffusion has been through international donor
organizations, as well as other donors, that have required recipient countries to
adopt certain rationalist methods of budgeting. In addition, forms of economic
analysis such as cost-benefit have become ubiquitous as means of making policy
choices, especially for capital investments, and that has also been driven in no
small part by donor organizations.
The other question which must be considered here is whether the (relatively)
high level of influence of the United States in policy analysis is declining. That may
be expected simply through the development of larger communities of scholars
around the world, especially in Europe, and the significant contributions in all
fields of the social sciences coming from those communities. The developments
in other areas of the world may also reflect perspectives on policy analysis that
are more closely aligned with the needs of countries in those areas than are the
ideas coming from the United States (see Fontaine 2015).
As well as the growth of the policy analytic communities in other settings,
the influence of American policy analysis may be declining simply because
governments in the United States have been less innovative. The combination
of gridlock and the control of Congress by the Republican Party has meant that
there is little or no capacity for innovative policy making. To the extent that there
are important policy changes, these are occurring at the state level, and these
innovations may be less visible to the international community.
The policy analysis community in the United States has been central to the
development of this field, both in the academic and intellectual perspective and
in the applied perspective. This applied use of the social sciences has a long
tradition in the United States and there are a number of dimensions of the
influence of American approaches to policy. Despite that significant influence
over time, there are a number of important—an increasing—developments in
other parts of the world that are supplementing and perhaps supplanting that
long American tradition.
Notes
1 Joseph advising Pharaoh in the book of Genesis is one of the earliest examples.
2 As well as some European scholars who very clearly have been interested in public policy per
se, perhaps most notably Richard Rose (although he, unlike Ted Cruz, is American by birth).
3 For an interesting review and critique see Zeisel (1981).
4 The US federal budget is, for example, line-item, annual and fails to separate capital from
current expenditure. The budgetary process at the state level tends to be more modernized but
Congress, in particular, has been a barrier to any reform in Washington.
5 During the 1980s the Alfred P. Sloan Foundation provided funding for a small number of
universities to initiate undergraduate programs in public policy, intended not only to provide
professional education, but also to generate greater policy awareness among the population.
On the latter perspective on the role of policy education see MacRae and Wilde (1979).
349
6 Perhaps the only real competitor was the GI Bill, providing educational benefits for veterans
of World War II, and then subsequent conflicts.
7 These universities include some of the most distinguished American public universities including
Wisconsin, Minnesota, Texas A and M, and Michigan State.
8 But see Fischer (2016).
9 This version of isomorphism is not dissimilar to the logic found in the more general literature
on the diffusion of innovations.
References
Aaron, H. J. (1973) Politics and the Professors: The Great Society in Perspective,
Washington, DC: The Brookings Institution.
Cohen, M. D., March, J. G., and Olsen, J. P. (1971) ‘A Garbage Can Model of
Organizational Choice’, Administrative Science Quarterly, 17: 1–25.
Dimaggio, P. J. and Powell, W. W. (1983) ‘The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields’, American
Sociological Review 48: 147–60.
Dunn, W. N. (2010) Public Policy Analysis: An Introduction (5th edn), Englewood
Cliffs, NJ: Prentice-Hall.
Durning, D. (1993) ‘Participatory Policy Analysis in a Social Service Agency:
A Case Study’, Journal of Public Policy Analysis and Management 12: 287–312.
Fischer, F. (2016) Handbook of Critical Policy Studies, Cheltenham: Edward Elgar.
Fontaine, G. (2016) El Analisis de Politicas Publicas: Conceptos, Teorias y Métodos,
Quito: FLACSO.
Hajer, M. and Wagenaar, H. (2003) Deliberative Policy Analysis: Understanding
Governance in the Network Society, Cambridge: Cambridge University Press.
Hawkesworth, M. E. (1989) Theoretical Issues in Policy Analysis, Albany, NY: State
University of New York Press.
Hira, A. and Parfitt, T. (2004) Development Projects for a New Millennium, London:
Praeger.
Ingram, H., de Leon, P., and Schneider, A. L. (2016) ‘Public Policy Theory and
Democracy: The Elephant in the Corner’, in B. G. Peters and P. Zittoun (eds),
Contemporary Approaches to Public Policy, London: Macmillan.
Jacobssen, B., Pierre, J. and Sundeström, G. (2015) Governing the Embedded State:
The Organizational Dimension of Governance, Oxford: Oxford University Press.
Jones, C. O. (1984) An Introduction to the Study of Public Policy (3rd edn), Monterey,
CA: Brooks-Cole.
Joyce, P. O. (2011) The Congressional Budget Office: Honest Numbers, Power and
Policymaking, Washington, DC: Georgetown University Press.
Kingdon, J. W. (2003) [1985] Agendas, Alternatives and Public Policies (2nd edn),
New York: Longman.
Knoepfel, P., Larrue, C., Varone, F., and Hill, M. (2011) Policy Analysis, Bristol:
Policy Press.
Koppenjan, J. and Klijn, E. H. (2004) Managing Uncertainties in Networks: A
Network Approach to Problem-Solving and Decision Making, London: Routledge.
350
Lasswell, H. D. (1956) The Decision Process; Seven Categories of Functional Analysis,

College Park: College of Business and Public Administration, University of
Maryland.
Lasswell, H. D. and Lerner, D. (1951) The Policy Sciences: Recent Developments in
Theory and Method, Stanford, CA: Stanford University Press.
Laumann, E. O. and Knoke, D. (1987) The Organizational State: Social Change in
National Policy Domains, Madison: University of Wisconsin Press.
Leibfried, S. (2001) Welfare State Futures, Cambridge: Cambridge University Press.
Levy, G. (2006) ‘A Parliamentary Budget Officer for Canada’, Canadian
Parliamentary Review 21(2): 35–41.
Lord, G. (1973) The French Budgetary Process, Berkeley, CA: University of
California Press.
MacRae, D. and Wilde, J. A. (1979) Policy Analysis for Public Decisions, North
Sinuate, MA: Duxbury.
Marmor, T. R., Mayhaw, J. L., and Harvey, P.L. (1994) America’s Misunderstood
Welfare State: Persistent Myths, Enduring Realities, New York: Basic Books.
Merriam, C. (1945) Systematic Politics, Chicago, IL: University of Chicago Press.
Mizruchi, Mark S. and Lisa C. Fein (1999) ‘The Social Construction of
Organizational Knowledge: A Study of the Uses of Coercive, Mimetic, and
Normative Isomorphism’, Administrative Science Quarterly 44: 653–83.
Mosher, F. C. (1979) The GAO: The Quest for Accountability in American Government,
Mydal, G. (1960) Beyond the Welfare State, New York: John Wiley.
Palier, B. (2010) A Long Goodbye to Bismarck: The Politics of Welfare Reform in
Continental Europe, Amsterdam: Amsterdam University Press.
Peters, B. G. (2016) ‘Bureaucracy for Democracy: Administration in Support of
Legislatures’, paper prepared for Conference on Legislative Staffing, European
University Institute, Florence.
Pierson, P. (2005) The New Politics of the Welfare State, Oxford: Oxford University
Press.
Princen, S. (2013) ‘Punctuated Equilibrium Theory and the European Union’,
Journal of European Public Policy 20: 854–70.
Radin, B. (2013) Beyond Machiavelli: Policy Analysis Reaches Middle Age (2nd edn),
Washington, DC: Georgetown University Press.
Real-Dato, J. (2009) ‘Mechanisms for Policy Change: A Proposal for a Synthetic
Analytic Framework’, Journal of Comparative Policy Analysis 11: 117–43.
Sabatier, P. A. and Jenkins-Smith, H. (1993) Policy Change and Learning: An
Advocacy Coalition Framework, Boulder, CO: Westview Press.
Sabatier, P. A. and Weible, C. (2014) Theories of the Policy Process (3rd edn),
Sabatti, P. (1973) ‘Theory of Public Policy—European Contributions’, Policy
Studies Journal 2: 127–32.
Schneider, A. L. and Ingram, H. M. (1997) Policy Design for Democracy, Lawrence,
KS: University Press of Kansas.
351
Self, P. (1977) Econocrats and the Policy Process: The Politics and Philosophy of Cost-
Benefit Analysis, Boulder, CO: Westview Press.
Shadish, W. R., Cook, T. D., and Leviton, L. C. (1991) Foundations of Program
Evaluation: Theories of Practice (Thousand Oaks, CA: Sage).
Sørensen, E. and J. Torfing (2007) Theories of Democratic Network Governance,
Basingstoke: Macmillan.
Stone, D. A. (2012) Policy Paradox: The Art of Political Decision-making, New York:
W. W. Norton.
Stone, D. and Denham, A. (2004) Think Tank Traditions—Policy Research and the
Politics of Ideas, Manchester: University of Manchester Press.
Titmuss, R. M. (1969) Essays on the Welfare State, Boston, MA: Beacon Press.
Torfing, J. (2010) ‘Discourse Models in Policy’, in B. Badie, D. Berg-Schlosser
and L. Morlino (eds), IPSA Encyclopedia of Political Science, London: Sage.
Vedung, E. (2010) ‘Four Waves of Evaluation Diffusion’, Evaluation 16: 263–77.
Weiss, C. (1977) Using Social Science Research in Public Policy-making, Lexington,
MA: Lexington Books.
Wildavsky, A. (1978) ‘A Budget for All Seasons?: Why the Traditional Budget
Lasts’, Public Administration Review 38: 501–9.
Zeisel, H. (1981) ‘Social Science Hubris—A Review of Lindblom and Cohen’s
Usable Knowledge’, American Bar Foundation Research Journal, n.v. 273–81.
352
Index
Index
Note: Page numbers for tables appear in italics. ALEC (American Legislative Exchange
Council) 211–12
A Alliance for Innovation 133, 137–8, 141–2
Abel-Smith, Brian 347 survey 146
academia 48, 82, 83–4 Allison, Graham 331
academic journals 139 America COMPETES Act 177–8
academic research 146 American Competitiveness Initiative 177–8
accountability 259 American Enterprise Institute 11, 229, 285,
accreditation, NASPAA 320–2 288
ACF (Advocacy Coalition Framework) 341 American Pacific Corporation 233–4
Achen, Christopher H. 187–8 American Petroleum Institute 229
active learning 310 American Philosophical Society 74
Administrative Behavior: A Study of Decision- Americans for Prosperity 119
Making Processes in Administrative American Society for Public Administration
Organizations (Simon) 87 (ASPA) 323
Administrative Procedure Act 14 Center for Accountability and Performance
Adolescent Health’s Teen Pregnancy 117
Prevention Program, Office of 101 Analysis for Public Decisions (Quade) 81
adult corrections system 118 Annie E. Casey Foundation 251
Advanced Light Source (ALS) 176 anticipation, rational 190
Advisory Commission on Intergovernmental antipoverty programs 135
Relations (ACIR) 133 APPAM (Association for Public Policy and
Advisory Committee on Immunization Management) 85, 297–8, 312, 320, 328,
Practices 15–16 329, 336
advisory committees 12, 16, 173–9 doctoral programs 323, 332, 333, 334
advocacy 256–8 founded in 1979 24, 98, 322
advocacy coalition approach 186 Park City conference 299
advocacy tanks 11 appropriations bill 231–2
Aelst, Peter Van 270 APSA Committee on Political Parties 217
Aerojet 233–4 argumentative turn 55, 62–8
aerospace industry 233 Arizona State University 135, 142
Affordable Care Act (ACA) 113, 158, 167, Army Corps of Engineers 340
214–15, 217, 257 Arnold Foundation 133, 251
AFL-CIO 229 arsenic 226
African Americans 187, 283 The Art of Judgment: A Study of Policymaking
agencies 19–20, 86, 95, 154, 165, 173, 174–7 (Vickers) 87
see also CBO (Congressional Budget Office); ASPA Task Force on Educating for Excellence
GAO (Government Accountability Office) 328
agenda cutting 269 astronomy 177
Agendas, Alternatives, and Public Policies ASU’s Center for Urban Innovation 137
(Kingdon) 86 attitudinal data 187
agenda setting, media 266–7, 268, 270–1, Auditor General 124
273–4 Augustine, Norm 178
Agranoff, R. 46 Austin, TX 144
Aguilar, Luis 341–2 Austria 345
Aid to Families with Dependent Children Azoulay, P. 179
(AFDC) 96
353
B budgeting 85, 156, 158, 163, 165

Bachrach, Peter 184 public 344–5
BAD (Budget Analysis Division) 156, 158 Bundesrechnungshof 342
Bakken shale formation 230 Bureaucracy: What Government Agencies Do and
balance of powers 164–5 Why They Do It (Wilson) 87
Banfield, Edward 81–2 Bureau of Labor Statistics 96
Bang, S. S. 272 Bureau of the Budget 17, 22, 34, 75–6, 282
Bank of America 237 Bush, President George H. W. 234
banks and waivers 237–8 Business and Its Environment (Baron) 223
Baratz, Morton S. 184
Barber, Benjamin 147 C
Baron, David 223 Califano, Joseph 73
base rights 78 California 113, 234
Basic Energy Sciences Advisory Committee Campbell Collaboration 101
(BESAC) 176 Canada 273, 343
Battelle Memorial Institute 229 cap-and-trade system 113
Baumgartner et al. 230 capital punishment policy 269
Baumgartner, Frank 185, 267, 269–70, 274 capstone course 302
Bawn et al. 210, 217 carbon emissions 113
Beam, D. 31 career development 49
Belgian Parliament 271 Carlsbad, CA 144
Bellevue, WA 143, 144 Carnegie, Andrew 10
benefit-cost analysis 118, 120, 126, 223, 229, Carnegie Corporation 115, 247
231, 239 Carter Administration 13
Berkeley program 331 Carter, President Jimmy 22, 157, 268
Bernalillo County, NM 142–3 case-method teaching 310
Berry et al. 192 Case Western Reserve University 257
Bertelli, Anthony 82 Cato Institute 288, 290
big data 26, 137, 138, 145, 147 Caughey, Devin 193
Biological & Environmental Science, Office CBO (Congressional Budget Office) 76, 86,
of 177 97, 156–8, 164, 165–6, 167, 343
Bipartisan Campaign Reform Act (BCRA) budget scores 3
217 created in 1974 96
Birgeneau-Shen report 176 eight divisions 18–19
bisphenol A (BPA) exposure 270 and immigration policy 168
Blackett’s Circus 77 on insurance 215
Bloomberg Foundation 252 and Lienert 169
Bloomberg Philanthropies 133, 137 Census Bureau 96
Blueprints for Healthy Youth Development Census of Governments, 2012 132
126 Center for American Progress 285
‘Blue Ribbon’ advisory commissions 57 Center for Evidence-based Policy, Oregon
Bone, H. A. 208 Health and Science University 119–20
Boydstun, Amber 267, 268, 269, 270 Center for Performance Analytics 136
Bradley, Smith Richardson foundation 289 Center for Urban Innovation 135
‘Brain Trust’ 57 Center for Urban Policy Research 135
Bremberg, Andrew 238 CGEPA (Council on Graduate Education for
Brewer et al. 331 Public Administration) 320, 327
Brewer, G. D. 82 change, theory of 250, 251
British government 343–4 charities 245, 247, 256
Brookings Institution 11, 75, 282 Charity Organization Society (COS) 247
Metropolitan Policy Program 135 charter schools 113, 124, 218
Budget Act 156 chemical industry 231
Budget Analysis Division, CBO 18 Chicago 136, 270
Budget and Accounting Act 75, 96 child abuse 268
budgetary power 168–9 children’s programmes 270
budget deficits 50
354
Index
Children, Youth and Families Department and agencies 95

118 and CBO 157, 158
China 342 committees 153–5, 175
Chubb, J. E. 226 and CRS 163
Churchill, Winston 77 and GAO 160
citizen groups 226, 227, 228 and GPRA 76
citizen ideology score 192 and Maternal, Infant, and Early Childhood
citizen participation 143–3 Home Visiting Program 103
citizen preferences 192 and media coverage of legislation 269
citizens 227, 228 and OTA 175
Citizens’ Bureau in Milwaukee 10 and Twitter 272
Citizen’s Conference on State Legislatures Congressional Budget and Impoundment
115 Control Act 18, 76, 156
Citizens United 217, 246 Congressional Reference Service 342–3
city manager movement 133 Congressional Research Service (CRS)
Civic Federation in Chicago 10 17–19, 272
civil rights 187, 283 conservative ideology 283
Civil War 12 conservative think tanks 283–9, 290
Clark, William 74 Consolidated Appropriations Act 230
classroom response systems 310 ‘Constituency Influence in Congress’ (Miller
clearinghouses 100, 101, 102–3, 126 and Stokes) 186–7
Cleary, R. E. 334 consultancies 15, 86
Cleveland 257 consultants 134, 139, 140
Cleveland Foundation 257 Consumer Product Safety Commission
client definition 38 (CPSC) 14
Client Information System 115 contracting 15, 16
client-oriented policy analysis 85 contracting out 136
clients 41, 44, 45–6, 49 Contract with America Advancement Act 22
climate change 147, 218, 224, 238–9 Cook, Robert 95
Clinton, President 22, 234 COPRA (Commission on Peer Review and
CNN 275 Accreditation) 321
Coalition for Evidence-Based Policy 126 copying 348
Cobb, R. 185, 267, 274 Cordes et al. 332, 333
coercion 349 Corporation for National Community Service
Cohen, Bernard 267 104
Cohen, D. 61, 343 corporations 223, 224, 226–7, 228, 229–30,
collective bargaining rights 113 233–4, 237–8, 239
collective impact 252–3 cost-benefit analysis 21–3, 46–7, 60, 340–1
Collingridge, D. 60 cost estimating 156–7, 165–7
Colorado 118 council-manager governments 132, 133
Columbia, District of 118 Council of Economic Advisers 75
Commission on Economy and Efficiency 74 Council of State Governments 115, 122
committees, state legislatures 116 Cozzens, Susan 174–5
Common Evidence Guidelines for Education craft occupations 104
Research 103, 104 credence goods 236
community agencies 249 credentialing 85–6
Community Chest 248 Crimesolutions.gov 126
community programs 250 criminal justice policy 25
Community Renewal Program 134 critical policy analysis 65
community service clubs 247 critical theory 347
compensation policies 102 Croly, Herbert 56
complexity, need to manage 16–20 CRS (Congressional Research Service) 76,
Comptroller General 159, 160 86, 96, 161–4, 165, 167, 168
computation 26 C-Stat 118
Comte, August 56 CTAs (trade associations) 223–40
congregations 246 curation resources 126
Congress 39, 42, 164–9, 187, 342–3, 349
355
D E
data infrastructure investment 106 Eagleton Institute of Politics, Rutgers
Data Quality Act (DQA) 225, 230–3 University 115
death penalty policy 269 Eastern Evaluation Research Society (EERS)
De Boef, Suzanna 269 98
decentralization 50 Eastman, George 10
defense 163, 188 eco-labeling 236
Defense Department (DoD) 13, 33–4, 37, economic environment 46–7
73, 80, 233, 234, 282 The Economic Naturalist (Frank) 300
defense industry 233 Economic Opportunities, Office of 97
Delaware 132 Economic Opportunity Act 94
delegated responsibilities 15–16 economics 300, 311
deLeon, P. 60, 74 The Economics of Defense in the Nuclear Age
deliberation 66 (Hitch and McKean) 86
Deming, W. Edwards 312 Educating the Reflective Practitioner (Schön) 305
democratic deficits 195 education 101–2, 103, 193, 247, 297–315,
democratic elitism 184 319–36, 345
Democratic National Convention 134 Elementary and Secondary Education Act
democratic representation 190 94, 95
democratic values 51 nonprofit organizations 246
The Democratic Wish: Popular Participation and programs 100, 103
the Limits of American Government (Morone) Education Department (DOED) 93, 94–5
86 Edwards, Lee 283
the Depression 248–9 effectiveness 103, 104, 348
Desai, A. 301 efficiency 81, 104, 343, 348
desk officer 20 Ehrenhalt, A. 213
devolution 50 ‘eightfold path’ framework 305
Dewey, J. 57 Elder, Charles D. 185
DHHS (Departments of Health and Human Elder, T. 267, 274
Services) 93, 94–5, 100, 104 electioneering 246
Dillon’s Rule 132 electoral replacement 190
disaster response 186 Elementary and Secondary Education Act
doctoral policy education 313–14 94, 95
doctoral programs 323, 329–34, 336 elites 184, 186, 188, 191, 272–3, 274–5
DOED (Education Department) 93, 94–5 Ellwood, J. W. 298, 326, 328
DoE (Department of Energy) 100, 104, emergency relief funding 248
175–6, 177, 233, 237 Emerson, Jed 253
DOL (Labor) 38, 93, 94–5, 99, 100, 104 Emerson, Jo Ann (R-MO) 232
Dominguez, C. B. K. 216 employment
Downs, Anthony 185 evidence-based studies 103
DPAs (doctorates of public administration) government 132
331, 332, 333 public policy graduates 328
DQA 232–3, 234, 239–40 and training programs 100
dropout prevention programs 100 EMS (environmental management systems)
Dror, Yehezkel 33, 35 223, 224, 225–6, 235–6
drug courts 20 Energy Department (DoE) 100, 104, 175–6,
Drug Effectiveness Review Project 120 177, 233, 237
drugs 267–8 energy industry 230, 233
Drutman, L. 155, 212 energy policy 157
Duke Energy 134, 237 Engaging Local Government Leaders 133
Dunn, William 341 Engelbert, Ernest A. 23
Dye, Thomas 194 Enthoven, Alain 77, 78
dynamic scoring 166 entrepreneurs, policy 209
environment 273
environmental impact statements 21, 229
environmental protection 186
356
Index
EPA (Environmental Protection Agency) 14, Fischer, Frank 62, 64–5, 346
38, 223, 228, 231, 237 Fisher, Louis 168
and DQA 232 Fleisher, L. 173
and Keystone XL pipeline 230 Florida 114, 123–4
and perchlorate RfD 225, 234 Florida OPPAGA (Office of Program Policy
and SDWA 233 Analysis and Government Accountability)
EPN (extended party network) 206, 207–8, 116, 123–4
209–18 Florida State University 119
equilibrium theory, punctuated 185–6, 267 Fontaine, Guillaume 341–2
Erikson, Robert 190, 192, 194 Food and Drug Administration (FDA) 231,
ethics 304 232
EU (European Union) 235, 236 Food Stamp Act 94, 96
Europe 235, 339–40, 344, 346, 347 Ford Foundation 13, 83, 115, 249, 298, 331
Evaluating Public Policy (Fischer) 65 grants 23
evaluation 41–2, 44, 104, 106, 343–4 Ford, President 22
Every Child Succeeds Act (ESSA) 101 foreign policy 163, 187, 188
Evidence, Argument, and Persuasion in the Policy Forester, John 62
Process (Majone) 86 Fortune 1000 companies 225
evidence-based decision making 60, 100–4, foundations 247–8, 249, 251
106–7, 131, 133–47, 251, 257 fracking 113
evidence-based policy making 25, 101, Framework for Design Based Implementation
133–5, 343–4 Research 103, 104
evidence-based programs 104, 126 framing theory 269–70
evidence clearinghouses 101, 102 France 344, 347
evidence-guided funding streams 105 Frank, Robert 300
Excellence in Evaluation Award, 2015 118 French Union for a Popular Movement 209
executive actions 22 Friedmann, J. 86
executive agencies 86 Fritschler et al. 327
Executive Branch 95, 96, 104 Fritzen, S. A. 336
executive branch agencies 118, 121 fuels 143
Executive Order 12088 231 funding 104, 105
experimental design evaluations 100
experimental evidence 99–100 G
expertise, need for 12–16 Gallup, George 183
ex post evaluations 134 GAO (General Accounting/Accountability
Eyestone, Robert 185 Office) 18–19, 86, 96, 159–61, 164–5,
166, 342
F creation of 75, 76
FACA (Federal Advisory Committee Act) focus on whether programs are working
173–4, 175 167
Facebook 271 and staff training 17
Fachhochscule 345 during World War II 17
faculty 307–8 Gardner, H. 306
Family Support Information System 115 Genomes to Life (GTL) User Facilities 177
Federal Advisory Committee Act 12 George Washington University State- Local
Federal Communications Commission (FCC) Finances Project 13
14 Germany 178, 345
federal emergency relief funding 248 Geva-May, I. 320, 329
federal government, assumed leading role 114 Gill, N. 10
The Federalist 84 globalization 50
Federal Register 14, 22, 234 Global Parliament of Mayors 147
Federal Trade Commission (FTC) 14 Goldman Sachs 237, 252
Felbinger et al. 334 Goldman School 303, 308
fire department 270 Governance movement 326–7
First in the World 105 governing networks 209
first-level agenda 274 government
357
local 114, 131–47 Hird, J. 19, 116, 121–2

structure of 44 Hitch, C. J. 86
Government Accountability Office 115 Hogwood, Brian 43
Government Finance Officers Association Holland 58–9
136 Holzer, M. 331
Government Performance and Results Act home rule 132
(GPRA) 76 Hoover, President Herbert 75
Government Research Association 119 Hoppe, Robert 59
GPS 143 House of Representatives 190, 272
graduate programs 35 Housing and Urban Development
see also doctoral programs; MPAff (Master Department 13
of Public Affairs); MPA (Master of Public How Great Expectations in Washington Are
Administration); MPP (Master of Public Dashed in Oakland; Or, Why It’s Amazing
Policy) that Federal Programs Work at All ((Pressman
graduation rate 118 and Wildavsky) 298
graduation ratios 193 Human Genome project 177
Graham, John 232 human service agencies 246
grant programs, federal 250 hydraulic fracturing drilling 113
Grasso, P. G. 167
Gray, Virginia 195 I
Greater University Circle Initiative (GUCI) Ickes, Harold 75
257 ICMA (International City/County
‘Great Principle’ 207 Management Association) 133, 134, 136,
Great Society 73, 343 137, 142–3, 144, 323
greenhouse gas reductions 235 Innovation Survey 139–41, 145, 146
GRE test 332 ideology 50–1, 192, 209
Grey Areas program 249 IHS Cambridge Energy Research Associates
Grossmann, M. 209, 216 (IHS CERA) 229
GSA (General Services Administration) 173, Illinois 132
174 impact analysis 100
Gueron, J. M. 100 impact, collective 252–3
Gulick, Luther 324 implementation 42, 44
Gunn, Lewis 43 Implementation (Pressman and Wildavsky) 86
income tax
H cuts 113
Habermas, J. 65 experiments, negative 95–6
Hahn, Robert 239 Independent Budget Office, City of New
Hardy, A. 268 York 19
Harriman, W. Averell 78 inflation, rules on 22
Harvard Business School 302, 310 Ingram, H. 65
Harvard-Smithsonian Center for Astrophysics innovation 116, 345
310 Innovations in Government Awards
Harvard University 331 competition 299
Havens, H. 159 Innovation Survey 137, 139–41, 142, 145–6
Hawaii 132 Institute for Urban Research 135
Head Start 94, 95, 97 Institute of Education Sciences 101, 103, 104
Health and Human Services Department 38 intellectual ethos 86
health care 93–4, 119–20, 157, 167, 186, intellectual integrity 85
214–15, 344 interest groups 40, 195, 210–11, 216, 233
spending 273 Interior Department 75
Health Care for America Now 215 International Comparative Policy Analysis
health insurance experiments 95–6 (ICPA) Forum 323
health organizations 246 the Internet 26, 45
Heaney et al. 216 Investing in Innovation 105
Henry, L. 322, 324 IRS (Internal Revenue Service) 246, 256
Heritage Foundation 215, 283, 288, 290 ISO 14000 standards 235, 236, 240
358
Index
ISO (International Organization for Korean National Assembly 272

Standardization) 226, 235 Kramer, M. 252
isomorphism 348 Kyoto Protocol 235
Israeli Parliament 271
issue framing 191 L
issue priming 191 Labor Department 38, 93, 94–5, 99, 100,
item response theory models 193 104
land grant colleges 345
J Langan et al. 136
Jacobssen, B. 344 Lasswell, Harold 57, 75, 81, 84, 184–5, 341
Jacoby, William G. 193–4, 195 Las Vegas, NV 143, 145
Jefferson, Thomas 74 Latin America 341–2
Jenkins-Smith, Hank 186, 332, 341 law course 303
Jensen, E. R. 304 Law, J. 62
Job Corps 95 Lax, Jeffrey R. 195
Job Training Partnership Act 99, 100 leadership of think tanks 286
John D. and Catherine T. MacArthur learning
Foundation 126 active 310
Johnson, President Lyndon B. 13, 34, 73, 93, organizational 255, 259
134–5, 343 lecturing 309–10
Joint Committee on Taxation 166 Lee, Francis E. 213
Jones, B. 185, 267, 274 Legislative Analyst’s Office 19
Journal of Policy Analysis and Management Legislative Committee on Intergovernmental
(JPAM) 24, 86, 297, 299–300, 312, 323, Relations 124
332 legislative policy analysis offices 122
Journal of Public Affairs Education (JPAE) 312, Legislative Reference Service (LRS) 162
322 Legislative Reorganization Act 17, 159–60,
Journal of the American Medical Association 268 162
Joyce, P. 157, 158 legislative research offices 121–2
JPMorganChase 237 Legislative Research Service, Library of
juvenile justice clients 118 Congress 17
legislative service 123
K legislatures, state 111–15, 116–18
Kania, J. 252 Le Grand, Julian 347
Kansas 113, 144 Leibfried, Stephan 347
Kaufman, H. 18 Lepawsky, Albert 329
Keefe, W. J. 213 Lerner, Daniel 75, 81
Kellogg Foundation 251 LeRoy Collins Institute 119
Kennan, George F. 75 Lester, J. 122
Kennedy School 298, 302, 303 Lewis, Merriweather 74
Kentucky Long-Term Policy Research Center Liberal Reform Strategy 57
124 liberal think tanks 284, 285–7, 288–91
Kerr-McGee Chemical 233–4 Lienert, I. 168–9
Keynes, John Maynard 57 Lim, E. 273
Keystone XL pipeline 225, 229–30 Lindblom, Charles 61, 79, 85, 343
Key, V. O 183 Lions Club 247
Kingdon, John 86, 185, 211, 267, 341 Lippmann, Walter 183, 274
Kiss, Simon J. 270 lobbying 211, 240, 246
Klein, Rudolf 347 lobbyists 155, 225, 226
Klingemann et al. 208 local governments 113–26, 131–47
Knowledge 59 Lockheed Martin 233–4
knowledge curators 125 logic models 143, 250, 251, 257
knowledge, ordinary 61 London School of Economics and Political
Koch, David 119 Science 347
Koger et al. 210 Los Angeles 134
Kolodny, R. 212–13 Louisiana Purchase 74
359
Low Countries 347 Meltsner, Arnold 35, 41, 47

Lowery, David 195 Memorandum on Transparency and Open
Lowi, Theodore 184, 207 Government 136
Luke 14: 28–31 73 Mendez, Jose Luis 341–2
Merriam, Charles 340
M ‘messy problems’ 59
MacArthur Foundation 19 Michigan 123
Mackinaw Institute for Public Policy 290 the military 13, 80
MacKuen, Michael B. 190 Miller, Warren E. 186–7, 188
macropolity 190 Mills, C. Wright 183
Madisonian politics 76 mimesis 348
Madison, James 84 minimum wage laws 218
Majone, G. 64, 86 Minnesota 113
managed competition 136 Moffitt, S. L. 173
management courses 301–2, 305–6 Moncrief, G. 114
Manhattan Institute 290 monitoring 104
the market 343 Monroe, Alan D. 188
Marmor, Theodore 347 mood, public 190, 192
Marxist approaches 347 ‘Moon/Ghetto’ metaphor 58
Maryland 118 The Moral Sense 304
National Parks and Planning Commission Morone, James A. 86
143 Morrill Act 345
Maslove, A. 320, 329 Mosher, F. 160
Massachusetts 113, 234 MPAff (Master of Public Affairs) 324–5, 327,
matching 102 329, 330
materials science 176 MPA (Master of Public Administration) 297,
Maternal, Infant, and Early Childhood Home 298, 323, 327, 328, 329, 330, 336
Visiting Program 103, 105 MPP (Master of Public Policy) 297, 298,
Mathematica 246 299–308, 323, 325, 327, 328–9, 330, 335,
Maxwell School of Citizenship and Public 336
Affairs 324 MRP (multilevel regression and
Mayhew, D. R. 208 poststratification) 192–3
Mayne et al. 118 multiple streams theory 267, 341
mayor-council governments 133 multi-state organizations 115–16
mayors 135, 147 Mulvaney, Mick 158
McCombs, Max 266–7, 269, 273 municipal research bureaus 9–10, 74, 133,
McCubbins, M. 155 247, 324
McGann, James G. 11 Mydal, Alva 347
McGuire, M. 46 My Teaching Partner 102
McIver, John 192, 194
McKean, R. N. 86 N
McNamara, Robert 13, 58, 73, 75, 76, 77, nanotechnology 225
78 NASA (National Aeronautics and Space
MDRC 252 Administration) 176–7, 233, 234
Mead, T. 270 NASPAA (Network of Schools of Public
media 265–76 Policy, Affairs, and Administration)
media storms 268 297–8, 312, 320–3, 327, 328, 329, 336
Medicaid 94, 96 doctoral programs 331, 332, 333, 334
Medicaid Evidence-based Decisions Project founded in the 1970s 98
119–20 and SDGs 335
Medicaid Management Information System National Abortion Rights Action League
115 (NARAL) 247
Medicare 94 National Academy Committee 97–8
Medvetz, T. 12 National Academy of Public Administration
Meier et al. 193 (NAPA) 174, 323
360
Index
National Academy of Sciences (NAS) 12, Nixon, President Richard 22, 268
174, 234 nonprofit advocacy sector 286, 287
decadal survey in astronomy and astrophysics nonprofit sector 20, 98, 245–59
176–7 normative isomorphism 348
National Academy Panel 97 North Dakota 132, 230
National Association of Chief Information Northern Europe 235
Officers 115 NRC (National Research Council) 12–13,
National Association of Counties 135 174, 334, 336
National Association of Drug Court NSF (National Science Foundation) 103,
Professionals 20 104, 178, 179
National Association of Manufacturers 229 Nurse–Family Partnership (NPF) 103, 257
National Association of Schools of Public Nuytemans, M. 271
Affairs and Administration 24
National Association of State Budget Officers O
118 Obamacare 113, 217
National Association of State Fiscal Officers Obama, President Barack 22–3, 158, 230,
115 232
National Bureau of Economic Research 11 Obergefell v Hodges 164
National Civic League 133 objectivity 168
National Committee on Child Labor 247 object salience 273–4
National Conference of State Legislatures Occupational Safety and Health
115, 122 Administration (OSHA) 14
National Environmental Policy Act 21 Office of Basic Energy Sciences 176
National Governor’s Association 115 Office of Policy Analysis 20
National Health Insurance Experiments 95–6 Office of Science and Technology Policy
National League of Cities 133, 135, 136 (OSTP) 175
National Legislative Program Evaluation Office of Science, Department of Energy
Society 116, 117, 118 175–6
National Resources Planning Board (NRPB) Office of the Solicitor 20
75 Ogallala Reservoir 229, 230
National Science Board 174 O’Hare, M. 310
natural resources 163 Ohio Legislative Office of Education
negative income tax experiments 95–6 Oversight 124
Neisser, P. T. 65 OIRA (Office of Information and Regulatory
Nelson, R. 36, 80, 81–2 Affairs) 226–7, 231, 232, 234, 235, 239
neoliberalism 343 OMB (Office of Management and Budget)
Netherlands 58–9 22, 76, 96–7, 165, 177, 231
network agenda setting 274 and DQA 232
network governance 347 Office of Evidence-based Initiatives 103,
networks 46, 49, 51, 209 104
neutrality 167–8 Omnibus Drug Bill 268
neutron source 176 open data 136–7, 147
New Deal 14, 57, 75 operations research 13, 298, 303
New Jersey Income Maintenance Experiment OPPAGA (Office of Program Policy Analysis
14 and Government Accountability), Florida
New Mexico LFC (Legislative Finance 117, 123–4
Committee) 117–18, 125, 126 Orange County (CA) Sheriff’s Department
New Public Management 299 145
newspapers 271 ordinary knowledge 61
Newsweek 267 Oregon Health and Science University
New Teacher Induction Model 102 119–20
New York 134, 252 Oregon Progress Board 124
New York Bureau of Municipal Research 10, organizational learning 255, 259
133, 324 Organ Procurement and Transplantation
New York Times 267, 268, 275 Network (OPTN) 16
NIH (National Institutes of Health) 174, 178, ORR (Office of Regulatory Review) 239
179, 225
361
OTA (Office of Technology Assessment) 18, polarization 216, 217

165, 175 police department 270
over-critical model 60–1 policy agenda 185
oversight staff 176–7 Policy Analysis for the Real World (Hogwood
and Gunn) 43
P Policy Analysis, Office of 20
Page, Benjamin L. 188–9 policy analysis profession 31–51
Palier, Bruno 347 policy analysis programs 23–4
Paperwork Reduction Act 23, 231 policy analysis schools 24
Paris Climate Accord 147, 218 ‘Policy Analysts: A New Professional Role in
Park City conference 299–300, 301, 302 Government’ (Dror) 35
Parliamentary Budget Office (Canada) 343 Policy Analysts in the Bureaucracy (Meltsner) 35
participation 66–7 policy argumentation 64–6
participatory policy analysis 26 policy bargaining 38
partisanship 206, 209, 210 policy design 305
party committees 208–9 policy epistemics 67–8
Patient Protection and Affordable Care Act policy evaluations 100
16, 211 policy liberalism 193
see also Affordable Care Act (ACA) policy research firms 83–4
patronage politics 207 policy research institutions 83
pedagogy 308–10 Policy Sciences 81, 329
peer review panels 178–9 The Policy Sciences (Lerner and Lasswell) 75,
Pennsylvania, University of 135 81
pension reforms 134 policy-specific measures 193
perchlorate 225, 233–5 policy stream 185
Perchlorate Study Group (PSG) 233–4 Policy Survey 137–9, 140, 146
performance-based contracts 252 political conflict 50–1
performance data 140–1, 142, 145, 147 political culture 191
performance management 136, 251, 252 political entrepreneurs 209
performance measurement 137, 343 political parties 205–18
pesticide industry 230 political science programs 325
PET (Punctuated Equilibrium Theory) 341 politicians 21–3
Petracca, M. 12 politics
Pew Charitable Trusts 19, 126 course 300–1
Pew-MacArthur Results First Initiative 121, and the media 266–71
126 The Politics and Economics of Public Spending
PFS (pay-for-success) 252 (Schultze) 79, 87
pharmaceutical industry 230, 231 ‘Pork Chop Gang’ 114
pharmacy programs 120 Portland 136
PhD in public policy 313–14 positivism 33
see also doctoral programs positivist knowledge 56
PhET project 311 poverty 58, 73, 93, 95, 269–70
philanthropic organizations 99, 103 powers, separation of 76
philanthropy, strategic 254 PPBS (Program Planning and Budgeting
Phillips, Justin H. 195 Systems) 13–14, 34, 36, 344
philosophy of science 67 Defense Department 37
Phoenix 134, 144 instituted by McNamara 76
Physical Science Studies Committee 311 and local government 134
Pierre, J. 344 and President Johnson 73
Pierson, Paul 347 and RAND Corporation 75
pipeline, Keystone XL 225, 229–30 Pragmatism 57
Planning in the Public Domain: From Knowledge predictive analytics 137, 145
to Action (Friedmann) 86 predictive modeling 147
pluralism 184 Presidential Advisory Commission on
Podesta, John 285 Election Integrity 218
PODSCORB skills 324 President’s Council of Advisors on Science
and Technology (PCAST) 174
362
Index
Pressman, J. 86 Reagan, President Ronald 22, 100, 230–1,

Primer for Policy Analysis (Stokey and 234, 268–9
Zeckhauser0 303 recidivism 25, 252
Prince George’s County Parks and Recreation REDF 253, 254
Department 143 Reed, Kasim 147
‘Principles for a Graduate School of Public Reeve, C. 60
Policy’ (Wildavsky) 42 reference dose (RfD) 233
private foundations 246 reflexive deliberation 64
A Private Universe 310–11 reform, desire for 9–12
problem stream 185 reformers 247
professional associations 98, 135, 138, 139, regulatory agencies 173, 226–7, 231, 239
140, 320–3 Regulatory Flexibility Act 23
professions 104 regulatory waivers 237–8, 240
program evaluations 99 Reiter, H. L. 208
Progressive Era 207 renewable energy 145
Progressive movement 10, 57 Reorganization Act 164
Progressive Policy Institute 211 republican governance 76
progressive reformers 10, 56, 247 Republican Party 349
‘Project for France’ 209 research and think tanks 289, 290
Project RAND 75 research bureaus, municipal 74, 133
Public Administration movement 324 research centers, university 125, 135
public administration programs 23–4, 326 research clearinghouses 126
Public Administration Review 35 Research Committee on Social Trends 75
Public Affairs movement 324 research generators 125
the public and the media 272–3 research intermediaries 125–6
public benefit organizations 246 research offices, central 116
public budgeting 344–5 research organizations
public charities 245, 256 for-profit and nonprofit 98, 246
public education 247 as liberal think tanks 288
Public Management movement 326 research units, executive branch agencies 118
public opinion 183–96 Resources for the Future 11, 211
Public Opinion (Lippmann) 274 Results First Initiative 19, 126
Public Policy movement 325 Results Washington 118
public–private partnerships 252 retrospective analytic techniques 41
public service programs 324 returns on investment 103, 104
public universities 342, 345 Rhode Island 239
see also universities RIAs (regulatory impact analyses) 22
punctuated equilibrium theory 185–6, 267 Riker’s Island Correctional Facility 252
Riksrevisionsverket 342
Q ‘Rising Above the Gathering Storm’ 177–8
Quade, Edward 81 risk assessments 229, 230
Quality-Adjusted Life Years (QALYs) 167 risk-benefit analysis 341
quality assurance 312 Rittel, H. W. 59
Quality of Life reviews of rules 22 Rivlin, Alice 37, 78, 82, 83, 86, 157
Robinson, W. 39, 162
Rochester Bureau of Municipal Research 10
R Rockefeller Foundation 247
Radin, B. 81, 82 Rockefeller, John D. 10
RAND Corporation 13, 75, 77, 96, 134, 282 roll call votes 187
‘the Rand experience’ 325 Rolston, H. 100
randomization 102 Rom, Mark Carl 193
randomized controlled trials 99, 179 Romzek, B. S. 213
Rao, Naomi 232 Roosevelt, President Franklin 268
rational anticipation 190 Roosevelt, President Theodore 56, 74
RCB (rationalisation des choix budgetaires) 344 Rorty, R. 67
Reagan administration 249–50, 343 Rose, M. 269–70
Rosenwald 247
363
Rossi, J. 237 Shape Up Somerville 252

the Rotary 247 Shapiro, Robert Y. 188–9
Roundtable on Community Change, Aspen Sharkansky, I. 167
Institute 250 Shaw, Donald 266–7, 273
Rubin, I. 166 Shelby, Senator Richard (R-AL) 232
Ruder, David 238 SIBs (social impact bonds) 252
rulemaking 22–3, 25, 227–8 Sierra Club 247
rule of law 85 Simon, H. A. 87
rules, agency 22–3 Sinclair, Barbara 214
Russell, A. 272 skills 47–8, 49
Russell Sage Foundation 247, 248 Skogstad, G. 212
Russia 342 Smart Cities Council 135
Rutgers University 135 Smith, B. L. R. 12
Smith, J. A. 290
S Smith, K. W. 77
Sabatier, Paul 186, 341 Social and Behavioral Research Institute,
Safe Drinking Water Act (SDWA) 233 California State University-San Marcos
salience 273–4 144
San Antonio Light 270 social construction theory 186
San Francisco 136 social constructivism 67
Santa Monica, CA 143 social experiments 14
Sarah Scaife foundation 289 Social Innovation Fund 105
Scandinavian countries 345, 347 social media 26, 271–2
Schattschneider, E. E. 184, 267, 271, 274 social policy research 162–3, 347
Schlesinger, James 59 social reform associations 247
Schneider, A. 65 social-science-based public policy research 85
Schneider, Sandra K. 193–4, 195 Social Security Act 94
Schön, D. A. 305 social security benefits 94
school districts, independent 132 social services 97, 251–2
schools Social Services Information system 115
charter 113, 124, 218 social welfare 103, 187
of public policy 24 organizations 245
Schram, S. 65 Society 58
Schultze, Charles 59, 73, 79, 82, 83, 85, 87 Society for Benefit-Cost Analysis 24
Schwartz, T. 155 Society for Risk Analysis 24
science socioeconomic factors 194
expertise 12–13 solar projects 145
policy 3, 4, 163, 174, 175 Solicitor, Office of the 20
teaching 311 SOPs (standard operating procedures) 237,
science agencies 174–7 240
Science and Technology Policy, Office of Soroka, Stuart 189–90, 269, 271, 273
(OSTP) 175 special districts 132
Science, Office of, Department of Energy specialized offices 116–17
175–6 special purpose governments 132
‘The Science of Muddling Through’ Squire, P. 114
(Lindblom) 79 SROI (Social Return on Investment) 253–4
Science Speaks to Power (Collingridge and Staats, General Elmer 75, 160
Reeve) 60 staff
scientific management 56, 133 in the CRS 161–2
Seattle, WA 142 in the GAO 159, 160
Securities and Exchange Commission (SEC) in think tanks 286
14, 237–8 stage heuristic model 184–5
Seitz-Eastman plan 176 stagflation 283
self-worth 49 State Department 230
Sentencing Commissions 118 state governments 113–26, 131–47
shadow prices 78 Statehouse Democracy (Erikson, Wright, and
McIver) 194
364
Index
state legislators 121–2 Thatcher government 343

State of the Union address 268–9 Theory C/Theory T 305, 310
state partisanship 192 theory of change 250, 251
states, as laboratories of democracy 191–5 thermostatic model 189–90
StateStat 118 think tanks 9–12, 39, 86, 121, 125, 139, 216,
Statistical Analysis Centers 118 281–92, 342, 346
statistics course 301, 311 and lobbying 210–11
steady-state neutron source 176 national 134–5
Stimson, James A. 190, 192 as nonprofit organizations 246
Stokes, Donald 186–7, 188, 324, 326, 336 in the Progressive Era 207
Stone, Deborah 346 universities 119–20
STPI (Science & Technology Policy Institute) Thrive Washington 257
175 Titmuss, Richard 347
strategic philanthropy 254 tobacco industry 230
strategic planning 254–6 Toulmin, S. 65
Straus et al. 271 toxic chemicals 233
Straussman, J. D. 300–1, 302 Tozzi, Dr. James 230–1, 232
streams 185 trade 163
street-sweeping operations 143 Trade Adjustment Assistance Community
student surveys 312–13 College and Career Training 105
subsystems, policy 186 trade associations (CTAs) 223–40
Sundeström, G. 344 training 100, 103, 104, 106
Sunshine Foundation 175 TransCanada Corporation 225, 229, 230
Sunstein, Cass 232 transparency 51
Super PACs 217 Tribe, Lawrence 80
support agencies 154, 165 Trivelpiece, Al 176
see also CBO (Congressional Budget Office); Troutdale, OR 144
CRS (Congressional Research Service); Trump administration 106, 158, 259, 334–5
GAO (General Accounting/Accountability Trump, President Donald 215, 218, 230, 232,
Office) 238–9, 275–6
Supreme Court 114–15, 190 trust, lack of 106
Survey Commission in Atlantic City 10 Twitter 271, 272, 275, 276
Survey of Earned Doctorates 332
surveys 10, 143–4, 192, 312–13 U
Sustainable Development Goals (SDGs) 335 UK Case Centre 302
Sweden 344, 347 under-critical model 60–1
Switzerland 345 undergraduate policy programs 313, 314
Syracuse University 324 Unemployment Insurance 96
Systematic Thinking for Social Action (Rivlin) Unfunded Mandates Reform Act 23, 157
86 Uniform Crime Reporting system 115
Systems Analysis, Office of 77 United Nations 335
United Way 245, 248, 250, 252, 255
T universities 96, 98–9, 119–20
Taft, President William Howard 74 courses 302, 330
Tatel, David 85 former land grant colleges 345
Taylorism 56 private nonprofit 246
Teacher Advancement Program 102 public 342
Teach for America (TFA) 102 and training of public administrators 83
teaching 311, 312–13 working with cities 135
training 101–2 working with local government 134,
technocracy 56 138–9, 146
technological change 25–6 see also individual universities
Technology Review Workgroup 124 university research centers/institutes 125, 134
teenage mothers welfare policies 100 Urban Institute 11, 134–5, 136, 246
term limits 123 U.S. military 13
Texas Austin, University of 326
365
U.S. News and World Report (USNWR) Wieman, Carl 311

rankings 329 Wildavsky, Aaron 34, 42, 78–9, 86, 298, 331,
U.S. Public Interest Research Group 119 344
utilization 59 Willard, C. A. 67
Utter, J. A. 213 Wilson, James Q. 87, 304
Wilson, Woodrow 56, 207, 324
V Wisconsin 113
Vaccines for Children Program 15–16 Wisconsin Legislative Reference Bureau 19
Valley Benchmark Cities project 144 Wissenschaftsrat 178
value maximization 81 Wlezien, Christopher 189–90
Van Aelst, P. 271 Wolfe, Michelle 267, 269, 274
VanLandingham, G. 122 women in faculty 308
Veblen, Thorsten 56 Workforce Innovation Fund 105
Vedung, Evert 343 World War II 13
vehicle fleet 143 Wright, Gerald 192, 194
veil of ignorance 21 Wyatt, Edward 237
Vera Institute 252 Wyoming 113, 132
Vickers, S. G. 87
Vietnam War 58, 73, 283, 298–9 X
Vining, A. 9 x-ray light sources 176
vote fraud 218 Xu, H. 331
W Y
waivers, regulatory 237–8, 240 Yates, D. 325, 331
Walgrave, Stefaan 268, 270, 271
Wang et al. 146 Z
Wang, T. 331 Zero-Based Budgeting (ZBB) 13, 344
War Department 75 Zigler, Ed 95
Warner, Amos 247
war of ideas 285–7, 290
War on Poverty 58, 73, 93, 95, 97, 98, 249
Warshaw, Christopher 193
Washington 118, 342
Washington Post 268
Washington, President 12
Washington State Institute for Public Policy
(WSIPP) 19, 120
Washington, University of 302
water resource projects 21
Weaver, K. 11, 283
Webber, M. D. 59
Webb, P. 212–13
Weible, Chris 341
Weimer, David 9, 74, 85
Weiss, Carol 39, 41, 45, 61, 250, 343
welfare 93–4, 100, 103, 187, 193
reform 186
welfare state 75, 347
Welfare-to-Work waiver experiments 100
Wells Fargo 237
Westmoreland, T. 167
‘What Works Cities’ 133, 137
What Works Clearinghouse 101, 102–3, 103
Whiskey Rebellion Commission 12
Whiteman, D. 19
‘wicked problems’ 43, 59
366
Vol 12
“This collection offers a unique and valuable set of perspectives on the history of
policy analysis, changes in the field over recent decades, and current challenges facing
practitioners.” Michael E. Kraft, University of Wisconsin-Green Bay INTERNATIONAL LIBRARY OF POLICY ANALYSIS
SERIES EDITORS:

OF POLICY ANALYSIS
INTERNATIONAL LIBRARY
Policy analysis in the United States brings together contributions from some of the world’s
leading scholars and practitioners of public policy analysis including Beryl Radin, David Weimer,
Rebecca Maynard, Laurence Lynn, and Guy Peters.
This volume represents an indispensable companion to other volumes in the International
Library of Policy Analysis series, enabling scholars to compare cross-nationally concepts and
practices of public policy analysis in the media, sub-national governments, and many more
institutional settings.
The volume represents an invaluable contribution to public policy analysis and can be used
widely in teaching at both graduate and undergraduate levels in schools of public affairs and
public policy as well as in comparative politics and policy.
Features of the ILPA series

• a systematic study of policy analysis systems by government and non-governmental actors
• a history of the country’s policy analysis, empirical case studies and a comparative overview
POLICY ANALYSIS IN
Edited by John A. Hird

POLICY ANALYSIS IN THE UNITED STATES
• a key reference for research and teaching in comparative policy analysis and policy studies
Recent volumes published and forthcoming
• Policy analysis in Mexico, edited by Jose-Luis Mendez and Mauricio I. Dussauge-Laguna (2017)
• Policy analysis in Belgium, edited by Marleen Brans and David Aubin (2017)
the United States

• Policy analysis in France, edited by Charlotte Halpern, Patrick Hassenteufel and Philippe Zittoun
(2018)
• Policy analysis in Canada, edited by Laurent Dobuzinskis and Michael Howlett (2018)
JOHN A. HIRD is Dean of the College of Social & Behavioral Sciences, University of Massachusetts Amherst,
where he is Professor of Political Science and Public Policy. The founding director of the Center for Public
Policy and Administration, his research focuses on the use of expertise in public policymaking.
www.policypress.co.uk
@policypress PolicyPress
In association with
International Comparative Policy Analysis Forum Edited by John A. Hird

and Journal of Comparative Policy Analysis ISSN 2059-0326
Policy analysis in the United States [HB] [PRINT].indd 1 25/01/2018 16:23

John A. Hird (Editor) - Policy Analysis in The United States-Policy Press (2018)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

John A. Hird (Editor) - Policy Analysis in The United States-Policy Press (2018)

Uploaded by

Copyright:

Available Formats

INTERNATIONAL LIBRARY OF POLICY ANALYSIS

Edited by John Hird

Each volume includes a history of the country’s policy analysis which

Early volumes in the series will cover the following countries:

In association with the ICPA-Forum and Journal of Comparative Policy Analysis.

International Library of Policy Analysis, Vol 12

List of tables and figures vii

Part One: History, styles, and methods of policy analysis in 7

one Policy analysis in the United States 9

Part Two: Policy analysis by governments 91

Part Three: Policy analysis outside of government 171

nine Policy advisory committees: an operational view 173

twelve Policy analysis by corporations and trade associations 223

Part Four: Policy analysis education and impact internationally 295

sixteen Public policy education in the United States 297

Scott H. Ainsworth is Professor and Head of Political Science in the School of

Zachary Albert is a PhD student in political science at the University of

Frank Fischer is Professor Emeritus of Politics and Global Affairs at Rutgers

Erik Godwin is a lecturer at Brown University’s Taubman Center for Public

Kenneth Godwin is Professor Emeritus of Political Science and Public

William G. Jacoby is Professor in the Department of Political Science at Michigan

Raymond J. La Raja is Professor of Political Science at the University of

Rebecca A. Maynard is University Trustee Chair Professor of Education and

Maxwell McCombs is the Jesse H. Jones Centennial Chair in Communication

Karen Mossberger is Professor of the School of Public Affairs at Arizona State

Michael O’Hare is Professor of Public Policy in the Goldman School of

Nicholet Deschine Parkhurst is a PhD student in public administration and

B. Guy Peters is the Maurice Falk Professor of American Government at

Beryl A. Radin is a member of the faculty at the Georgetown Public Policy

Andrew Rich is the Executive Secretary of the Harry S. Truman Scholarship

Nadia Rubaii is Associate Professor of Public Administration at Binghamton

Annelise Russell is a PhD student in Government at the University of Texas

Saundra K. Schneider is Professor in the Department of Political Science at

Sector: International Perspectives, and Nonprofits and Advocacy: Engaging Community

Kuang-Ting Tai is a Graduate Research Assistant at the School of Public Affairs

Gary VanLandingham is Professor and Reubin O’D. Askew Senior Practitioner

David L. Weimer is the Edwin E. Witte Professor of Political Economy at the

the work done by policy analysts working in different institutional settings?

Lupia, A. and McCubbins, M. D. (1994) ‘‘Who Controls? Information and

Policy analysis in the United States

A narrow definition of policy analysis as a professional activity emphasizes client

Demand arising from the desire for reform (advocacy)

increasing ideological differentiation sometimes brings competing reforms into

established in Milwaukee by its socialist administration in 1910 (Gill 1944). The

Advocacy tanks combine a strong policy, partisan or ideological

Advocacy tends to be along identifiable ideological lines (McGann 2005,

Demand arising from the need for expertise

Medicine, undertakes projects requested by Congress as well as by administrative

Demand arising from the need to manage complexity

General analytical support for executives and legislatures

requests; hosted over 7,400 congressional participants at seminars, briefings and

Analytical support within agencies

Politicians wishing to tie their own (or their colleagues’) hands

The imposition of procedural rules on decision processes by politicians has been

regulatory commissions to conduct the analyses mandated for other agencies

Supply response to policy analysis demands

training within public administration schools resulted in its recognition by the

alternatives. In some cases, despite our differences in party or ideology, we may

most valuable tool for gathering information needed in short-term analyses.

Sherwin, E. (2006) ‘The Cost-Benefit Analysis of Financial Regulation: Lessons

The evolution of the policy analysis

Policy analysis in three eras

Era one: 1960–89