Complexity Economics and Sustainable Development A Computational Framework For Policy Priority Inference Omar A Guerrero Full Chapter PDF

Complexity Economics and Sustainable
Development: A Computational
Framework for Policy Priority Inference
Omar A. Guerrero
Visit to download the full and correct content document:
https://ebookmass.com/product/complexity-economics-and-sustainable-development-
a-computational-framework-for-policy-priority-inference-omar-a-guerrero/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Fiscal Policy for Sustainable Development in Asia-

Pacific Lekha S. Chakraborty
https://ebookmass.com/product/fiscal-policy-for-sustainable-
development-in-asia-pacific-lekha-s-chakraborty/
Visions and Strategies for a Sustainable Economy:

Theoretical and Policy Alternatives Nikolaos
Karagiannis
https://ebookmass.com/product/visions-and-strategies-for-a-
sustainable-economy-theoretical-and-policy-alternatives-nikolaos-
karagiannis/
Complexity Economics: Building a New Approach to

Ancient Economic History Koenraad Verboven
https://ebookmass.com/product/complexity-economics-building-a-
new-approach-to-ancient-economic-history-koenraad-verboven/
Towards the Next Revolution in Central Banking: A

Radical Framework for Monetary Policy Wehner
https://ebookmass.com/product/towards-the-next-revolution-in-
central-banking-a-radical-framework-for-monetary-policy-wehner/
(eTextbook PDF) for Health Economics and Policy 7th
Edition
https://ebookmass.com/product/etextbook-pdf-for-health-economics-
and-policy-7th-edition/
Complexity Economics: Building a New Approach to

Ancient Economic History 1st ed. Edition Koenraad
Verboven
https://ebookmass.com/product/complexity-economics-building-a-
new-approach-to-ancient-economic-history-1st-ed-edition-koenraad-
verboven/
Bus Transportation: Demand, Economics, Contracting, and

Policy 1st Edition David A. Hensher
https://ebookmass.com/product/bus-transportation-demand-
economics-contracting-and-policy-1st-edition-david-a-hensher/
Information as a Driver of Sustainable Finance: The

European Regulatory Framework Nadia Linciano
https://ebookmass.com/product/information-as-a-driver-of-
sustainable-finance-the-european-regulatory-framework-nadia-
linciano/
Water and Climate Change: Sustainable Development,

Environmental and Policy Issues Edited By Trevor M.
Letcher
https://ebookmass.com/product/water-and-climate-change-
sustainable-development-environmental-and-policy-issues-edited-
by-trevor-m-letcher/
Complexity Economics and Sustainable Development
The Sustainable Development Goals are global objectives set by the UN.
They cover fundamental issues in development such as poverty,
education, economic growth, and climate. Despite growing data across
policy dimensions, popular statistical approaches offer limited solutions
as these datasets are not big or detailed enough to meet their technical
requirements. Complexity Economics and Sustainable Development
provides a novel framework to handle these challenging features,
suggesting that complexity science, agent-based modelling, and
computational social science can overcome these limitations. Building
on interdisciplinary socioeconomic theory, it provides a new framework
to quantify the link between public expenditure and development while
accounting for complex interdependencies and public governance.
Accompanied by comprehensive data of worldwide development
indicators and open-source code, it provides a detailed construction of
the analytic toolkit, familiarising readers with a diverse set of empirical
applications and drawing policy implications that are insightful to a
diverse readership.
omar a. guerrero is the Head of Computational Social Science

Research at The Alan Turing Institute, London. He was an Oxford
Martin Fellow at the University of Oxford and a senior research
fellow at the University College London, as well as consulted for
governmental bodies and international organisations.
gonzalo castañeda is Professor in Economics at CIDE, Mexico

City. He has been a visiting scholar in various universities and
research centres in the US and Europe. His previous research
includes the book: The Paradigm of Social Complexity, which
covers diverse approaches to economic thinking and modelling.
Published online by Cambridge University Press

Complexity Economics
and Sustainable
Development
A Computational Framework for
Policy Priority Inference
omar a. guerrero
The Alan Turing Institute
gonzalo castañeda
Centro de Investigación y Docencia Económicas

Shaftesbury Road, Cambridge CB2 8EA, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
Cambridge University Press is part of Cambridge University Press & Assessment,

a department of the University of Cambridge.
We share the University’s mission to contribute to society through the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781316516980
DOI: 10.1017/9781009022910
© Omar A. Guerrero and Gonzalo Castañeda 2024
This publication is in copyright. Subject to statutory exception and to the
provisions of relevant collective licensing agreements, no reproduction of any
part may take place without the written permission of Cambridge University
Press & Assessment.
First published 2024
A catalogue record for this publication is available from the British Library.
A Cataloging-in-Publication data record for this book is available from the
Library of Congress
ISBN 978-1-316-51698-0 Hardback
ISBN 978-1-009-01654-4 Paperback
Cambridge University Press & Assessment has no responsibility for the
persistence or accuracy of URLs for external or third-party internet websites
referred to in this publication and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.

Omar’s dedication: To Sanna, Rita, and Omar
Gonzalo’s dedication: To Verónica, Almudena, and Valeria
For their unconditional and loving support and the loss of joyful
family time they had to endure.

Contents
List of Figures page xiii

List of Tables xvi
Foreword by Luis F. López Calva and Robert Axtell xvii
Acknowledgements xxiii
List of Abbreviations xxv
PART I A COMPLEXITY APPROACH TO

SUSTAINABLE DEVELOPMENT
1 Introduction 3
1.1 Motivation for This Book 3
1.2 Cutting-Edge Methods for Challenging Goals 6
1.3 The ‘Policy Priority Inference’ Research Programme 9
1.4 Target Audience 14
1.5 Structure of the Book 16
2 Policy Prioritisation, Complexity, and Agent Computing 22

2.1 Modelling the Expenditure–Development Link 24
2.2 Generative Causation and Social Mechanisms 30
2.3 On Causal Inference and Agent Computing 34
2.3.1 The Identification of Counterfactuals 36
2.3.2 The Workings of the Dependency and
Generative Accounts 38
2.3.3 The Validity of Agent-Computing Counterfactuals 43
2.3.4 The Benefits of Using Agent Computing for
Policy Evaluations 47
2.4 Summary and Conclusions 51
vii

viii contents
3 Relevant Data and Empirical Challenges 53

3.1 A Worldwide Look at Sustainable Development through
Data 54
3.1.1 SDGs and Indicators 54
3.1.2 Pre-processing Indicators and Descriptive
Statistics 62
3.1.3 Countries and Government Spending 68
3.2 Popular Modelling Frameworks and Their Limitations 84
3.2.1 Benchmark Analysis 85
3.2.2 Regression Analysis 86
3.2.3 General Equilibrium Models 88
3.2.4 System Dynamics 91
3.2.5 Network Analysis 92
3.3 Empirical Challenges 95
3.3.1 Adapting to Coarse-Grained Indicators 95
3.3.2 Moving beyond Associations 96
3.3.3 Handling Complex Expenditure Linkages 97
3.3.4 Embedding Vertical Mechanisms 99
3.3.5 Estimating Interdependency Networks 101
4 A Computational Model 104

4.1 Policy Instruments 105
4.2 Indicator Dynamics 107
4.3 Public Servants 110
4.4 Central Authority 117
4.5 Development Outcomes 122
5 Calibration and Validation 129

5.1 Calibration Strategy 130
5.2 Optimisation Algorithm 132
5.3 Goodness of Fit 137
5.4 On Statistical Confidence and Testing 139

contents ix
5.4.1 Confidence Intervals 140

5.4.2 Hypothesis Testing 142
5.5 Validation 144
5.5.1 External Validation 144
5.5.2 Internal Validation 147
5.5.3 Soft Validation 150
5.5.4 Stakeholder Validation 151
5.6 Statistical Behaviour 152
5.6.1 Testing for Synthetic Counterfactuals 153
5.6.2 Parameter Recovery 154
5.6.3 Overfitting 158
5.6.4 Time Equivalence 160
5.7 On Interdependency Networks 160
PART II A GLOBAL VIEW OF SUSTAINABLE

DEVELOPMENT
6 The Feasibility of the Sustainable Development Goals 167

6.1 On Quantifying the Feasibility of the SDGs 167
6.2 Simulation Strategy 169
6.3 Expected Gaps 171
6.3.1 Gap Closures 176
6.4 Sensitivity to the Budget Size 181
6.4.1 Proportional Changes in the Budget 184
7 Government Spending and Structural Bottlenecks 188

7.1 On the Concept of a Structural Bottleneck 189
7.1.1 A Formalisation: The Budgetary Frontier 190
7.3 Results 192
7.3.1 Insensitivity on the Budgetary Frontier 192
7.3.2 Poor Performance + Insensitivity = Bottleneck 195

x contents
7.3.3 Not All Bottlenecks Are Created Equal: A

Flagging System 200
8 Public Governance and Sustainable Development 207

8.1 On the Study of Corruption and the Rule of Law 209
8.1.1 Two Conceptual Frameworks: Principal–Agent
versus Systems Thinking 209
8.1.2 The Dominant Empirical Approach: Econometrics 210
8.2 Data with an Endogenous Rule of Law 214
8.4 Results of Counterfactual Analyses 219
8.4.1 Non-linear Responses to Expenditure in the
Rule of Law 219
8.4.2 Rugged Policy Landscapes 223
9 The Impact of International Aid 231

9.1 Studies on Aid Effectiveness 233
9.2 Data 236
9.2.1 Countries and Indicators 236
9.2.2 Government Expenditure 237
9.2.3 Aid Flows 240
9.3.1 Expenditure, Aid, and Counterfactuals 243
9.3.2 Impact Metric 244
9.3.3 Statistical Significance 248
9.4 Results 250
9.5 Beyond Conventional Methodologies 256

contents xi
PART III A FOCALISED VIEW OF SUSTAINABLE

DEVELOPMENT
10 Subnational Development and Fiscal Federalism 265

10.1 On Fiscal Federalism 267
10.1.1 Fiscal Decentralisation in Mexico 268
10.2 Data 271
10.2.1 Development Indicators 271
10.2.2 Development Clusters 271
10.2.3 Expenditure Data 277
10.4 Results 282
10.4.1 The Impact of Contributions 282
10.4.2 Optimising Contributions 282
10.4.3 Policy Priorities and Contributions 285
11 Accelerators and Systemic Bottlenecks 289

11.1 Accelerators, Bottlenecks, and Their Empirical
Quantification 291
11.2 Data 293
11.2.1 Government Expenditure 293
11.2.2 Development Indicators 294
11.3.1 Counterfactual Budgets 310
11.3.2 Detection of Bottlenecks and Accelerators 311
11.4 Results 313
11.4.1 Identification of Systemic Bottlenecks
and Accelerators 313
11.4.2 Comparison against Naïve Approaches 315
11.4.3 Disaggregation of Systemic Bottlenecks and
Accelerators 319

xii contents
12 Deprivation, Income Shocks, and Remittances 323

12.1 Socioeconomic Deprivation in the Mexican Context 324
12.1.1 The Importance of Remittances and Research
Design 325
12.2 Data 327
12.2.1 Indicators 329
12.2.2 Social Expenditure 333
12.2.3 Household Spending and Remittances 335
12.2.4 The Complex Structure of Government
Spending and Development 336
12.4 Results 340
12.4.1 Impact Evaluation by Expenditure Source 340
12.4.2 Shock Mitigation via Government Expenditure 346
13 Lessons and Reflections 350

13.1 Lessons Learnt 350
13.2 From Analysis to Policy Guidelines 357
13.2.1 Workflow for Strategic Planning 359
13.3 A Call for Computational Social Scientists 362
13.3.1 Necessary Infrastructure 363
13.3.2 Upgrading Skills in Technical Teams 365
13.3.3 Updating Social Science Programmes 366
Bibliography 370
Index 390

Figures
2.1 Factors affecting the inference of policy prioritisation. page 26

2.2 Social mechanisms in the expenditure–development link. 32
2.3 The epistemological triad of causation accounts. 37
2.4 Causal inference through agent computing. 41
2.5 Causal impact distributions with inadequate
counterfactuals 45
3.1 The 17 Sustainable Development Goals. 56
3.2 Average indicator levels and development gaps by
country group 65
3.3 Countries and regions covered by the SDR dataset 69
3.4 Government expenditure per capita (real USD) 72
4.1 Illustrative indicator dynamics 109
4.2 Learning dynamics and the emergence of social
norms under different public governance regimes 116
4.3 Bottom-up and top-down structure linking
government expenditure and indicators. 126
5.1 Error minimisation behaviour 136
5.2 Distribution of goodness-of-fit metric by SDG and
country group 138
5.3 External validation 146
5.4 Internal validation 149
5.5 Well-behaved impact estimates 155
5.6 Parameter recovery 157
5.7 Overfitting detection (‘folding statistic’) 159
6.1 Expected gaps in 2030 172
6.2 Gap closures 177
6.3 Potential impacts of increasing government annual
expenditure in $100 USD per capita 182
xiii

xiv figures
6.4 Change in convergence time to the goals by country

group and SDG 186
7.1 Illustration of idiosyncratic bottlenecks as a result
of structural constraints 191
7.2 Reductions in expected gaps under the budgetary frontier 194
7.3 Distribution of indicators according to historical
performance and potential gap reductions 197
7.4 Distribution of structural bottlenecks across SDGs
by country groups 199
7.5 Bottleneck flags by country and SDG 202
8.1 Association between corruption and public governance 215
8.2 Six cases of within-country sensitivity to
expenditure in the rule of law 220
8.3 Policy surfaces of six countries 224
8.4 Roughness scores of the policy surfaces and their
association with development 227
9.1 Countries recipients of SDG-classified aid between
2000 and 2013 236
9.2 Total aid received as a fraction of government
expenditure during 2000 to 2013 241
9.3 Hypothetical scenarios to illustrate the workings of
the impact metric 246
9.4 Country-level impact of international aid 252
9.5 SDG-level impact of international aid by group 253
9.6 Indicator-level impact of international aid 255
9.7 Disaggregated impact of aid related to access to
basic sanitation services 259
10.1 State budgets 270
10.2 SDG-level impact metric of federal contributions by
state cluster 283
10.3 Results from optimising the distribution of federal
contributions across states 284
10.4 Optimal distributions of federal contributions 286
11.1 Budgetary links between SDG targets and indicators 308

figures xv
11.2 Systemic bottlenecks 313

11.3 Accelerators 314
11.4 Weak correlation between clogging/acceleration
potential and data on budgets and centrality 317
11.5 Networks of indirect impacts 320
12.1 Development indicators by right 329
12.2 Expenditure programmes by right 334
12.3 Expenditure patterns and remittances among
lower-income households 336
12.4 Bipartite network structure of expenditure and
development indicators 338
12.5 Impacts by spending source with different
aggregation levels 342
12.6 Government expenditure increments needed to
mitigate income shocks 347
13.1 Workflow to specify policy guidelines with PPI 361
13.2 Computational social science and related fields. 367

Tables
3.1 Indicators sampled from the Sustainable

Development Report page 57
3.2 Countries sampled from the Sustainable
Development Report 76
4.1 Variables of the model 125
9.1 Average indicator level by SDG and country group
(2000–2013) 237
9.2 Total per capita aid flows and government
expenditure per country group (2000–2013) 239
9.3 Percentage of countries with statistically significant
aid impact in water and sanitation 260
10.1 Indicators for the subnational analysis of Mexico 272
11.1 SDG targets that have corresponding indicators and
budgetary data 296
11.2 Development indicators 300
12.1 Social development indicators 330
12.2 Top prolonged programmes by direct impact after an
expenditure reduction of 20% 344
12.3 Top prolonged programmes by system-wide impact
after an expenditure reduction of 20% 345
xvi

Foreword
Recent global events have emphasised the uncertain and intricate

nature of our world. Various unexpected shocks have disrupted our
development pathways, and each of them requires attention and
resources from policy makers. Determining where to allocate scarce
resources is not straightforward. The traditional epistemological and
methodological approaches, which rely on a linear and mechanistic
understanding of causes and effects, are insufficient for comprehend-
ing social challenges and offering effective policy advice to achieve
desirable societal outcomes. In light of these challenges, the authors
of this book have stepped up to address this gap by providing a broad
understanding of the multiplicity, interdependency, and ambiguity of
forces involved in development practice.
Of particular note is their focus on the intertwining nature of
development objectives. Policies aimed at addressing one objective
can have an impact on other objectives, creating feedback effects,
either complementarities or trade-offs. This interconnected reality,
along with the fiscal constraints that most governments face, under-
scores the importance of policy prioritisation. How can governments
identify those policy actions that will have the most impact across
development goals, thereby accelerating systemic progress or lifting
systemic bottlenecks? In this book, the authors provide a power-
ful framework, drawn from years of research at the Policy Priority
Inference (PPI) programme, to help decision-makers navigate this
effort.
Since the Fall of 2018, when I started my tenure as Regional
Director at the United Nations Development Programme (UNDP), I
aimed to bring new and diversified approaches to enhance the effec-
tiveness of the development work we support, as the most important
development agency of the United Nations, in the pursuit of the
xvii
https://doi.org/10.1017/9781009022910.001 Published online by Cambridge University Press

xviii foreword
achievement of the Sustainable Development Goals by 2030. Having

had the chance to see the agent-based modelling and complexity work
done by Omar A. Guerrero and Gonzalo Castañeda, I immediately
thought of how much that approach could contribute to the UNDP’s
integrator role at the UN. We thus supported the launch of the Policy
Priority Inference research programme, coordinated by Almudena
Fernandez. The evolution of this research programme over time
demonstrates the importance of efforts that intentionally seek to
bridge the gap between academic work and policy engagement. By
working together with local and national governments to apply the
PPI to specific policy questions, the authors were able to strengthen
the tool by ensuring that its cutting-edge methodologies were also
grounded in real-world assumptions – from fiscal rigidities to limited
information. It was not surprising that as soon as word began to spread
about the framework and potential results presented by the authors
in previous papers, colleagues and public officials began to take notice
and seek out its insights. The demand for this kind of knowledge was
immediate and organic, reflecting the hunger of policy makers for
reliable and rigorous guidance on how to prioritise public spending
to achieve better outcomes.
What sets this work apart is that its analytical framework
is not only theoretically rich but empirically feasible and rigorous.
This is achieved through an interdisciplinary approach, which draws
upon insights mainly from economics, political science, computer
science, and complexity science. Importantly, however, this book is
more than just an academic exercise. It has the potential to influence
policy by providing evidence and empirically driven knowledge of the
interactions in the complex SDG space. This can be useful to inform
the prioritisation of public spending to achieve better outcomes, given
real-world budgetary challenges. Beyond offering guidance on how to
reallocate expenditure to close development gaps in specific policy
contexts, the computational tool described in the book also prompts
a broader reimagining of how we can think about achieving the 2030
Agenda (with its 17 Sustainable Development Goals supported by 169

foreword xix
targets and over 200 indicators) from a more integrated and holistic
perspective.
In short, this book is a must-read for anyone interested in under-
standing development as a complex process and the challenges of
policy prioritisation in an uncertain world. It represents an excellent
contribution to the field. I have no doubt that its insights will be of
immense value to scholars, practitioners, and policy makers alike.
—Luis F. López Calva

Global Director for Poverty and Equity Practice
The World Bank
Washington, D.C.
This book by Professors Guerrero and Castañeda is an important con-

tribution to the social science of economic development, generally,
and to sustainable development, specifically. Their goal, as described
in the first chapter, is to provide a framework for formulating sus-
tainable development policies in a multidimensional setting and from
a complex systems perspective. The authors have backgrounds in
economics and computational social science and interests in devel-
opment economics, and each has experience working within the
large ecosystem associated with international development activities,
consisting of government agencies, NGOs, foundations, and inter-
national organizations like the World Bank and the United Nations.
Their broad exposure to both academic and practical aspects of eco-
nomic development is one reason this work is so exciting, for it takes
on real-world challenges with new ideas and tools from the research
frontier to first understand extant development strategies and then
create new and better approaches.
The set of tools they build on are grounded in the emerging
science of complexity economics. A well-known economist once
quipped of the field of environmental economics that its relation to
the overall profession of economics was like the relation of military
music to music in general: its repertoire limited, its instruments a
subset of all instruments. In certain ways, development economics

xx foreword
is similar: the methodologies used by development economists may

be years or even decades behind current practice at the forefront of
the profession, such as heavy reliance on regression models in the
desire to be empirically relevant. But we are living through the great
flowering of data in the social sciences, from small-scale laboratory
experiments with human subjects to larger-scale field studies and
internet-enabled experimentation, to administrative data that con-
stitute complete universes – every relevant person represented – in
a wide variety of fields within economics and beyond, for example,
finance. There exists today a far-ranging and rapidly growing set of
tools and techniques for working with and analysing such data, from
machine learning (ML) to artificial intelligence (AI) approaches. The
focus of Guerrero and Castañeda on ‘agent computing’ is one of these,
with so-called multi-agent systems usually considered a branch of AI.
It is from these rich and hopeful advances that the field of complexity
economics has arisen, focusing on social networks, direct agent–agent
interactions, out-of-equilibrium dynamics, and realistic behavioural
rules, among other things. While today this new approach has been
used to model markets, firms, trade, and many other sub-fields within
economics, its application to development questions is really in its
infancy, which brings us to the focus of this book – Policy Priority
Inference (PPI).
While the roots of this idea lie in the abstract ideas of com-
plexity economics, its goal is to deal with real-world development
challenges, to take capabilities from the ‘know-how’ stage to actual
knowledge about specific countries and their development trajecto-
ries. As described in this book, particularly Chapters 2–4, PPI is a way
to use all the data that are available for purposes of first understanding
what is going on in a country, from the development perspective,
assessing the performance of existing policies and programmes, and
only then moving on towards the creation of more effective policies.
I am not an expert on economic development. However, I
have spent much of my career trying to apply the new tools and
techniques from complexity economics to real-world problems, from

foreword xxi
firms to fisheries, housing to Hayek, the formation of economic

classes and retirement economics. It seems to me that Guerrero and
Castañeda have leveraged the particular strengths of the new approach
in order to advance development thinking, and for that they are to
be commended. Specifically, economics graduate students today are
immersed in a highly idealised theoretical world in which assump-
tions about the attainment of equilibrium, the rationality of economic
actors, the utility of representative agents, and myriad others are
taken as given, despite the fact that most of these specifications
are vitiated by the facts, especially in developing countries where
markets are often more volatile than in the developed world, people
face higher transaction costs, production may be less efficient, and
so on. Therefore, to the extent that the relaxation of such heroic
assumptions is the main goal of complexity economics, to build more
realistic models, and if this can be accomplished with agent-based
modelling, for example, it must be even more true that problems of
sustainable development can benefit from such an approach relatively
more. If complexity economics can help us understand economies
from the bottom up, as largely self-organised, spontaneous ecologies
of people and groups, partially guided and regulated from the top
down, surely it can aid in our understanding of why some parts of
the world are stymied from comparable achievements.
Crucially, Guerrero and Castañeda also have specific, deep
domain knowledge of the Mexican development situation. This helps
them frame PPI in a realistic way, focusing on the kinds of policies
that have some chance of success, for they have seen good intentions
lost to bad policies in their home country. But any developing country
is vulnerable to ineffective development attempts, stemming from
partially or largely incorrect diagnoses of the problems in the first
place. The novelty of PPI is that it does not separate its diagnosis
of extant problems from its recommendations about what should be
done. It is an integrated perspective and a welcome breath of fresh
air over the crude statistical and econometric analyses that dominate
conventional development discussions and policy recommendations.

xxii foreword
This book serves as the best, most comprehensive introduction

to PPI. I think it is fair to surmise that the authors do not think of
it as the ‘last word’ on the subject, but perhaps something closer to
the ‘first word’. As such, it represents a new approach to sustainable
development, one with great promise, a modern approach for the
computational age in which we live. This is not a book that could have
been written 10 or 15 years ago. Back then we simply did not either
have sufficient data or computing for this approach to be feasible.
But now we are entering a whole new world, in which increasing
amounts of both intra- and inter-country data are becoming available
and reshaping our empirical understanding of our world. What is in
short supply now is new thinking about how economic development
works, how it changes lives, and how crafting country-specific poli-
cies can lead to sustainable development. This book offers us hope
that such new ideas, fueled by new data, can be brought to bear
on development economic questions, under the rubric of complexity
economics.
This book offers us a kind of ‘chemical’ mixture of new data
and new computational tools – the data having little use without the
models, the computing vacuous without the data – out of which, in
true complexity style, the whole is more than the sum of the parts.
And for this we can be very grateful to the authors.
—Robert Axtell
Professor of Computational Social Science
George Mason University and Santa Fe Institute
McLean, Virginia

Acknowledgements
Developing the Policy Priority Inference (PPI) research programme

has been the result of various years of work, collaborations, and
engagement with different stakeholders. We are grateful for the
insights and support of the people and institutions involved in
PPI throughout this process. We would like to thank The Alan
Turing Institute for its unconditional support and, in particular,
Helen Margetts and Cosmina Dorobantu from the Public Policy
Programme for embedding PPI as a cornerstone initiative in the
Institute; Pauline Kinniburgh and Beth Wood for the management
and communications strategy support; and Daniele Guariso for his
work in recent collaborations that are setting the stage for the future
generation of PPI studies using large language models. We are thankful
to the Centro de Investigación y Docencia Económicas (CIDE) and
its National Laboratory of Public Policies for collaborating in various
research and policy-oriented projects, especially to Florian Juárez-
Chávez, Georgina Trujillo, Lucy Hackett, Cristina Galindez, and
Eduardo Sojo. Also from CIDE, we would like to thank Guillermo
Cejudo, Iván Bautista, and Sofia Huidobro for their collaborations
in assembling the data on expenditure for social policies. From
the Saïd Business School at the University of Oxford, we thank
Felix Reed-Tsochas for his advice during the very initial stages of
this research. We would also like to thank UCL Public Policy for
facilitating external engagement and communication strategies.
We are also grateful to the United Nations Development Pro-
gramme (UNDP) and its various regional offices. From the UNDP
for Latin America and the Caribbean, we would like to thank the
support of Luis Felipe López Calva, Alejandro Pacheco, Almudena
Fernandez, and all those who participated in the early workshops of
PPI. We are also grateful to the country-level UNDPs with whom we
xxiii

xxiv acknowledgements
worked during these years: Mauricio Ruiz Vega, Oscar Sánchez, and
Luis Alberto Palacios from UNDP-Colombia; Annabelle Sulmont,
Stephanus Visser, and Maite García from UNDP-Mexico; Gabriela
Elgegren, Gabriela Salinas, and Mirian Piscoya from UNDP-Peru; and
Stefan Liller and Virginia Varela from UNDP-Uruguay.
From the Global Initiative for Fiscal Transparency, we would
like to thank Juan Pablo Guerrero and Aura Martínez for introducing
us to a vibrant and relevant community of public finance managers
and helping us understand better the intricacies of how public admin-
istrations link large-scale budgetary data to the SDGs. The Centro
de Estudios Espinosa Yglesias has also been an enthusiastic promoter
of our research programme, so we are grateful to Roberto Vélez and
Marcelo de la Jara. We are also grateful to Lorena Rivero del Paso from
the International Monetary Fund for opening new doors in the Public
Financial Management community. We thank Michael Weber from
the World Bank for introducing our research to the Human Capital
Project working group.
We would like to thank all the scholars, officials, analysts,
and consultants from the various universities, journals reviewer
pools, research centres, governments (national and subnational),
and international organisations with whom we have engaged during
this process. Last but not least, we are grateful for the sponsorship
from the United Kingdom’s Economic and Social Research Council
(through grant ES/T005319/1) and the Engineering and Physical
Sciences Research Council (through grant EP/N510129/1); as well
as the National Council for Science and Technology from Mexico.
Without the contributions and hard work of all these people and
institutions, the PPI research programme would not have reached the
maturity and real-world relevance and impact that it has today.

Abbreviations
ABM agent-based model

AI artificial intelligence
Banxico Banco de México (Mexico’s Central Bank)
CGE computable general equilibrium
CIDE Centro de Investigación y Docencia Económicas (Center for
Research and Teaching in Economics)
CONEVAL Consejo Nacional de Evaluación de la Política de Desar-
rollo Social (National Council for the Evaluation of Social Develop-
ment Policy)
CSS computational social science
DiD difference-in-differences
ENIGH Encuesta Nacional de Ingresos y Gastos de los Hogares
(National Survey on Household Income and Expenditure)
EPSRC Engineering and Physical Sciences Research Council
ESRC Economic and Social Research Council
GDP gross domestic product
HDI Human Development Index
INEGI Instituto Nacional de Estadística y Geografía (National Insti-
tute of Statistics and Geography)
LAC Latin America and the Caribbean
LPI line of poverty due to insufficient income
MENA Middle East and North Africa
ODA Official Development Assistance
OECD Organisation for Economic Co-operation and Development
PPI Policy Priority Inference
RCT randomised controlled trial
RoL rule of law
SCG subnational central government
SDG Sustainable Development Goal
xxv

xxvi list of abbreviations
SDR Sustainable Development Report

SHCP Secretaría de Hacienda y Crédito Público (Secretariat of
Finance and Public Credit)
UK United Kingdom
UN United Nations
UNDP United Nations Development Programme
UNSDG United Nations Sustainable Development Goals (data plat-
form)
US United States
USD United States dollar
VAT value added tax

PART I
A Complexity Approach to
Sustainable Development

1 Introduction
1.1 motivation for this book

“A priority not reflected in the budget is pure demagogy”. This
saying has gained popularity in the past decade among economists,
political analysts, and mainstream media outlets in Mexico.1 This
expression reflects not only the political and economic circumstances
of a particular country but fits the reality of budgetary decisions
made by many governments worldwide. Examples of these practices
abound, either for populist reasons or for lacking knowledge of how
budgetary decisions affect the economy’s performance. When inco-
herent spending plans emerge in democratic societies, they tend to
elicit a strong negative response from opposition parties or society
at large. Unfortunately, budgetary proposals of this sort are not rare,
so countries and their societies have to endure disappointing results
down the road.
Today, it is not difficult to find cases of societal dissatisfaction
with a government’s budgetary proposal. For instance, in recent years,
a government in the UK collapsed in less than 50 days due to a
disastrous budgetary proposal. The so-called mini-budget, as termed
by the media, did not sit well with a large sector of the population,
financial markets, and international organisations since it lacked a
sound commitment to achieve a set of promised goals (BBC News,
2022). In another case, the minister for agriculture of India promised
in 2021 to raise the target of agricultural credits. Yet, when presenting
the national budget, the potential beneficiaries did not perceive any
1 The phrase stems from an older aphorism commonly used in the Mexican political
parlance between the 1960s and 1990s: A friendship not registered in the payroll is
pure demagogy. This phrase was popularised by the satirical writer Carlos Monsiváis.
Both versions refer to the asymmetry between stated intentions and real financial
commitments.
3

4 1 introduction
real intent to improve their livelihoods. Protest leader Kirankumar

Visa declared: “Forget about the targets […] There is not even one
measure to either raise farmers’ income or generate jobs” (Bhardwaj,
2021).
While expenditure dissatisfaction may not always be a
society-wide movement, specific sectors – often represented by
lobbying groups – frequently attempt to reflect their agenda in the
government’s budgetary allocation. For example, in early 2022, the
South African Federation of Trade Unions (with 800,000 members)
mobilised protests in Pretoria with the aim of affecting the national
budget. Its general secretary, Zwelinzima Vavi, demanded the
government not only a set of policies to hike workers’ income but
also to commit more resources to service delivery, schools, hospitals
and police stations (Business Tech, 2022). Thus, there was a clear
request to modify policy priorities through government expenditure
across a large set of issues.
Perhaps, when writing this book, the most mediatic example
of social protest to modify governments’ priorities is the Just Stop
Oil movement. These protests have taken place in several European
countries and became highly visible in mainstream media due to
polemic actions such as defacing world-renowned paintings and clos-
ing motorways. Independently of whether this movement is right
or wrong, their motivation has a clear budgetary target: eliminating
subsidies and tax breaks for new fossil fuel extraction (Lu, 2022).
Thus, these movements justify their actions by the need to make
their voice heard against demagogic politicians claiming that climate
change is a high priority when, in fact, they do not tackle this issue
in their national budgets.
Like these examples, we can find similar others around the
world, in both rich and poor countries, and across different develop-
ment issues. Overall, they commonly exhibit three related features.
First, they have a set of (development) goals in mind. Often, these are
goals expressed by the government in official documents or campaign
promises that a large societal sector agrees upon. Second, there are

1.1 motivation for this book 5
specific actions by the government to achieve those goals, often

materialised in terms of a budget. Protests like the ones mentioned
above emerge when these expenditure actions do not seem conducive
to the goals broadly agreed upon. Thus, there is a perception of
policy incoherence as the state’s priorities are not reflected in the
budget. Third, as a response to this incoherence, lobbying groups
exert pressure on governments to align their stated priorities with
the budget and, sometimes, push their own agendas into the govern-
ments’ priorities. That being said, different societal actors can read,
from government budgets, how seriously committed a government is
to achieving its development goals or campaign promises. However,
such reading is not always straightforward.
While priorities on big issues such as climate change and agri-
cultural subventions may be easily identifiable by stakeholders and
the media, there exist many other development dimensions where
such identification is not possible if they do not receive enough media
attention. For instance, budgets (or a large part of them) may not
be disclosed truthfully by governments; expenditure data may be
too aggregate to make a clear connection between the expenditure
and specific policy issues; certain policy interventions are inherently
more expensive, which is not always easy to observe from just looking
at data; expenditure patterns may carry an inertial or historical
component that does not reflect the current government’s priorities;
and there may be fiscal, political, and bureaucratic rigidities limiting
the flexibility and scope with which a government can reallocate
resources. Moreover, the opacity in government expenditure data
has been historically a political tool for governments attempting to
obfuscate their true priorities, as unclear data limit society’s auditing
capabilities.
Nevertheless, in the last decades, there has been a shift
towards fiscal transparency, at least in democratic societies. With
the aim of making budgetary data more accessible, multiple
international initiatives have emerged to strengthen the growing
‘Open Government’ agenda. Hence, today, we can find large

6 1 introduction
repositories of open-spending data for national and subnational

governments, with various degrees of quality and resolution. Despite
these remarkable efforts, the challenge of understanding policy
prioritisation remains daunting since it is not just about reading
true intentions through budgets but also about how such expenditure
translates into development. Thus, budgetary transparency is only the
first step towards a bigger challenge: understanding the expenditure–
development relationship. This inquiry is precisely our motivation
for elaborating an analytic computational framework, applying it to
answer diverse questions related to sustainable development, and
writing this book.
1.2 cutting-edge methods for challenging

goals
Let us switch from the expenditure side to the development end. Here,
we can say that efforts to quantify societal progress predate and are
more notorious than the advances in fiscal transparency. Nowadays,
many development indicators support evidence-based policymaking
on specific issues. Historically, different communities of consultants
and academics use specific indicators, in isolation from each other,
according to their domain of expertise. In the last decades, however,
the development community has shifted towards integrating several
domains into a multidimensional and complexity perspective. The
leading initiative is the United Nations 2030 Agenda of the Sus-
tainable Development Goals (SDGs), which has catalysed the con-
struction of more development indicators across a broader spectrum
of policy issues (e.g., poverty, inequality, social inclusion, environ-
mental sustainability, and public governance). Several academics and
consultants have jumped into the SDG bandwagon, giving place to
numerous studies analysing interrelationships between SDGs.
While these efforts have been instrumental in shifting the over-
all discussion of development towards a systemic framework, they
remain disconnected from budgets and, hence, from policy prioritisa-
tion. That is so because they tend to focus on the ‘output’ side of the

1.2 cutting-edge methods for challenging goals 7
expenditure–development relationship. Without understanding the

role of government spending (the input), insights derived exclusively
from indicator-based studies are somehow limited, especially when
advising governments on the prioritisation of development issues.
One would imagine that understanding the expenditure–development
link more explicitly and the policy prioritisation process would be
a primary task in an agenda that seeks to persuade governments to
commit to budgetary actions in the pursuit of global goals. Unfor-
tunately, this is not yet the case, at least not when it comes down
to quantitative frameworks that operate in large multidimensional
settings.
From our experience, we have come to realise that there is a lack
of analytic tools to address policy prioritisation (from a systemic point
of view). Because of that, efforts in promoting data transparency seem
wasted. In general, governments do not get serious about promoting
policy evaluations for measuring the impact of budgetary allocations
on development outcomes. More specifically, there are five popular
practices among governments and consultants that, in our opinion,
need to be addressed to make substantial progress in policy coherence.
First, the process through which governments arrive at their
development strategies (reflected in documents such as development
plans, agendas to promote industrial transformation, and campaign
platforms) tends to be, at best, ad hoc and, in the worst cases,
completely arbitrary. Second, even if development plans result from
legitimate democratic or professionalised processes such as national
consultations and diagnostic frameworks, the data indicate that bud-
gets rarely reflect the declared priorities. Third, traditional evaluation
tools do not address systemic problems. In particular, studies employ-
ing techniques such as micro-econometrics and field experiments
need to be extremely narrow to produce causal statements with the
available data. Fourth, even with good data, conventional studies
do not consider the myriad political-economy factors that mediate
the expenditure–development relationship (such as inefficiencies and
corruption). Not accounting for the political economy precludes an

8 1 introduction
accurate mapping between budgets and outcomes. Fifth, to make

things even more difficult, such political-economy mediation often
takes place at various scales (micro, meso, macro), so the use of aggre-
gate data in regression analyses is insufficient to provide convincing
advice.
Oddly, when we started this research programme, mainstream
economists and political scientists in academia did not seem inter-
ested in tackling these issues. In development economics, a social
norm, where the gold standard was to perform micro-level studies of
causal impacts, had emerged. In political science, the most relevant
research used aggregate expenditure data to estimate the distribution
of budgetary changes. Unfortunately, the latter studies showed no
intention of establishing a connection with the outcomes generated
by expenditure changes. Overall, a systemic focus seemed to be
irrelevant in the minds of conventional analysts, as if complications
associated with inadequate policy prioritisation did not exist. For us,
the data, the published research, the methods employed, the proposed
solutions, and the policy practices indicated quite the opposite.2
Outside economics and political science, different research
communities have long been interested in problems related to
a systemic view of development. However, they did not usually
share the right combination of tools and theories. For example,
development studies journals started publishing some analyses using
network metrics and complexity ideas, but their usual readership
lacked the skills to properly embrace cutting-edge ideas coming
from network and complexity sciences. In contrast sustainability
scholars and network scientists were already investigating networks
2 As scholars, we are aware that it is common to observe conformity with dominant
ideas and a tendency to dismiss proposals attempting to defy the status quo. For
example, complexity economics has been subjected to such dismissals for more than
three decades (Arthur, 2021). Nevertheless, today, many ideas and methods
stemming from this community have transformed practices in key economic affairs
such as financial regulation (through the measurement of systemic risk), industrial
strategy (by better understanding firm dynamics), and international trade (by
quantifying the economic sophistication of nations).

1.3 the ‘ policy priority inference’ research programme 9
of SDGs using state-of-the-art methods but without incorporating

the critical expenditure–development connection, or its political-
economy elements. Other fields that have historically focused on
systems thinking (such as system dynamics) were also studying SDGs
but their tools could only work at the macro level, missing causal
mechanisms involving the political-economy process.
Both of us had worked for some time in the intersection between
economics and complexity science. Hence, we were familiar with
the methodological gaps between intriguing problems raised by social
scientists and their unawareness of cutting-edge methods. Thus, it
became clear that, if we wanted to push the boundaries of under-
standing the expenditure–development link, we had to take on some
of these challenges and formulate a new analytical framework that
would combine ideas and methods from various disciplines. Impor-
tantly, due to the urgency of providing well-grounded advice for
closing the development gaps of the 2030 Agenda, any proposal would
have to be empirically usable, easy to scale, and flexible enough
to accommodate budgetary data with different levels of granularity.
Furthermore, our framework would need to become a contributing
factor in the capacity-building of policymakers and their analysts.
1.3 the ‘ policy priority inference’ research

programme
In 2018, and in collaboration with Florian Juárez-Chávez, we pub-
lished a paper entitled ”How do governments determine policy pri-
orities? Studying development strategies through spillover networks”
(Castañeda et al., 2018). This work was our first attempt to discern the
expenditure–development relationship from a systemic point of view.
There, we opted for a computational approach because it allowed us
to explicitly model some critical political-economy mechanisms that
mediate the expenditure–development link while, at the same time,
allowing us to work in a multidimensional setting. On the empirical
front, we discarded using statistical methods, such as regression
analysis and machine learning, since they were inadequate when

10 1 introduction
working with many development indicators (perhaps hundreds) and

relatively short time series of development indicators.
Accordingly, we had to develop a framework in which a the-
oretically rich computational model would supplement the lack of
data. This approach would allow us to generate synthetic series that
preserved essential features of the empirical data, bypassing the need
for gigantic development datasets and enabling scalability. The lack
of disaggregated budgetary information for many countries meant
that we had to model how a government allocated resources in a
system where bureaucrats mediated their use and where development
indicators impacted each other through spillovers. For this reason,
precisely, we decided to refer to our research programme as ‘Policy
Priority Inference’ (PPI).
Following this paper, we kept enriching our model, access-
ing new datasets, developing better calibration methods, and tack-
ling more specific empirical challenges such as quantifying policy
resilience (Castañeda and Guerrero, 2018), measuring policy coher-
ence (Guerrero and Castañeda, 2021b), and estimating the effective-
ness of the rule of law (Guerrero and Castañeda, 2021a). While starting
as a typical research programme within the boundaries of academia, it
suddenly turned into an ambitious agenda with policy implications.
This take-off occurred when members of the Bureau for Latin America
& the Caribbean of the United Nations Development Programme
(UNDP) read our first publication and invited us to their New York
offices to give a workshop. Around the same time that we were
working on our first PPI paper, the United Nations had given the
UNDP the mandate of coordinating much of the activities related
to the 2030 Agenda. Such coordination involved facilitating analytic
tools that governments could use to inform policy prioritisation
towards the SDGs. Thus, our initial work seemed relevant to the
UNDP’s new tasks.
We quickly realised that the need for new analytic frameworks
was more pressing than we originally had imagined. It was clear, from
these interactions, that the models and empirical methods provided,

at that time, by economists and other communities were insuffi-

cient, as the SDGs suddenly shifted the discussions to the domain
of complex systems. Our adoption of a non-conventional approach
implied that we had many methodological challenges ahead, concep-
tual, technical, and data-related. The evolution of PPI resulted from
cross-fertilisation between our academic work and our engagement in
policy projects. In collaboration with the UNDP, we conducted more
than ten projects where we had to adapt PPI to address highly specific
empirical questions that local and national governments wanted to
address.
Recurrent questions inspired by real policy problems created the
need to make more realistic assumptions, which is not straightfor-
ward with more conventional frameworks. Some examples of these
assumptions include the differentiation between indicators where a
government has relevant intervention programmes and where it does
not; allowing for fiscal rigidities that limit the government’s ability
to reallocate funds; enabling systemic impacts for the identification
of development accelerators; and incorporating the substitutability
between private and public (internal and external) sources of funding
to produce advances in different development issues. Many of these
elements were absent in the scholarly literature, at least in the
quantitative one, so it was up to us to push the frontiers and propose
ways to account for them.
During this exciting process, we received the support of
different institutions that complemented the efforts of the UNDP.
The Alan Turing Institute (Omar’s home institution) provided, all
along this process, financial means and other resources; the United
Kingdom’s Engineering and Physical Sciences Research Council and
Economic and Social Research Council funded various parts of this
research; the National Laboratory of Public Policy, at the Centre for
the Research and Teaching of Economics (CIDE, Gonzalo’s home
institution), organised workshops with public servants, which were
helpful to inform, validate, and socialise PPI; the Mexican National
Council of Science and Technology funded, in part, Gonzalo’s

12 1 introduction
sabbatical in the academic year 2018–2019; the Global Initiative

for Fiscal Transparency brought us closer to potential beneficiaries,
and the Centro de Estudios Espinosa Yglesias and the International
Monetary Fund opened new dissemination spaces.
This book is the synthesis of the PPI research programme.
It provides a computational framework for studying sustainable devel-
opment and policy prioritisation in a multidimensional and complex
setting. In it, we present the latest version of PPI and its theoret-
ical underpinnings. We walk the reader through various empirical
applications using datasets on development indicators and govern-
ment expenditure with different levels of granularity. While much
of the content builds on our academic publications, this book is
not a collection of papers but an integrated guide through a novel
methodology. All chapters are interconnected, sometimes by data,
others by the simulation strategy, and others by the scope of the
study. In these chapters, we employ the same model and the latest
version of the data available when we started working on this book.
In trying to reach the broadest possible audience, we aim to achieve
a healthy balance between technical level and policy readability.
Therefore, the book touches only on the technical elements needed
to understand the analysis, leaving more sophisticated details to the
academic publications on which it builds.
In this book, we do not intend to make a comprehensive review
of other existing frameworks or studies for analysing SDGs,3 but only
to provide a straightforward narrative encompassing challenges in the
study of sustainable development and to present our analytic tools to
tackle them. There are five main contributions of this book. First,
it provides a comprehensive and flexible quantitative framework
that uses modern computational methods and the most recent data
available. Second, it presents a collection of empirical works tack-
ling essential inquiries from the sustainable development literature.
Third, it is a technical and detailed example of how computational
3 The reader may refer to our scholarly work for that purpose.

models can support social scientists and development consultants in

approaching complex problems. Fourth, it facilitates an empirically
relevant introduction to computational social sciences. Fifth, it serves
as a manual or tutorial on the PPI framework.
Throughout the book, the reader is familiarised with diverse
methods; for example, agent computing, gradient-search optimisa-
tion, Gaussian processes, and network analysis. However, agent com-
puting is the principal method employed for developing the PPI
framework. Agent computing – or agent-computing modelling – is an
artificial intelligence (AI) framework known by other names across
various fields.4 A popular name to refer to agent computing is agent-
based modelling. However, the term agent-based modelling is some-
times confused with neoclassical rational models from economics.
The reason is that, before digital computers, economists were already
elaborating agent-level rational-equilibrium models. Hence, today,
some economists that are not familiarised with a complexity view
(especially in the public sector) tend to understand the term ‘agent-
based’ as rational agents in an equilibrium world, which, ironically,
are assumptions that the agent-computing literature leaves aside.
Axtell (1999) proposed the term agent computing, as it reflects closer
to what these models do: using the agents to compute limited infor-
mation and interact in a decentralised fashion in an uncertain envi-
ronment. Nevertheless, the term agent-based model is also widely
accepted as a standard for agent computing. For this reason, we
use agent computing and agent-based model (ABM) interchangeably
throughout the book.
In disciplines such as sociology and epidemiology, agent com-
puting is used extensively but not in others. In economics, despite
a large body of literature employing agent computing (Axtell and
4 Sociologists often use the term “actor-oriented models”, ecologists and
epidemiologists call them “individual-based models”, computer scientists say

“multi-agent systems”, and sustainable-development scholars refer to “bottom-up
multi-agent models”. While each of these names conveys different epistemological
elements, they use similar analytic tools.

14 1 introduction
Farmer, 2022), it is still a method that, according to several scholars,

does not fit the methodological cannons of conventional thinking.
While discussing the reasons for the marginalisation of agent com-
puting in economics is beyond the scope of this book, it is worth
mentioning that a common argument is the lack of reliable/robust
empirical applications. We disagree with this argument, especially
since the past decade has seen tremendous progress in using this
framework for empirical studies. Notwithstanding, we do see a lack
of seminal books that walk the reader through the entire process of
developing a theoretically rich model, demonstrating how to use it
empirically with statistical rigour, and deriving policy implications
from the resulting inferences.5
A key contribution of this book is precisely filling the afore-
mentioned gap. In other words, this is not a book about a ‘toy model’
nor a collection of models. The book takes the reader through a series
of theoretical arguments motivating the use of agent computing in
the study of countries’ development, followed by the construction
and validation of a model of policy prioritisation. Then, it demon-
strates how the model can be employed to devise different empirical
strategies to deal with a diverse set of development-related problems.
In addition, and in the spirit of promoting open science and trans-
parency, the book comes with companion code and data that allow
the reader to replicate and modify all the analyses and figures.6
1.4 target audience

The target audience of this book is quite diverse. Let us divide it into
two groups: researchers and practitioners. In the research camp, three
major communities gather scholars from several fields. First, there
5 The seminal books on the subject were written in the 1990 and consisted mainly of
theoretical models. Among the most important contributions, we can find Epstein
and Axtell’s Sugarscape (Epstein and Axtell, 1996) and Axelrod’s prisoner’s dilemma
tournaments (Axelrod, 1997a). Other posterior books consist of collections of papers.
6 The repository of data and code can be accessed at https://github.com/oguerrer/
CESD.

1.4 target audience 15
is the development community with three subdivisions: develop-

ment economics, development studies, and sustainable development.
Each of these communities is large and includes a diverse set of schol-
ars. The second group encompasses academicians in political science
and public administration. In particular, policy change and public
financial management are highly relevant fields as they study gov-
ernment expenditure and its societal implications. Third, it includes
a community of interdisciplinary studies that spans beyond social
scientists and brings together researchers from fields such as physics,
computer science, and applied mathematics. Here, we can find the
growing communities of complexity economics, complexity science
(which includes network science), systems analysis, and computa-
tional social science. Attracting the attention of all these heteroge-
neous communities is not an easy endeavour. For this reason, the book
strives for a balance between theory, technical detail, and empirical
analyses.
In the practitioner audience, we can also find a diverse read-
ership. First, we have governments, local and national. More specif-
ically, treasuries, budget and planning offices, and their technical
experts could greatly benefit from this tool as it can easily take
advantage of much of the data they have built. Second, multilateral
organisations and consultants seeking to support capacity building
or to perform global assessments are also potential users. Third,
NGOs and civil society at large could use PPI to evaluate the gov-
ernment’s priorities and strategies, or to propose alternative ones.
Fourth, legislative institutions could use this tool as part of the budget
approval process. Fifth, political parties can employ PPI to assess the
feasibility of political campaigns (of themselves or their competi-
tors). Sixth, think tanks could benefit from integrating this frame-
work as part of the analytic toolkit of their consultants. Seventh,
international partnerships promoting open data and transparency
agendas could use PPI as an example of how such data could be
exploited and used for the benefit of society (i.e., for advocating better
budgeting).

16 1 introduction
1.5 structure of the book

We organise the book into three parts. The first introduces the theo-
retical framework and the empirical datasets that serve as the back-
bone for the computational model of PPI. The second part presents
applications with a global perspective. That is, we demonstrate how
to apply PPI to various problems typically discussed on a compar-
ative cross-country basis. In the third part, we descend into more
disaggregate levels of analysis, both conceptually and data-wise. In
these chapters, we introduce the frontier in expenditure data and
demonstrate how PPI can exploit their nuanced structures. Overall,
the book structure establishes, in the first stage, sound conceptual
and theoretical grounds. Then takes the reader, gradually, through a
sequence of increasingly more nuanced and sophisticated applications
of PPI. Next, let us summarise the content of each of the remaining
chapters.
In Chapter 2, we start with a brief exposition of how PPI anal-
yses multidimensional development in a setting of interdependent
policies and uncertainty. Then, we explain how to produce causal
inference through agent computing. This part is highly relevant for
the analyses presented throughout the book, and for contrasting the
capabilities of any ABM with those of popular econometric methods.
After clarifying that there are different accounts of causality in the
literature, we argue that agent computing combines data and theory
(social mechanisms that connect the micro and macro levels) to
produce artificial data-generating processes. Through simulations,
the method provides counterfactuals suitable for assessing the per-
formance of a system that exhibits a policy intervention. Lastly,
we mention ten advantages that agent-computing models offer for
measuring causal impact in contrast to econometric approaches.
In Chapter 3, we introduce a popular dataset on development
indicators and aggregated data on government expenditure across a
large set of countries. These two types of data constitute a com-
prehensive picture of the current state of development across mul-
tiple dimensions and geographies, as well as our first glimpse of

1.5 structure of the book 17
governments’ heterogeneous budgeting capabilities for improving

these indicators. Then, we review some of the most popular method-
ological frameworks to analyse these data and to try to understand
the expenditure–development relationship. We highlight their lim-
itations to explain why there has been little progress in studying
development from a multidimensional perspective and even less in
estimating expenditure impacts from a systemic point of view. This
presentation sets the ground for a more thorough discussion of what
we see as empirical challenges that need addressing to make progress,
both in research and policy.
Taking the aforementioned challenges as motivation for a new
framework, Chapter 4 presents the PPI model in full detail. We
elaborate further on data considerations and the empirical features
or regularities that our model tries to reproduce. Then, we describe
causal mechanisms binding government spending, political economy,
and development indicators together. Along the way, we elaborate on
the socioeconomic intuition of each component. Finally, we present
an overview of the model structure through its parameter space, a
structural diagram, and pseudo-code describing its dynamics.
Chapter 5 is for the more technically knowledgeable reader.
It is not necessary to go over it to understand the applications of PPI;
nonetheless, a reading is convenient if one has more specific inquiries
on how to make empirically relevant agent-computing models. It cov-
ers various aspects regarding the indirect calibration of parameters
for cases lacking data for direct calibration. Alongside the develop-
ment of a gradient-search calibration algorithm, we discuss important
aspects of statistical inference from computational models, such as
confidence intervals and hypothesis testing. Then, we elaborate on
the various validation aspects of agent computing. Finally, we discuss
the virtues of this type of tool in enabling counterfactual analysis and,
hence, causal inference.
We begin Part II with Chapter 6, which presents the first appli-
cation of PPI. In this chapter, we perform a prospective analysis of the
potential trajectories of various development indicators across a large

18 1 introduction
sample of countries. The analysis estimates development gaps that

would remain open by 2030 should similar government expenditure
levels remain in place. Then, we perform counterfactual simulations
that vary the expenditure usage and estimate the sensitivity of the
different development dimensions to budgetary changes in a large set
of countries.
Next, in Chapter 7, we build on these estimates and introduce a
new concept: the budgetary frontier. This idea consists in simulating
the effects of unrestricted government expenditure (the frontier) and
studying the performance of the indicators under such a setting. If
some gaps remain open by 2030 while operating on the budgetary
frontier, this result indicates the potential existence of idiosyncratic
structural bottlenecks. In other words, we infer the presence of struc-
tural factors – unrelated to government expenditure – that dampen
development and render government spending ineffective. We call
them idiosyncratic bottlenecks because they are specific to individual
policy issues (in Chapter 11 we analyse another type of bottleneck).
Finally, we develop a flagging system to pinpoint the potential pres-
ence of these bottlenecks and differentiate whether poor performance
is due to underfunded policies or idiosyncratic bottlenecks.
In Chapter 8, we maintain a global perspective but shift to a
more specific type of expenditure: the one dedicated to improving
the implementation of public governance programmes. Political econ-
omy is a central element of PPI, and it uses data about the quality of
public governance to inform some of its parameters. In this chapter,
we investigate how public expenditure can impact the efficiency
of government programmes when those parameters are part of the
endogenous development indicators. We perform this study by devel-
oping a novel conceptualisation of the expenditure channels that can
affect public governance, distinguishing between relative and total
changes in the relevant spending. We also devise a simulation strategy
to characterise the policy landscape (across the different expenditure
channels) governments face when trying to reform public governance.
Roughness in this landscape quantifies the uncertainty of potential

Another random document with
no related content on Scribd:
Nell’ora stessa della notte all’altra estremità del palazzo vegliava il
Duca in convegno coll’astrologo Ebreo.
La camera ove essi stavano sorgeva a guisa di torre all’angolo
orientale della Rocca, e non si poteva colà pervenire che per mezzo
di un ponte coperto e chiuso, il quale veduto dal basso s’aveva
forma d’un arco altissimo che congiungeva due parti dell’edifizio.
Quella camera conteneva ogni specie di macchine, stromenti e
arnesi ch’erano stati sino a quell’epoca inventati per segnare la
misura del tempo, e per lo studio delle sfere celesti; era insomma un
osservatorio astronomico, quale si può immaginare ginare che fosse
al principio del secolo decimoquinto; e ciò che meglio caratterizzava
il tempo e le idee erano gli utensili alchimistici che si vedevano
ovunque frammisti a quelli che unicamente servivano alle operazioni
dell’astrologia.
Fra i quadranti, i lambicchi, i cerchii, le clessidre e i gnomoni,
distinguevasi sopra larghi sostegni d’oro un ampio globo stellato e
dipinto a figure d’uomini e d’animali. Il Duca lo aveva comperato per
ingente somma da un mercante saraceno, e pretendevasi fosse il
celebre Planetario arabico, stato mandato in dono dal Califfo di
Bagdad ad Abderamo re di Granata.
Una gran lampada rifletteva la sua viva luce su quel globo, di cui gli
anni avevano alquanto annerito lo splendido azzurro. Il Duca stava
seduto in atto attentivo, tenendo fisi gli occhi sul Planetario, mentre il
vecchiardo Elìa con una verga d’ebano nella destra, toccando i segni
rappresentanti lo zodiaco, andava spiegandogli i nomi, i moti, gli
influssi delle varie costellazioni, le quali erano ripetute in un grosso
libro ch’ei sosteneva coll’altra mano.
Un colpo dato al battitojo di bronzo di quella camera fece
sospendere le parole alll’Astrologo; il Duca porse orecchio, e avendo
udito succedersi due altri tocchi leggierissimi, quindi uno più risentito
— Entra — gridò con impazienza.
La porta s’aprì, ed avanzossi un uomo pressochè interamente
avvolto nel mantello; s’accostò al Duca e gli parlò all’orecchio.
Filippo Maria ai detti di colui mostrò prima sdegnarsi, poi sogghignò
fieramente; dopo pochi istanti di secreto colloquio tra loro, fecegli un
cenno, quegli uscì, e la porta si serrò di nuovo.
Elìa era intanto rimasto immobile cogli occhi sul suo libro, nella
lettura del quale sembrava interamente assorto.
«Proseguite, maestro (disse con calma il Duca). Non parlavate voi
delle stelle che compongono la coda allo Scorpione?
«In cauda venenum» — profferì lentamente il Filosofo israelita come
se ripetesse le parole che stava leggendo; poscia alzò la testa e
divisi sulle labbra i peli della bianca barba, ritoccando colla verga sul
globo la nera figura, proseguì in sua nasale cantilena — «Quest’è il
celeste Scorpio che s’abbranca al Sagittario e colla coda percuote la
Libra. Efraim Afestolett Mammacaton ne’ precetti del decimo mese,
insegna essere tre volte sette il numero degli effetti nefasti che piove
sul mondo questo freddo animale. Esso è propizio a chi annoda
occulte trame, e attenta colpi proditorii; siccome d’indole sua penetra
nelle case e sta celato presso le coltri ove ferisce nel sonno...
«Un mostro di tal natura, uno scarabeo avvelenito in sembianza
umana, abita presso di noi (disse interrompendolo e con subitaneo
rancore Filippo Maria).
«Non vi prendete di ciò pensiero (rispose l’Ebreo); quando la sua
traccia verrà scoperta tutti si affretteranno a schiacciarlo.
«Eppure non è così. Una donna lo accoglie, lo accarezza e si lascia
da lui aizzare contro di me (replicò il Duca misteriosamente, fatto più
truce nell’aspetto). Ma essi non sanno che queste mura s’infuocano
e fanno contorcere le membra ai traditori come se fossero collocati
sopra lastre roventi.
«Le tenebre non lo terranno lungamente avvolto. Guai se lo
scellerato si palesa!
«Io li conosco già i suoi delitti: essi sono troppo gravi (profferì Filippo
con feroce freddezza). Gettate per lui le sorti, o maestro, questa
notte medesima. Domani allo svegliarmi entrerete a riferirmi ciò che
avrà prescritto il destino; rammentatevi che attendo voi pel primo.
Elìa chinò il capo in segno d’obbedienza. Il Duca alzossi; poscia ad
una sua chiamata si spalancò di nuovo la porta, ed ei ne uscì
preceduto per le scale ed i corritoi da due paggi che recavano i
doppieri.
Da quanto fu detto colà è agevole comprendere che i progetti di
Macaruffo non erano rimasti ignoti. L’intrattenersi ch’ei faceva
soventi ora con uno, ora coll’altro dei capi delle antiche bande di
Facino; il trarli seco a convegno nei battifredi più appartati del
Castello mentre mostravasi taciturno e selvatico con tutte l’altre genti
di Corte, aveva eccitati i sospetti e destata la vigilanza della turba dei
delatori del Duca. Ogni suo passo fu quindi numerato, sorvegliate
diligentemente le sue azioni.
La notte susseguente a quella in cui avvenne il colloquio da noi
riferito, il Venturiero passando meditabondo sotto il portico che dal
cortile interno della Rocca metteva all’andito della torre, sentì
afferrarsi per un braccio. Rivoltosi riconobbe Scaramuccia, valletto di
confidenza del Duca, con cui aveva stretta conoscenza militando
insieme sotto le insegne del Conte.
«Rendi grazie a’ tuoi santi protettori ch’io t’abbia ritrovato — disse
pianissimo Scaramuccia traendolo in un canto dietro le spalle
dell’arco, fuori della lista di luce che mandava la lampada. Il
Venturiero con voce aspra rispose:
«Renderei grazie sì, ma quando potessi al tuo padrone....
«Zitto, zitto (proseguì l’altro) non è tempo da far parole. Ascolta. Se
fra poche ore non sei lontano le molte miglia da queste mura tu
finirai di mala morte. Hanno girato per te la luna, il sole e le stelle: il
tuo nome sta in mano al Giudeo, e la gola del pozzo in fondo alle
vôlte fu aperta e t’aspetta. Pensa a’ tuoi casi. Addio. — Ciò detto lo
lasciò frettolosamente e scomparve nell’ombra.
Macaruffo benchè non suscettivo di timidi pensieri e omai
indifferente ad ogni sventura, non dubitò a tale inaspettato
avvertimento, che in realtà la sua morte fosse stata ordinata da
Filippo Maria, sia per avere scoperto i di lui tentativi, sia per togliere
un amico fedele alla Duchessa. Quindi non volendo cadere vittima
invendicata dell’abborrito Visconte determinò di cercare salvezza
nella fuga.
Deposta ogni arma e tramutate le vesti, presso l’albeggiare potè
uscire inosservato dal Castello. Comunque grande però fosse il suo
pericolo rimanendo in queste vicinanze, non sapeva staccarsi dai
luoghi ove l’infelice sua Signora, serbando solo i titoli e le apparenze
della sovranità, gemeva prigioniera d’un inesorabile tiranno.
Per lunghi giorni andò errando nelle terre prossime a questa città, e
la notte accostavasi guardingo alla tremenda ducale dimora, spiando
se qualche lume apparisse nelle finestre dal lato occidentale della
Rocca, e s’affisava in quello come in una luce amica, consolatrice,
poichè sembravagli illuminasse la camera della Duchessa, ch’ei si
rappresentava assisa a quel mesto chiarore in atto pensivo e col
volto irrigato di lagrime. Chi potrebbe ridire quanta fosse la potenza
che l’immagine di lei esercitava su quell’anima, chiusa in ributtanti
spoglie, ma sì nobile e generosa che avrebbe con gioja, e senza
ch’ella pure il sapesse, sagrificata l’esistenza per procurarle un
istante di contento e di pace?
Dovette però convincersi alfine Macaruffo ch’era vano ogni tentativo
per rivederla, e sarebbe stata follìa l’intraprendere di sottrarla suo
malgrado alle mani del Duca. Pensando d’altronde che se si fosse
scoperto ch’ei s’aggirava quivi d’intorno avrebbe potuto far cadere su
di lei il dubbio che per suo mezzo tramasse congiure o tradimenti, si
decise con pena indescrivibile ad abbandonare questo suolo, e
riprese cammino verso la patria.
Allorchè calando da una delle Alpi che fiancheggiano il mare di
Liguria, distinse tra il verde della valle le torri del castello di Tenda,
vide il lago de’ palombi, e poco lungi scorse tra il folto degli alberi le
merlate mura del maniero de’ Gualdi, non dolci affetti si sollevarono
in lui con soave tumulto, non esclamò, non sorrise; solo un grave
sospiro uscì dal suo petto affranto dalla fatica e dalla doglia, e
s’asciugò due amare stille di pianto che gli caddero sulle arsiccie
guancie.
Visse colà inconsolabile, solingo.
Quando nelle paterne mura ribombò con terrore e desolazione
l’orrendo annunzio che Beatrice, dannata per scellerata sentenza dal
Marito, aveva lasciata la testa sul patibolo nel castello di Binasco, il
Venturiero quivi più non si rinvenne.
Alcuni giorni dopo apparve un Pellegrino in vicinanza al castello del
supplizio, e fu veduto starsi ogni notte immobile per lunghe ore,
pregando alla ferriata della cappella dei morti, ove i resti della
Contessa erano stati deposti. Nè andò guari che chiuse gli occhi
esso pure alla vita, e nessuno scoprì mai la sua storia o il suo nome.
Un Cadavere antico [7]
.... Orrendo e vero

Simulacro di morte!...
H.
Era il cielo cinericcio; oscurava. Ad ogni istante rendevasi più fosco il

colore delle alte mura della Basilica, che la Longobardica Regina
eresse in Monza al divino Precursore, e più visibile traspariva dalle
sue arcuate finestre il chiarore delle lampane solitarie.
Io camminava a lenti passi sotto l’atrio contiguo a quella vetusta
Chiesa attendendo l’un de’ custodi. Il silenzio universale, la tenebria
della sera che s’avanzava m’avevano reso mesto e meditabondo,
onde le impressioni della mia mente consuonavano all’intutto
coll’idea del lugubre oggetto che una viva curiosità mi aveva tratto
quivi ad ammirare. Venne alfine la guida recando un torchio acceso
e m’accennò di seguirla; le tenni dietro: ed essa arrestatasi ov’era
un’imposta alla parete, la spalancò ed offrì al mio sguardo un
cadavere, una specie di mummia, quivi serbata in una nicchia.
Fatto immobile, affisai avidamente gli occhi in quella salma antica, e
i molti gravi pensieri che sorsero in me, attutarono il profondo
ribrezzo che suol sempre assalire alla vista di umane carni
inanimate. Sta ritto quel cadavere, rigido, giallognolo: il diseccare de’
muscoli, de’ tendini, l’indurare delle cartilagini l’ha alcun poco
contratto e impicciolito, ma mantiene tuttavia intatte le forme,
conserva i denti, i capelli, le ciglia, e mostra illesa ovunque la cute,
fuorchè alla nocca d’un piede. Lo esaminai con tutta attenzione
facendo appressare mano mano il lume ad ogni sua parte, e provava
il mio spirito certo qual solenne diletto nel contemplare un corpo, che
senza bende egizie o balsami trinacrii, si sottrasse alle possenti
consuntrici leggi della natura; un corpo, che serbando per quattro
secoli le primiere sembianze, giunse da spente e trapassate
generazioni sino a noi come l’unico resto d’un gran naufragio sopra
ignoti lidi.
Quella salma non fu d’uomo volgare, poichè s’ascrisse anch’esso fra
i dominatori dei popoli. Oltre d’avere padroneggiato Monza, venne
salutato signore di Milano, e sebbene qui non tenesse il comando
che per brevissimo spazio di tempo, collegò milizie, impose tributi,
stampò monete, attribuzioni esclusive della sovranità. Portò il nome
di Estore, e vanta per padre Bernabò Visconte, principe temuto e
crudele, che perì di veleno nel Castello di Trezzo.
Estore non condusse una fiacca o codarda vita. Pugnò
possentemente contro i due duchi suoi cugini, Giovanni e Filippo
Maria, figli di Galeazzo. Seppe tenere il campo a fronte di Facino
Cane, di Lancillotto Beccaria e di Valperga ch’erano fra i più valenti
condottieri dell’epoca. A Filippo Maria contrastò con vigore il
possesso dello Stato; assalito in Milano da forte schiera d’uomini
d’armi, oppose nelle stesse contrade della città la più ostinata
resistenza. Cedette alfine e ritirossi in Monza nel di cui Castello
sostenne un lungo assedio, respingendo più volte gli assalti di tutto
l’esercito ducale. Per ordine di Filippo, che sterminato voleva un sì
audace rivale, non cessavano mai le offese contro le assediate
mura, le quali dall’assiduo lavoro de’ mangani e delle bricolle già
scoperchiate e cadenti in più luoghi, non offrivano ai difensori che
debole e incerto riparo.
Un giorno (era in settembre del 1412) Estore Visconte se ne stava
nel mezzo del cortile del suo Castello presso al pozzo ove si
abbeveravano i cavalli. Nel momento forse che all’udire le esterne
grida de’ nemici meditava sdegnato una tremenda uscita, o che
l’anima sua sorpresa dall’idea dell’instante periglio cominciava a
vacillare invilita, una grossa pietra slanciata con tutta veemenza
venne dall’alto e lo colse in una gamba che gli spezzò presso
l’attaccatura dei piede, onde cadette, e perdendo sterminata quantità
di sangue, in capo a poco tempo rimase esanime.
Eccolo innanzi a me coi segni del fatal colpo, che ha lacerate le
carni, infranto l’osso, e lasciò su quello le traccie del sangue
aggrumolato.
— O arido ed annerito carcame, tu dunque fosti un guerriero
d’intrepido cuore, e ricoperto di ferro vivesti tra l’armi e le battaglie?
Ah perchè schiavando i denti che serrò morte, non puoi narrare tu
stesso i fatti de’ tuoi giorni, o rivelare ai mortali i segreti delle tombe
tra cui sì lungamente dimorasti! Ma l’ironia ferale dei tuoi tratti è fisa
e impassibile, e nel rimirarti mi fai sentire più amaro e truce il
pensiero che mentre ogni oggetto vivente con somma rapidità
trapassa e si solve, un cadavere s’innoltra incorrotto verso le ridenti
età future.
FINE.
IL BACIO FATALE
....... Ei nell’amata
Donna s’affigge, ode uno squillo: il suono
Quest’è che serra le stridenti porte.
Un istante gli resta, un bacio invola
A quella fronte gelida, una croce
Alle sue mani impallidite, e come
Luce nell’aer per le mute logge
Inosservato e celere dispare.
Tealdi-Fores.
Chi ignorava la beltà di Evelleda, la prigioniera d’Oriente divenuta

sposa del cavaliero Unfredo de’ Rodis?... Dal lago alle Alpi tutta la
valle dell’Ossola risuonava delle lodi di lei e si portavano a cielo non
solo le avvenenti sue forme, ma le virtù e la dolcezza soavissima
dell’animo. Nel mirarla era un’estasi che infondevano in petto la
leggiadrìa e la nobiltà delle sue movenze, l’armonìa della voce che
serbava ne’ suoni alcun che di straniero e la luce celeste di che
erano animati i suoi sguardi. Oh! gli sguardi di Evelleda superavano
quanto mai l’immaginazione più ardente sa figurarsi d’incantevole e
d’angelico: quegli su cui quelle nere pupille si posavano con
tenerezza o con mesto sorriso provava in cuore un ineffabile
commovimento e sentiva circondarsi da un’aura più pura.
Questo fiore di bellezza era nato sotto altri soli e dalle falde del
Libano era stato trapiantato presso quelle delle Alpi. Il cavaliero
Unfredo valente di braccio quanto d’animo ardente e vendicativo,
offeso in cuore da secrete ingiurie, determinò sino dall’età sua
giovanile d’abbandonare la patria; radunò una schiera de’ più prodi
suoi vassalli Ossolani, fece voto di combattere per la liberazione di
Gerusalemme e raggiunse in Oriente l’esercito dei Baroni Crociati.
Ebbe parte nelle imprese più ardue e famose; venne ferito e si
ritrasse a Bisanzio sotto la protezione de’ greci Imperatori.
Ricuperata colà la salute e il vigore, tornò in Palestina ove
capitanando una parte dell’esercito prese d’assalto una ricca città
de’ Saraceni, di cui gli furono cedute in premio le spoglie. Egli
trascelse per sè le più preziose; abbandonò l’altre a’ soldati, e dei
vinti non tenne in suo potere che una donna bellissima fra tutte,
madre d’unica fanciulletta, vezzosa come l’amore, la quale fu trovata
dai guerrieri cristiani nel solitario harem custodita da due schiavi muti
e neri al paro della pece.
Vinta Nicea ed Antiochia, Unfredo, a cui le ferite benchè rimarginate
rendevano l’armeggiare penoso, volle far ritorno alle patrie terre, e
caricate su una nave Pisana le conquistate ricchezze, afferrò le
spiaggie d’Italia. Morì attrita dai lunghi affanni, anzichè toccasse i
nostri lidi, la bella prigioniera saracena, e il Cavaliero le rese meno
penosi gli ultimi istanti giurandole sulla croce che a lui segnava il
petto, che avrebbe con ogni studio vegliato al bene dell’orfana
fanciulla ch’ella abbandonava nelle sue mani.
Toccava questa appena il tredicesimo anno, nè altri che la propria
madre conosceva sulla terra che potesse intenderla, guidarla e che
le fosse di sostegno e d’aita. Vedendo la genitrice languire per
mortale angoscia gemeva profondamente, sinchè giunta al punto
estremo ne raccolse disperata l’ultimo sospiro e si dovette strapparla
a forza dalla fredda salma di lei.
Per lunghi giorni le sgorgò incessante un pianto inconsolabile: alla
fine però le tenere e più che paterne cure del generoso Unfredo le
ridonarono la calma; cessarono le lagrime d’irrigarle le pallide
delicate guancie, ed ei si dispose a condurla alla propria valle nelle
mura dell’avito castello.
La fama delle sue gesta lo avevano preceduto: accorsero i vassalli
esultanti ad incontrarlo ed ei ricalcò festeggiante dopo tanti anni di
lontananza l’antico ponte del suo fiume nativo. Nel guerresco
corteggio che lo seguiva attraevano gli sguardi di tutti i due schiavi
Etiopi abbigliati nella loro barbarica foggia; ma ciò che destava più
vivamente la curiosità generale era la fanciulla che sedeva sopra un
placido e bellissimo palafreno guidato a mano da un paggio,
ricoperta da fitto velo il quale l’avviluppava pressochè interamente.
Allorchè dopo molti mesi il dolore della perdita della madre fu
alquanto più mitigato nell’animo della giovinetta, Unfredo che sentiva
nascere in seno per lei ardentissima fiamma, la fece istruire nei sacri
misteri di nostra religione e poscia rigenerare nelle acque del
Battesimo. Profuse quindi tesori per rendere il proprio castello il più
sontuoso che mai si vedesse e per prevenire ed appagare ogni lieve
brama dell’adorata fanciulla, un di cui sorriso lo rendeva felice.
Riconoscente essa pure a tante affettuose dimostrazioni del suo
guerriero vincitore, benchè non lo amasse che quale amoroso padre,
cedere dovette alle lunghe ripetute istanze e condotta da Unfredo
all’altare con pompa regale divenne sua sposa.
Sorgeva il castello di Unfredo sulle sponde della Toce là dove questo
fiume abbandonati i nativi dirupi, scende limpido e tranquillo ad
irrigare l’esteso piano della valle dell’Ossola. Il ponte levatojo di quel
castello rimaneva sempre abbassato, e sebbene numerosa schiera
d’armati vi stesse a guardia continuamente, erane però a tutti libero
l’ingresso, poichè colà venivano accolti con eguale cortese ospitalità
il povero pellegrino, il ricco barone, il questuante eremita e lo
sfarzoso Abate che vi giungeva cavalcando con gran seguito di
monaci e di laici. Infiniti erano quivi entro gli scudieri, i paggi, i servi,
tutti abbigliati con vaghe e ricche assise. Nei portici, negli atrii, sulle
scale miravasi scolpito in marmo o dipinto lo stemma della possente
famiglia de’ Rodis, ch’era una stella d’oro con due ali in campo
azzurro, circondato da una nera fascia.
Le stanze superiori nelle quali abitava il Signore del castello erano
tutte magnificamente addobbate; ma ove si poteva dire veramente
esausto quanto mai il lusso de’ tempi sapeva creare di più
sorprendente e ricercato, era la grande aula di ricevimento e
l’oratorio di Evelleda. Nella sala entravasi per due ampie porte alle
quali corrispondevano vaste finestre, divisa ognuna in due archi
acuti sostenuti da sottilissima colonna spirale: ne chiudeva il varco
una vetriata a colori su cui si diramavano simetrici arabeschi. Le
pareti erano coperte da purpurei arazzi trapunti in oro: marmoreo era
il pavimento ed istoriata la volta: i larghi sedili finamente intagliati, e
sulle tavole, ricoperte di lastre di preziosi marmi, posavano gemmati
doppieri. Sulla parete frammezzo alle porte d’ingresso stavano
sospese a modo di trofeo le armi più ricche d’Unfredo: nel mezzo era
collocato l’usbergo coi guanti, i bracciali e gli schinieri; a sinistra lo
scudo collo stemma rilevato a cesello; a destra la spada e la lancia,
ed al di sopra l’elmo di massiccio argento con cimiero d’altissime e
candide penne.
Quell’appartata camera che nella dimora d’una ricca dama viene a
lei unicamente consacrata e sta presso la stanza di riposo, servendo
così ai misteri dell’addobbamento, come alle solitarie letture ed alle
meditazioni, la quale ora noi chiamiamo Gabinetto, appellavasi nei
bassi tempi Oratorio, poichè conteneva una specie di domestico
altare avanti a cui soleva la Dama profferire le serali e mattutine
preghiere. L’oratorio d’Evelleda non era spazioso ma rinserrava
tesori. V’avevano due entrate, l’una da una porta che s’apriva
nell’atrio vicino alla sala, e l’altra più ristretta che riusciva nella
camera contigua ove era eretto il talamo nuziale. Di contro all’arcata
finestra d’egual forma di quelle della sala, stava nell’oratorio una
nicchia, dentro la quale sorgeva sopra un piedestallo il simulacro
della Vergine col divino infante, coronati l’uno e l’altro di un serto di
gemme: sul petto della celeste Madre pendeva appeso ad un serico
nastro un’anforetta in un cerchio d’oro che conteneva un frammento
del velo di Lei, reliquia rarissima acquistata per cento bisanti dallo
stesso Unfredo in Palestina da un Maronita di Betlemme. Davanti al
simulacro stava un ginocchiatojo tutto rivestito da ricco e morbido
drappo. In giro alla camera vedevansi arche ed armadietti d’ebano e
d’avorio, elegantemente intarsiati con fili d’oro e tempestati di pietre
preziose: alcuni di essi rimanendo aperti, mostravansi ripieni di vasi
lucenti, di cassette d’aromi, di odorosi unguenti; altri di fermagli
d’oro, vezzi di perle, spille, colanne, braccialetti e di quanto può
concorrere ai più sontuoso e variato femminile adornamento. Le
seggiole andavano ricoperte di velluto azzurro frangiato in argento, e
ad una di esse co’ bracciuoli, i quali avevano la forma di morbidi colli
di cigno, pure d’argento, stava dinanzi un tavoliere su cui posava un
vaso di cristallo cilestrino con fogliature in oro che conteneva i più
vaghi fiori, e vicino v’erano varj libri in pergamena con leggiadre
miniature. Da un lato del tavoliere stava un tripode in bronzo con
coperchio a traforo che serviva ad ardere profumi, dall’altro lato eravi
un elegante leggìo a cui stava sospeso un arpicordo saracinesco
con bischeri d’oro. Dalla volta pendeva una lampada alabastrina
sostenuta da tre catene in figura di serpi. La luce che dalla finestra
entrava in quella camera era mitigata a piacere, poichè le ampie
tende bianche e turchine che la fiancheggiavano potevansi
variamente panneggiare, ed ora si simulava con esse il soave
chiarore dell’aurora, ora la luce moribonda del crepuscolo e per sino
il bianco irradiare della luna.
Varia poi e spaziosa era la veduta che s’appresentava da quella
finestra, se ne venivano spalancate le imposte. Vedevasi l’intera
corona degli alti monti che formano parete alla valle, e tutta la
chiudono fuorchè a mezzodì ove ne lambiscono il confine le acque
del Lago Maggiore; miravasi più da presso la merlata roccia di
Vogogna eretta sopra scoscesa rupe, e scorgevasi nel piano il lucido
esteso serpeggiare della Toce che toccava mormorando a quelle
mura. Al di là del fiume quasi a prospetto sorgeva un edificio di
semplice architettura ma che s’aveva del castello insieme e del
convento: constava di massiccie mura, aveva porte e finestre ad
archi acuti, ma non era merlato nè munito di torri. Tale edificio
chiamavasi la Masone ed era ospizio de’ cavalieri Templari, i quali
solevano ivi stanziare ogni qual volta recavansi in Francia o ne
redivano.
Prediletto ad Evelleda era quell’oratorio ed ella passava in esso le
più lunghe ore del giorno o con qualche fida ancella occupata ai
lavori della spola e dell’ago, o da sola leggendo i canzonieri degli
amorosi Trovatori, o traendo dalle corde melodiosi suoni. Talvolta
nell’ora più tacita della sera ella univa a que’ suoni la sua voce:
arrestavansi negli atrii i paggi ed i donzelli ad ascoltarla, sospendeva
il passo per fino il rude arciero che stava a guardia a piè delle mura.
Eravi in quel canto un non so che di nuovo che rapiva, era una
melodìa ispirata da un altro cielo, da una più ridente natura.
Il raggio candidissimo della luna brillava sulle acque del fiume, ed
illuminava la fronte della Masone dei Templari. Ritto nel varco
dell’arcuata porta si stava uno dei guerrieri dell’Ordine appoggiato
alla sua lunga spada; la bianca sopravveste eragli serrata ai lombi
dal pendone della spada stessa, e in mezzo al suo petto si scorgeva
un’ampia croce rossa. Teneva scoperto il capo, il quale aveva da
nera inanellata capellatura rivestito, bruno e regolare era il giovanile
suo viso. In atto mesto e pensieroso lasciava errare le pupille ora
sulle correnti acque, ora sulla pallida verdura, ed ora le alzava al
disco della luna. Ad un tratto un irrompere di dolcissime note tratte
da sonoro stromento gli ferisce l’orecchio; guarda al castello di
prospetto da cui quel suono partiva e quasi tratto da magica forza
s’accosta alla sponda del fiume, onde meglio bearsi in quell’armonìa.
S’alza una voce... ma qual gioja inaspettata, qual soave sorpresa
manifesta il Cavaliero del Tempio!... quella voce canta
nell’armoniosa lingua dei poeti dell’Alambra, essa ripete gli accenti
che richiamano al Yemen felice la memoria dell’avventuroso
guerriero. Ecco come canta quella voce celebrando il suolo nativo.
«Mia sfera è l’Oriente, splendida regione, ove sorge magnifico

il sole come un possente monarca e procede per le vie del
giorno sempre serene: così una nave d’oro voga sull’onde
azzurre portando l’Emiro di vasta contrada.
«I doni tutti del cielo furono versati sulla zona orientale: in
ogni altro clima il fatale destino fa germogliare amari frutti a
lato ai saporosi. Ma Iddio che guarda sorridendo le terre
dell’Asia, la riveste de’ fiori più puri e accorda maggiori stelle
al suo cielo, maggiori perle al suo mare.
«Quivi sono le ampie città che l’universo ammira. Laora dai
campi fiorenti: Golconda, Cascemira, Damasco la guerriera,
la reale Ispahan; Bagdad da baluardi coperta come da ferrea
armatura, e Aleppo il mormorìo delle cui immense contrade
sembra al lontano pastore il fremito dell’oceano.
«Misora è qual regina collocata sul trono. Medina dalle mille
torri irte d’aguglie colle punte d’oro rassembra al campo d’un
esercito nel piano che inalza sulle tende una selva di lucicanti
saette.
«Chi non brama contemplare sì grandi maraviglie? Chi non
desìa sedere su quei terrazzi simiglianti a canestri di fiori; o
seguire nei prati l’Arabo vagabondo? Al cader del sole
quando i cammelli s’arrestano spossati presso le fresche
acque dei pozzi, la giovinetta bajadera intreccia la sua danza
voluttuosa.
«Anch’io un giorno con passi infantili errando pensosa presso
al chiosco solitario sotto i rami delle palme beveva l’aure
imbalsamate che scendevano dagli azzurri monti! Ma ohimè!
io non potrò mai più rivedere nè le palme, nè quei monti
quantunque la mia anima voli incessantemente alle beate
regioni orientali.
Armando di Nerra, tal era il giovine Cavaliero, fu scosso da quel

canto sin nell’intime fibre del cuore. L’oriente era pure il suo sospiro:
in oriente egli aveva appreso ad amare; quando l’oggetto de’ suoi
deliri perì, egli da libero combattente divenne Cavaliero dell’ordine
del Tempio, consacrando sè stesso e la sua spada alla Religione ed
assoggettandosi ai voleri del gran Maestro.
Attese ansiosamente la sera successiva: una melodia parimenti
soave lo venne dal castello a beare sulla sponda della Toce.
L’incanto fu irresistibile. Seppe chi era Unfredo, lo riconobbe ed
entrò nel suo castello da lui stesso accoltovi ed onorato.
Unfredo era oltre modo bramoso che distinti personaggi
contemplassero il lusso e la magnificenza da lui spiegati entro le
proprie mura; e siccome andava superbo di possedere una
bellissima sposa, gioiva che venisse ammirata ed elevata a cielo da
tutti: fiero e contento che gli altri invidiassero a lui quella beltà
famosa, a lui già d’età provetto, a lui d’ispidi lineamenti, a lui che
giovane in quella patria aveva dovuto subire l’umiliazione d’un rifiuto
quando pretese alla mano di donzella uscita da un lignaggio ch’ei
stimava paro al suo. Aveva abbandonata la terra nativa giurando di
vendicarsi di quel disprezzo o morire: e la sua vendetta era completa
quando alcuno proclamava non esservi nell’Ossola castello più ricco,
nè sposa più leggiadra di que’ d’Unfredo. Raggiante di gioja, dopo
avere fatto osservare gli atrii fastosi e le stanze più addobbate;
condusse il giovine Templario nella gran sala ove fece dare annunzio
ad Evelleda di presentarsi.
Esiste un’arcana relazione fra i diversi sentimenti dell’uomo, per cui
allo svilupparsi di un solo, più altri s’intraveggono con secreto
presentimento. Armando di Nerra al primo mirare avanzarsi dalla
spalancata porta la Dama del castello, sentì con certezza che da
nessun altri che da lei sola potevano essere partite quelle
maravigliose note che avevano richiamate tante dolci e dolorose
memorie al suo spirito. Unfredo nominò alla moglie il Cavaliero,
magnificandolo per la nobiltà del sangue e le illustri sue gesta. Ella lo
salutò con sorriso gentile, e allorchè si fu assiso in prossimità di lei e
del marito, le chiese se recavasi allora nei campi della Palestina o ne
retrocedeva. Rispose il Cavaliero che di là veniva e ritornava nelle
sue terre di Francia per riabbracciare il padre cadente, che più non
aveva veduto dal giorno che s’indossò la bianca sopravveste dei
Templari.
— Oh voi felice (esclamò con trasporto Evelleda), che avete la bella
sorte di ricalcare quel suolo ove apriste gli occhi alla luce
coll’indiscrivibile consolazione di esservi atteso dall’autore dei vostri
giorni! Quanti e quanti hanno posto il piede fuori della patria terra e
non la rivedranno mai più! —
Queste ultime parole furono pronunciate con tutta l’espressione della
soavità e della melanconìa, ed Armando assorto nel contemplare
quel volto e quell’angelico sguardo che s’abbassò con tristezza, vi
lesse la storia della profonda piaga d’un cuore senza amore e senza
speranze. — Oh figlia di una terra prediletta dal sole, perchè non ho
io pel tuo spirito languente un balsamo più dolce del frutto della
palma, più del ditamo fragrante? — così susurrò a bassa voce in
favella orientale il giovine Cavaliero e una gioja inaspettata si diffuse
sul volto alla bella. Ma Unfredo s’alzò, onde fu forza ad Armando
seguirlo, e ad Evelleda ritirarsi nelle proprie stanze.
Chi può descrivere i sogni d’una mente colpita dallo spettacolo
incantatore della bellezza, d’una bellezza mesta e pensierosa a cui
si sente il potere d’infondere nel cuore il sorriso della felicità? A tale
immagine la fantasìa vagando fra il sereno e le rose, dà forma alle
beatitudini eterne e si crede la favorita del cielo. Ahi troppo
ingannata! poichè non sa che il destino alla coppa dei beni aggiunge
irremissibilmente quella delle più crudeli amarezze.
Unfredo accolse più volte Armando nel proprio castello, sicchè
questi divenne famigliare a segno, che pure allorquando il Signore
n’era assente, o per sedere nel consiglio dei capi della valle o per
seguire le alpestri caccie, entrava liberamente tra quelle mura e vi
stanziava a suo talento.
Fragile è l’uman cuore e troppo possente incanto esercitano su di
esso le grazie, gli amorosi sospiri e le dolci animate parole che
giungono sommesse all’orecchio con un alito fragrante, carezzevole
e quasi affannoso, allora quando si è liberi da ogni altro umano
sguardo e il sole stesso non manda che timido il suo raggio a
traverso i cristalli colorati e le tende!
Nell’oratorio di Evelleda troppo felici scorrevano le ore per lei, per
Armando. Allorchè essa s’accompagnando con flebili o lieti suoni
cantava a piana voce canzoni piu tenere dell’usato, il Cavaliero in
un’estasi voluttuosa stava immobile contemplandola, sorreggendo il
mento col braccio appoggiato sul morbido velluto del sedile di lei, e
quando egli narrava le proprie imprese o ripeteva le novelle apprese
sulle rive del Giordano dagli Arabi pastori, ella pendeva beata dalle
labbra di lui, immemore per sino di sè medesima.
L’invidia rabbiosa e la vigilante maldicenza non concessero però che
lunghi corressero quei giorni di felicità. Tronche parole, maligni
sorrisi, inattese interpellazioni stillarono il veleno della gelosia nel
cuore d’Unfredo: si fece cupo e taciturno, parlava rado e sulla moglie
più non alzava che severo lo sguardo. Intese tremando Evelleda la
giusta causa di quel cangiamento, e risolvette di sacrificare anche sè
stessa al proprio dovere. Da fido messo fece recare un foglio ad
Armando in cui dicevagli «Non doversi essi rivedere mai più; averli il
cielo riuniti un istante per disgiungerli per sempre; solo giurava che
vivrebbe nell’anima sua eternamente la memoria di lui, che pregava
non dimenticarsi d’una infelice per la quale erano estinti tutti i beni
della terra.
La disperazione s’impossessò d’Armando. Il pensiero di non più
rivedere Evelleda era per lui tremendo come quello della morte: egli
sentiva di non potersi staccare da quei luoghi, di non poter
sopportare la vita se non le porgeva un addio, un ultimo addio, e se
non udiva ripetere dalla bocca stessa di lei il giuramento di
mantenere sempre impressa nel seno la sua immagine. Fece ogni
cosa disporre per la propria partenza, e messo frattanto uno
scudiero in agguato, quando seppe che Unfredo erasi allontanato a
cavallo dal castello, ei vi si recò e penetrò nell’oratorio di Evelleda.
Scorse però breve spazio di tempo da che egli aveva posto piede in
quelle soglie e già Unfredo, benchè discosto, n’aveva avuto avviso:
rivolge a furia il destriero, galoppa per una via fra’ boschi, rientra nel
castello e sale nella camera di riposo di Evelleda, da cui a passi
sospesi s’affaccia alla porta dell’oratorio, e vede... oh che vede egli
mai!... Il Cavalier del Tempio, un ginocchio a terra innanzi ad
Evelleda, con ambe le proprie mani premevasi al cuore una mano di
lei, ed essa seduta e colla faccia inclinata verso la sua lo inondava
singhiozzando di lagrime e faceva forza per rilevarlo.
A sì tenero spettacolo la pietà imbrigliò il furore, e le dita di Unfredo
rimasero un momento arrampinate al pugnale senza trarlo dalla
vagina. Ma ohimè! non fu che un lampo: una crescente foga d’affetti
vinse gl’incauti amanti, le loro labbra s’accostarono, s’unirono ed
essi si perdettero in un bacio di delirio... Era il primo... e fu celeste
quanto fatale. Il pugnale d’Unfredo s’infisse fino alla guardia nel
cuore d’Armando, Evelleda acciecata con un ferro rovente perì fra gli
spasimi: ruina e desolazione regnarono in quel castello dal quale
Unfredo disparve senza che più traccia si trovasse di lui.
FINE.

Complexity Economics and Sustainable Development A Computational Framework For Policy Priority Inference Omar A Guerrero Full Chapter PDF

Uploaded by

Copyright:

Available Formats

You might also like

Complexity Economics and Sustainable Development A Computational Framework For Policy Priority Inference Omar A Guerrero Full Chapter PDF

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

Complexity Economics and Sustainable Development A Computational Framework For Policy Priority Inference Omar A Guerrero Full Chapter PDF

Uploaded by

Copyright:

Available Formats

Complexity Economics and Sustainable

Fiscal Policy for Sustainable Development in Asia-

Visions and Strategies for a Sustainable Economy:

Complexity Economics: Building a New Approach to

Towards the Next Revolution in Central Banking: A

Complexity Economics: Building a New Approach to

Bus Transportation: Demand, Economics, Contracting, and

Information as a Driver of Sustainable Finance: The

Water and Climate Change: Sustainable Development,

omar a. guerrero is the Head of Computational Social Science

gonzalo castañeda is Professor in Economics at CIDE, Mexico

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of Cambridge University Press & Assessment,

Published online by Cambridge University Press

Published online by Cambridge University Press

List of Figures page xiii

PART I A COMPLEXITY APPROACH TO

2 Policy Prioritisation, Complexity, and Agent Computing 22

Published online by Cambridge University Press

3 Relevant Data and Empirical Challenges 53

4 A Computational Model 104

5 Calibration and Validation 129

Published online by Cambridge University Press

5.4.1 Conﬁdence Intervals 140

PART II A GLOBAL VIEW OF SUSTAINABLE

6 The Feasibility of the Sustainable Development Goals 167

7 Government Spending and Structural Bottlenecks 188

Published online by Cambridge University Press

7.3.3 Not All Bottlenecks Are Created Equal: A

8 Public Governance and Sustainable Development 207

9 The Impact of International Aid 231

Published online by Cambridge University Press

PART III A FOCALISED VIEW OF SUSTAINABLE

10 Subnational Development and Fiscal Federalism 265

11 Accelerators and Systemic Bottlenecks 289

Published online by Cambridge University Press

12 Deprivation, Income Shocks, and Remittances 323

13 Lessons and Reﬂections 350

Published online by Cambridge University Press

2.1 Factors affecting the inference of policy prioritisation. page 26

Published online by Cambridge University Press

6.4 Change in convergence time to the goals by country

Published online by Cambridge University Press

11.2 Systemic bottlenecks 313

Published online by Cambridge University Press

3.1 Indicators sampled from the Sustainable

Published online by Cambridge University Press

Recent global events have emphasised the uncertain and intricate

https://doi.org/10.1017/9781009022910.001 Published online by Cambridge University Press

achievement of the Sustainable Development Goals by 2030. Having

https://doi.org/10.1017/9781009022910.001 Published online by Cambridge University Press

—Luis F. López Calva

This book by Professors Guerrero and Castañeda is an important con-

https://doi.org/10.1017/9781009022910.001 Published online by Cambridge University Press

is similar: the methodologies used by development economists may

https://doi.org/10.1017/9781009022910.001 Published online by Cambridge University Press

ﬁrms to ﬁsheries, housing to Hayek, the formation of economic