Bernard C. Beaudreau - Science and The Wealth of Nations-Cambridge Scholars Publishing (2021)

Science and the Wealth
of Nations
Science and the Wealth
of Nations:
The Physics of Economic Growth
By
Bernard C. Beaudreau
Science and the Wealth of Nations: The Physics of Economic Growth
By Bernard C. Beaudreau
This book first published 2021
Cambridge Scholars Publishing
Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library
Copyright © 2021 by Bernard C. Beaudreau
All rights for this book reserved. No part of this book may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording or otherwise, without
the prior permission of the copyright owner.
ISBN (10): 1-5275-6581-5

ISBN (13): 978-1-5275-6581-4
My own sense is that New Growth Theory never really had the elements
needed to turn it into an intellectual success story; too much of it involved
making assumptions about how unmeasurable things affected other
unmeasurable things. In short, it was an intellectual bubble that eventually
deflated of its own accord.
—Paul Krugman, The New Growth Fizzle
TABLE OF CONTENTS
List of Tables and Figures ......................................................................... xi
Preface ...................................................................................................... xv
Introduction ............................................................................................ xvii
1 The Enigmatization of Economic Growth ............................................... 1

1.1 Introduction ..................................................................................... 1
1.2 Consilient Formalizations of Growth .............................................. 2
1.3 Literature review: steady-state growth models ............................... 4
1.4 The Exceptions: Mainstream Incursions into the Material
Sciences ......................................................................................... 14
1.5 Summary and Conclusions ........................................................... 15
1.6 References ..................................................................................... 16
2 Thinking About Production: From Smith to Romer .............................. 19

2.1 Thinking About Production Prior to the 18th Century ................... 19
2.2 Thinking About Production in the Newtonian Era........................ 20
2.3 Thinking About Production in the Thermodynamic Era ............... 24
2.4 Thinking About Production in the Electrical Era .......................... 26
2.5 Thinking About Production in the OPEC Era ............................... 32
2.6 Summary and Conclusions............................................................ 35
2.7 References ..................................................................................... 36
3 The Impact of Electric Power on Productivity: A Study of U.S.

Manufacturing 1950-84 ............................................................................ 39
3.1 Introduction ................................................................................... 39
3.2 The Marginal Product of Electric Power ...................................... 43
3.3 Accounting for the Sources of Growth in US Manufacturing
1950-84 ......................................................................................... 45
3.4 Electric Power Consumption and Productivity ............................. 46
3.5 Conclusions ................................................................................... 47
3.6 References ..................................................................................... 48
viii Table of Contents
4 The Energy-Organization Framework ................................................... 51

4.1 Introduction ................................................................................... 51
4.2 The Energy-Organization Approach to Modeling Material
Processes ....................................................................................... 52
4.3 A Physical Model of Production ................................................... 53
4.4 A General Model of Production Processes: The Energy-
Organization Framework............................................................... 57
4.5 The General Energy-Organization Production Function............... 60
4.6 Energy-Organization Productivity Measures ................................ 63
4.7 The Energy-Organization Approach to Modeling Production
Processes: The Evidence ............................................................... 65
4.8 Summary and Conclusions ........................................................... 71
4.9 Bibliography ................................................................................. 72
5 Engineering and Economic Growth....................................................... 77

5.1. Introduction .................................................................................. 77
5.2 Growth Processes: Engineering versus Economics ...................... 79
5.3 The Energy-Organization Framework: Theory and Evidence ...... 81
5.4 The Role of Capital and Labor in Growth Accounting ................. 86
5.5 Conclusions ................................................................................... 86
5.6 References ..................................................................................... 87
6 The Dynamo and the Computer: An Engineering Perspective

on the Modern Productivity Paradox ........................................................ 89
6.1 Introduction ................................................................................... 89
6.2 Production Processes as Seen by Engineers.................................. 91
6.3 Implications for the Modern Productivity Paradox ....................... 95
6.4 Conclusions ................................................................................. 100
6.5 References................................................................................... 101
7 The Physical Limits to Economic Growth by R&D Funded

Innovation ............................................................................................... 105
7.1 Introduction ................................................................................. 105
7.2 The Productivity Slowdown, Policy Responses and R&D
Expenditure ................................................................................. 107
7.3 Residual Growth and R&D ......................................................... 108
7.4 The Evidence .............................................................................. 108
7.5 Wealth: The Underlying Physics ................................................ 110
7.6. Estimates of Potential Increases in Energy Efficiency ............... 112
7.7 Commercial ................................................................................. 121
7.6. Results: The Upper Bound of R&D Energy Efficiency ............. 123
Science and the Wealth of Nations: The Physics of Economic Growth ix
7.7 Conclusions ................................................................................. 124

7.8 References................................................................................... 125
8 The Economies of Speed, KE=1/2mv2 and the Productivity

Slowdown ............................................................................................... 129
8.1. Introduction ................................................................................ 129
8.2. Speed, Energy Deepening and Control Technologies ................ 131
8.3 Model .......................................................................................... 135
8.4. Empirical Evidence .................................................................... 137
8.5. Further Evidence of the Law of Kinetics in Manufacturing ...... 143
8.6. The Limits to Speed, Control Technologies and the Productivity
Slowdown.................................................................................... 146
8.7 Case studies................................................................................. 150
8.8 Implications ................................................................................ 155
8.9. Summary and Conclusions......................................................... 159
8.10 Appendix ................................................................................... 160
8.11 References ................................................................................. 161
9 A Kinetics-Based Approach to Production: Theory and Evidence...... 165

9.1 Introduction ................................................................................. 165
9.2 Non-Mainstream Contributions .................................................. 167
9.3 The Critics: Big on Principles, Small on Specifics ..................... 172
9.4 Model .......................................................................................... 176
9.5 Production Activity Indicators .................................................... 182
9.6 Empirical Evidence ..................................................................... 182
9.7 Applications and Implications .................................................... 189
9.8 Summary and Conclusions ......................................................... 191
9.9 References ................................................................................... 192
10 Ideas That Do and Don’t Matter….For Growth ................................ 195

10.1 Introduction ............................................................................... 195
10.2 Innovations/Developments That Increased the Demand
for Speed ..................................................................................... 199
10.3 Innovations/Ideas that Don’t Matter ......................................... 201
10.4 Prognosis for Future.................................................................. 202
10.5 Exponential R&D Growth and Anemic Productivity Growth .. 203
10.6 Summary and Conclusions........................................................ 207
10.7 References ................................................................................. 207
Summary and Conclusions ..................................................................... 209

LIST OF TABLES AND FIGURES
Tables
Table 1.1 Material Processes, Energy and Organization ............................ 2
Table 1.2 Neoclassical and Other Enigmas ................................................ 5
Table 1.3 Output and Input Growth Rates: U.S., German and Japanese
Manufacturing ..................................................................................... 12
Table 1.4 Incursions into the Material Sciences: Missed Opportunities ... 13
Table 2.2 Key Developments in Thinking About Production .................. 36
Table 3.1 Manufacturing Output and Input Data 1950-84a ...................... 42
Table 3.2 Factor Income Shares, U.S. Manufacturing Value-Added ....... 44
Table 3.3 KLEP Regression Results, 1950-84' ......................................... 44
Table 3.4 Simple Growth Rates for Output and Inputs in US
Manufacturing, Selected Periods ........................................................ 45
Table 3.5 Changes in Labor Productivity, Total Factor Productivity, and
the Labor Intensity Ratios for the Pre-1973 and post-1973 Periods ... 46
Table 4.1 A Tool Taxonomy .................................................................... 57
Table 4.2 NPA Energy Productivity Measures ......................................... 64
Table 4.3 NPA Organization Productivity Measures ............................... 65
Table 4.4 KLEP Regression Results U.S., Canadian and Japanese
Manufacturing ..................................................................................... 68
Table 4.5 Kummel, Henn and Lindenberger’s Output Elasticities: U.S.,
Germany and Japanese Manufacturing ............................................... 68
Manufacturing ..................................................................................... 69
Table 4.7 Productivity Growth: U.S., German and Japanese
Manufacturing ..................................................................................... 70
Manufacturing ..................................................................................... 84
Manufacturing ..................................................................................... 85
Table 6.1 Contributions to Aggregate Productivity Growth of Industry ICT
Capital Deepening, EU-4 and USA .................................................... 97
Manufacturing ..................................................................................... 99
Table 7.1 Global R&D spending in 2011-billion U.S.$ ......................... 106
xii List of Tables and Figures
Table 7.2 World Energy Consumption by End Use Sector, 1990 .......... 112
Table 7.3 World Electricity Generation: Primary Energy Intensity
Relative to 1990 ................................................................................ 115
Table 7.4 World Transportation: Primary Energy Intensity Relative
to 1990 .............................................................................................. 116
Table 7.5 World Residential Primary Energy Intensity Relative to 1990 ... 119
Table 7.6 World Industrial: Primary Energy Intensity Relative to 1990 .. 121
Table 7.7 World Commercial: Primary Energy Intensity Relative
to 1990 .............................................................................................. 122
Table 7.8 World Primary Energy Intensity Relative to 1990. ................ 123
Table 8.1 Estimates of the Electricity-Use Output Elasticity-
Manufacturing ................................................................................... 139
Table 8.2 OLS-AR(1) Estimates-Log Linear Specification.................... 139
Table 8.3 OLS-AR(1) Estimates-Linear Specification ........................... 140
Table 8.4 OLS-AR(1) Estimates-Log Differences Specification ........... 140
Table 8.5 2-Digit SIC Industry Electric Power Output Elasticities
1947-1984 ......................................................................................... 142
Table 8.6 Multifactor Productivity Growth in Major U.S. Industries,
1948-1985 ......................................................................................... 145
Table 8.7 U.S. Annual Productivity Growth 1947-1980 ........................ 147
Table 8.8 Chronology of Maximum Fourdrinier Newsprint Paper
Machine Speeds ................................................................................ 155
Manufacturing ................................................................................... 160
Table 9.1 Manufacturing Processes and Corresponding Kinetics .......... 177
Table 9.2 Production per Factor Indexes-PFI ......................................... 182
Table 9.3 Estimates of the Electricity-Use Output Elasticity-
Manufacturing .................................................................................. 184
Table 9.4 OLS-AR Estimates-Log Linear Specification ........................ 186
Table 9.5 OLS-AR (1) Estimates-Linear Specification .......................... 186
Table 9.6 OLS-AR (1) Estimates-Log Differences Specification .......... 187
Table 9.7 2-Digit SIC Industry Supervision Input Elasticities
1947-1984 ......................................................................................... 188
Table 10.1 Second-Law Efficiency Increasing (increasing work per unit
energy) .............................................................................................. 196
Table 10.2 Innovations That Increased Energy Use Per Unit Capital Per
Unit Time .......................................................................................... 197
Table 10.3 Innovations/Ideas That Increased Energy Availability ......... 198
Table 10.4 Processes That Have Reached Their Kinetic Limit .............. 203
Science and the Wealth of Nations: The Physics of Economic Growth xiii
Figures
Figure 1.1 Actual and Predicted Growth of U.S. GDP 1960-1978 ........... 11
Figure 5.1 Structure of Production Processes ........................................... 79
Figure 6.1 Manufacturing Production Processes ...................................... 92
Figure 6.2 U.S. Net Stocks of Fixed Assets.............................................. 98
Figure 6.3 IT and non-IT Capital Services ............................................. 100
Figure 8.1 Productivity Growth Slowdown and Past Productivity
Growth, by Manufacturing Industry ................................................. 144
Figure 8.2 Productivity Growth by Manufacturing Industry
1948-1985 ........................................................................................ 147
Figure 8.3 Value-Added per kwh 1948-1980 by 2-Digit SIC Industry ... 149
Figure 8.4 Electric Power to Capital Ratios-Manufacturing Pre-1973 ... 151
Figure 8.5 History of Clock Rate and Power for Intel x86
Microprocessors ................................................................................ 156
Figure 10.1 U.S. R&D by Performing Sectors and Source of Funding
1953-2015 ......................................................................................... 204
Figure 10.2 Ratio of U.S. R&D to Gross Domestic Product by Source
of Funding 1953-2015....................................................................... 205
PREFACE
This volume is a response of sorts to the papers presented in a session at

the 2018 American Economic Association meetings in Philadelphia, entitled
“Domestic and International Dimensions of Innovation Spillovers” which
provided an update on the progress and new directions in New Growth
Theory, specifically in the area of innovation spillovers. The results were
sobering: after over four decades of idea-based growth theories and policies,
growth in Western industrialized democracies remains anemic and in many
instances is on the wane. Clearly, the time for alternatives is nigh. While
some have characterized the approach taken here as bold, it is our view that
the current set of ideas and theories regarding growth is bold in its
orthogonality to our understanding of growth in other material process-
based fields. In short, it is in keeping with the geist of the pluralism of ideas,
but firmly seated in the science of material processes and their growth.
INTRODUCTION
The ultimate irony of idea-based growth theory is the total

absence of new ideas about the growth process itself.
This quotation captures the purpose of the present volume, namely of

expanding the set of ideas regarding the growth processthat is, beyond the
current idea-based growth theory (also referred to as New Growth Theory).
As both Robert Solow and Paul Romer, two Nobel laureates, maintained,
economic growth is the result of new ideas. In short, economic growth is
increasing in capital, labor and new ideas. The more ideas there are, the
greater is/will be the rate of growth. Or so it is/was believed.
New ideas in the current era are the result of research and development
activity, formal and informal. Today more than ever, a greater and greater
proportion of worldwide GDP is allocated to R&D. For example, in 2017,
$1.7 trillion (USD) was spent on invention and innovation. Yet, despite four
decades of such spending, growth rates have remained and continue to
remain anemic. Clearly, something is amiss.
This raises the obvious question, namely in light of the poor growth
record of Western industrialized economies, why has the thinking about the
growth process itself not witnessed new ideas, new directions–that is, an
increase in the set of ideas. Or put more directly, why is it that an approach
that celebrates diversity, novelty and evolution (i.e. of ideas), appears to
have been unable to evolvespecifically, beyond a view that dates back to
the immediate post-WWII period?
Part of the answer, we believe, lies with the open-ended nature of the
associated growth process. In short, Solow- and Romer-based growth is a
function of known and measurable factor inputs (capital and labor) as well
as an open-ended factor input, namely the Solow residual. After WWII, a
handful of economists, including Edward Denison, Zvi Griliches, and
Moses Abramovitz, turned their attention to the growth process. Armed with
little more than the decades-old neoclassical production function, output,
labor and capital data (the fruit of Keynesian macroeconomic policy) were
used for the first time to study the growth process. Their findings were eye-
opening to say the least, as capital and labor, the two cornerstones of modern
production theory, were unable to explain the observed growth rates. This
left what Charles Cobb and Paul Douglas (Cobb and Douglas 1928) had
xviii Introduction
referred to as the A parameter, or technology as the leading cause. And so

was born the Solow residual, which as its name implies, is a residual factor
in the growth process. Heretofore, growth was attributed to new knowledge,
to new ideas, to new inventions, and to new innovations.
And so was born, modern growth theory with its emphasis on ideas and
knowledge. Fast forward to the early 1980s and the productivity slowdown,
which saw growth rates plummet, from an average of five percent to little
over one percent. This cataclysmic event raised a whole host of questions,
including why? Why did growth rates suddenly plummet? After all, they
had been robust for decades and there was no reason to believe that this
would not continue.
Far-reaching in its effects and after-effects, the productivity slowdown
mobilized the economics profession, drawing in the likes of Nobel laureate
and macroeconomists Robert Lucas and Peter Howitt. The task was as simple
as it was complex, namely understanding the Solow residual. In other words,
opening the black box of technology/ideas/innovation/invention. Long story
short, this resulted in a number of new approaches to innovation, from Nobel
laureate Paul Romer’s Ak model, to Philippe Aghion and Peter Howitt’s
formalization of Joseph Schumpeter’s notion of creative destruction.
If nothing else, the profession’s all-out, frontal attack on the problem of
slow-to-no growth in productivity served to refocus the public debate, not
to mention public policy. Soon, the growth slowdown became public enemy
No. 1, resulting in a new set of growth-increasing policies, from free trade,
to deregulation, to patent laws, to R&D, and to education. In short, all policy
guns were aimed at growth, with the hope of restoring post-WWII levels.
The results were less than spectacular–in fact, they were downright
disappointing. Four decades of pro-growth policies had failed to turn things
around, so to speak. Trillions and trillions were spent on R&D and
education, yet growth remained as elusive as ever. Romerites pointed to,
among other things, slowing population growth as a potential cause, while
Schumpeterites invoked various metaphors such as the low-hanging fruits
view according to which all high-growth ideas had been exploited leaving
only marginal ones.
The fact that the residual was and continues to be open-ended has, in our
view, prevented any progress on the bigger question of the growth process–
and indeed, understanding the underlying production processes–itself.
Today, the profession continues to view further refinements of the idea-
generating process as holding the key to fostering growth, past, present and
future. Interestingly, at no time has anyone so much as questioned the
wisdom of the underlying approach.
Science and the Wealth of Nations: The Physics of Economic Growth xix
It would be an understatement to point out that this whole episode, from

the productivity slowdown to the low growth rates that have become a
fixture in the West, has been a resounding failure. Clearly, ideas–lots and
lots of ideas–have not been and are in all likelihood not the key to restoring
growth. Which is why we feel that it is time to “go back to the drawing
board,” so to speak. Put differently, we feel that what is needed is/are new
approaches to understanding the growth process itself, ones that will, in the
end, result in a more diversified, a more multifaceted, a more consilient and
a more complete set of theories/approaches.
This is where this volume, entitled Science and the Wealth of Nations: The
Physics of Economic Growth, fits in, namely as a springboard for alternative
approaches to formalizing economic growth. In short, it consists of ten
chapters, each being a contribution in its own right. They range from
critiques of modern growth theory, to century-old attempts at refocusing and
updating production theory, the basis of modeling wealth creation, and thus
of growth theory.
The upshot of this volume is twofold. First, there is the reaffirmation of
the primordial role of fundamentals in the growth debate, specifically of the
physical underpinnings of material processes. Growth theory has, from its
inception, been plagued by weak fundamentals in the form of the post-
WWII neoclassical production function, which was and is little more than
formalized correlations, with no basis whatsoever in classical mechanics or
applied physics.
The second is the irrelevance of the very definition of new ideas as found
in New Growth Theory, a definition that blurs the line between process-
based new ideas and product-based ones. Theoretically, the new products
referred to by Romer and others will alter the makeup of GDP without,
however, increasing its overall level. In short, emphasis should be on
process-based innovation.
The take-away from this volume is relatively simple and straightforward,
namely that economic growth, like the growth of all other known material
processes, obeys a set of laws that are immutable, and which place
force/work/energy at its very core. Conventional factors inputs, labor and
capital, are not sources of work, but rather serve to define and organize
energy-using material processes. Furthermore, new process-related ideas
must necessarily conform to these same immutable laws. To believe
otherwise is to violate the laws of physics, something the profession has
been guilty of, and unfortunately, continues to be guilty of.
The proof of the pudding is in the eating. And in the case of this volume,
the proof of non-idea-based growth lies in its ability to rationalize the
various puzzles that have come to plague the literature, including the
xx Introduction
productivity slowdown and the information paradox. Both are shown to

result from basic physics, with the former being the result of the law of
kinetics, and the latter, the result of the very nature of information as being
non-physically productive.
Chapters-Brief Summary
This section briefly describes each of the ten chapters.
1-The Enigmatization of Economic Growth. Real World Economic

Review. 86. 2018:92-105
This chapter sets the tone for the rest of the volume. It points to what the
author views as the increasing enigmatization of economic growth, a
process whereby each generation has, in response to the failure of previous
generations to understand the underlying mechanisms, proceeded to further
enigmatize the process in the hope of piercing the underlying mystery. The
author goes from the classical production theory of Adam Smith, to the
neoclassical production theory of Marshall, Jevons and Walras, to the Solow
residual of the 1950s, and finally to the New Growth Theory of the 1980s.
He documents the emergence in the early 20th century of a radical school
of thought, one that sought a rapprochement between economics and
physics, and one that was eschewed by the mainstream, only to be
rediscovered in the 1980s and 1990s.
2-Thinking About Production: From Smith to Romer

While energy continues to be conspicuous by its absence in economic
theory, there have been, over the course of the past two hundred years, a
number of attempts by both mainstream and non-mainstream writers, to
incorporate it into the analysis. This chapter provides an account of these
attempts, starting with Adam Smith’s reference to “fire power” and ending
with the myriad contributions in the 1920s and 1930s, a period marked by a
manifold increase in energy consumption.
The chapter is organized around a number of themes, including
mainstream and non-mainstream, academic versus professional economists,
outside influences/critics, and the impact or lack thereof.
3-The Impact of Electric Power on Productivity: The Case of U.S.

Manufacturing 1950-1984. Energy Economics. 1995 17, no. 3: 231-236.
In 1995, Bernard Beaudreau estimated a modified KLEMS model, the
KLE model, using data on U.S., German and Japanese manufacturing from
the 1950s to the 1980s. However, unlike previous work, the output
Science and the Wealth of Nations: The Physics of Economic Growth xxi
elasticities were estimated directly, as opposed to indirectly. These were

then used to explain growth in the post-WWII period, the results of which
were conclusive: when energy was formally incorporated into the analysis,
the Solow residual literally disappears.
4-The Energy-Organization Approach to Modeling Material Processes

in Energy and Organization: Growth and Distribution Reexamined.
Westport. CT, Praeger 1998.
According to Chapter 1 of Energy and Organization: Growth and
Distribution Reexamined, the basic underlying premises of both Kummel
(1982) and Beaudreau (1995) stood in direct violation of the laws of
classical mechanics. For one, according the classical mechanics, what
economists refer to as capital (plant and equipment) was not physically
productive. The same was true of labor since the dawning of industrialization.
Workers were machine operatives or supervisors, not being sources of work
(force).
This led the author to develop a consilient approach, namely the energy-
organization model according to which there are two universal factor inputs,
namely the broadly-based energy input, and the broadly-based organization
input, with the former being physically productive, and the latter organized
the former. Not only was the result consistent with growth as understood in
other material sciences, it answered the key question of the day, namely why
did growth slow down?
5-Engineering and Economic Growth. Structural Change and

Economic Dynamics. 16, no. 2. 2005: 211-220.
This chapter reexamines the debate over economic growth from an
engineering perspective, arguing that from the productivity slowdown
onwards, the economics profession has isolated itself and attempted to
tackle the question of growth without any input from related fields whose
main purpose it is to understand the mechanics of material processes. The
upshot is that when an engineering approach is adopted, many of the puzzles
and conundrums disappear, including the mystery surrounding the
productivity slowdown.
6-The Dynamo and the Computer: An Engineering Perspective on the

Modern Productivity Paradox. International Journal of Productivity
and Performance Management. 59, no. 1. 201: 7-17.
In the slipstream of the productivity slowdown was the debate over
computers and computed and their role in raising productivity, and
ultimately, restoring post-WWII levels of growth. In 1987, Robert Solow
xxii Introduction
went on record raising what he felt was a paradox, namely the ubiquitous
presence of computers everywhere from manufacturing plants to homes, but
their apparent absence in productivity statistics. Put differently, the much-
anticipated growth from computing had failed to materialize.
Using the energy-organization approach, he argues that because
information is not a source of force/energy, it cannot increase productivity,
thus rationalizing the information paradox as a prediction of the underlying
principles of basic physics.
7-The Physical Limits to Economic Growth by R&D Funded

Innovation. Energy, 84. 2015: 45-51 (with Douglas Lightfoot).
There are ideas and there are ideas. This statement summarizes Bernard
Beaudreau and Douglas Lightfoot’s Energy article entitled The physical
Limits to Economic Growth by R&D Funded Innovation, which examines
in detail the set of ideas that could potentially increase growth via their
impact on second-law efficiency–that is, on energy efficiency. A good
example is James Watt’s external condenser which increased the efficiency
of the atmospheric steam engine by 200 percent, thus increasing the amount
of useful work per ton of coal.
Over the course of the past century, many innovations have contributed
to increasing second-law efficiency. In this work, the authors examine, on
an industry basis, potential increases based on industry data. They come to
the conclusion that at best, second-law efficiency-related innovations can
increase the average rate of growth by 0.75 per year.
8-The Economies of Speed, KE=1/2mv2, and the Productivity

Slowdown. Energy. 124. 2017: 100-113.
For roughly two decades, the accepted view among those in the energy-
growth sub-field held that higher energy prices in the 1970s and 1980 had
reduced the rate of growth of energy consumption, thus decreasing the rate
of economic growth. Bernard Beaudreau, in his 2017 Energy article entitled
“The Economies of Speed, KE=1/2mv2, and the Productivity Slowdown,
challenged this view, pointing to various inconsistencies, notably the fact
that when energy prices returned to their pre-OPEC levels, growth rates did
not rebound.
This led him to pursue a novel research program aimed at better
understanding the role of energy in material processes. The result was the
kinetics approach to production, based on machine/process speed. Drawing
from the historical record, he argued that rising energy consumption
throughout the industrial era translated into greater machine speeds and
thus, greater output/wealth.
Science and the Wealth of Nations: The Physics of Economic Growth xxiii
The end result was a framework that shed considerable light on events
such as the Great Depression and the Productivity Slowdown. In short, the
latter was attributed to the fact that maximum machine speeds had been
reached in most industries/sectors in the late 1960s/early 1970s. As such,
the lower rates of growth of energy consumption pointed out by Reiner
Kummel and others owed not to higher prices, but to kinetics, namely that
machine speeds could not be increased further. In short, the productivity
slowdown concorded with what is generally regarded as the “end of the age
of speed.”
9-The Kinetics Approach to Production: Theory and Evidence, 2018

As the title indicates, this article develops in more detail the kinetics
approach to production, an approach that combines the laws of physics with
the concerns of economists, engineers and managers. In addition to
formalizing the role of energy in production, it provides estimates of the
crucial parameters, the most notable of which is the output elasticity of
energy, whose theoretical value of 0.5 is confirmed in both aggregate and
disaggregate data. The result is a consilient approach to understanding both
the creation of wealth and its growth through time. It also provides a guide
to understanding a number of historical phenomena, including growth in the
1920s and 1930s, the Productivity Slowdown, as well as the Information
Paradox.
10-Ideas That Do and Don’t Matter…..for Growth.

In this chapter, it is argued that while ideas in general don’t matter, some
ideas do, notably those that pertain to increasing the amount of useful work,
, whether by increasing second-law efficiency or by increasing the amount
of energy consumed per unit tools/capital, the latter commonly known as
speed-ups. Implicitly, this provides a criterion for distinguishing between
ideas that do and don’t matter in so far as growth is concerned.
As such, product-related ideas, while prevalent in the world today, are
not growth increasing, as they do not contribute to increasing the supply of
useful work, but instead will apportion a share of the existing supply. For
example, the introduction of Sony’s new PS5 will take resources away from
its production of the PS4.
1
THE ENIGMATIZATION OF ECONOMIC

GROWTH
Abstract
In virtually every material science, the process of growth is not only well
understood, it has been systematically reduced to its thermodynamic and/or
kinetic equivalent. That is, growth is a function of resource and energy
availability, whether it be within a stationary or non-stationary environment.
This begs the question, why is economics the outlier, the exception? Why
are models of economic growth decoupled from the basic science of material
processes? This chapter attempts to answer this question by focusing on the
formalization of growth. It will be argued that for a number of reasons, the
economics profession has enigmatized material processes, introducing
concepts that were orthogonal to the laws found in the material sciences,
leading to the current situation where a whole new generation of enigmatic
approaches (quality ladders, institutions, etc.) has emerged to understand
previous engimas.
1.1 Introduction
In virtually every material process-based field/discipline, the process of
growth is not only well understood, it has been and is systematically reduced
to its thermodynamic and/or kinetic equivalent (biology, ecology,
demography). In short, growth is either a function of growth in energy
availability/use–whether it be within a stationary or non-stationary
environment–or an increase in second-law efficiency. The quintessential
example is photosynthesis where the growth of biomass is a function of
solar radiation, the latter being the force that acts on carbon dioxide and
water to produce carbohydrates/sugars. This begs the question, why is
economics the outlier, the exception? This chapter attempts to answer this
question by focusing on the very way in which the profession has formalized
material processes. It will be argued that for a number of reasons, the
economics profession, by enigmatizing a simple energy-based material
2 1. The Enigmatization of Economic Growth
process, has generated findings (i.e. the Solow residual) that have
prompted/led to the increased enigmatization of the growth process.
The chapter is organized as follows. To begin with, we present a
consilient approach to modeling material processes in general, namely the
energy-organization approach according to which output is increasing in
terms of two universal factor inputs, namely broadly-defined energy and
broadly-defined organization. This will then provide the basis for a
comprehensive review of the literature organized around two themes,
namely steady-state growth and non-steady-state growth (technology
shocks). This will be followed by a discussion of consilient approaches to
growth–that is, approaches that incorporate elements or aspects of physics.
We end with a set of external–read: scientific–guidelines for future work on
growth.
Table 1.1: Material Processes, Energy and Organization
Material Process Energy Input Organization Inputs

Chemical Processes Heat Kettles, Ladles
Manufacturing Kinetic Energy Simple and Complex
Processes Tools, Supervisors
Photosynthesis Solar Radiation Molecular Structure of
Raw Materials
Mitochrondria Glucose Molecular Structure of
Raw Materials
1.2 Consilient Formalizations of Growth

While this may come as a surprise to some, economics is not the only
scientific discipline in which growth and the growth process are integral
parts. In fact, essentially all material sciences focus on growth and hence
have developed analytical frameworks to describe and understand it. Take,
for example, biology, specifically plant biology which has modeled growth
in terms of photosynthesis, where solar radiation powers a series of
chemical reactions that result in the production of glucose. As in all other
material sciences, energy is the essential factor input. Unlike material
processes as studied by engineers, there are no tools (simple or complex)
involved. Similarly, unlike cell growth where the set of instructions is
contained in the organism’s RNA or DNA, there is no specific set of
instructions nor supervision. Table 1.1 presents a list of material processes
and the energy- and organization-related factor inputs.
Science and the Wealth of Nations: The Physics of Economic Growth 3
Beaudreau (1998) provided a consilient approach to understanding

material processes and growth in general–that is, across disciplines. The
energy-organization (hereafter EO) approach models material processes in
terms of two universal factor inputs, namely broadly-defined energy and
broadly defined organization, the former being physically productive, while
the latter being organizational. In keeping with basic mechanics and
thermodynamics, energy and energy alone can accomplish work, the
implication being that all other factors are organizational in nature.
ܹ (‫)ݐ( ܶ[ߟ = )ݐ‬, ܵ(‫)ݐ‬, ‫)ݐ(ܧ])ݐ(ܫ‬ 1.1
The latter, in turn, is a function of S(t) the supervisory input, T(t), tools,
and I(t) information. In keeping with basic physics, the latter three factor
inputs are not physically productive, but rather are organizational in nature,
affecting second-law efficiency. Better tools (e.g. Watt’s external condenser,
the Boulton-Watt dual-action steam engine, electric unit drive) increase
energy efficiency by minimizing losses. As Ș is bounded from above, it
stands to reason that organizational innovations will have limited effect on
output and output growth (Beaudreau and Lightfoot 2015). Equation 1.1
provides a simple description of the EO approach to material processes, with
E(t) being the energy input and Ș being the thermodynamic concept of
second-law efficiency. This can be seen as a measure of energy
productivity, which in this case, is a function of the relevant organizational
variables, including tools T(t), supervision S(t), and information I(t).
Beaudreau (1998) maintained that this simple model was universal in scope,
being applicable to all material processes.
The EO approach to growth is straightforward, namely that growth of
the output is an increasing function of growth of the energy input as well as
growth/innovations in Ș, second-law efficiency. The key as far as we are
concerned is the universality of Equation 1.1. Any and all growth processes
in the material sciences is/are predicated on growth in the energy input, and,
the case that concerns us here, growth in the organizational context. For
example, in the case of economic material processes, growth requires an
increase in energy as well as an equivalent increase in tools and supervision
–conventional capital and labor.
This raises the question of productivity or, put differently, the
contribution of factor inputs to output and growth. In keeping with basic
mechanics and thermodynamics, the only physically productive factor input
is energy/force. All others are organizational inputs, which together define
the material process, but are not productive in the traditional sense. Put
differently, they increase with output but are not the ultimate cause.
Frederick Soddy captured the essence of material processes–animate and

inanimate–in the following parable.
“At the risk of being redundant, let me illustrate what we mean by the
question ‘How do men live?’ by asking what makes a railroad train go. In
one sense or another, credit for the achievement may be claimed by the so-
called ‘engine-driver, ’ the guard, the signalman, the manager, the capitalist,
the share-holder, -or again, by the scientific pioneers who discover the
nature of fire, by the inventors who harnessed it, by labour which built the
railroad and the train. The fact remains that all of them, by their collective
effort could not drive the train. The real engine-driver is the coal. So, in the
present state of science, the answer to the question how men live or how
anything lives, or how inanimate nature lives, in the sense in which we speak
of the life of a waterfall or of another manifestation of continued liveliness,
is, with few and unimportant exceptions, ‘By sunshine.’ Switch off the sun
and a world would result lifeless, not only in the sense of animate life, but
also in respect of by far the greater part of the life of inanimate nature”
(Soddy, 1924, 4).
1.3 Literature Review: Steady-State Growth Models

In this section, we review the steady-state growth literature, focusing on
the underlying microfoundationsthat is, the implied formalizations of
material processes, with the EO approach as our guide. For the most part,
this literature consists of models/approaches that are based on neoclassical
production theory where the emphasis is on labor and capital, and in more
recent cases, other factor inputs such as materials, services and energy
(Berndt and Wood 1975).
The Harrod-Domar (HD) and Solow-Swan (SS) models are steady-state
neoclassical growth models, as both focus on labor and capital, and both
derive the equilibrium steady-state growth rate. As such, both see labor and
capital as being physically productive. Where they differ is in terms of factor
substitution with HD being based on fixed proportions and thus no
possibility of labor-capital substitution, and SS allowing for unlimited
substitution. Ultimately, the steady-state growth rate is defined as the sum
of the rate of growth of the labor input and the rate of technological change.
Technological change is incorporated in both, but not modeled explicitly
(i.e. exogenous). In short, technological change is of the manna-from-
heaven type, with no specific structure. It is fair to say that the HD and SS
approaches to growth are the de facto gold standards of the growth literature
as virtually all subsequent work is a variant of these models.
The enigmatization of growth that is inherent in these models, we
maintain, owes in large measure to the enigmatization of the underlying
material processes, which in turn can be traced back to the earliest attempts
on the part of moral philosophers and political economists to understand
wealth and its creation. For example, in Chapter 1 of Adam Smith’s An
Inquiry into the Nature and Causes of the Wealth of Nations, industrial
material processes are modeled as being labor-based, with labor
productivity (a scaler) being a function of (i) worker learning (ii) reduced
downtime, and (iii) the introduction of machinery. Ironically, while labor
had been reduced to a marginal factor input, overseeing steam-powered
machines, Smith put it at the center of his analysis, a decision that would be
heavy in consequences. For roughly a century, labor was front and center,
while the steam engine was couched in a parameter.
Table 1.2: Neoclassical and Other Enigmas
Enigma Violation
Capital is physically productive Principles of basic mechanics; tools
(simple and complex) are not
productive.
Labor is physically productive Machine operatives; production
requires both, yet worker-less
factories exist
Physically productive labor and Neither is physically productive
capital can be substituted
Energy/force is absent
Production functions exist Neither labor nor capital is physically
productive.
Solow Residual Roughly half of growth is attributed to
unknown factors. Metaphysical in
nature.
As it turned out, this became the central theme of Karl Marx’s labor
theory of value, namely that labor was the only productive factor input and
as such was entitled to the entirety of the product. Ironically, throughout
Marx’s life, labor was little more than an organizational factor input,
overseeing the workings of machines. Its brawn no longer powered the
material processes of the industrial revolution, yet it remained at the center
of the discourse.
The resulting crisis in classical economics (after all, Marx should be
considered to be a de facto classical economist), led eventually to the
neoclassical rejoinder, one that evacuated the problem of (unearned) profits
by simply decreeing capital to be physically productive. No justification

was, nor could be given based on the role of tools in classical mechanics
and applied physics. While a stop-gap measure intended to calm the waters,
it would go on to muddle them even further. By the end of the century,
material processes in economics were defined in terms of two, organizational,
non-physically productive factor inputs, namely labor (supervisors) and
capital (non-productive tools).
While both are necessary factor inputs, the problems stemmed from and
continue to stem from the underlying implications, namely that both are
assumed to be physically productive with an average and marginal product.
Ironically, neoclassical stalwart Alfred Marshall, in his 1890 magnum opus,
Principles of Political Economy, referred to workers as “machine operatives,”
yet continued to view them as being physically productive. The list of
associated enigmas is provided in Table 1.2, where we see that unlike
elsewhere the material sciences, organizational inputs (i.e. tools) are
assumed to be physically productive. Similarly, labor or what is essentially
a supervisory factor input is also assumed to be physically productive.
Together, these two assumptions are the metaphorical equivalent of a fuel-
less or energy-less automobile complete with driver–or of glucose-free
mitochondria. Another interesting enigma is the concept of labor-capital
substitution and the resulting implications, namely that output can be
maintained by giving up one and getting one of the other. How this came to
being is a mystery given that neither is physically productive–hence the enigma.
However, the greatest enigma is and will always remain the Solow
residual. Ironically, the neoclassical approach to understanding material
processes spawned, in the post-WWII period, this the greatest enigma of all
times. In short, using a completely inappropriate accounting framework
(Divisia and Tornquist, indexes), roughly half of the observed growth was
attributed to capital and labor, leaving the other half as a residual. Put
differently, half of the observed levels of growth was attributed to non-
productive, organizational units, and the rest was a mystery. One could
argue that this is a second-order enigma, having been born of the first, more
basic enigma, namely neoclassical production theory.
When growth rates plummeted in the 1970s and 1980s, the standard pat
response was that the rate of technological change had, for all intents and
purposes, fallen to zero. As little was known of the underlying dynamics
and causes of the residual in the post-WWII period, myriad hypotheses were
advanced, ranging from the welfare state, to fiscal policy, to import
substitution, to the OPEC-induced energy crisis, to unionization, etc.
Unfortunately, all of these were little more than hastily-crafted, often
times, ideologically- inspired ex-post rationalizations, with little-to-no basis
in science. The proof is that some four decades later, not one of these
hypotheses has been confirmed empirically. In time, however, the
profession responded with a veritable flurry of activity aimed at modeling
the residual, understanding the productivity slowdown and ultimately
affecting the policy debate (i.e. increasing growth). The result has since
become known as New Growth Theory, which we now examine.
1.3.1 Endogenous or New Growth Theory

The birth of the residual in the post-WWII period gave rise to yet another
series/generation of enigmatic approaches to understanding material
processes and growth, also known as New Growth Theory (NGT). Not
surprisingly, these approaches were orthogonal to the science of material
processes and their growth, owing in large measure to the enigmas referred
to in Table 1.2. In essence, the profession set out to understand the enigma
that was the Solow residual with what turned out to be a new generation of
enigmas, including notions such as creative destruction, Ak models, and
institutional economics.
In this section, we choose not to review the many contributions to this
literature since Romer’s path-breaking work in the 1980s. Rather, our focus
will be on its implications and more importantly, on its success or lack
thereof of NGT in shedding light on the processes underlying economic
growth. With this in mind, we start with its implications. In short, there are
three basic implications, namely that in so far as growth is concerned,
history matters, institutions matter and geography matters. By the latter, it
should be understood that time and place are integral components of
technological change, and as such, growth. In this regard, it has much in
common with the field of evolutionary economics pioneered by Nelson and
Winter (1973). Within this framework, NGT maintains that markets in
general underinvest in knowledge (Romer 1986), that monopolistic
competition is more conducive to innovation and that multiple equilibria are
not only possible, but likely. In short, it sees innovation as a succession of
monopolistically competitive technologies, instigated by existing and new
firms (sometimes referred to as quality ladders).
Given the paucity of knowledge about the Solow residual, NGT was a
welcomed addition to the literature. After all, it sought to shed light on the
greatest enigma of the 20th century, in addition to providing a framework
for understanding the process of growth in general. The problems, however,
were many. As far as we are concerned, it unwillingly or unknowingly
contributed to further enigmatizing the question of economic growth. It did
so by increasing the dimensionality of the problem in a number of
directions. For example, instead of focusing on the Solow residual which is

a material process-based residual, it couched the discussion in the larger
question of innovation in general–that is, innovation involving processes,
products and institutions. In fact, much of the discussion and examples
found in the theoretical and empirical literature are taken from the realm of
product innovation, not process innovation. While there is nothing
intrinsically wrong with this, it detracts from the question/problem at hand,
namely understanding output growth. Put differently, quality ladders add
little to our understanding of material process-based growth.
While the original NGT framework defined a whole new world for
innovation and technological change, there was no subsequent attempt to
narrow the focus to material process innovation.
This has made for a situation in which process and product innovation
are used interchangeably when in actual fact, the latter has little to no
bearing on growth (i.e. on the underlying material process). A good example
of this is the R&D literature where process and product innovation are
lumped into one, overriding variable, namely R&D. Clearly, product
innovation cannot and will not increase the growth of output.
1.3.2 Non-Steady State Models (shocks)

This sub-field of the growth literature is as enigmatic as its steady-state
counterparts. In essence, it attempts to identify the factors that contributed
to the first (18th century) and second industrial revolutions (late 19th/early
20th century), what most agree were singular occurrences. Unlike the steady-
state literature, work in this field is shock-specific, with virtually no attempt
to provide a general theory of industrial revolutions–that is, a general theory
of industrial revolutions.
McCloskey (2004) pointed to what have been two approaches to
industrial revolutions (mostly the first), namely material and non-material.
The former refers to the various known and documented technological
advances that led to the first industrial revolution, while the latter refers to
the institutions–including culture–that spawned these changes. In this
regard, non-material approaches have much in common with NGT where
institutions are front and center.
In general, material approaches have focused, for the most part, on the
process-based innovations that led to the well-documented increase in
output in the early 19th century (Boulton-Watt double-acting steam engine,
Paul’s power carding and spinning machines, Arkwright’s spinning frame).
Surprisingly, this literature is decidedly neoclassical in nature, with the
various technological innovations being seen as affecting the technology (A)
scaler. The underlying mechanics are, for the most part, not specified. As
such, subsequent developments such as the development in the 1840s of
high-pressure steam engines, and the development in the 1880s of the steam
turbine, both of which paved the way for greater machine speeds and
productivity, are ignored.
While these models or theories of industrial revolutions are enigmatic
with regard to the basic neoclassical production function (i.e. they affect A,
the technology scaler), they are, in essence, consistent with the underlying
principles of the material sciences. This owes in large measure to the
emphasis placed on the role of energy in the various aspects/dimensions of
first industrial revolution. For example, the Watt atmospheric steam engine
(with external condenser) in coal mines, the Boulton-Watt dual-action,
reciprocating engine in spinning, carding and weaving material processes.
In short, the first industrial revolution witnessed a massive increase in
energy use/consumption, resulting in an equally massive increase in
wealth. In other words, they are consistent with the laws of
physics/mechanics/thermodynamics. What could be regarding as an
important advance in the understanding of economic growth was, however,
lost in the ensuing analysis of its effect on productivity and output.
Specifically, it was seen as increasing both labor and capital productivity,
two inert factor inputs. In short, the steam engine was modeled as a scaler
affecting, in a one-shot manner, capital and labor productivity.
Dissatisfaction with material approaches to industrial revolutions led, in
the 1980s and 1990s, to a new approach, namely non-material where the
emphasis was on the institutional and cultural underpinnings of the
technological and institutional change that characterized the industrial
revolution eras. Drawing largely from institutional economics, it sought to
identify the exact set of circumstances or causes. Among the leading non-
material approaches is Joel Moykr’s Republic of Letters according to which
the enlightenment in England–what he refers to as the Baconian program–
combined with a growing interest in practical knowledge, led to the
industrial revolution (Moykr 2016). A good example of this, he argued, is
the Birmingham Lunar Society where men of science rubbed shoulders with
entrepreneurs/practical men. Another is Deirdre McCloskey’s notion of
Bourgeois Dignity, according to which societal values towards business in
general and endeavoring to make a return on one’s investment, laid the
ground for the industrial revolution (McCloskey 2010). In both cases, non-
economic factors were responsible for the cataclysmic change that was the
first industrial revolution.
From a scientific point of view, these approaches only serve to further
muddy the waters for the simple reason that they are virtually not testable.
In short, the data set is limited to a single observation, thus eliminating any
possibility of testing for regularity. Both Moykr and McCloskey are aware
of this, and have responded with a barrage of anecdotes filling volumes. For
example, McCloskey develops her thesis in a trilogy of works, including
Bourgeois Dignity, Bourgeois Values and Bourgeois Equity. The sheer
volume, however, does not take away from the fact that her theory is not
testable.
1.3.4 Policy Implications

As the endogenous growth literature (steady-state and episodic shocks)
has grown in size and scope, the list of possible causes has exploded, as has
the breadth of the enigma. Instead of providing specific answers to specific
episodes of growth, it has increased the dimensionality and potentiality of
the various causes of growth.
Nowhere is this more apparent than in the policy implications or lack
thereof. Take, for example, Mokyr’s notion of the “Republic of Letters” or
McCloskey’s notion of “Bourgeois Dignity.” The obvious question is what
are associated policy implications, if any? The last century has witnessed an
unprecedented increase in useful knowledge, yet no industrial revolutions
have occurred. The same holds for bourgeois dignity. Presently, the world
is spending upwards of $1.7 trillion per year on research and development,
yet growth rates remain disappointingly low–compared to the post-WWII
period.
It is our view that this is not surprising in the least bit, in light of the
enigmatization described here. In short, policy measures are set against
enigmas that are intended to explain enigmas that themselves, had resulted
from the original enigmatic representation of material processes. Is it any
wonder that the results have been and continue to be less than ideal? One
could argue that, despite the efforts of the past forty years, the profession is
further today from understanding growth than it ever was, owing to the
continued divergence from the physics of material processes, or what we
refer to here as enigmatization. While growth is understood to the point of
no longer being an issue/question of concern in the other material sciences,
it remains more of a mystery in economics than ever.
1.3.5 Consilient Approaches

In this section, the discussion will be limited to two approaches, namely
LINEX (Kummel 1982, Kummel et al 1998, Lindenberger and Kummel
2002) and Energy-Organization EO (Beaudreau 1998). Other approaches
such as the ecological approach and the biophysical approach, while

relevant, are not considered for lack of a complete theory of material
processes by which they should be understood, an approach that considers
the role of all factor inputs (i.e. capital and labor), not just energy.
Figure 1.1: Actual and Predicted Growth of U.S. GDP 1960-1978
Source: Kummel (1982), 196.
Let us begin with the question of post-WWII growth. A consilient

approach would attribute the high rate of growth to a high rate of growth of
energy use/consumption. As it turns out, the LINEX and EO approaches
corroborate this result. Referring to Figure 1.1, we see that manufacturing
output in the U.S. tracks almost perfectly energy use/consumption, with a
dip in the 1970s. According to Kummel (1982), this corresponds to the
productivity slowdown, where energy use decreased. The EO approach also
attributes post-WWII economic growth to energy use/consumption. Table
1.3 shows how growth in manufacturing output in the U.S., Germany and
Japan (USVA, GERVA, and JAPVA, respectively) tracks energy
use/consumption (USEP, GEREP, and JAPEP, respectively)–specifically
how output and energy growth both fell precipitously from 1973 onwards.
Both approaches maintain that the record growth in labor and multifactor
productivity owed to an increase in energy-use intensity–that is, the increase
in energy use/consumption per unit of labor/capital. Implicitly, the enigma
that is the Solow residual is resolved, with productivity growth being
attributed to greater energy use per unit of labor/capital.
This brings us to the question of the productivity slowdown. In keeping
with the laws of classical mechanics, both attribute it to the fall in the rate
of growth of energy use/consumption, itself the result of the OPEC-induced
price increases. In other words, higher energy prices and the specter of even
higher future prices, reduced the rate of growth of energy use/consumption.
Once again, the enigma of the productivity slowdown is resolved, being
attributed to the decrease in the rate of growth of energy use/consumption.
In more recent work, Beaudreau (2017) re-examined the underlying
hypothesis, namely higher energy prices leading to a lower rate of growth
of energy use/consumption. Specifically, he showed that while fossil-fuel
prices increased in the mid-1970s, the price of electricity (primary source
of energy in manufacturing) remained relatively constant and moreover, the
real price of energy (and electricity) had, by the 1980s, returned to its pre-
OPEC crisis level, yet the rate of energy use/consumption did not rebound.
Drawing from his work on the economies of speed, he invoked the laws of
kinetics to attribute the decrease in energy use/consumption to the problems
inherent in speeding up material processes. More to the point, he argued that
maximum machine speed had, by the late 1960s/early 1970s, been reached
in most industries, making for a slowdown in the rate of increase of machine
speeds and consequently in productivity.
Table 1.3: Output and Input Growth Rates: U.S., German and
Japanese Manufacturing
U.S. 1950–1984 1950–1973 1974–1984

USVA 2.684 3.469 0.121
USAI* 2.674 3.472 0.310
USEP 4.052 5.371 0.246
USN 0.662 0.900 -0.091
USK 3.694 3.614 3.4008
Germany 1963–1988 1962–1973 1974–1988
GERVA 2.462 6.522 1.486
GERAI* 2.433 5.190 1.080
GEREP 2.894 5.883 1.366
GERN -0.785 0.592 -0.938
GERK 2.945 5.620 1.406
Japan 1965-1988 1965-1973 1974-1988

JAPVA 3.826 8.844 3.099
JAPAI* 3.566 9.856 1.538
JAPEP 3.559 11.320 0.965
JAPN -0.082 2.297 -0.367
JAPK 7.520 13.536 5.182
ଵ ௗா௉ ଵ ௗே ଵ ௗ௄
*‫ߚ = ܫܣ‬ா௉ ா௉ ௗ௧
+ ߚே
ே ௗ௧
+ ߚ௄
௄ ௗ௧
, ‫ ߚ ݄݁ݐ ݁ݎ݄݁ݓ‬ᇱ ‫ݏ݁݅ݐ݅ܿ݅ݐݏ݈ܽ݁ ݐݑ݌ݐݑ݋ ݄݁ݐ ݁ݎܽ ݏ‬.
To recapitulate, over two centuries ago, political economists approached

the question of understanding how the steam-engine powered the industrial
revolution, by enigmatizing the process for the first time, attributing
physical productivity properties to capital and labor, two organizational,
non-energy-based factor inputs. This then resulted in a second round of
enigmas in the form of the Solow residual, where the brunt of post-WWII
growth was attributed to what Moses Abramovitz referred to as our “our
measure of ignorance.” By then, waist deep in enigmas, the profession set
out to understand the enigma caused by enigmas with yet another round of
enigmas that included Romer’s Ak model, Schumpeterian quality ladders,
etc.–in short, New Growth Theory. This has made for the current situation
where the profession now finds itself inundated with enigmas when the
answer, according to material scientists, is as simple as the basic principles
of mechanics and thermodynamics. This has resulted in a situation in which
the study of material processes in economics is a virtual island onto itself,
with enigmatic notions and concepts, and with two centuries of growth still
to be explained.
Table 1.4: Incursions into the Material Sciences: Missed Opportunities
Author Missed Opportunity

Adam Smith (1776) Fire Power
Robert Owen (1820) Scientific Power
Karl Marx (1867) Classical Mechanics
William Stanley Jevons (1865) Coal is the Mainspring of Modern
Material Civilization
Alfred Marshall (1890) Labor “as Machine Operatives"
Thorstein Veblen (1921) Power Resources
Frederick Soddy (1922) Cartesian Economics
Adam Smith (1776) Fire Power
1.4 The Exceptions: Mainstream Incursions

into the Material Sciences
In this chapter, we have argued that the many formalizations of material
processes and their growth throughout the 19th and 20th centuries have led
to a series of enigmas, the sum total of which has left the profession with an
understanding of material processes that is orthogonal to the laws and
principles that govern all other material sciences. We would, however, be
remiss to maintain that there were no exceptions. After all, the 19th century
witnessed important developments in the science behind the steam engine,
namely thermodynamics. As it turns out, a number of ranking political
economists did make incursions into the material sciences. However, most
of these were either (i) in direct contradiction with their more fundamental
contributions, or (ii) of secondary interest or concern.
Table 1.4 provides a non-exhaustive list of these references to elements
of material sciences, mostly regarding the role of energy in production.
Take, for example, Adam Smith whose magnum opus, the Wealth of
Nations, was inspired by Matthew Boulton’s experience with steam power
at his Hockley Brook factory. In 1776, the science of steam or fire was
inexistent, prompting him to refer to it in naive terms, namely as fire power.
Why he chose to see it as increasing labor productivity is a question open
for debate. Was it to assuage/reassure labor, or was it a vestige of a bygone
era when labor was not only the source of energy/power (i.e. brawn).
Perhaps the most perspicacious of these writers in so far as references to the
material sciences is concerned was German economist Karl Marx, whose
1867 Das Kapital contains a surprising account of the basic elements of
process engineering, complete with references to the role of power and force
as the ultimate drivers of material processes (Chapter 15 entitled On
Machines and Machinery). What is astounding is its orthogonality to
Chapters 1-7 where he develops the labor theory of value (surplus value)
based in large measure on classical production theory.
Another surprise is neoclassical pioneer William Stanley Jevons who in
The Coal Question, published in 1865, trumpeted the essential role of coal
in material civilization, going as far as arguing that it constituted the
“mainspring”. Contrast this with what would become the neoclassical
theory found in his 1874 classic The Theory of Political Economy where
coal/energy is entirely absent (Jevons 1874). What is also surprising is the
fact that by then, the laws of thermodynamics were well established.
1.5 Summary and Conclusions

One could argue that the history of economics or the science of wealth
is the story of a profession which in spite of itself, has attempted to
understand material processes in what was and an intellectual vacuum,
choosing to ignore developments in related material sciences–and in science
in general. The result has been a series of enigmas which as we have shown
have engendered subsequent rounds of enigmas, with the result that today,
the question of growth is more misunderstood than ever. Over the course of
its history, notions of capital and labor (physical) productivity were
advanced, while the energy input, the cornerstone of the science of material
processes, was ignored completely. Instead, enigmatic, oftentimes orthogonal
rationalizations were advanced, resulting in even more enigmas that not
surprisingly engendered a whole new generation of enigmas to explain
them.
In this chapter, a roadmap to the enigmatization of economic growth was
provided, one that goes a long way explaining why, as Paul Krugman
remarked in a 2013 New York Times editorial, the promise of New Growth
Theory has fizzled out. Our starting point was that material processes are
well understood outside of economics, where there has been and continues
to be no need for arcane notions like the Solow residual, or other “measures
of our ignorance.” In keeping with basic mechanics, all work is ultimately
the result of the use of force/energy. More importantly, there can be no
exceptions, nor violations to the laws of physics. The notion that generic
technological change can miraculously increase output is an affront to basic
scientific knowledge, one that borders on the sublime.
It is our view that economics in general and growth theory in particular
have suffered as a result. For one, economics is the only material process-
based discipline where growth is largely not understood, and what is
assumed to be understood (i.e. the role of labor and capital) is, in reality, an
illusion, an enigma. Second, despite four decades of what was a concerted
attempt to understand growth, the verdict is one of complete and utter
failure–or what Krugman refers to as “fizzle”.
Lastly, it is our view that the blame for what we refer to as the
enigmatization of economic growth lies squarely on the shoulders of the
profession. While the problem of growth in the other material process
sciences has been resolved, it remains very much an open issue in
economics. And the reason is clear, namely the concatenation and
multiplication of enigmas. When one tells a lie, one is oftentimes forced to
tell another or others to cover it up. The enigmatization of basic material
processes in the 19th century led to more enigmas and even more in an
attempt to understand them.
1.6 References
Alting, Leo. 1994. Manufacturing Engineering Processes. New York, NY:
Marcel Decker Inc.
Beaudreau, Bernard C. 1998. Energy and Organization: Growth and
Distribution Reexamined. Westport, CT: Greenwood Press.
Beaudreau, Bernard C. and Douglas Lightfoot. 2015. “The Physical Limits
to Economic Growth by R&D-Funded Innovation,” Energy, 84: 45-52.
Beaudreau, Bernard C. 2017. “The Economies of Speed, KE=1/2mv2, and
the Productivity Slowdown,” Energy. 124: 100–113.
Beaudreau, Bernard C. 2018. “A Pull-Push Theory of Industrial
Revolutions,” Université Laval, manuscript.
Beiser, Arthur. 1983. Modern Technical Physics. Menlo Park, CA: The
Benjamin/Cummings Publishing Company.
Berndt, Ernst and David Wood. 1975. “Technology, Prices and the Derived
Demand for Energy.” Review of Economics and Statistics, 57, no.
3:259–268.
Denison, Edward F. 1962. The Sources of Economic Growth in the United
States and the Alternatives Before Us. New York, NY: Committee for
Economic Development.
Denison, Edward F. 1985. Trends in American Economic Growth 1929–
1982. Washington, DC: The Brookings Institution.
Jevons, W.S. 1865. The Coal Question. London: MacMillan and Co.
Jevons, W.S. 1965[1874]. The Theory of Political Economy. New York,
NY: Augustus M. Kelley, Bookseller.
Krugman, Paul. 2013. “The New Growth Fizzle.” The New York Times,
August 18, 2013.
Kummel, Reiner. 1982. “The Impact of Energy on Industrial Growth.”
Energy. 7, no. 2: 189- 201.
Kummel, Reiner, Dietmar Lindenberger, and Wolfgang Eichhorn. 1998.
“The Productive Power of Energy and Economic Evolution,” Indian
Journal of Applied Economics, Special Issue on Macro and Micro
Economics.
Lindenberger, Dietmar and Reiner Kummel. 2002. “Thermodynamics and
Economics.” Post-Autistic Economics Review, 14.
Lloyd-George David. 1924. Coal and Power. London: Hodder and
Stoughton.
Marshall, Alfred. 1890. Principles of Economics. London: MacMilllan.
Marx, Karl. 1906[1867]. “Machinery and Modern Industry.” in Marx, Karl.

Das Kapital. Chapter 15. New York, NY: The Modern Library.
Mokyr, Joel. 2016. A Culture of Growth: The Origins of the Modern
Economy. Princeton, NJ: Princeton University Press.
McCloskey, Deirdre. 2004. The Bourgeois Values: Ethics for an Age of
Commerce. Chicago, IL: University of Chicago Press.
McCloskey, Deirdre. 2010. Bourgeois Dignity or Why Economics Can’t
Explain the Modern World. Chicago, IL: University of Chicago Press.
Nelson, Richard and Sidney Winter. 1973. “Toward an Evolutionary Theory
and Economic Capabilities.” American Economic Review. May. Vol. 63,
no. 2: 440-449.
Owen, Robert. 1820. Report to the County of Lanark. London: Equitable
Labor Exchange.
Piketty, Thomas. 2014. Capital in the 21st Century. Cambridge, MA:
Harvard University Press.
Romer, Paul M. 1986. “Increasing Returns and Long-Run Growth.” The
Journal of Political Economy. 94, no. 5:1002-1037.
Smith, Adam. 1990[1776] “Of the Division of Labour.” Book I, Chapter 1,
An Inquiry into the Nature and Causes of the Wealth of Nations.
Chicago, IL: Encyclopaedia Britannica:1–12.
Soddy, Frederick. 1922. Cartesian Economics, The Bearing of Physical
Sciences upon State Stewardship. London: Henderson.
Veblen, Thorstein. 1921. The Engineers and the Price System. New York:
B.W. Huebsch.
2
THINKING ABOUT PRODUCTION:

FROM SMITH TO ROMER
Abstract
While energy/physical work and the basic laws of physics continue to be

conspicuous by its absence in economic theory, there have been, over the
course of the past two hundred years, a number of attempts by both
mainstream and non-mainstream writers, to incorporate them into the
analysis and in so doing provide a more consilient, a more scientific
approach to understanding production and growth. This chapter provides an
account of these attempts, starting with Adam Smith’s reference to “fire
power” and ending with the myriad contributions in the 1920s and 1930s, a
period marked by a manifold increase in energy consumption.
The chapter will be organized around a number of themes, including

mainstream and non-mainstream, academic versus professional economists,
outside influences/critics, the impact or lack thereof. The presentation will
be chronological starting with the pre-industrial period and ending with the
electrical era.
2.1 Thinking About Production Prior to the 18th Century

Anthropomorphic material processes have existed from time immemorial.
In fact, they define our species as evidenced by the use of tools to define
what are epochs in our history. For example, the Paleolithic, Mesolithic and
Neolithic eras are defined by the nature of the tools employed in the various
material processes, from hunting to cooking to eating. Tools in the
Paleolithic era were fashioned from stone (flint, agate) and were less durable
and less efficient (second-law efficiency) than later as bronze and
ultimately, iron provided better and more durable surfaces.
In general, little was written or understood of material processes
throughout this period, if only that wealth was increasing in human toil,
augmented in some cases by water, wind and animal power. Wind and water
20 2. Thinking About Production
powered a number of material processes without giving rise to a theory of

production. The same was true of slaves who, throughout history, contributed
to increasing wealth, again without giving rise to a theory of production.
That this be the case is, in many ways, perfectly understandable. Very
little changed over the course of the millennia. Theories of production and
theories, in general, arise in response to various challenges, notably change
and our desire to understand and in many instances to adapt. From the
dawning of civilization in the Middle East (Mesopotamia), material
processes have, for the most part, not changed, hence the absence of a
science of material processes–in short, a theory of production.
2.2 Thinking About Production in the Newtonian Era

The late 18th/early 19th century witnessed one of two massive increases
in energy availability of the modern era. By energy availability, it should be
understood both the presence of an energy source as well as a means
whereby energy can be converted into useful work–in short, can increase
output/wealth.
This resulted from a number of developments, including the Papin-
Savary-Newcommen atmospheric steam engine, the Watts atmospheric
steam engine, the Watts-Boulton reciprocating steam engine, and the
Trevithick high-pressure boiler, among others. These developments increased
energy availability, thus setting the stage for a manifold increase in
operating speeds and output per capital/labor.
The first to attempt to understand the consequences of this new power
drive technology was Scottish moral philosopher Adam Smith whose
magnum opus, An Inquiry into the Nature and Causes of the Wealth of
Nations, published in 1776, was based largely on Matthew Boulton’s
experimentation with a steam engine at his Hockley Brook plant in
Birmingham, England. The overall tone of the work was set in Chapter 1
where he outlines what can be viewed as an impressionistic theory of the
“causes of wealth,” consisting of enumerating the advantages of specialization,
which is nothing more than the mechanization of what were artisanal
processes. In short, labor productivity, defined as the ratio of output to labor
input, increases for three reasons, namely increased dexterity and reduced
idle time from worker specialization (essentially becoming a machine
operative), and third, the introduction of machinery which he refers to as
“fire power,” a reference to steam. In other words, Smith, like most
contemporaries, were aware that fire or combustion was the ultimate source
of power, not the steam.
It can therefore be surmised that the Wealth of Nations was first and
foremost about the new power drive technology that was the steam engine.
Processes that had been the realm of artisans and craftsmen were now
mechanized, increasing output manifold. Yet, despite this, the resulting
view of production was long on labor and short on force/energy. More
specifically, classical production theory posited a monotone increasing
relationship between labor and output. Put differently, Smith who
admittedly was a moral philosopher had failed to incorporate the basic
principles of natural philosophy (i.e. classical mechanics) into his analysis.
The great increasing of the quantity of work which, in consequence of the
division of labor, the same number of people are capable of performing, is
owing to three different circumstances; ¿rst, to the increase in dexterity in
every particular workman; secondly, to the saving of the time that is
commonly lost in passing from one species of work to another; and lastly,
to the invention of a great number of machines which facilitate and abridge
labor, and enable one man to do the work of many. (Smith 1776, 10)
While surprising given his myriad acquaintances in learned circles, it

was to become the standard in the emerging field of political economy. His
heir apparent, David Ricardo, popularized the labor theory of value, and the
underlying classical theory of production in his 1817 Principles of Political
Economy and Taxation. By then, the Watts atmospheric steam engine had
given way to the Boulton-Watt reciprocating steam engine which found the
favor of industrialists throughout England. Among its many advantages was
the increased machine speed and its near-continuous operation. Processes
that had been operated by craftsmen and artisans could now operate
uninterrupted 24/7 at greater speeds, thus making for manifold increases in
output per unit of capital.
Yet, there is little mention of fire or steam power in On the Principles of
Political Economy and Taxation, which is astonishing given the revolution
underway in England. He did, however, make reference to the mechanization
of industry. In Chapter 20, entitled “Value and Riches, Their Distinctive
Properties,” he referred, albeit indirectly, to the effects of steam power on
overall wealth.
Value, then, essentially differs from riches, for value depends not on
abundance, but on the difficulty or facility of production. The labour of a
million of men in manufactures will always produce the same value, but will
not always produce the same riches. By the invention of machinery, by
improvements in skill, by a better division of labour, or by the discovery of
new markets, where more advantageous exchanges may be made, a million
of men may produce double, or treble the amount of riches, of “necessaries,
conveniences and amusements,” in one state of society that they could

produce in another, but they will not, on that account, add anything to value.
(Ricardo 1817, 182)
This raises the obvious question, why were early moral philosophers
unable to provide a more thorough description of the steam engine and its
role in generation wealth or riches? In the grand scheme of things, the focus
remains on labor, with machinery as an aid of sorts, increasing conventionally-
measured labor productivity.
The first writer/political economist to refer to science, or the science of
the steam engine, was Robert Owen who in his Report to the County of
Lanark, pointed to “scientific improvements and arrangements” in reference
to advances in steam-driven power drive technology.
It is well known that, during the last half-century in particular, Great Britain,
beyond any other nation, has progressively increased its powers of
production, by a rapid advancement in scientific improvements and
arrangements, introduced, more or less, into all the departments of
productive industry throughout the empire. The amount of this new
productive power cannot, for want of proper data, be very accurately
estimated; but your Reporter has ascertained from facts which none will
dispute, that its increase has been enormousʊthat, compared with the
manual labour of the whole population of Great Britain and Ireland, it is, at
least, as forty to one, and may be easily made as 100 to one; and that this
increase may be extended to other countries; that it is already sufficient to
saturate the world with wealth and that the power of creating wealth may be
made to advance perpetually in an accelerating ratio. (Owen 1820, 19)
Forty-to-one or one hundred-to-one are references to the increase in

work that could potentially result from the application of continuous-flow,
rotary steam-engine power to the various material processes of the early 19th
century. Clearly, continuity of and speed of material processes were behind
these estimates.
There were, however, several more in-depth attempts at understanding
the underpinnings of steam-powered machines. Perhaps the most thorough
was that of Charles Babbage in 1832 who in his On the Economy of
Machinery and Manufactures provided detailed descriptions of the new
technology. Consider, for example, the following excerpt where classical
mechanics is used to illustrate the contribution of wind, water, and steam.
Of those machines by which we produce power, it may be observed, that

although they are to us immense acquisitions, yet in regard to two of the
sources of this power, the force of wind and of water, we merely make use
of bodies in a state of motion by nature; we change the directions of their
movement in order to render them subservient to our purposes, but we

neither add to nor diminish the quantity of motion in existence. When we
expose the sails of a windmill obliquely to the gale, we check the velocity
of a small portion of the atmosphere, and convert its own rectilinear motion
into one of rotation in the sails; we thus change the direction of force, but
we create no power….The force of vapour is another fertile source of
moving power; but even in this case it cannot be maintained that power is
created. Water is converted into elastic vapour by the combination of fuel.
(Babbage 1832, 15)
Interestingly, he devoted a whole chapter to speed or what he referred to

as “velocity.” Chapter 4, entitled “Increase and Diminution of Velocity,”
showcases using industry-speci¿c examples the role of increased speed as a
key feature of mechanization.
In turning from the smaller instruments in frequent use to the larger and
more important machines, the economy arising from the increase in velocity
becomes more striking. In converting cast into wrought iron, a mass of
metal, of about a hundredweight, is heated almost to white heat and placed
under a heavy hammer moved by water or steam power. This is raised by a
projection on a revolving axis; and if the hammer derived its momentum
only from the space through which it fell, it would require a considerably
greater time to give a blow. But it is important that the softened mass of red-
hot iron should receive as many blows as possible before it cools, the form
of the cam or projection on the axis is such, that the hammer, instead of
being lifted to a small height, is thrown up with a jerk, and almost the instant
after its strikes a large beam, which acts as a powerful spring, and drives it
down on the iron with such velocity that by these means about the double
the number of strokes can be made in a given time. (Babbage 1832, 26)
Whereas previous writers referred to specialization, Babbage provides a

detailed account of the role of power in material processes in general, and
the role of steam power in U.K. manufacturing. Further, he perspicaciously
was the first to formalize the role of rotary motion/power in material
processes, alluding to the importance of velocity or put differently, machine
speed.
Unfortunately, Babbage’s work did not usher in a new era in political
economy, one characterized by a more clear and complete understanding of
high-throughput, continuous mass production–in short, of Ricardo’s 40-1 or
100-1 increase in output. Rather, the thinking about production remained
relatively primitive, with output being a function of labor input–in short,
classical production theory.
2.3 Thinking About Production in the Thermodynamic

Era
While political economy made little headway in its attempt to describe
and understand the revolution underway in England, physicists were on the
cusp of laying the foundations of a new sub-field, namely thermodynamics
or the science of heat. By mid-century, Rudolf Clausius and William
Thomson (Kelvin) had formulated both the First and Second Laws of
Thermodynamics.
Political economy, on the other hand, was embroiled in an internecine
debate over the legitimacy of profits, itself the result of the rise of radical
political economy. Taking their cue from classical production theory where
labor and labor alone is productive, they held that profits were unjustified
and unjustifiable. Karl Marx and Friedrich Engels formalized this in terms
of the rate of surplus-value in reference to the share of wealth that was
unjustifiably (according to classical production theory) appropriated by the
owners of the means of production.
Classical political economists responded by simply decreeing capital to
be physically productive, thus entitled to a share of the plunder, so to speak.
This, one could argue, only made matters worse. By mid-century, labor had
become a supervisory factor input, hence not physically productive. As
capital (simple and complex tools) is also not physically productive, the
result was a theory of production in which neither of the factor inputs were
physically productive. Such were the inauspicious origins of modern
neoclassical production theory.
By the mid-to-late 1800s, the thinking about production had become
polarized with the neoclassicals at the one end, and the Marxists at the other.
At stake was the future of society as a whole. However, what is interesting
is the fact that despite the vitriolic and increasing polarization, science in
general and thermodynamics in particular found their way into the thinking.
Take, for example, William Stanley Jevons, the father of neoclassical
production theory, who in The Coal Question: An Inquiry Concerning the
Progress of the Nation, and the Probable Exhaustion of Our Coal-Mines,
published in 1865, described the role of coal in England’s wealth in the
following terms.
Day by day it becomes more evident that the Coal we happily possess in
excellent quality and abundance is the mainspring of modern material
civilization. As the source of ¿re, it is the source at once of mechanical
motion and of chemical change. Accordingly, it is the chief agent in almost
every improvement or discovery in the arts which the present age brings
forth. It is to us indispensable for domestic purposes, and it has of late years
been found to yield a series of organic substances, which puzzle us by their

complexity, please us by their beautiful colours, and serve us by their
various utility. And as the source especially of steam and iron, coal is all
powerful. This age has been called the Iron Age, and it is true that iron is the
material of most great novelties. By its strength, endurance, and wide range
of qualities, this metal is ¿tted to be the fulcrum and lever of great works,
while steam is the motive power. But coal alone can command in suf¿cient
abundance either the iron or the steam; and coal, therefore, commands this
age - the Age of Coal. Coal in truth stands not beside but entirely above all
other commodities. It is the material energy of the countrythe universal aid
- the factor in everything we do. With coal, almost any feat is possible or
easy; without it we are thrown back into the laborious poverty of early times.
(Jevons 1865, xi)
Undoubtedly, this citation could easily find itself into the first chapter of
any thermodynamics textbook or manual, given its perspicacity. Yet, less
than a decade later, all of this was forgotten in what is considered to be
Jevon’s magnum opus, namely The Theory of Political Economy published
in 1874 where wealth is defined as an increasing function of labor and
capital. In short, coal or energy was conspicuous by its absence.
One could argue that Jevons was of two minds, one positive, and the
other normative. His early work predated the publication of Das Kapital by
Karl Marx by a few years. Clearly, the thinking about production had been,
in many ways, subverted by the larger issue of legitimacy, the legitimacy of
profits.
Surprisingly, this duality is also present in Marx’s writings, specifically
in Chapter 15 of Das Kapital where he noted:
Mathematicians and mechanicians, and in this they are followed by a few

English economists, call a tool a simple machine, and a machine a complex
tool. They see no essential difference between them, and even give the name
of machine to the simple mechanical powers, the lever, the inclined plane,
the screw, the wedge, etc. As a matter of fact, every machine is a combination
of those simple powers, no matter how they may be disguised. From the
economic standpoint this explanation is worth nothing, because the
historical element is wanting. Another explanation of the difference between
tool and machine is that in the case of a tool, man is the motive power, while
the motive power of a machine is something different from man, as, for
instance, an animal, water, wind, and so on. According to this, a plough
drawn by oxen, which is a contrivance common to the most different epochs,
would be a machine, while Claussen’s circular loom, which, worked by a
single labourer, weaves 96, 000 picks per minute, would be a mere tool. Nay,
this very loom, though a tool when worked by hand, would, if worked by
steam, be a machine…. All fully developed machinery consists of three
essentially different parts, the motor mechanism, the transmitting mechanism,
and ¿nally the tool or working machine. The motor mechanism is that which
puts the whole in motion. It either generates its own motive power, like the
steam-engine, the caloric engine, the electromagnetic machine, etc., or it
receives its impulse from some already existing natural force, like the water-
wheel from a head of water, the wind-mill from wind, etc. The transmitting
mechanism, composed of Ày-wheels, shafting, toothed wheels, pullies,
straps, ropes, bands, pinions, and gearing of the most varied kinds, regulates
the motion, changes its form where necessary, as for instance, from linear to
circular, and divides and distributes it among the working machines. These
two ¿rst parts of the whole mechanism are there, solely for putting the
working machines in motion, by means of which motion the subject of
labour is seized upon and modi¿ed as desired. The tool or working machine
is that part of the machinery with which the industrial revolution of the 18th
century started. And to this day it constantly serves as such a starting-point,
whenever a handicraft, or a manufacture, is turned into an industry carried
on by machinery. (Marx 1867, 357)
Again, such an account of material processes in the industrial age would

be worthy of a chapter in a process engineering handbook. There can be no
doubt that Marx understood–and appreciated–the role of steam and power
in the industrial revolution. Moreover, it is clear that the labor theory of
value, the cornerstone of Marxian economics, was a ruse, having no fact in
science, but constituting a key axiom in his theory of surplus-value and his
critique of society in general.
Unfortunately, the normative came to dominate the debate and indeed
the science. By the end of the century, the thinking about production was
couched in inaccuracies and violations of the basic laws of physics. The
insights provided by the like of Charles Babbage, William Stanley Jevons
and Karl Marx were ignored. That is, until the second energy shock would
befall the Western world, namely the introduction of electric unit drive in
the early 20th century which ushered in the second industrial revolution, an
industrial revolution which like the first was based on energy, velocity and
speed.
2.4 Thinking About Production in the Electrical Era

Political economy, the science of wealth, entered the 20th century with a
theory of production that dated back to Adam Smith’s writings in the mid-
18th century, augmented in the neoclassical case by the assertion that capital
was physically productive. Despite the many references to engineering, to
classical mechanics and to energy in general, the science of wealth continued
to ignore the advances made in its sister field of thermodynamics, which
like political economy, was born of the steam engine.
As it turned out, the early 20th century witnessed a second energy surge,
as it were, with the widespread use of a new power drive technology, namely
electric motors/dynamos. Like other energy-related innovations, it would
open the way to even greater energy use per unit of capital/labor, again in
the form of greater machine speeds, often referred to as machine speed-ups.
In short, shafting and belting would be replaced by electric unit drive,
untethering material processes and subprocesses from the constraints of
shafts and belts, resulting in autonomy and, more importantly, faster
machine speeds, the latter being achieved by increasing kinetic energy.
Like the Boulton-Watt reciprocating steam engine, this development/
innovation would go on to define an era, known commonly as the second
industrial revolution. Not only were belting and shafting-driven material
processes speeded up, processes that had not been mechanized, were. The
overall result was a massive increase in productivity and output. As the
General Electric Company was to describe it, ”Today is the Age of Speed.”
As Alfred D. Chandler pointed out perspicaciously, speed not physical scale
was of the essence.
Unfortunately, this massive, even paradigmatic shock failed to appear
on the economics radar. While productivity records were being shattered,
the profession remained impervious. After all, nothing had changed. In fact,
if anything, employment and capital spending were down. As it turned out,
electric unit drive was less cumbersome and less costly than belting and
shafting, resulting in a decrease in capital intensity, the latter being
measured in terms of costs. It is our view that this was perhaps the most
egregious failing of neoclassical production theory as it de facto preempted
a deeper understanding of the changes that were transforming not only the
U.S., but the World in general.
Not surprisingly, discordant voices began to appear. Clearly, a revolution
was underway. Productivity records were being shattered. New, electric-
powered goods were changing life as we knew it. These included household
appliances, electric-powered tools, electric-powered transportation, etc.
As it turned out, this prompted a response on the part of a handful of
economists, one that like Marx and Jevons’ forays in the 19th century,
focused on the role of energy/force/power in production. For example,
consider the cases of Columbia University professor Rexford G. Tugwell
and Yale University professor Irving Fisher. In a book entitled Industry's
Coming of Age, published in 1927, Tugwell described various changes to
industry, including “the bringing into use of new and better power resources
more suited to our technique, more flexible and less wasteful.”
The electrification of industry has now progressed to the extent of between

55 and 60 per cent completion. So widespread an adoption of this new
flexible means of moving things cannot have taken place without numerous
secondary results in lowering costs, improvements in quality, and a
heightened moral among workers. For the new power is not only cheaper to
use; it is also cleaner, more silent and handier. On the whole, the
electrification of industry must be set down as the greatest, single cause of
the industrial revolution. (Tugwell 1927, 182)
In The Stock Market Crash and After, Irving Fisher ranked the
electrification of U.S. industry as one of the contributing factors behind the
stock market's rise in the late 1920s. Changing economic fundamentals, he
maintained, were the underlying cause.
But after 1919, something happened. The implications of which are not yet
sufficiently gauged. It was of enough significance to cause President
Hoover's Committee on Recent Economic Changes to remark that
“acceleration rather than structural change is the key to an understanding of
our recent economic developments.” The committee added: “But the breath
of the tempo of recent developments gives them new importance.” What
happened was indicated by the fact that in the United States, eight million,
three hundred thousand workers produced in 1925 one quarter more than
nine million wage workers turned out in 1919. The new indexes of the
Federal Reserve Board measuring production record this gratifying advance
which reflects an increase in the American standard of living. The indexes
cover, directly and indirectly, four-fifths of the industrial productivity of the
nation directly in about thirty-five industries, and collaterally, in many more.
The general volume of production had increased between 1919 and 1927 by
46.5 percent, primary power by 22 percent, and primary power by wage
earner by 30.9 percent (between 1919 and 1925) and productivity per wage
worker by 53.5 percent between 1919 and 1927. (Fisher 1930, 120)
While insightful, their descriptions of the new power-drive technology

and its effects on wealth failed to make any headway in a science that was
devoid of energy, focusing instead on capital and labor. Despite the fact that
analyst after analyst pointed to electric unit drive as the key to the
revolution, the science of wealth remained unresponsive.
Some argued in favor of a new paradigm, one in which energy/force/work
would have its rightful place. This was the case of F.G. Tryon of the
Institute of Economics (Brookings Institution), who in 1927 denounced the
little attention paid to power by economists, and Woodlief Thomas of the
Federal Reserve Board, who attributed the “Increased Efficiency of
American Industry" to the increasing use of machinery and power.
According to Tryon:
Anything as important in industrial life as power deserves more attention
than it has received by economists. The industrial position of a nation may
be gauged by its use of power. The great advance in material standard of life
in the last century was made possible by an enormous increase in the
consumption of energy, and the prospect of repeating the achievement in the
next century turns perhaps more than anything else on making energy
cheaper and more abundant. A theory of production that will really explain
how wealth is produced must analyze the contribution of this element of
energy……These considerations have prompted the Institute of Economics
to undertake a reconnaissance in the field of power as a factor of production.
One of the first problems uncovered has been the need of a long-time index
of power, comparable with the indices of employment, of the volume of
production and trade, of monetary phenomena, that will trade the growth of
the factor of power in our national development. (Tryon 1927, 281)
In an article in the American Economic Review entitled “The Economic

Significance of the Increased Efficiency of American Industry,” Thomas
attributed the “striking changes” in American industry” to power-related
developments:
Large-scale production is dependent upon the machine process, and the

increasing use of machinery and power and labor-saving devices has
accompanied the growth in size of productive units. The growing use of
power in manufacturing, for example, is reflected in the increase in
horsepower of installed prime movers. This does not tell the whole story,
moreover, for owing to increased use of electricity, the type of power used
is now more efficientrequiring less fuel and labor for its production. Out
of a total installed horsepower in factories of thirty-six million in 1925,
twenty-six million or 72 per cent was transmitted to machines by means of
electric motors, as compared with 55 percent in 1919, 30 per cent in 1909,
and only 2 per cent in 1899. Between 1899 and 1925 horsepower per person
employed in factories increased by 90 percent and horsepower per unit of
product increased by 30 percent.... Power has been substituted for labor not
only through machines of production but also in the form of automatic
conveying and loading devices. (Thomas 1928, 130)
Unfortunately, Tryon and Woodlief were the exceptions. Mainstream

political economists were, for the most part, unaware of the profound
changes thrust upon U.S. industry by the application of a new power drive
technology, namely electric unit drive. Which explains the rise of alternative
paradigms both within the realm of political economy and from fields such
as engineering. In short, its failings led to what we refer to as the
“Economics of Abundance,” spearheaded by MIT political economist Stuart
Chase, and to the Technocracy Movement, led by Howard Scott. Among
the leading authorsand there were manywere Edward A. Filene, Stuart
Chase, Henry Ford, John Atkinson Hobson, Howard Scott, Frederick
Soddy, Rexford Tugwell, and Thorstein Veblen. Defining the economics of

abundance were (i), the role of energy in general, and electric power in
particular in production processes, and consequently in early 20th century
U.S. economic growth, (ii) the purported flawed nature of the for-profit
economy, distribution system, and/or money supply process, and (iii) the
call for change, whether it be in the conduct of business or government
regulation. A good example is the following excerpt taken from Chapter 1
of Stuart Chase's The Economics of Abundance, entitled “Forty to One":
Suppose that the thirteen million people living in the United States in 1830
had awakened on the morning of January 1, 1831, with forty times the
physical energy they had gone to bed with the night before. An active picture
meets the mind's eye; a very active picture. A lumberman can fell forty times
as many trees in a week, a housewife sweeps forty times as many square feet
of floor; forty barns can be built in the time hitherto required for oneand
forty chests and forty chairs. Porters can transport forty times their
accustomed load in a day; weavers ply their shuttles forty times as fastif
the shuttles can brook the strain; and children raise forty times their normal
rumpus.
Assuming no increase in the invention of labor-saving devicesand
where would be the point with such an exuberance of labor availablewhat
might we logically expect in the way of economic changes in a culture
essentially handicraft? From an economy of scarcity, with barely enough to
go around, the young republic would almost immediately enter an economy
of abundance. (Chase 1933, 1)
In the same chapter, Chase defines “Abundance" in terms of a series of

propositions.
1. A condition where the bulk of the economic work is performed not by

men, but by inanimate energy, drawn from coal, oil and water power.
Such a condition was reached in the United States towards the close of
the nineteenth century. Circa 1880.
2. A point at which living standards per capita reach an average which is,
at least potentially, twice as high as ever obtained under scarcity
conditions. Reached. Circa 1900.
3. A point at which the curve of invention, following, as it does, a
geometric increase, becomes the dominant factor in economic
lifeprecisely as the Nile was the dominating factor in the economic life
of Egypt. Circa 1870.
4. A point at which the scientific method supersedes the use and want of
the craftsman in the production of most material goods. Circa 1900.
5. A point where output per man hour becomes so great that total
productive labor must thereafter decline, even as output grows. A point
at which labor ceases to be a measure of outputʊas it always has been

in preceding ages. Circa 1920.
6. A point at which overproduction carries a more serious threat to the
financial system than shortage. Circa 1880.
7. A point at which specialization has destroyed all practicable local self-
sufficiency and made economic insecurity for all classes latent, growing,
and ultimately intolerable, given no change in financial methods. Circa
1900, with the closing of the American frontier.
8. A point at which consumption becomes a greater problem than
production. Circa, 1920. “Our economy,” says F.L. Ackerman,”is so set
up that it produces goods at a higher rate that it produces income with
which to purchase them."
9. A point at which the industrial plant is, substantially, constructed,
requiring relatively smaller outlays for capital goods in the future, and
where pecuniary savings are not only unnecessary in their old volume,
but seriously embarrassing. Circa 1925.
10. A point where, due to the presence of the technical arts, costs, prices,
interest rates, debts, begin a descent with zero as their objective. Circa
1920. (Chase 1933, 12)
Clearly, frustration levels were on the rise as the thinking about production
had not kept pace with technology, making for dissidence. As it turned out,
this spilled over in the related field of engineering where the failure of
economics to adequately describe the paradigm changes underway at the
time, led to the emergence of a new approach not only to wealth, but to the
very workings of an industrial economy. This was Technocracy, which
offered up a new approach to understanding wealth, and a blueprint for
guaranteeing the transition to the new higher equilibrium growth path.
Consider, for example, the following excerpt from Howard Scott's Introduction
to Technocracy, published in 1933:
A century ago, these United States had a population of approximately 12,

000,000 whereas to-day our census figures a total of 122, 000, 00a tenfold
increase in the century. One hundred years ago, in these United States, we
consumed less than 75 trillion British thermal units of extraneous energy per
annum, whereas in 1929, we consumed approximately 27, 000 trillion
British thermal unitsan increase of 353-fold in the century. Our energy
consumption now exceeds 150,000 kilogram calories per capita per day;
whereas in the year 1800 our consumption of extraneous energy was not less
than 1600 or more than 2000 kilogram calories per day. The United States
of our forefathers, with 12,000,000 inhabitants, performed the necessary
work in almost entire dependence upon the human engine, which, as its chief
means of energy conversion, was aided and abetted only by domestic
animals and a few water wheels. The United States to-day has over one
billion installed horsepower. In 1929, these engines of energy conversion,
though operated only to partial capacity, nevertheless had an output that

represented approximately 50 percent of the total work of the world. When
one realizes that the technologist had succeeded to such an extent that he is
to-day capable of building and operating engines of energy conversion that
have nine million times the output capacity of the average single human
being working an eight hour day, one begins to understand the acceleration,
beginning with man as the chief engineer of energy conversion and
culminating with these huge extensions of his original one-tenth of a horse
power. Then add the fact that of this 9,000,000-fold acceleration, 8,766,000
has occurred since the year 1900. (Scott 1933, 42)
While well-intentioned, their forays into the science of wealth were

largely unsuccessful, owing in large part to the chasm which separated their
thinking, from that of political economy. The latter attributed wealth to
labor and capital, while Technocracy focused, almost exclusively on energy.
Put differently, Technocracy was based in large measure on classical
mechanics and thermodynamics, where the role of tools (capital) and
supervision (labor) are minimized, while neoclassical production theory
focused exclusively on what, to engineers, were non-physically productive
factor inputs. Clearly, convergence was virtually impossible, which
explains why Technocracy and the Economics of Abundance failed to make
inroads in the thinking about production.
2.5 Thinking About Production in the OPEC Era

Having successfully thwarted any and all attempts at a rapprochement
with physics and engineering, the thinking about production emerged from
WWII relatively unscathed. Wealth continued to be seen as an increasing
function of physically-productive capital and labor. However, a number of
challenges lay ahead, one of which being growth accounting. In the afterwar
years, new data on labor and capital, a by-product of the Keynesian
revolution, showed that labor and capital growth could not explain output
growth. In other words, the standard factor inputs were unable to account
for the post-WWII growth rates. And so was born, the Solow residual. The
thinking now was that labor and capital explain roughly half the growth,
with the other half being the result of technological change. Just what it was
that caused the historically-high growth rates of this period was unknown.
And so began a new era in the thinking about production, namely new
growth theory, where the emphasis is on ideas. As the Solow residual is, as
its name implies, a residual, it is unbounded and unstructured, by which it
should be understood, lacking a basis in science.
Still, throughout this period energy continued to haunt as it were the

thinking about production. The first salvo was a book by Nicholas
Georgescu-Roegen, entitled The Entropy Law and Economic Processes,
which caught the eye of the profession. In short, Georgescu-Roegen argued
that economic processes are not reversible – in short, the arrow of time.
While heralded by leading figures in the profession, notably Paul
Samuelson, his work fell flat. With the benefit of hindsight, this was not
altogether surprising as the basic thinking about production namely
neoclassical production theory was devoid of the energy input, focusing on
labor and capital.
The second instance was the OPEC-induced energy crises of the 1970s
which witnessed a fourfold increase in energy prices. Given the absence of
energy from mainstream production theory, most analysts downplayed the
role of the price hikes on economic growth. However, with the passage of
time, it became obvious that growth rate in the West had decreased
substantially, raising questions of causality. “Could higher energy prices be
the cause?” many asked.
Clearly, the time was nigh: the thinking about production would have to
somehow incorporate energy if only to test a number of hypotheses,
including could higher energy prices, by reducing the demand for energy,
be the cause of the productivity slowdown, and second, could the latter be
the result of what could be referred to as energy-capital complementarity.
That is, higher energy prices would have reduced growth by reducing the
rate of growth of capital.
Ernst Berndt and David Wood, in 1975, presented the KLEMS production
function which consisted, in large measure, of a standard capital and labor
neoclassical production function augmented by three new, physically-
productive inputs, namely energy, materials and services. This was heralded
as an important achievement as evidenced by the fact that it has since
become the standard in empirical work.
The upshot of the voluminous empirical literature that followed was
clear: energy doesn’t matter as evidenced by output elasticities less than
0.05. Capital, labor, materials and services–all non-physically productive
factor inputs–were deemed to be more important than the only physically-
productive input, energy. In fact, history was repeating itself as like William
Stanley Jevons in the mid-19th century who had ascribed much importance
to energy, eventually eliminated it from his thinking about production,
KLEMS acknowledged the presence of energy, but minimized its
importance.
Berndt and Wood’s and indeed all KLEMS-based results (output elasticities
and cross-elasticities) are obtained in what can be qualified as a roundabout
manner. In general, the various elasticities are obtained by assuming both

perfect competition and factor market equilibrium. That is, the price of a
factor is, in equilibrium, equal to its marginal product multiplied by its price.
With data on factor shares and prices, estimates of the output elasticities are
obtained. In the case of electric power and other forms of industrial energy,
this is not the case as prices are typically regulated – that is, set by regulatory
agencies.
Interestingly but not surprisingly, this was overlooked by German
theoretical physicist, Reiner Kummel, who in 1982 estimated capital, labor
and energy elasticities directly using LINEX, a technique he developed.
Using data for the U.S., Germany and Japan, he reported energy output
elasticities in the 0.50 range. Similar results were reported in Beaudreau
(1995) using a simple logarithmic regression for the U.S., Germany and
Japan.
Unfortunately, these findings were proverbial drops of water in the
larger bucket of the KLEMS-based empirical literature which resoundingly
rejected and rejects the physics-based view that energy was and is the key
driver of manufacturing and thus of growth. Add to this the view, shared by
the U.S. Bureau of the Census, that energy is not a productive factor input,
but rather a material input in the same category as raw materials and semi-
finished inputs and you get the current view of energy as an essential but
unimportant factor input.
Contrast this with the view from ecologists and ecological economists
who, perhaps in reaction to such findings, developed a whole new sub-field
based in large measure on applied physics. Much like the Technocrats of
the late 1920s and 1930s, they felt that standard neoclassical model was
incomplete and thus inadequate, preferring to couch production in the
universal context of entropy/negentropy. Whether an intended consequence
or not, this choice served to widen the chasm between it and the mainstream,
thus sealing its fate so to speak.
2.5.1 Beyond the Obvious

As shown in this chapter, it has taken the science of wealth roughly three
centuries for any progress to be made in the direction of recognizing the
fundamental principles of classical mechanics, or put starkly, Newtonian
physics. And what progress that has been made has been limited by flawed
empirics–KLEMS-based output elasticities.
While this constitutes an encouraging development, the future remains
in doubt for at least two reasons. First, estimates of energy output elasticities
continue to be on the low side, and second, the lack of solid underlying
fundamentals. Specifically, while most physicists, ecologists, ecological

economists and engineers recognize the keyin fact fundamentalrole
energy plays in all material processes and indeed in the rise of Western
civilization, just how increasing energy use affects productivity remains a
mystery. This, we argue, owes to methodological issues in economics,
notably the use of unstructured production functions. Adding the energy
input to capital and labor, and parsing how an increase in the energy/capital
or energy/labor input increases the output/capital or output/labor ratio, is
poor science.
Lacking are the underlying mechanics. For example, how, in actual fact,
does an increase in energy use per unit capital or labor result in greater
output? In other words, just how is the increase achieved on the plant floor?
Surprisingly, few have addressed what we feel is the key issue.
One such attempt is Beaudreau (2017, 2020) where the author develops
a kinetics approach to production, one in which output is increasing in
process speed, which in turn is increasing in energy use per unit of capital.
The rise of Western industrial civilization, he maintains, owes to the nearly
constant increase in process speeds starting in the late 18th century and
continuing unimpeded until the mid-to-late 20th century. The march to
modernity, he maintains, started with the atmospheric steam engine,
followed by the reciprocating steam engine, steam turbines, belting and
shafting, and finally electric unit drive, the latter providing the wherewithal
to increase process speeds to their current upper limits.
As he shows, references to speed or velocity in the production process
abound in the 19th and 20th century literature, despite its absence in
theoretical constructs. Which is where his work fits in, namely by providing
a consilient framework in which to analyze the role of energy in general and
energy deepening in growth.

As we have shown in this chapter, the thinking about production started
in earnest in the 18th century, peaked in the 19th century and has remained
relatively unchanged throughout the 20th century. Table 2.1 presents a
summary of the most important thoughts regarding material processes in
general and material processes in economics in particular, starting with
Adam Smith and ending with New Growth Theory. Importance in this case
does not necessarily connote inclusion and status. Instead, in many
instances, it refers to scientific relevance, defined as being consistent or
consilient with material processes in other fields.
The upshot of this chapter is the presence of a form of intellectual stasis,

one that began with Adam Smith and continues to this day, one that has
resisted many well-intentioned efforts to steer the debate on to more
scientific grounds, and one that has been and continues to be more
impressionistic than scientific. Surprisingly, the puzzles, the mysteries and
the dead-ends in production, have not prompted new directions in modeling
material processes, but rather have contributed to increasing enigmatization
via the Solow residual, as evidenced by the past four decades of New
Growth Theory.
Table 2.1: Key Developments in Thinking About Production
Thought Author(s) Year

Classical Theory Adam Smith 1776
Marxian Production Karl Marx 1848
Theory
Coal as the Source of William Stanley Jevons 1865
Wealth
Neoclassical Production William Stanley Jevons, 1872
Theory Alfred Marshall, John Bates Clark
Energy Theory of Frederick Soddy 1914
Production
Technocracy Howard Scott 1933
The Solow Residual Robert Solow 1955
Entropy Theory of Nicholas Georgescu- Roegen 1972
Production
Endogenous Growth Paul Romer 1986
Theory
As was argued, this owes in large measure to what we refer to as the

open-ended nature of standard production theory which, in turn, can be
attributed to the lack of any formal structure. Despite being a material
process which implies that it is physical in nature, the laws of physics are
conspicuous by their absence. This leaves the door wide open to any and all
thoughts, fleeting and confounding. In short, the lot/fate of a science that
allowed and continues to allow a fudge factor.
2.7 References
Beaudreau, Bernard C. 1995. The Impact of Electric Power on Productivity:
The Case of U.S. Manufacturing 1958–1984. Energy Economics. Vol.
17:231–236.
Beaudreau, Bernard C. 2017. “The Economies of Speed, KE=1/2mv2, and

the Productivity Slowdown.” Energy.
Beaudreau, Bernard C. 2020. The Economics of Speed: Machine Speed as
the Key Factor in Productivity. Switzerland: Springer.
Babbage, Charles. 1832. On The Economy of Machinery and Manufactures.
London: Charles Knight.
Berndt, Ernst and David O. Wood. 1975. “Technology, Prices and the
Derived Demand for Energy.” The Review of Economics and Statistics:
259–268.
Chandler, Alfred D. Jr. 1977. The Visible Hand, The Managerial Revolution
in American Business. Cambridge, MA: Harvard University Press.
Chase, Stuart. 1934. The Economy of Abundance. New York, NY:
MacMillan Company.
Engels, Frederick and Karl Marx. 1848. The Communist Manifesto.
London: The Communist League.
Georgescu-Roegen, Nicholas. 1972. The Entropy Law and the Economic
Process. Cambridge, MA: Harvard University Press.
Jevons, William S. 1865. The Coal Question. London: MacMillan and Co.
Jevons, William S. 1871. The Theory of Political Economy. London: Pelican
Books.
Kummel, Reiner. 1982. “The Impact of Energy on Industrial Growth.”
Energy. 7, no. 2: 189- 201.
Marshall, Alfred. 1890. Principles of Economics. London: MacMillan.
Marx, Karl. 1867. Das Kapital. Chicago: Encylcopaedia Britannica (1992).
Owen, Robert. 1817. The Life of Robert Owen. London: Cass, 1967.
Owen, Robert. 1820. A New View of Society and Other Writings, London:
J.M. Dent and Sons, Ltd. 1927.
Ricardo, David. 1817. The Principles of Political Economy and Taxation.
New York: Everyman's Library.
Scott, Howard et al. 1933. Introduction to Technocracy. New York, NY:
The John Day Company.
Smith, Adam. 1776. An Inquiry into the Nature and Causes of the Wealth
of Nations. Chicago: Encyclopaedia Britannica, 1990.
Sciences upon State Stewardship. London: Hendersons.
Solow, Robert M. 1974. “The Economics of Resources or the Resources of
Economics.” American Economic Review. 64, no. 2: 1-14.
Tryon, F.G. 1927. “An Index of Consumption of Fuels and Water Power,”
Journal of the American Statistical Association. 22: 271-282.
Tugwell, R.G. 1927. Industry’s Coming of Age. New York, NY: Columbia
University Press.
3
THE IMPACT OF ELECTRIC POWER

ON PRODUCTIVITY:
A STUDY OF U.S. MANUFACTURING 1950-84
Abstract
According to the historical record, the increased use of energy in general

and electric power in particular stands as one of the key factors in
productivity growth. Growth accountants, however, find little support for
this view, rejecting claims that the productivity slowdown can be attributed
to the lower rate of growth of the energy input in U.S. manufacturing. In this
study, an attempt at reconciling the historical record with the empirical facts
is made. Specifically, the commonly used factor income shares in the
relevant Divisia input-growth index, are replaced by the estimated output
elasticities. The latter attribute a considerably larger role to electric power
than previous studies indicated. These are then used to reevaluate the
sources of growth in US manufacturing.
3.1 Introduction
It is generally agreed that the electrification of U.S. industry contributed
greatly to raising productivity and output (National Bureau of Economic
Research 1929, Schurr 1983, Rosenberg 1983, Jorgenson 1983, 1984). For
example, as early as 1929, the Committee on Recent Economic Changes,
chaired by Herbert Hoover, identified the increased use of electric power as
the most significant change in U.S. industry (National Bureau of Economic
Research 1929). More recently, Nathan Rosenberg observed that “the
spreading use of electric power in the 20th century has been associated with
the introduction of new techniques and new arrangements which reduce
total costs through their saving of labor and capital” (Rosenberg 1983, 24).
He concludes that “there has been a very wide range of labor-saving
innovations throughout industry which have taken an electricity-using form.
As a consequence, greater use of electricity is, from a historical point of
40 3. The Impact of Electric Power on Productivity
view, the other side of the coin of a labor-saving bias in the innovation
process” (Rosenberg 1983, 36).
In a study of 35 industrial sectors in the United States, Dale Jorgenson
found that technical change was electricity using in 23 of the 35. The decline
in real electricity prices prior to 1973, he argues, prompted increased
electrification via the substitution of electricity for other forms of energy
and through the substitution of energy for other inputsespecially labor
(Jorgenson 1983, 21). Electrification, he concludes, plays a fundamental
role in productivity growth.
Studies of the various sources of output growth, however, paint an
altogether different picture of the role of energy in general and electric
power in particular in the growth process (Denison 1985, Gollop and
Jorgenson 1980, Gullickson and Harper 1987). Nowhere is this more
apparent than in the recent debate over the productivity slowdown. Time-
series data show that output growth, however, measured, has fallen
dramatically from the mid-1970s on. For example, Gullickson and Harper
(1987) report that manufacturing output increased at an average annual rate
of 4.2% from 1949-73, but decreased to 0.6% from 1973 to 1984. A number
of writers pointed to the energy crisis as a cause. After all, everything
seemed to fit: energy prices, including electricity, began their upward climb
in 1973 which corresponded to the time at which productivity growth began
its downward slide.
Growth accountants, however, dismissed such hypotheses (Denison
1985). While they concurred that the rate of growth of energy consumption
had fallen dramatically, it alone could account for no more than 0.10 of the
decrease in output growth (i.e. 4.2%) (Gullickson and Harper 1987). Central
to their argument is the weight attributed to energy in deriving measures of
total factor productivity, notably 0.02-0.04. Put differently, the share of
energy in total factor income is largely insignificant. The energy crisis, they
surmise, could not explain the productivity slowdown.
This leaves us in a conundrum. The bulk of the historical evidence points
to energy in general, and electric power in particular, as an important cause
of productivity and output growth, yet, on the other hand, most, if not all,
studies of the sources of growth find energy to be marginally important.
Thus, either the historical record is incorrect, or previous studies have
underestimated the role of energy in general and electric power in particular
in the process of economic growth. In a recent study, Hisnanick and Kymm
(1992) provide new evidence which suggests that part of the productivity
slowdown can be attributed to the energy crisis. Specifically, they show that
by disaggregating broadly defined energy into a petroleum component and
a non-petroleum component, decreased energy use in U.S. manufacturing

has contributed to the productivity slowdown (Hisnanick and Kymm 1992).
In this chapter, the role of electric power in US economic growth is
reexamined critically. Our starting point is the set of weights used in
constructing the relevant Divisia index of factor inputs.1 Specifically, most
studies use factor shares in lieu of actual estimates of the relevant output
elasticities. Implicit in this approach is the assumption that factor markets
are perfectly competitive (Christensen and Jorgenson 1970, Gollop and
Jorgenson 1980). However, since the market for electric power in the USA
is largely regulated, there is no reason to believe that the marginal value
product of electric power would be equal to its price. This led us to estimate
the various input elasticities directly using output and input data (Table
3.1).2 Specifically, manufacturing value-added for the period 1950-84 was
regressed against electric power consumption, employment and capital.3
Our results show that indirect estimation techniques, by implicitly assuming
perfectly competitive factor markets, generate downwardly biased estimates
of the value-added electric power elasticity and upwardly biased estimates
of the value-added employment and value-added capital elasticities.
The resulting set of estimates provided the basis for a wholesale revision
of the role of electric power in output and productivity growth. First, we
show that by replacing factor shares with actual values of the elasticities,
the Solow residual disappears, as does the total factor productivity residual
in conventionally defined productivity growth (Gullickson and Harper
1987). The productivity slowdown, we find, is explained totally by the
energy crisis.
1 The relevant Divisia input indices are defined as follows: σ௡௜ୀଵ ߚ௜ ‫ݔ‬௜ where Ei is the
relevant factor output elasticity and xi, is the relevant rate of growth of factor i.
2 Previous studies (e.g. Berndt and Wood 1975) estimate the relevant elasticities
indirectly using cost functions.

3 Jorgenson (1983) also disaggregates energy into an electric component and a non-
electric component.
Table 3.1: Manufacturing Output and Input Data 1950-84a
Year Q EP L K
1950 79.06 56.79 93.80 74.00
1951 84.25 62.74 99.26 77.28
1952 88.14 65.62 102.01 80.63
1953 97.34 74.69 108.23 83.78
1954 92.28 77.57 101.44 86.64
1955 104.95 93.73 106.05 89.48
1956 108.92 101.28 108.24 93.82
1957 107.13 101.94 107.77 97.89
1958 100.00 100.00 100.00 100.00
1959 112.33 109.81 104.15 101.67
1960 112.17 113.38 104.71 104.29
1961 110.96 114.91 101.99 106.69
1962 119.58 121.60 104.74 109, 27
1963 126.59 127.31 105.24 112.09
1964 133.90 136.89 106.89 116.01
1965 144.58 142.17 111.85 122, 09
1966 155.62 150.31 118.01 129.95
1967 157.39 158.44 119.90 137.24
1968 164.68 169.66 121.12 143.84
1969 167.78 181.79 124.20 150.74
1970 156.89 182.95 118.59 156.95
1971 157.40 187.13 112.99 161.97
1972 171.69 201.92 116.92 166.97
1973 185.22 218.46 122.36 173.05
1974 186.18 218.51 121.33 181.22
1975 166.83 206.75 111.35 187.56
1976 182.32 220.51 114.65 193.90
1977 195.95 227.51 120.05 200.81
1978 204.49 230.96 124.74 209.20
1979 208.94 234.58 128.10 217.79
1980 190.57 224.55 125.21 226.49
1981 186.87 225.72 122.67 234.84
1982 173.26 209.76 115.53 241.05
1983 179.66 214.18 113.16 244.95
1984 192.09 233.15 115.77 250.94
aThe output and input indices used in this study were derived as follows. The
manufacturing value-added index (i.e. Q) was constructed using real value-added

data for manufacturing, 1950-84 (U.S. Department of Commerce 1985, 1986); EP
using total electric power consumption (purchased and generated), 1950-84 (U.S.
Department of Commerce 1985, 1986), Fuels and Electric Energy; L using total
employment data for manufacturing (U.S. Department of Commerce 1985, 1986);
and K using revised capital stock data for manufacturing reported in U.S.
Department of Commerce 1986 Survey of Current Business.
3.2 The Marginal Product of Electric Power

As is customary in the growth literature, we assume the existence of a
well-behaved, twice differentiable, monotonic and quasi-concave production
function, such as Equation 3.1, where Q is value-added, EP is electric
power, L is employment and K is capital.4 Christensen and Jorgenson (1970)
and Gollop and Jorgenson (1980) have shown that the rate of growth of total
factor productivity, tfp = dTFP/TFP, can be defined as tfp =qvepep+vkk+vll
, where q = dQ/Q, ep = dEP/EP, 1=dL/L, and k=dK/K, and vi = the weighted
average of the ith factor share over the discrete time interval i = EP, L, K
(Gullickson and Harper 1987 ):
ܳ௧ = ݂[‫ ݐܲܧ‬, ‫ ݐܭ‬, ‫] ݐܮ‬ 3.1
The use of factor shares in lieu of the relevant elasticities is based on a

number of assumptions, not the least of which is the existence of perfectly
competitive factor markets.5 Referring to Table 3.2, we see that broadly
defined energy accounts for 3.0-12.0% of manufacturing value-added.
Broken down into the relevant component parts, because electric power
accounts for approximately 50% of the total cost of purchased fuels and
electricity (U.S. Department of Commerce 1985, 1986), it follows that its
share of total manufacturing value-added stands at approximately 1.5-
6.0%.6 Electric power's factor share (i.e. 1.5-6.0%), it therefore follows,
should reflect its contribution to overall manufacturing value-added.7 Put
differently, the relevant output elasticity for electric power should lie
somewhere between 0.015 and 0.06. As pointed out above, this stands in
stark contrast with the bulk of the historical evidence described above.
4 Implicitly, we assume that total manufacturing output and materials are weakly
separable, which is also known as the Leontief aggregation condition. This can be
justified on a number of grounds. For example, by contrast with Berndt and Wood
(1975), our purpose here is not to predict the demand for energy, specifically,
electric power. Second, manufacturing output and value-added were found to be
highly correlated in the chosen sample period (U= 0.9652).
5 Among the others is a linear homogeneous production function (i.e. constant
returns to scale).
6 In 1967, 1974 and 1981 total energy costs in US manufacturing stood at USD$7,
691.7 million, USD$19, 468.3 million and USD$55, 255.1 million respectively. The
corresponding costs of electric power (purchased and generated) are USD$4, 398.11
million, USD$9576.26 million and USD$27, 609.85 million. Thus, the ratio of
electric power to total energy in US manufacturing stood at 0.5717 in 1967, 0.49188
in 1974 and 0.4996 in 1981.
7 Formally, V
EP = O.5VG.
Table 3.2: Factor Income Shares, U.S. Manufacturing Value-Added
Period vn vk ve
Berndt and Wood (1975)a 1947-71 0.6736 0.2902 0.0360
Gullickson and Harper (1987) 1949-83 0.7162 0.1590 0.1246
Denison (1962)b 1967-82 0.8263 0.1021
Kendrick and Grossman (1980) 1948-73 0.6552 0.3448
US Dept of Commerce (1985) 1967-81 0.4282 0.5233' 0.0459
aCost-based estimates of the relevant output elasticities. The values reported are net
of materials (i.e. value-added).

bEnergy and/or electric power were not included.
cIndirect taxes were netted from value-added.
Table 3.3: KLEP Regression Results, 1950-84'
Independent variables (1) (2)

Electric power (EP) 0.533046 0.658879
(10.791) (34.728)
Employment (L) 0.418822
(19.750)
Capital (K) 0. 064250
(1.132)
Constant 5.294295
(17.595)
R2 0.98474 0.97337
F (2, 32) 1032.52 1206.03
Dependent variable: manufacturing value-added (Q)
Mean of dependent variable: 140.34
Number of observations: 35
In light of this, we decided to estimate the relevant elasticity directly.

Specifically, data obtained from the U.S. Department of Commerce's
Annual Survey of Manufactures and Survey of Current Business, on value-
added, electric power consumption, employment and capital in manufacturing
from 1950 to 1984 were used to estimate the relevant Cobb-Douglas KLEP
production function, given as ܳ = ‫ܲܧ‬ఉభ ‫ܮ‬ఉమ ‫ܭ‬ఉయ .
The results, presented in Table 3.3, show a value for E1 the value-added
electric power elasticity, of 0.533043, which differs significantly from the
electric power shares of 0.015-0.06 reported above.8 That is, for the period
8 lmplicitly, we assume that E , the scalar is equal to unity ((E =1). When E is
0 0 0
included, restrictions on E1, E1, and E1 are needed. Specifically, by imposing linear
homogeneity, we obtained the following E0, E1, E1, and E1 estimates (t-statistic): E0=
extending from 1950 to 1984, as a 1% increase in electric power

consumption in manufacturing resulted in a 0.53% increase in
manufacturing value-added.9 When electric power is included as the sole
regressor, the relevant elasticity is 0.658879 (see column 2). In both cases,
all coefficients have the relevant signs; however, E3, the estimated value-
added capital elasticity, was found to be statistically insignificant at the 95%
level. Clearly, electric power as a factor input in manufacturing is more
important than previously believed. Lastly, the estimated elasticities
confirm the presence of constant returns to scale. Specifically, the three
elasticities sum to 1.016 which is not statistically different from unity.
Table 3.4: Simple Growth Rates for Output and Inputs in US

Manufacturing, Selected Periods'
1950-84 1950-73 1974-94

Value-added 2.995 4.217 0.330
Aggregate Input (AI)a 2.932 4.129 0.321
Electric Power (EP) 4.455 6.226 0.591
Labor (L) 0.784 1.375 -0.503
Capital (K) 3.564 3.651 3.378
a ଵ ௗா௉ ଵ ௗ௅ ଵ ௗ௄
‫ߚ = ܫܣ‬ா௉ + ߚ௅ + ߚ௄ , where the Ⱦᇱ s are the output elasticities..
ா௉ ௗ௧ ௅ ௗ௧ ௄ ௗ௧
3.3 Accounting for the Sources of Growth in US

Manufacturing 1950-84
Our findings led us to reexamine the sources of growth in US
manufacturing. Clearly, judging from our estimates of Ei, i = EP, L and
K, the importance accorded to electric power consumption (EP) in growth
accounting, will increase, at the expense of both employment (L) and capital
(K). Electric power consumption and hence growth in electric power
consumption play a more important role output growth than previously
thought (Gullickson and Harper 1987). Table 3.4 reports the relevant growth
rates for manufacturing value added (Q), electric power (EP), labor (L) and
capital (K), as well as the relevant Divisia index for aggregate input
(Hisnanick and Kymm 1992, Gullickson and Harper 1987) for three-time
intervals: 1950-84, 1950-73 and 1974-84.
1.073 (11.639); E1= 0.5372 (12.883); E2= 0.3997 (23.377); and, E3 =0.0630 (1.313).
The corresponding R2 is 0.984.
9 This result was found to be robust with regard to the sample period. Estimates for
the period 1964-84 were not statistically different from those reported in Table 3.3.
Unlike previous studies which found a sizable gap between the actual
rates of growth of output and aggregate input, our results show that growth
in U.S. manufacturing value-added is fully explained by growth in the
relevant Divisia index of factor input growth. For the complete period of
1950-84, manufacturing value-added increased at an average annual rate of
2.99%, while the aggregate input increased at 2.939%. For the first
subperiod 1950-73, manufacturing value-added increased at 4.217%, while
the aggregate input increased at 4.129%. Lastly, in the second subperiod,
1974-84, manufacturing value-added increased at an average annual rate of
0.3309, while the aggregate input increased at 0.3215.10
Table 3.5: Changes in Labor Productivity, Total Factor Productivity,

and the Labor Intensity Ratios for the Pre-1973 and Post-1973 Periods
1950-84 1950-73 1974-84

1p 2.211 2.842 0.834
tfp 0.077 0.111 0.002
vEP(ep 1) 1.956 2.585 0.583
vK(k 1) 0.178 0.146 0.249
Chief among the causes of growth in manufacturing value-added is

electric power consumption, increasing at an average annual rate of 4.455%
over the entire period 1950-84. Per worker consumption of electric power
in this period goes from 12, 534 kilowatt hours in 1950 to 41, 688 kilowatt
hours in 1984, a total increase of 232% (U.S. Department of Commerce
1985, 1986). Prior to the energy crisis (i.e. 1973), electric power consumption
in manufacturing increased at an average annual rate of 6.226%; however,
over the ensuing decade (1974-84), it had slowed to 0.591% per annum,
bringing with it the growth rate in value-added.
3.4 Electric Power Consumption and Productivity

Gullickson and Harper (1987) and Hisnanick and Kymm (1992)
evaluate the sources of productivity growth in US manufacturing using
1p=q1 (see Equation 3.2) as the appropriate measure of labor productivity.
In this section, we report a series of revised estimates of the role of electric
power and capital in labor productivity in U.S. manufacturing. Table 3.5
summarizes our results. As was the case above, we find that the rate of
growth of labor productivity is entirely explained by increasing capital and
10 The corresponding values for tfp, defined earlier, are 0.063 (1950-84), 0.088
(1950-73) and 0.0093 (1974-84).
electric power intensities in US manufacturing, defined here as ep1, the

shift away from labor to electric power, and k1, the shift away from labor
to capital:
݈‫ ݍ = ݌‬െ ݈ = ‫ݒ‬௘௣ [݁‫ ݌‬െ ݈] + ‫ݒ‬௞ [݇ െ ݈] + ‫݌݂ݐ‬ 3.2
For the entire sample period (i.e. 1950-84), labor productivity increased
at an average annual rate of 2.211%. In this period, the electric powerlabor
ratio increased at an average annual rate of 3.6704%, which when multiplied
by the relevant elasticity (i.e. 0.53304) yields a value of 1.956 percent which
measures the effects on labor productivity of the substitution of labor for
electric power referred to by Dale Jorgenson (Jorgenson 1984). In this
period, the capitallabor ratio increased at an average annual rate of 2.7801,
which when multiplied by the relevant elasticity yields a value of 0.178,
which measures the effects on labor productivity of the substitution of labor
for capital. The sum of these two effects accounts for 97% of the growth in
labor productivity in US manufacturing.
Prior to the energy crisis, labor productivity increased at an average
annual rate of 2.842% of which 2.585% can be attributed to laborelectric
power substitution and 0.146% can be attributed to laborcapital
substitution. Together, these two effects account for 96% of the overall
increase in labor productivity. In the ensuing decade, labor productivity
increased at an average annual rate of 0.834% of which 58.3% can be
attributed to laborelectric power substitution and 24.9% can be attributed
to laborcapital substitution.
3.5 Conclusions
For years, writers have pointed to the increased use of electric power as
a key factor in the growth of productivity. As such, to most observers, it
seemed perfectly reasonable to assume that the productivity slowdown
which began in 1973-74 was the direct result of the energy crisis. Oil prices
had increased dramatically, driving up all energy prices, including the prices
of electricity. Growth accountants, however, dismissed such claims, arguing
that while energy consumption in general and electric power in particular
had indeed slowed down markedly, neither could explain more than 1-2%
of the productivity slowdown.
Our findings, however, reconcile the historical record with the empirical
facts. By using direct as opposed to indirect estimates of the relevant output
factor input elasticities, we were able to reconcile the historical record as
described by Schurr (1983), Jorgenson (1983, 1984), and Rosenberg (1983)
with the findings of growth accountancy. Growth in electric power

consumption accounts for 79% of the growth of manufacturing value-added.
This provided the necessary basis to reevaluate the role of energy in general
and electric power in particular in the productivity slowdown. We were able
to show that the decline in electric power growth accounted for the bulk of
the decline in output and productivity growth. In fact, our results show that
when energy in general and electric power in particular are attributed their
rightful place in manufacturing production processes, the Solow residual
disappears, as does the total factor productivity residual (i.e. tfp) in
conventionally defined productivity growth.
3.6 References
Berndt, Ernst and Wood, David O. 1975. “Technology, Prices and the
Derived Demand for Energy.” The Review of Economics and Statistics.
August: 259-268
Christensen, L. R. and Dale Jorgenson. 1970. “US Real Product and Real
Factor Input” The Review of Income and Wealth, March.
Denison, Edward F. 1985. Trends in American Economic Growth 1929-
1982. The Brookings Institution, Washington, DC (1985).
Gollop, F. M. and Dale Jorgenson. 1980. “US productivity Growth by
Industry, 1948-1973” in Kendrick, J. W. and Vaccara B. (eds.) 1980
New Developments in Productivity Measurement and Analysis.
Chicago: National Bureau of Economic Research.
Gullickson, W. and M. J. Harper, 1987. “Multifactor Productivity in US
Manufacturing, 1949-1983.” Monthly Labor Review. October:18-28
Hisnanick, J. J. and Kymm, K. O. 1992. “The Impact of Disaggregated
Energy on Productivity.” Energy Economics. October: 274-278.
Hounshell, David A. 1984. From the American System to Mass Production
1800-1932: The Development of Manufacturing Technology in the
United States. Baltimore, MD: The Johns Hopkins University Press.
Jorgenson, Dale W. 1983. “Energy Prices and Productivity Growth,” in
Schurr, Sam. et al (eds.) Energy, Productivity, and Economic Growth.
Cambridge, MA: Oelgeschlager, Gunn and Hain.
Jorgenson, Dale W. 1984. “The Role of Energy in Productivity Growth,” in
Kendrick, J. W. (ed) International Comparisons of Productivity and
Causes of the Slowdown. Ballinger, Cambridge, MA: Ballinger.
Kendrick, J. W. and Grossman, E. S. 1980. Productivity in the United States,
Trends and Cycles. Baltimore, MD: The John Hopkins University Press.
National Bureau of Economic Research. 1929. Recent Economic Changes
in the United States. New York, NY: McGraw-Hill.
Rosenberg, Nathan. 1983. “The Effects of Energy Supply Characteristics on

Technology and Economic Growth.” in Schurr, Sam. et al (eds.) Energy,
Productivity and Economic Growth. Cambridge, MA: Oelgeschager,
Gunn and Hain.
Schurr, Sam. et al (eds.) 1983. “Energy, Productivity and Economic
Growth,” Cambridge, MA: Oelgeschlager, Gunn and Hain.
U.S. Department of Commerce. 1975. Historical Statistics of the US:
Colonial Times to 1970, Bicentennial Edition. Washington, DC: Bureau
of the Census.
U.S. Department of Commerce. 1985. Annual Survey of Manufacturers.
Washington, DC: Bureau of the Census.
U.S. Department of Commerce. 1986. Survey of Current Business.
Washington, DC: Bureau of Economic Analysis.
4
THE ENERGY-ORGANIZATION FRAMEWORK
The term process can in general be defined as a change in the properties of

an object, including geometry, hardness, state, information content (form
data), and so on. To produce any change in property, three essential agents
must be available: (i) material, (ii) energy, and (iii) information.
—Leo Alting, Manufacturing Engineering Processes
4.1 Introduction
As pointed out in the Introduction, 19th and 20th-century political
economists failed to provide a theory or model of production that was consilient
with classical mechanics and/or the emerging field of thermodynamics. The
fallout was to be fatal, to say the least. Labor, by then a supervisory input,
was seen as physically productive; capital (tools) was simply decreed to be
physically productive, in abject contradiction with classical mechanics. To
begin with, we present a consilient model of material processes as they
apply to political economy, namely the energy-organization model
developed in Beaudreau (1998). Drawing from the pure and applied
sciences, the energy-organization approach focuses on two universal inputs,
namely broadly-defined energy and organization. Broadly-defined energy
includes both animate and inanimate forms of energy. By animate energy,
it should be understood muscular (human and animal) power; by inanimate
energy, it should be understood wind, fossil fuel-based, hydraulic and
nuclear power. All production processes, Beaudreau (1998) argued, involve
the consumption of energy of one type or another.1 Organization will be
defined as the conception/design of, and the overseeing (i.e. supervision) of
energy-consuming (i.e. entropic) production processes. The development of
the steam engine by Papin, Savary, Newcommen and Watt in the 17th and
18th centuries is an example of the former, while its day-to-day operation is
1 By consumption of energy, it should be understood the consumption or use of the
available work of a particular energy source, keeping in mind that energy cannot be
created or destroyed (First Law of Thermodynamics). As such, energy consumption
is equivalent to the concept of entropy.
52 4. The Energy-Organization Framework
an example of the latter. In the natural world, the design of and supervision
of energy-consuming processes is governed by forces that are not fully
understood. We shall refer to such processes as naturally-occurring/spontaneous
entropic processes. Where these differ from man-made (i.e. anthropomorphic)
entropic processes is in their organization, broadly defined. From the
Paleolithic era to the present, Homo sapiensneanderthalensis and
sapienshave designed and redesigned man-made entropic processes (i.e.
anthropic entropic processes). For example, the development of stone tools
in the Paleolithic era altered the very nature of work. By reducing waste (i.e.
increasing efficiency), primitive tools such as hammers and knives
increased the amount of work that could be carried out by a given quantity
of muscular energy. In fact, the development and improvement of
anthropomorphic entropic processes are what define various pre-historical
and historical eras (e.g. the stone age, the bronze age, the machine age).2
As we shall see, the resulting model of production has important
implications for distribution theory. Specifically, given that energy and
energy alone is productive in the physical sense, existing theories of
distribution in factory production (i.e. classical, radical-Marxian,
neoclassical) are incomplete, not to mention misspecified. For example, the
notion of marginal product of labor and capital is theoretically irrelevant,
labor and capital not being physically productive. What’s more, they violate
basic mechanics. In their place, we offer a bargaining approach to
distribution in which the owners of energy and organization bargain over
the final product, in this case, value-added, one in which only energy is
physically productive. The notion of energy rents is introduced. Energy
rents are, by definition, the difference between the value product of energy
and its price (cost). The owners of the energy and organization inputs, it is
argued, bargain over these rents, the result (solution) of which is the
observed functional distribution of income.
4.2 The Energy-Organization Approach to Modeling

Material Processes
The cornerstone of the energy-organization approach to modeling
production processes (Beaudreau 1998) is Sir Isaac Newton’s second law
of motion, namely F=ma, where F = force, m = mass, and a = acceleration.
Put differently, a=F/m. That is, acceleration is simply force divided by
mass. The corresponding definition of work is W = fd, where W = work, f =
2 Those interested in the evolution of production processes over time are referred to
Chapter 2 of Energy and Organization: Growth and Distribution Reexamined.

force, and d = distance: work equals force times distance. Thus, the greater
is f, ceteris paribus, the greater is W. The greater is d, ceteris paribus, the
greater is W. In short, the more force exerted and the longer the distance (or
time period) over which the force in question is exerted, the more work
performed. Redefining work as output permits us to write the basic axiom
of the energy-organization approach, namely that for any given, well-
defined man-made entropic process, output is an increasing function of
energy consumption.
Contrast this with the standard, seemingly time-invariant definition of
work found in political economy, namely that work is an increasing function
of capital and labor (i.e. W=f(K, L)). Capital and labor, it is argued, produce
output (value-added). Both are assumed to be productive. Just what it is that
capital and labor do in production processes is unspecified. Terms such as
capital productivity and labor productivity, however, connote the idea that
both somehow work. Inanimate forms of energy such as oil, gas and
electricity consumption are assumed to be intermediate inputs, and, hence,
are not productive in the conventional sense. Put differently, they are not
factors of production (value-adding). In short, capital and labor are assumed
to add value to energy and other raw materials. Broadly-defined organization
is also ignored. Production processes are assumed to exist. Management
issues are, in general, ignored.
Clearly, the physical and economic definitions of work are worlds apart.
For three centuries, physicists have focused on force and energy as the basis
of work; economists, on the other hand, have focused on capital and labor.
In physics, tools and machines (i.e. capital) modify and transmit force/
energy, but are not, as such, sources of energy. In political economy energy
(inanimate energy forms) is viewed as an intermediate good, and, thus, is
not productive in the traditional sense (value-adding).
4.3 A Physical Model of Production

Here, we begin by examining the purely physical aspect of production,
namely the relationship between work and energy. To this end, we start by
defining production as the following functional relationship between work
(i.e. output) and the force (i.e. energy) expended in the process.
ܹ(‫])ݐ(ܧ[ܨ = )ݐ‬ 4.1
where: W(t) = work in period t; and E(t) = energy (i.e. force) in period t.
4.3.1 Animate and Inanimate Forms of Energy

In general, energy can be disaggregated into two categories, namely
Ea(t), animate (i.e. muscular) energy, and Ei(t), inanimate energy
(E(t)=Ea(t)+Ei(t)). Examples of animate energy include human and animal
force (i.e. muscular force), while examples of inanimate energy include
internal combustion, steam power, wind power, hydraulic power and
electrical power.3 Work will be modeled as an increasing function of
animate and inanimate force/energy.
4.3.2 The Role of Tools/Machines

From the Paleolithic era (i.e. the Stone Age) onwards, work as defined
above, has invariably involved the use of tools/machines. These include
axes, adzes, levers, presses, drills, screws, hammers, screwdrivers, saws,
etcetera. This raises a number of questions. For example, what is the role of
tools in broadly-defined work? Are they a source of energy? Are they
productive in the classical sense? Similar questions were raised by Nobel
prize laureate (Chemistry) Frederick Soddy in the 1920s. Who or what, he
asked, was ultimately responsible for producing goods and services? As the
following quote indicates, in Soddy’s mind, there was no doubt: energy was
the primary factor.
At the risk of being redundant, let me illustrate what I mean by the question
How do men live? by asking what makes a railroad train go. In one sense or
another, credit for the achievement may be claimed by the so-called engine-
driver, the guard, the signalman, the manager, the capitalist, the share-
holderor, again, by the scientific pioneers who discovered the nature of
fire, by the inventors who harnessed it, by labour which built the railroad
3The view that conventionally-defined labor can be broken down into a force (i.e.
energy) and supervisory component is as old as thermodynamics itself. German
physicist and physiologist Hermann von Helmholtz argued that the forces of nature
(mechanical, electrical, chemical, etcetera) are forms of a single, universal energy
or Kraft, that cannot be either added to or destroyed. According to Ansom
Rabinbach, As Helmholtz was aware, the breakthrough in thermodynamics had
enormous social implications. In his popular lectures and writings, he strikingly
portrayed the movements of the planets, the forces of nature, the productive forces
of machines, and, of course, human labor power as examples of the principle of
conservation of energy. The cosmos was essentially a system of production whose
product was the universal Kraft, necessary to power the engines of nature and
society, a vast and protean reservoir of labor power awaiting its conversion to work
(Rabinbach 1990, 3).
and the train. The fact remains that all of them by their collective effort could
not drive the train. The real engine-driver is the coal. So, in the present state
of science, the answer to the question how men live, or how anything lives,
or how inanimate nature lives, in the sense in which we speak of the life of
a waterfall or of any other manifestation of continued liveliness, is, with few
and unimportant exceptions, By sunshine. Switch off the sun and a world
would result lifeless, not only in the sense of animate life, but also in respect
of by far the greater part of the life of inanimate nature. (Soddy 1924, 4)
Accordingly, because tools are not a source of energy, they are not
productive in the physical sense. That is, they do not work. Nowhere is this
better seen than in the physicist’s definition of a machine which consists of
an instrument used to transmit or modify force/energy.
Machinery is used to change the magnitude, direction and point of

application of required forces in order to make tasks easier. The output of
useful work from any machine, however, can never exceed the total input of
work and energy. (Betts 1989, 172)
Arthur Beiser, in Modern Technical Physics, provides a similar definition:
A machine is a device which transmits force or torque to accomplish a

definite purpose. (Beiser 1983, 208)
By using primitive hammers and knives, early man (i.e. Paleolithic and
Neolithic man) was able to better direct and apply his force. Analytically
speaking, however, while tools allowed him to minimize energy loss (i.e.
wasted energy), they did not allow him to increase the total amount of work
beyond the initial level of force. That is, they were not a source of additional
energy.4 Implicit here is the basic notion of thermodynamic efficiency,
defined as work out versus work in. By better transmitting muscular force,
tools improved early man’s thermodynamic efficiency. With the same
amount of energy, more work could be done (e.g. skinning animals, cutting
firewood).
This raises the question of the relevant measure of tool/capital
productivity. Specifically, how do we measure and define the productivity
of tools/machines (i.e. capital)? The answer, we believe is straightforward.
Since tools/machines/capital are not a source of energy, they cannot be
regarded as productive in the classical sense (i.e. performing work/working).
4 This has important consequences for distribution. Clearly, the owner(s) of the
tools/machines cannot lay claim to a portion of the output on the basis of work.
Instead, their claim has to be based on tools/machines contribution to second-law
efficiency. Tools/machines improve second-law efficiency, thus increasing output.
Rather, capital productivity must be measured in terms of its ability to

transform force/energy into useful work, a concept known as second-law
efficiency. As such, the notion of capital productivity is qualitative (i.e. a
scalar), and not quantitative in nature. Increasing the amount of capital,
ceteris paribus, will not increase output. However, increasing the quality of
capital, say by improving thermodynamic efficiency, will increase output.
For example, James Watts steam engine, by reducing the heat loss of the
Savary and Newcommen steam engines, increased the overall efficiency
(i.e. exergy) of what at the time was known as “fire power.” One important
fact remains, namely that it was not and is not an energy source.
This allows us to rewrite Equation 4.1 as follows:
ܹ(‫)ݐ(ܧ )ݐ(ߟ = )ݐ‬ 4.2
where Ș(t) is defined as second-law efficiency at time t.5

Equation 4.2 defines the energy-organization approach to production
processes. W(t), output/work is an increasing function of E(t), energy/force,
and Ș(t), thermodynamic efficiency. Increasing Ș(t), thermodynamic
efficiency, ceteris paribus, leads to higher output/work. Increasing energy,
ceteris paribus, leads to higher work/output. Ș(t) is tool-specific. Better,
more efficient tools will have higher Ș(t)’s, and vice versa.
4.3.3 A Tool/Machine Taxonomy

According to classical mechanics, there are three basic tools/machines:
the lever, the inclined plane and hydraulic press.6 Each directs and applies
a specific form of energy. Table 4.1 lists the three basic tools/machines as
well as the associated composite tools, which are combinations of basic
tools. We see, for example, that the lever is the basic tool/machine which
transforms force in the case of a crane, a wheelbarrow, and various pulleys.
All tools/machines, it therefore follows, can and should be viewed as
combinations of the three basic tools, the lever, the inclined plane and the
hydraulic press (see Betts (1989), Beiser (1983)).
5 By definition, second-law efficiency consists of the ratio of the minimum
theoretical amount of energy required to perform a task, to the actual of amount of

energy used in any given production process.
6 We forego a discussion of the notion of mechanical advantage for the simple reason
that while tools/machines can better distribute the overall amount of work to be
done, they do not, in any way, reduce it. In other words, because they are not a source
of energy, they cannot increase the overall amount of work being performed.
Table 4.1: A Tool Taxonomy
Basic Tool Composite Tool

Lever Scissor, Plier, Wheelbarrow
Hydraulic Press Hydraulic Jack
Inclined Plane Wedge, Jack, Screw
Source: Beaudreau (1998).
4.4 A General Model of Production Processes:

The Energy-Organization Framework
The discussion of the role of tools in modern production processes brings
us to the second universal factor input, namely, organization. Machines define
the framework in which force/energy is transformed into work. In physics as
in economics, it is implicitly assumed that (i) well-defined work processes
exist (i.e. Equation 4.2) and (ii) that they are self-regulating. In this section,
we examine the broader question of organizationspecifically, the
organization of the use of energy to accomplish specific goals, in this case,
the creation of wealth. The discussion begins with a definition of what we
shall refer to as anthropomorphic entropic processesthat is, man-made
energy-using processes. This is then followed by two issues, namely (i) the
design of and (ii) the supervision of anthropomorphic energy-consuming
processes. Together with tools, these constitute broadly-defined organization.
4.4.1 Spontaneously-Occurring and Anthropomorphic Entropic

Processes
We begin this section with a discussion of entropic processes in general.
Entropic processes can be defined as environments in which energy is
transferred from one system to another or others in the form of work. For
the sake of discussion, we define two basic types of entropic processes:
natural entropic processes which occur spontaneously in nature, and
anthropomorphic entropic processes which, as their name implies, are
conceived of and overseen by man (i.e. managed, supervised).7 For
example, solar radiation, the wind, the tides, river currents, unicellular and
multi-cellular organisms are examples of natural spontaneously-occurring
entropic processes based on solar energy, and its derivatives. The
production of food, tools, shelter, and culture, however, are examples of
anthropomorphic entropic processes. What distinguishes these two sets of
7 By spontaneous, it should be understood not having to do with man.

processes is the nature of the relevant set of instructions. In the former, it

evolved randomly (or so it appears) over the course of the last 15 billion
years, while in the latter, it evolved over the course of the last 2, 000 years.
In the case of anthropomorphic entropic processes, Homo sapiens-sapiens
designs and supervises (i.e. oversees) the work process. Take, for example,
the simple windmill, which converts a temperature gradient (i.e. the wind)
into work. The human body, however, is an example of a natural entropic
process whose blueprint is the result of millions of years of evolution/mutation
(i.e. the human genome), and whose supervision is auto-regulated (autonomic
nervous system, paleopallium).
Clearly, tools are the defining element in anthropomorphic entropic
processes. They are to production processes what the human body is the
process of life. They transmit and/or modify force/energy. Given the
presence of friction, it follows that the better are the tools (i.e. the less
friction), the higher the efficiency, and the greater the work.
The presence of energy and tools, however, is not a sufficient condition
for work as both occur spontaneously in nature. Also required is an activity
which we shall refer to as the overseeing of or supervising of entropic
processes. For example, take the case of a simple gasoline engine-powered
water pump. With an endless supply of refined hydrocarbons (gasoline), it
can work continuously ad infinitum. Suppose, however, that, for some
reason, the pump arm spontaneously dislodges itself from the drive shaft.
Clearly, in this case, the energy provided by the engine will be lost (Ș(t)=0),
unless of course the problem is rectified. Supervision, as this example
illustrates, is an important aspect of organization. Anthropomorphic
entropic processes are subject to breakdown.8 Supervision, one could argue,
minimizes the resulting energy loss. Nineteenth-century political economist
Alfred Marshall described the role of supervision in manufacturing in the
following terms:
New machinery, when just invented, generally requires a great deal of care
and attention. But the work of its attendant is always being sifted; that which
is uniform and monotonous is gradually taken over by the machine, which
thus becomes steadily more and more automatic and self-acting; till at last,
there is nothing for the hand to do, but to supply the material at certain
intervals and to take away the work when it is finished. There still remains
the responsibility for seeing that the machinery is in good order and working
smoothly; but even this task is often made light by the introduction of
8One could argue that natural entopic processes are also subject to breakdown. For
example, take the human body and the numerous diseases that prevent energy from
being transformed into work.
automatic movement, which brings the machine to a stop the instant

something goes wrong. (Marshall 1890, 218)
To capture the role of tools and supervision in production processes,

Equation 4.2 was rewritten as follows:
ܹ(‫)ݐ(ܵ[)ݐ(ߟ = )ݐ‬, ܶ(‫)ݐ(ܧ ])ݐ‬ 4.3
where S(t) and T(t) correspond to the level of supervision at time t and tools
at time t, respectively. For the time being, we simply assume that second-
law efficiency is an increasing function of tools and supervision. Thus, for
a given quantum of energy, the more/better tools and supervision, the
greater is second-law efficiency (i.e. Ș(t)), and hence, the greater is output
(i.e. W(t)). Capital and supervision, it therefore follows, are not directly
productive; rather, they contribute to work via their effect on second-law
efficiency.
Here, we stop short of explicitly modeling supervisory activity, except
to point out the obvious, namely that, historically, it has been carried out by
individuals (animate supervision), and secondly, has been organized
hierarchically, with conventionally-defined workers (lower-level supervisors)
at the bottom, line supervisors above, and senior managers (e.g. CEOs,
CFOs and the Board of Directors) at the top. It is clear that while not a
source of energy, the supervisory input is nonetheless a sine quo non of
virtually all production processes. Without supervisors, output becomes
probabilistic (including the null set).9 Tools and machines break down,
resulting in a loss of energy and output.
4.4.2 Leibenstein’s Managerial Taxonomy

The view of the organization input (broadly-defined management) as a
two-dimensional activity (e.g. supervision and design) is not entirely new.
In his work on the role of entrepreneurs in economic development, Harvey
Leibenstein identified two types of managers, namely routine and innovative.
Put differently, some managers are concerned with the day-to-day workings
of the firm, while others are concerned with altering either the production
process or some part thereof, or the product itself (value-added)what Leo
Alting refers to the hardness, state, or information content.
I may distinguish two broad types of entrepreneurial activity: at one pole,

there is routine entrepreneurship (R-entrepreneurship), which is really a type
9 Theoretically, this can be expressed as lim K(‫)ݐ(ܵ[ )ݐ‬, ܶ(‫ = ])ݐ‬0.

ௌ(௧)՜଴
of management, and for the rest of the spectrum we have Schumpeterian or

new type entrepreneurship. (We shall refer to the latter as N-entrepreneurship.)
By routine entrepreneurship we mean the activities involved in coordination
and carrying on a well-established, going concern in which the parts of the
production function in use (and likely alternatives to current use) are well
known and which operates in well-established and clearly defined markets.
By N-entrepreneurs, we mean activities necessary to create or carry on an
enterprise where not all markets are well established or clearly defined
and/or in which the relevant parts of the production function are not
completely known. (Leibenstein 1968, 72)
The former are the equivalents of the supervisors referred to above,

while the latter are the equivalent of the designers/conceivers, those who
conceive of or improve upon existing entropic processes or products. Put
differently, progress requires dynamic innovative managers; otherwise,
supervisors will be condemned to manage the existing set of anthropomorphic
entropic processes ad infinitum. Recall that anthropomorphic entropic
processes, unlike natural entropic processes, are man-made, subject to
change if and only if changed.
4.5 The General Energy-Organization Production

Function
As in Beaudreau (1998), we contend that this simple two-factor model
(energy and organization), based on the laws of classical mechanics and
thermodynamics, and the principles of organization (i.e. design and
supervision), describes virtually all production processes, past, present and
future. Take, for example, tree harvesting. Prior to the advent of the
mechanical saw, animate energy was transmitted and applied by tools such
as axes and saws to fall trees. The required supervision was typically
provided by the owners of animate energy (i.e. lumberjacks). With the
advent of the chain saw, animate energy (i.e. the internal combustion
engine) was combined with inanimate energy, with essentially the same
tools and supervision (i.e. saw teeth), to accomplish the task. More recently,
with the advent of the mechanical harvester, inanimate energy alone is used
in conjunction with tools to accomplish the task. The requisite supervision
is provided by the operator. As this case clearly demonstrates, technological
change has resulted in a case in which labor goes from providing energy and
organization to providing only organization.
Another example is weaving. In the Paleolithic era, fibers or reeds were
woven together to produce clothing, baskets and sieves using animate,
muscular energy. With the advent of the loom (i.e. tools), human energy
was better transmitted and applied, resulting in higher output. Primitive

looms, it therefore follows, are examples of anthropomorphic entropic
processes. The application of water, steam, and electric power to the hand-
and foot-operated loom led to the power loom where human energy was
replaced by water, steam and ultimately electric power, transforming labor’s
role from that of inanimate energy source and supervisor to that of only
supervisor with the result that today, little to no human energy is deployed
in the weaving process. In the following passage, 19th-century British
economist Alfred Marshall describes just such a change:
We may now pass to the effects which machinery has in relieving that
excessive muscular strain which a few generations ago was the common lot
of more than half the working men even in such a country as England…. in
other trades, machinery has lightened man’s labours. The house carpenters,
for instance, make things of the same kind as those used by our forefathers,
with much less toil for themselves….. Nothing could be more narrow or
monotonous than the occupation of a weaver of plain stuffs in the old time.
But now, one woman will manage four or more looms, each of which does
many times as much work in the course of a day as the old handloom did;
and her work is much less monotonous and calls for much more judgment
than his did. (Marshall 1890, 218)
The operative word here is “manage.” Analytically speaking, factory

workers in the 19th century no longer worked; rather, they managed.10 Simon
Newcomb, the 19th-century physicist and part-time economist, described the
far-reaching change in labor’s role in production brought about by steam-
powered machinery as follows:
We may readily apply the principle here illustrated to the actual historical
facts. The introduction of machinery during the last hundred years has to a
certain extent changed the directions of man’s occupations. Instead of
making things with their own hands, as they formally had to do, they are
now managing machines or assisting in various ways in working them. The
pin-makers are no longer at work; a few of them are feeding pin-making
machines, but the majority of them have learned other employments. A large
class of carpenters no longer push the plane; a portion of them feed the
planning machines, and the remainder are fully occupied in executing work
that increased refinement which demand has encouraged. The same thing
may be traced all through the channels of industry. (Newcomb 1886, 390)
10This explains the increasing use of child labor (i.e. child supervision) in factories
in the early 19th century.

This was a recurrent theme in the 19th century, extending beyond

political economy into such fields as engineering and mechanics. For
example, James Martineau, a professional engineer, in a lecture to the
Liverpool Mechanics Institute, spoke of “machinery rapidly supplanting
human labour and rendering mere muscular force...worthless... That natural
machine, the human body, is depreciated in the market. But if the body has
lost its value, the mind must get into business without delay” (Tylecote
1957, 262).
The depreciation of human power continued in the early 20th century
with the introduction of electric power. More flexible than steam power, it
resulted in the mechanization of sectors of the economy that had resisted
earlier mechanization. Consider, for example, the following quote describing
the effects of electric-powered materials-handling equipment on the nature
of work in the mining industry, taken from David Nye’s work on the
electrification of America (Nye 1990):
Thus, for some, electrification meant unemployment as a few skilled jobs

replaced unskilled labor. At Green Ridge Colliery, for example, a station
engineer, motorman, and helper could run an electric locomotive that
replaced six mule drivers, four boy helpers, and seventeen mules. At New
York and Scranton Coal, three men and a locomotive replaced seven boys
and fourteen mules. Down in the mines other electrical machines replaced
hand labor, increasing productivity but reducing the need for artisanal skill.
Such innovations were contributory factors in the United Mine Workers
strikes of 1900, 1902, and 1912. A similar kind of labor replacement
disturbed the steel workers, as General Electric designed a set of electrical
motors and controls for rolling mills. The corporation proudly announced,
one man surrounded by a dozen or more operating levers controls every
motion of steel from the ingot furnace to the completed rail.... Every motor
replaced a man, but the work is done better and more quickly than formerly.
In many mills, a motor [was] installed on a charging and drawing machine,
arranged so as to automatically grip and withdraw the hot bloom from the
furnace, and release it when clear of the furnace door. The motorman has
simply to start and stop the motor. (Nye 1990, 206)
4.5.1 Substitutes or Complements

The resulting two-factor (energy and organization) model of production
processes provides a rich framework in which to study production-related
questions. For example, it provides a theoretical basis for studying the role
of conventionally-defined energy, capital, labor and organization in
production. To what extent are energy and organization substitutable? Can
managers be substituted for energy, and can labor be substituted for capital?
Further, it provides the necessary framework to study technological change

and its effects on the various factor inputs over time. For example, has
second-law efficiency increased or decreased over time?
Another relevant question, especially in light of the energy crisis in the
1970s and 1980s, is whether tools (i.e. capital) and broadly-defined energy
are substitutes or complements? Most analysts felt that they were substitutes
(Solow 1974, Berndt and Wood 1975). Hence, the accepted view at the time
held that firms would substitute away from energy over to capital and labor.
The energy-organization approach to production analysis provides an
altogether different view on this issue. The answer is a qualified no.
Theoretically, for a given value of Ș(t), an increase in capital (i.e. tools)
cannot make up for a loss of energy. Referring to Equation 4.3, we see that
the only way in which more capital could compensate for less energy is if
the result was higher second-law efficiency. That is, only if the additional
capital increases Ș(t) will output increase. As pointed out earlier, because
machines and tools are not a source of energy, but rather, apply and transmit
energy, it follows that, ceteris paribus, quantities of capital and energy
cannot possibly be substitutes, but rather, are complementary inputs. Tools
and machines cannot create energy. More tools for a given level of energy
cannot and will not result in more work.
This leaves the possibility of second-law efficiency-increasing investment.
In this case, the level of second-law efficiency, is assumed to be increasing
in broadly-defined capital, as measured by the value of tools and machines.
That is, the more capital, the higher is Ș(t). Were this to hold empirically,
then capital and energy would be substitutes in the sense that more capital,
by providing for higher Ș(t) values, reduces the amount of energy required
to perform a fixed, given amount of work. The point is that capital and
energy can only be considered substitutes if and only if the former is
heterogeneous in nature. More efficient capital can, at least theoretically,
compensate for less energy; otherwise, they are complementary.
4.6 Energy-Organization Productivity Measures

The energy-organization approach to modeling production processes
provides important insights into both the definition, the construction and the
use of productivity measures. Typically, productivity is measured as the
ratio of output (value-added) to one or many inputs. For example, in the
past, political economists have defined and measured labor and capital
productivity. Labor productivity is the ratio of output (value-added) to labor
input; capital productivity is the ratio of output (value-added) to capital
input. As pointed out above, implicit in these notions is the fact that both
capital and labor are physically productive. This raises an important

question. Does the fact that an input is physically present necessarily imply
that it is physically productive? Take, for example, the case of managers.
Are managers physically productive? If so, then how should their
productivity be measured? After all, managers are a sine qua non of
production. The point is that productivity measures differ both with regard
to their content and, of course, with regard to their meaning. Those which
refer to energy and force ought to be viewed as physically productive, while
those which refer to organization ought to be viewed as organizationally
productive. The fact that conventionally-measured labor productivity
increases does not imply that labor is responsible for the increase. As argued
above, the only physically-consistent measure of productivity is the ratio of
output to total energy.
4.6.1 Energy Productivity Measures

Three energy-related productivity measures follow. Referring to Table
4.2, the first is the ratio of output to animate energy (i.e. W(t)/Ea(t)),
commonly known as labor productivity. For this ratio to be meaningful (i.e.
in terms of thermodynamics), it must be the case that the labor in question
is a source of energy. Otherwise, labor productivity as conventionally
defined is devoid of any meaning. As we shall argue later, labor in modern
production processes is more appropriately viewed as a form of lower-level
organization (i.e. supervisor).
The second productivity measure is the ratio of output to inanimate
energy (i.e. W(t)/Ei(t)), Ei(t) being broadly defined to include electrical,
thermo-mechanical, and all other inanimate forms of energy. By virtue of
Equation 4.3, this measure explicitly defines Ș(t), thermodynamic
efficiency. Lastly, the third law of thermodynamics provides us with a third
measure of productivity, namely the ratio of output to total energy (i.e. E(t)).
This measures the relationship between output and the broadly-defined
energy input. Clearly, energy-conversion tables (i.e. labor-BTU, electric
power-BTU) are required to render this measure operational. However, for
production processes requiring no animate energy (i.e. Ea(t)), this ratio is
equivalent to the output-inanimate energy ratio.
Table 4.2: NPA Energy Productivity Measures
W(t)/Ea(t) Ratio of Work to Animate Energy

W(t)/Ei(t) Ratio of Work to Inanimate Energy
W(t)/E(t) Ratio of Work to Total Energy
4.6.2 Organization Productivity Measures

As pointed out, continued energy-deepening throughout the 19th and 20th
centuries altered fundamentally the supervisory input (i.e. S(t)). In the pre-
industrial revolution period, the organizational input (i.e. S(t)) and the
source of energy (i.e. E(t)) were one-in-the-same, namely the worker.
Weavers, stonecutters, shoemakers, and artisans in general provided both
energy and organization. In time, as inanimate energy replaced animate
energy, the organizational aspect of production changed. A new class of
production-related personnel came into being, namely managers. Together
with workers who had been relieved of the physical demands of work,
managers supervised production processes. With this was born the fully-
functional organization hierarchy, with workers (lower-level supervisors) at
the base, and senior managers (upper-level supervisors) at the apex. To
capture the developments in supervision technology, we define two sub-
classes of supervisors: Sl(t), lower-level supervisors (i.e. workers) and Su(t),
upper-level supervisors, where S(t)=Sl(t)+Su(t). The former find themselves
at the base of the organizational hierarchy, while the latter find themselves
at the apex.
Table 4.3 defines the resulting three supervision-related productivity
measures, namely the ratio of work to total supervisors, the ratio of work to
lower-level supervisors, and lastly, the ratio of work to upper-level supervisors.
Table 4.3: NPA Organization Productivity Measures
W(t)/S(t) Ratio of Work to Supervisors

W(t)/Sl(t) Ratio of Work to Lower-Level Supervisors
W(t)/Su(t) Ratio of Work to Upper-Level Supervisors
W(t)/T(t) Ratio of Work to Tools and Machines
4.7 The Energy-Organization Approach to Modeling

Production Processes: The Evidence
Testing the energy-organization approach to production processes
involves testing a series of propositions, including (i) that energy is the only
physically-productive factor input, (ii) capital and labor (homogeneous) are
not physically productive, and (iii) that second-law efficiency is increasing
in the quality of capital and labor (information). In short, this entails testing
the basic principles of classical mechanics as they apply material processes
in economics. Unfortunately, the nature of the available data (i.e.
aggregate), in conjunction with the very nature of production, make exact
tests of these hypotheses difficult given current statistical techniques.11 For

example, as tools (or supervision) and output are collinear, it stands to
reason that by performing simple regression analysis, a positive output
elasticity will be found. The question then becomes, what is the meaning of
such a result? Does it mean that labor is physically productive? It is
important to note that collinearity does not necessarily imply causality nor
physical productivity.
Despite these shortcomings, conventional testing of production
relationships can nonetheless yield interesting insights, notably with regard
to the relative contribution of energy, capital and labor. Standard production
theory (e.g. the KLEMS approach) views energy as a relatively minor factor
inputthat is, relative to capital and labor. The energy-organization
approach maintains the converse, namely that capital and labor are minor
factor inputs (productivity-wise), and energy as the prime factor input.
Here, the energy-organization model of production is tested indirectly,
and the resulting output elasticities are used to reexamine the causes of the
productivity slowdown. Drawing from earlier work (Beaudreau 1995, 1998),
we show that by widening the set of productive factor inputs to include
energy, specifically electric power consumption, and estimating the output
coefficients directlyas opposed to indirectly via cost datafundamentally-
different output elasticities result, which, when used in standard growth
accounting exercises reduces the unexplained variation (i.e. the Solow
residual), to the point of explaining nearly all of the productivity slowdown.
In earlier work (Beaudreau 1995, 1998), the energy-organization
production function was estimated using a modified KLEMS production
function, namely the KLEP (capital, labor and electric power) production
function. Electric power was used as a proxy for energy consumption given
the prevalence of electricity as the prime mover in modern manufacturing
(see Sonenblum 1990). Capital and labor were used as proxies for tools and
supervision.12 The resulting output elasticities, estimated directly using
11 This is especially true at the aggregate level. Production processes differ markedly
within and across sectors.
12 Unfortunately, qualitative measures or indexes (information) of capital and labor
in manufacturing (and for other sectors) are unavailable at the aggregate level.
Ideally, such measures could be used to proxy innovations in second-law efficiency.
For more on the role of information in production, see Jorgenson (2001), where
investment in information technology (IT)-based capital is used as a proxy for
information. Unfortunately, such a measure would be of limited value in our work,
given the complementary nature of IT-capital and supervision (automation), and the
lack of a reliable measure of information quality.
post-WWII input and output data for U.S., German and Japan manufacturing,
were then used in standard growth accounting.
The relevant output elasticities were estimated directly (i.e. as opposed
to indirectly using cost data). Chief among the reasons for proceeding this
way was the absence of competition in both energy and labor markets
(Blanchflower, Oswald and Sanfey 1996, Van Reenen 1996). Specifically,
electric power markets are, in general, regulated, via public utility
commissions, while labor markets, according to Blanchflower et al and Van
Reenen, are, in general, non-competitive. This violates one of the key
assumptions of cost-based growth accounting, namely that factor markets
are assumed to be competitive. In the presence of non-competitive factor
markets, factor shares do not mirror physical productivity. To get around
this problem, output elasticities were estimated directly. Data on value-
added, electric power consumption, total employment and capital for U.S.,
German and Japanese manufacturing were used to estimate the Cobb-
Douglas KLEP output elasticities.
ܳ = ‫ܲܧ‬ఉభ ‫ܮ‬ఉమ ‫ܭ‬ఉయ 4.5
The estimated output elasticities (ȕ’s) for all three countries are presented
in Table 4.4. What is striking are the similarities across countries. In all three
cases, electric power consumption is, by far, the most important factor input,
as evidenced by output elasticities for U.S. manufacturing, German
manufacturing and Japanese manufacturing of 0.537244, 0.747482 and
0.605599, respectively. Capital and labor output elasticities are lower than
in previous studies (i.e. Berndt and Wood 1975).
These results, we argue, provide some support, albeit limited, for the
energy-organization model of production presented above. ȕ1, the electric
power output elasticity, is greaterby a magnitude of tenthan estimates
reported elsewhere in the literature. The labor output elasticity is smaller
than typically reported, while the capital output elasticity is somewhat
comparable. Does the fact that ȕ2 and ȕ3 take on positive and statistically
significant values weaken the case for the energy-organization approach?
The answer, we believe, is no, as such elasticities only highlight the
weaknesses of conventional regression analysis in this context. Correlation
does not imply causality. It is important to recall that organization is as
important a part of the energy-organization approach as energy, but that
energy and only energy is physically productive. Organization is a
necessary input, one that is organizationally productive, but not physically
productive.
Physicists Reiner Kummel, Julian Henn and Dietmar Lindenberger

estimated a model similar to Equation 4.5 using the Linux technique
(Kummel, Henn and Lindenberger 2002). Whereas the energy-organization
approach to modeling material processes considers capital and labor to be
organizational inputs, and hence not physically productive, the Linux
approach considers all three to be physically productive. Just how capital
(tools and equipment) is physically productive is not addressed. Their
estimates, presented in Table 4.5, are comparable to those reported in the
previous table.
Table 4.4: KLEP Regression Results U.S., Canadian and Japanese

Manufacturing
Inputs U.S. Canada Japan

1950-1984 1962–1988 1965-1988
EP 0.533046 0.728269 0.609444
(10.079) (4.493) (2.067)
L 0.418822 0.249191 0.193766
(18.231) (2.332) (1.847)
K 0.064250 0.0339365 0.193766
(2.768) (0.543) (1.608)
R2 0.984 0.968 0.981
F 1032.52 367.85 265.98
Source: Beaudreau (1999), 137.
Table 4.5: Kummel, Henn and Lindenberger’s Output Elasticities:

U.S., Germany and Japanese Manufacturing
Inputs U.S. Germany Japan

1960-1993 1960-1989 1965–1992
E 0.45 0.50 0.45
L 0.21 0.05 0.21
K 0.36 0.45 0.34
Source: Kummel, Henn, and Lindenberger (2002), 423.
Using these estimates, Beaudreau (1995, 1998) and Kummel, Henn and
Lindenberger (2002) reexamined growth of manufacturing output in the
post-WWII period, with particular emphasis on the pre- and post-energy
crisis subperiods. Specifically, they set out to demonstrate that contrary to
the accepted wisdom, the two energy crises were a leading cause of the
productivity slowdown. In short, the energy crises (1973 and 1979) ushered
in a period of zero energy-consumption growth, which, given the prominent
role of energy in material processes, contributed to a marked decrease in

overall growth rates, one that continues to this very day.
U.S. 1950–1984 1950–1973 1974–1984

USVA 2.684 3.469 0.121
USAI* 2.674 3.472 0.310
USEP 4.052 5.371 0.246
USN 0.662 0.900 -0.091
USK 3.694 3.614 3.4008
Germany 1963–1988 1962–1973 1974–1988
GERVA 2.462 6.522 1.486
GERAI* 2.433 5.190 1.080
GEREP 2.894 5.883 1.366
GERN -0.785 0.592 -0.938
GERK 2.945 5.620 1.406
Japan 1965-1988 1965-1973 1974-1988
JAPVA 3.826 8.844 3.099
JAPAI* 3.566 9.856 1.538
JAPEP 3.559 11.320 0.965
JAPN -0.082 2.297 -0.367
JAPK 7.520 13.536 5.182
*‫ߚ = ܫܣ‬ா௉ ா௉ ௗ௧
+ ߚே
ே ௗ௧
+ ߚ௄
௄ ௗ௧
Using standard growth accounting techniques, Beaudreau (1998)

showed that using the output elasticities reported in Table 4.4, the decrease
in the rate of growth of energy consumption from 1974 onward, was able to
account for the slowdown in manufacturing output growth in the United
States, Germany, and Japan. Referring to Table 4.6, we see that AI, the
aggregate factor input index, grew at an annual rate of 2.932 percent in U.S.
manufacturing, which explains roughly 98 percent of the average annual
rate of growth of manufacturing value-added of 2.995 percent. Electric
power consumption went from 4.129 percent in the pre-energy crisis period
to 0.321 percent in the post-energy crisis period. Similar results were found
in the case of German and Japanese manufacturing, where electric power
consumption decreased markedly, going from an average of 5.466 percent
in the case of pre-energy crisis German manufacturing to 3.747 percent in
the post-energy crisis period, and from an average of 11.320 percent in pre-
energy crisis Japanese manufacturing to 0.965 percent in the post-energy
crisis period. In all three cases, the Solow residual virtually disappeared.
Similar results were reported by Kummel, Henn and Lindenberger using the
Linex approach.
Table 4.7: Productivity Growth: U.S., German and Japanese

Manufacturing
U.S. 1963-1988 1963-1973 1974-1988

lp 3.461 5.898 2.261
tfp -0.040 1.163 0.061
ȕ1(ep-l) 2.899 3.8535 1.809
ȕ2(k-l) 0.606 0.882 0.391
Germany 1963-1988 1963-1973 1974-1988
lp 3.461 5.898 2.261
tfp -0.040 1.163 0.061
ȕ1(ep-l) 2.899 3.8535 1.809
ȕ2(k-l) 0.606 0.882 0.391
Japan 1963-1988 1963-1973 1974-1988
lp 3.461 5.898 2.261
tfp -0.040 1.163 0.061
ȕ1(epl) 2.899 3.8535 1.809
ȕ2(kl) 0.606 0.882 0.391
These output elasticities were then used to examine labor productivity

growth in all three countries in the post-WWII period. Gullickson and
Harper (1983) and Hisnanick and Kymm (1992) examined the sources of
productivity growth in U.S. manufacturing using lp=ql to measure labor
productivity.13 Table 4.7 reports estimates of the respective roles of electric
power and capital in labor productivity in U.S., German and Japanese
manufacturing. We see that the rate of growth of labor productivity is, in all
three cases, explained by increasing electric power and capital intensities,
defined here as epl, the shift away from labor to electric power, and kl,
the shift away from labor to capital. Estimates of total factor productivity
(tfp) are insignificant, compared to the contributions of energy deepening
and capital deepening, a finding that is consistent with both the energy-
organization approach to modeling material processes and basic physics in
general.
13As pointed out earlier, this view of productivity is archaic. Labor is not productive
in the physical sense; instead, it is productive in the organizational sense. As such,
labor productivity is analogous to management productivity.

In 1923, Nobel-prize-winning chemist Frederick Soddy asked and
answered the rhetorical question, how do men live?
At the risk of being redundant, let me illustrate what I mean by the question
“How do men live?” by asking what makes a railroad train go. In one sense
or another, credit for the achievement may be claimed by the so-called
“engine-driver,”, the guard, the signalman, the manager, the capitalist, the
share-holder, -or, again, by the scientific pioneers who discovered the nature
of fire, by the inventors who harnessed it, by labour which built the railroad
and the train. The fact remains that all of them by their collective effort could
not drive the train. The real engine-driver is the coal. So, in the present state
of science, the answer to the question how men live, or how anything lives,
or how inanimate nature lives, in the sense in which we speak of the life of
a waterfall or of any other manifestation of continued liveliness, is, with few
and unimportant exceptions,”By sunshine.” Switch off the sun and a world
of by far the greater part of the life of inanimate nature. (Soddy 1924, 4)
Three years later, F. G. Tryon of The Institute of Economics, in a paper

published in the Journal of the American Statistical Association, went
further, arguing that:

than it has yet received by economists. The industrial position of a nation
may be gauged by its use of power. The great advance in material standards
of life in the last century was made possible by an enormous increase in the
next century turns perhaps more than on anything else on making energy
energy. (Tryon 1927, 281)
In this chapter, the energy-organization approach to modeling material

processes was presented, complete with measures of productivity, including
the concepts of physical productivity and organizational productivity, and
most importantly, corroborating empirical evidence. Material wealth was
modeled as an increasing function of broadly-defined energy and broadly-
defined organization, the former being physically productive, and the latter
being organizationally productive. Broadly-defined organization includes
tools, information and supervision, all of which combine to define second-
law efficiency, or put differently, the productivity of energy (exergy).
This approach, which is consilient with Soddy’s view of how men live,
and a response to Tryon’s critique of production theory, provides the
analytical framework that will be used throughout the next eight chapters of
the book. Soddy’s coal is the energy and his so-called “engine-driver,”
guard, signalman, manager, capitalist, share-holder, scientific pioneers who
discovered the nature of fire, inventors who harnessed it, labor which built
the railroad and the train, are the organization.
4.9 Bibliography
Abraham, Katherine G., James R. Spletzer, and Jay C. Stewart. 1999. "Why
Do Different Wage Series Tell Different Stories?" American Economic
Review. 89, no. 2: 34-39.
Adams, Gregory & Rausser, Gordon. & Simon, Leo. 1996. "Modelling
Multilateral Negotiations: An Application to California Water Policy,”
Journal of Economic Behavior & Organization. 30, no. 1: 97-111.
Aghion, Philippe and Peter Howitt. 1998. Endogenous Growth Theory.
Cambridge, MA: MIT Press.
Alting, Leo. 1994. Manufacturing Engineering Processes. New York:
Marcel Decker Inc. 1994.
Beaudreau, Bernard C. 1995. The Impact of Electric Power on Productivity:
The Case of U.S. Manufacturing 1958–1984. Energy Economics.
17:231–236.
Beaudreau, Bernard C. 1996. Mass Production, The Stock Market Crash,
and The Great Depression: The Macroeconomics of Electrification.
Westport, CT: Greenwood Press.
Beaudreau, Bernard C. 1998. Energy and Organization: Growth and
Beaudreau, Bernard C. 1999. Energy and the Rise and Fall of Political
Economy. Westport, CT: Greenwood Press.
Beaudreau, Bernard C. 2004. Making Markets and Making Money, Strategy
and Monetary Exchange. New York: iUniverse.
Beiser, Arthur. 1983. Modern Technical Physics. Menlo Park, California:
The Benjamin/Cummings Publishing Company.
Bernard, Jean-Thomas and Benoît Coté. 2002. “The Measurement of the
Energy Intensity of Manufacturing Industries: A Principal Components
Analysis.” Resources for the Future Discussion Paper 02-31.
Berndt, Ernst and David O. Wood. 1975. “Technology, Prices and the
Derived Demand for Energy.” The Review of Economics and Statistics:
259–268.
Betts, John E. 1989. Essentials of Applied Physics. Englewood Cliffs, NJ:

Prentice-Hall.
Binmore, Ken. Ariel Rubinstein and Asher Wolinsky. 1986. "The Nash
Bargaining Solution in Economic Modelling,” RAND Journal of
Economics. vol. 17, no. 2:176-188.
Blanchflower, David G., Oswald, Andrew J. and Sanfey, Peter. 1996.
“Wages, Profits, and Rent Sharing.” Quarterly Journal of Economics.
60, no. 1: 227-251.
Bose, Bismal K. 1987. “Introduction to Microcomputer Control,” in Bose,
Bismal K. ed. Microcomputer Control of Power Electronics and Drives.
New York: IEEE Press.
Bosworth, Barry, and Perry, George L. 1994. “Productivity and Real
Wages: Is There a Puzzle?” Brookings Papers on Economic Activity:
317-335.
Bullard, C. W., P. Penner and D. Pilati. 1978. “Net Energy Analysis
Handbook for Combining Process and Input-Output Analysis.”
Resources and Energy: 267-313.
Challoner, Jack. 1993. Energy. London: Darling Kindersley.
Chase, Stuart. 1934. The Economy of Abundance. New York, NY:
MacMillan Company.
Clark, John Bates, 1879. “The Nature and Progress of Real Socialism.” New
Englander and Yale Review: 565-582.
Kaufmann, Robert K. and Cutler Cleveland. 2001. Oil Production in the
Lower 48 States: Economic, Geological. and Institutional Determinants.
Energy Journal. 22: 27-49.
Costanza, Robert. 1980. Embodied Energy and Economic Valuation.
Science. 210:1219-1224.
Costanza, R. and R. A. Herendeen. 1984. “Embodied Energy and Economic
Value in the United States Economy: 1963, 1967, and 1972.” Resources
and Energy. 6:129-164.
Devine, Warren D. 1990. Electricity in information management: The
evolution of electronic control.” in Schurr, Sam. H. et al. (eds.),
Electricity in the American Economy. Westport CT: Greenwood Press.
Director, Aaron 1933. The Economics of Technocracy. Chicago, IL:
University of Chicago Press.
Dumagan, Jesus and Gurmukh Gill. 2002. “Industry-Level Effects of
Information Technology Use on Overall Productivity.” in Digital
Economy 2002. Economics and Statistics Administration.
Durkheim, Emile. 1915. The Elementary Forms of the Religious Life: A

Study in Religious Sociology. Translated by Joseph Ward Swain. New
York: Macmillan.
Eatwell, John et al. 1987. The New Palgrave: A Dictionary of Economics.
London: Basingstoke.
Engels, Frederick and Karl Marx. 1848. The Communist Manifesto.
London: The Communist League.
Fuller, Buckminister. 1982. The Critical Path. New York, NY: St. Martin’s
Press.
Gollop, F.M., and D. W. Jorgenson. 1980. U.S. productivity growth by
industry, 1948–1973,” in Kendrick, J.W., and Vaccara, B.N. (Eds.) New
Developments in Productivity Measurement and Analysis. Chicago, IL:
National Bureau of Economic Research.
Gullickson, William. and Michael J. Harper. 1988. “Multifactor Productivity
in U.S. Manufacturing, 1949-1983,” Monthly Labor Review. 1988: 18-
28.
Helpman, Elhanan. (ed.) 1998. General Purpose Technologies and
Economic Growth. Cambridge MA: MIT Press.
Helpman, Elhanan and Manuel Trajtenberg. 1994. “A Time to Sow and a
Time to Reap: Growth Based on General Purpose Technologies.”
National Bureau of Economic Research Working Paper No. 4854.
Hisnanick, John J. and Kymm, Kern. 1992. “The Impact of Disaggregated
Energy on Productivity,” Energy Economics: 274-278.
Hounshell, David A. 1984. From the American System to Mass Production
1800-1932: The Development of Manufacturing Technology in the
United States. Baltimore, MD: The Johns Hopkins University Press.
Howitt, Peter and Robert Clower. 2000. “The Emergence of Economic
Organization.” Journal of Economic Behavior and Organization. Vol.
41:55-84.
Jevons, William S. 1865. The Coal Question. London: MacMillan and Co.
Jevons, William S. 1871. The Theory of Political Economy. London: Pelican
Books.
Jones, Larry and A. Foster Chin. 1991. Electronic Instruments and
Measurement. Englewood Cliffs, NJ: Prentice Hall.
Jorgenson, Dale W. 1983. Energy Prices and Productivity Growth,” in
Schurr, S. et al. (eds.) Energy, Productivity, and Economic Growth.
Cambridge, MA: Oelgeschlager, Gunn, and Hain.
Jorgenson, Dale W. 1981.The Role of Energy in Productivity Growth. in
Kendrick, J.W. (ed.) International Comparisons of Productivity and
Causes of the Slowdown. Cambridge MA: MIT Press.
Dale W. Jorgenson and Kevin J. Stiroh. 2000. “Raising the Speed Limit:
U.S. Economic Growth in the Information Age.” Brookings Papers on
Economic Activity: 1, Brookings Institution.
Kahn, Alfred E. 1988. The Economics of Regulation: Principles and
Institutions. Cambridge, MA: MIT Press.
Krueger, Alan B. 1999. Measuring Labor’s Share. Cambridge, MA,
National Bureau of Economic Research, Inc. (Working Paper 7006.)
Kummel, Reiner, Dietmar Lindenberger and Wolfgang Eichorn. 1998. The
Productive Power of Energy and Economic Evolution. University of
Wurzburg Working Paper.
Kummel, Reiner, Julian Henn, Dietmar Lindenberger. 2002. “Capital,
Labor, Energy and Creativity: Modeling Innovation Diffusion.”
Economic Dynamics and Structural Change. 13: 415-433.
Landes, David S. 1980. “The great drain and industrialization: Commodity
flows from periphery to center in historical perspective,” in R.C.O.
Matthews (Ed.) Economic Growth and Resources, Trends and Factors.
London: MacMillan.
Leibenstein, Harvey. 1968. “Entrepreneurship and Development.” American
Economic Review. 58: 72-83.
Lloyd George, David. 1924. Coal and Power. London: Hodder and
Stoughton.
Maddison, Angus. 1987. “Growth and Slowdown in Advanced Capitalist
Economies: Techniques of Quantitative Assessment.” Journal of
Economic Literature. 25: 649–698.
Malthus, Thomas R. 1827. Principles of Political Economy Considered with
a View to Their Practical Application. New York: Augustus M. Kelley.
Marshall, Alfred. 1890. Principles of Economics. London: MacMillan.
Marx, Karl. 1867. Das Kapital. Chicago: Encylcopaedia Britannica (1992).
Newcomb, Simon. 1886. Principles of Economics. New York: Augustus
Kelley.
Nye, David E. 1990. Electrifying America: Social Meaning of a New
Technology. Cambridge, MA: MIT Press.
Odum, H.T. and E.C. Odum. 1976. Energy Basis for Man and Nature. New
York: McGraw-Hill.
Owen, Robert. 1817. The Life of Robert Owen. London: Cass, 1967.
Owen, Robert. 1820. A New View of Society and Other Writings, London:
J.M. Dent and Sons, Ltd. (1927).
Rabinbach, Anson. 1990. The Human Motor: Energy, Fatigue and the
Origns of Modernity. New York, NY: Basic Books.
Ricardo, David. 1817. The Principles of Political Economy and Taxation.
New York: Everyman's Library.
Rifkin, Jeremy. 1995. The End of Work. New York: G.P.Putnam’s Sons.
Rosenberg, Nathan. 1972. Technology and American Economic Growth.
Armonk, NY: M.E. Sharpe.
Rubinstein, Ariel and Delip Abreu. 1988. “The Structure of Nash
Equilibrium in Repeated Games with Finite Automata.” Econometrica.
56: 1259-1282
Scott, Howard et al. 1933. Introduction to Technocracy. New York: The
John Day Company.
Smith, Adam. 1776. An Inquiry into the Nature and Causes of the Wealth
of Nations. Chicago: Encyclopaedia Britannica, 1990.
Sobel, Robert M. 1972. The Age of Giant Corporations, A Microeconomic
History of American Business 1914-1970. Westport, CT: Greenwood
Press.
Sciences upon State Stewardship. London: Hendersons.
Solow, Robert M. 1974. “The Economics of Resources or the Resources of
Economics.” American Economic Review. 64, no. 2: 1-14.
Stahl, I. 1972. Bargaining Theory. Stockholm Research Institute, Stockholm.
Stiroh, Kevin J. 2001. “Information Technology and the U.S. Productivity
Revival: What Do the Industry Data Say?” Federal Reserve Bank of New
York Staff Reports. no. 115.
Tylecote, Mabel. 1957. The Mechanics Institutes of Lancashire and
Yorkshire before 1851. Manchester: Manchester University Press.
Tryon, F.G. 1927. “An Index of Consumption of Fuels and Water Power,”
Journal of the American Statistical Association. 22: 271-282.
United Nations. 1984. Industrial Statistics Yearbook 1984. New York,
United Nations.
U.S. Department of Commerce. 1975. Historical Statistics of the U.S.:
Colonial Times to 1970. Bicentennial Edition. Washington, DC: Bureau
of the Census.
Van Reenen, John. 1996. “The Creation and Capture of Rents: Wages and
Innovation in a Panel of U.K. Companies,” Quarterly Journal of
Economics. 61, no.1:195–226.
5
ENGINEERING AND ECONOMIC GROWTH
Abstract
Prior to the mid-1970s, Western industrialized nations’ GDP grew at an

average annual rate in excess of 3%. Since, growth rates have been halved
and productivity growth has plummeted, prompting renewed interest in
growth theory. This chapter examines the methodology of economic growth
theory from Moses Abramovitz to Paul Romer, focusing specifically on the
intellectual links between economics and engineering. It is argued that
growth models in general have ignored the basic principles of classical
mechanics, and, more importantly, that this omission has been costly. For
example, it is shown that by including energy, the cornerstone of production
processes as modeled in engineering, in simple growth accounting exercises,
the Solow residual is nearly eliminated.
5.1. Introduction
The economist’s purview is, to a large extent, circumscribed and
consists of studying consumer and producer behavior, tastes and technology
being provided by the psychologist (social psychologist) and the production
engineer, respectively. The firm’s problem, it therefore follows, consists of
combining factor inputs to either minimize costs, or maximize profits.
Technology, as such, is assumed to be parametric. For over three decades
(1945-1980), this view pervaded economics. The growth and productivity
slowdowns of the mid-1970s, however, changed this (Maddison 1987). By
the early 1980s, it was clear that the trend rate in growth had shifted
downward, prompting economists in general, and “growth economists” in
particular, to take a closer look at the A (technology) scaler in pro-duction
functions. It would be reasonable to expect that at this point in time, the
profession would have turned to the experts, namely, production engineers.
However, instead of calling upon production engineers for insights into the
“growth and productivity slowdowns,” the profession took it upon itself to
78 5. Engineering and Economic Growth
model (read: endogenize) A, the end result of which was “endogenous

growth theory,” or “new growth theory.”1
This raises a number of methodological issues. For example, why have
growth theorists in general and new growth theorists in particular not
considered other potentially fruitful approaches, such as production
engineering and applied physics? Why have production engineers and
applied physicists been absent from the debate over growth? Why was the
set of hypotheses regarding growth in general and the growth and
productivity slowdown not widened? As Richard Nelson has argued,
diversity in research strategies is, in general, a welfare-maximizing strategy
(Mansfield et al. 1971, Beaudreau 1996). History is replete with cases in
which research programs that appeared to be the least promising ex ante,
often turned out to be the most successful, ex post.
Given the problems with new growth theory (Jones 1995), perhaps it is
time to widen the set of ideas and hypotheses, not to mention actors, in the
debate over growth. This chapter is one such attempt. Specifically, a model
of growth based on production as seen by production engineers and applied
physicists is presented and tested using data from the manufacturing sectors
of three industrialized economies, namely, the United States, Germany and
Japan. The results are then used to examine the productivity slowdown in
each of these three countries. The chapter is organized as follows. In the
first section, we examine production and as such growth processes as seen
by engineers and by economists. Finding discordant views of the role of
energy, capital and labor in production processes, we proceed to present a
consilient model of output and growth based on energy and organization.
Output is assumed to be an increasing function of energy and organization,
the latter encompassing tools, supervision (labor) and information. The
predictions of the model are then tested against manufacturing data for the
United States, Germany and Japan for the post-World War II period. The
results corroborate the predictions of basic applied physics, notably that
energy is a key factor input, one whose role in production processes is more
important than is generally acknowledged.
1 The formalization notwithstanding, new growth theory offers little by way of new
ideas about growth per se. Early work by Paul Romer (Romer 1986, 1994) formalized
the notion of increasing returns to scale found in Edward Denison’s work on the
sources of economic growth. Recent work by Philippe Aghion and Peter Howitt
(Aghion and Howitt 1998) formalizes Joseph Schumpeter’s notion of creative
destruction.
5.2 Growth Processes: Engineering versus Economics

By definition, growth processes refer to the relationship between time
rate of change of outputs and inputs. In economics, output processes are
typically modeled in terms of the factor inputs, namely capital and labor and
variations thereof. As such, growth processes in economics refer to the
relationship between the rate of growth of output and the rate of growth of
input(s). Growth accounting is the study of the relative contributions of
various factor inputs to overall factor growth. Unexplained growth is
referred to as the “Solow” residual, or, the “measure of ignorance.”
Figure 5.1: Structure of Production Processes
Source: Alting (1994); p. 39.
In engineering, output processes are typically modeled in terms of

material, energy and information. According to Leo Alting: “the term
process can, in general, be defined as a change in the properties of an object,
including geometry, hardness, state, information content (form data), and
so” (Alting 1994, 38). To produce any change in property, according to
Alting, three essential agents must be available: (i) material, (ii) energy, and
(iii) information. Figure 5.1 illustrates the engineering approach to material
processes. Information is embodied in the various tools used in any given
production process.
The resulting growth process follows directly, namely that output
growth is an increasing function of material growth, energy growth and
information growth. It need be pointed out, however, that by information
growth, it should be understood growth in the quantity of tools and
equipment, as well as in the quality of tools and equipment (information).
That is, new, more productive tools and equipment contribute to increasing
energy efficiency.2
2The post-energy crisis emphasis on energy efficiency is an example of information

contributing to raising second-law efficiency. More complete and detailed information
on production processes and sub-processes increased efficiency.
Despite a common objective, namely of modeling production and

growth processes, the resulting analyses differ fundamentally. Economists
focus on capital and labor, while engineers focus on energy and information
(tools and equipment). Interestingly, energy is not considered to be a
productive factor input in most economic models. Also, interesting is the
fact that capital (tools) is not considered to be a physically-productive factor
input in classical mechanics.
This raises an interesting methodological question, namely, if process
engineering is, in fact, the source of knowledge regarding production
processes in economics, then why are production processes and growth
processes in engineering and economics fundamentally different, theoretically
and formally? Why is energy not included as a productive factor? Why is
capital (tools and equipment) assumed to be physically productive when in
classical mechanics, it is not?3 Why is labor altogether absent from process
engineering?
These questions and issues are not entirely new. At the beginning of the
20th century, writers such as Thorstein Veblen, Frederick Soddy, Walter
Rautenstrauch and Howard Scott, raised similar concerns. A common theme
running through this literature is the absence of energy from the analysis.
Others have attempted to address these concerns by including energy in the
standard, neoclassical two-factor model (Berndt and Wood 1975, Kummel
et al., 1998). In this literature, energy as a factor input is simply included as
a factor input alongside traditional capital and labor. The results have been
mixed. Using indirect estimation techniques, Berndt and Wood found that
energy was a minor factor, having an output elasticity of 0.06, while
Kummel, Lindenberger and Eichhorn found an output elasticity of 0.51.
While these findings are promising, they remain incomplete, at least
theoretically. Lacking is a theory of production sufficiently general to
include energy and information, as well as the traditional inputs of capital
and labor. The resulting production processes and growth processes
continue to be somewhat at odds with process engineering and classical
mechanics. For example, if tools and equipment are not physically
productive, then how should the corresponding output elasticities be
interpreted? If labor is a supervisory input, then how should its output
elasticity be interpreted?
In the next section, a consilient (engineering and economics) model of
production processes and growth processes is presented and tested against
3 According to Betts: “Machinery is used to change the magnitude, direction and
point of application of required forces in order to make tasks easier. The output of
useful work from any machine, however, can never exceed the total input of work
and energy (Betts 1989).” See also Beiser (1983).
manufacturing data for the United States, Germany and Japan. Unlike
previous attempts at integrating energy into production theory, the energy-
organization framework provides a theoretically-consistent model of
production that does not violate the basic principles of classical mechanics
and basic production engineering.
5.3 The Energy-Organization Framework:

Theory and Evidence
As pointed out, what is needed is an approach to modeling production
processes that is consistent with both engineering and economics, one that
provides a bridge between the two and, as such, allows for meaningful
exchange across disciplines. To this end, a number of conditions have to be
met. One such condition is that it include energy, information, capital and
labor. A second is that it be consistent with the laws of physics, specifically
the laws of classical mechanics. Third, the results must address the concerns
(preoccupations) of both economists and engineers.
One such approach is the energy-organization model of production and
growth processes (Beaudreau 1995, 1998). In short, work (value-added) is
modeled as an increasing function of energy consumption and second-law
efficiency, the latter being a measure of the productivity of energy. Second-
law efficiency captures the organizational (informational) aspects of
production, and, as pointed out in the engineering literature, is a function of
tools (capital) and supervision (information, labor). As pointed out above,
tools “change the magnitude, direction and point of application of required
forces in order to make tasks easier.” Supervision can be disaggregated into
higher- and lower-level supervision, with the former referring to managerial
functions, and the latter referring to machine operativesthat is, employees
that oversee the workings of machines. It follows that better tools and
supervision raise second-law efficiency, and, hence, increase the amount of
work that can be extracted from a given supply of energy. This view of
production is consistent with the principles of classical mechanics,
mechanical engineering and thermodynamics, as well as with material
processes found elsewhere in nature (e.g. photosynthesis). Moreover, it
improves on engineering models of production by explicitly incorporating
upper-level supervision (i.e. management) as a factor input.
This theory predicts that output growth is increasing in the rate of growth
of energy consumption growth, and variations in second-law efficiency.4
Energy and organization are, in general, complementary, and, as such,
collinear factor inputs. Increases in energy consumption will, in general, be
accompanied by increases in tools and supervision (organization). Variations
(growth) in second-law efficiency are the result of technological change.
That is, as a result of technological change, tools and supervision (organization)
are able to extract more work (value-added) out of a given amount of
energy.5 In general, output growth will be driven by energy consumption
growth, variations in second-law efficiency, and accompanied by a
corresponding increase in tools and supervision.
In the next section, this theory of production is tested against data for
U.S., German and Japanese manufacturing. As energy is assumed to be the
only physically productive factor input (as opposed to organizationally
productive), the theory predicts that the output elasticity for energy should
be greater than the output elasticity for either conventionally-defined capital
(tools) and/or conventionally-defined labor (supervision).
5.3.1 Energy and Organization: The Case of U.S., German

and Japanese Postwar Economic Growth
To test the energy-organization model, one would need to test a series
of propositions, including: (i) that energy is the only physically-productive
factor input, (ii) capital and labor (homogeneous) are not physically
productive, and (iii) that second-law efficiency is increasing in the quality
of capital and labor (information). In short, one would have to test the basic
principles of classical mechanics. Unfortunately, the nature of the data (i.e.
aggregate), in conjunction with the very nature of production, make exact
tests of these hypotheses difficult given current statistical techniques.6 For
example, as tools (or supervision) and output are collinear, it stands to
reason that by performing standard regression analysis, a positive output
elasticity will be found. The question then is, what is the meaning of such a
result? Does it mean that labor is physically productive? It is important to
4 For the case in which second-law efficiency and energy consumption enter
multiplicatively, the rate of growth of output is simply the sum of the rate of growth
of energy consumption, and the rate of growth of second-law efficiency.
5 A good example of this is James Watt’s external condenser, which raised the
second-law efficiency of the steam engine by roughly 200%.

6 This is especially true at the aggregate level. Production processes differ markedly
within and across sectors.

note that collinearity does not necessarily imply causality nor physical
productivity.
Despite these shortcomings, conventional testing of production
relationships can nonetheless yield interesting insights, notably with regard
to the relative contribution of energy, capital and labor. Standard production
theory (e.g. the KLEMS approach) views energy as a minor factor input,
relative to capital and labor. The energy-organization approach sees the
converse, namely that capital and labor are minor factor inputs (productivity-
wise), and energy as the prime factor input.
In this section, the energy-organization model of production is tested
indirectly, and the resulting output elasticities are used to reexamine the
productivity slowdown. Drawing from earlier work (Beaudreau 1995,
1998), it was shown that by widening the set of productive factor inputs to
include energy, specifically electric power consumption, and estimating the
output coefficients directlyas opposed to via cost datafundamentally-
different output elasticities are obtained, which, when used in standard
growth accounting exercises reduces the unexplained variation (i.e. the
Solow residual), to the point of explaining almost entirely the productivity
slowdown.
In earlier work (Beaudreau 1995, 1998), we estimated the energy-
organization production function using a modified KLEMS production
function, the KLEP (capital, labor and electric power) production function.
Electric power was used as a proxy for energy consumption given the
prevalence of electricity as the prime mover in modern manufacturing (see
Sonenblum 1990). Capital and labor were used as proxies for tools and
supervision.7 The resulting output elasticities, estimated directly using post
World War II input and output data for U.S., German and Japan
manufacturing, were then used in standard growth accounting exercises.
The relevant output elasticities were estimated directly (i.e. as opposed
to indirectly using cost data). Among the reasons for this is the absence of
competitive factor markets for both energy and labor (Blanchflower et al.
1996, Van Reenen 1996). Specifically, electric power markets are, in
general, regulated, via public utility commissions, while labor markets,
7 Unfortunately, qualitative measures or indexes (information) of capital and labor

in manufacturing (and for other sectors) are unavailable at the aggregate level.
Ideally, such measures could be used to proxy innovations in second-law efficiency.
For more on the role of information in production, see (Jorgenson 2001), where
investment in information technology (IT)-based capital is used as a proxy for
information. Unfortunately, such a measure is of limited value in our work, given
the complementary nature of IT-capital and supervision (automation), and the lack
of a reliable measure of information quality.
according to Blanchflower et al. and Van Reenen, are, in general, non-

competitive. Taken together, these violate one of the cardinal assumptions
of growth accounting, namely that factor markets are competitive. As such,
data on value-added, electric power consumption, total employment and
capital for U.S., German and Japanese manufacturing were used to estimate
the Cobb-Douglas KLEP production function: ܳ = ‫ܲܧ‬ఉభ ‫ܮ‬ఉమ ‫ܭ‬ఉయ .
The estimated output elasticities for all three countries are presented in
Table 5.1. What is striking are the similarities across countries.8 The
relevant energy elasticities for the U.S., West Germany and Japan are 0.54,
0.50 and 0.45, respectively. In all three cases, electric power consumption
is, by far, the most important factor input, as evidenced by output elasticities
for U.S. manufacturing, German manufacturing and Japanese manufacturing
of 0.537244, 0.747482 and 0.605599, respectively. Capital and labor output
elasticities are lower than in previous studies (e.g. Berndt and Wood 1975).

Manufacturing

1950-1984 1962–1988 1965-1988
EP 0.533046 0.728269 0.609444
(10.079) (4.493) (2.067)
L 0.418822 0.249191 0.193766
(18.231) (2.332) (1.847)
K 0.064250 0.0339365 0.193766
(2.768) (0.543) (1.608)
R2 0.984 0.968 0.981
F 1032.52 367.85 265.98
Source: Beaudreau (1998, 1999). t-values are in parentheses.
These estimates, one could argue, provide some support for the energy-
organization view of production presented above. Each of these output
elasticities, in conjunction with the laws of basic mechanics (i.e. regarding
tools), and the supervisory nature of labor, is consistent with the predictions
of the energy-organization model. As pointed out earlier, positive capital
and output elasticities have little theoretical relevance, beyond the collinear
nature of organization-related factor inputs.
These elasticities were then used to construct a fixed-weight aggregate
input index for each country, both before and after the energy crisis (circa
1974). Table 5.2 reports the relevant growth rates for manufacturing value
8Comparable estimates were obtained by (Kummel et al. 1998), using a generic

homogeneous production function of order one.
added (Q), electric power (EP), labor (L) and capital (K), as well as the
relevant fixed-weight aggregate input index (AI) (Hisnanick and Kymm
1992; Gullickson and Harper 1988) for three time intervals: 1950-1984,
1950-1973 and 1974-1984, for all three countries.
U.S 1950–1984 1950–1973 1974–1984

USVA 2.684 3.469 0.121
USAI* 2.674 3.472 0.310
USEP 4.052 5.371 0.246
USN 0.662 0.900 -0.091
USK 3.694 3.614 3.4008
Germany 1963–1988 1962–1973 1974–1988
GERVA 2.462 6.522 1.486
GERAI* 2.433 5.190 1.080
GEREP 2.894 5.883 1.366
GERN -0.785 0.592 -0.938
GERK 2.945 5.620 1.406
Japan 1965-1988 1965-1973 1974-1988
JAPVA 3.826 8.844 3.099
JAPAI* 3.566 9.856 1.538
JAPEP 3.559 11.320 0.965
JAPN -0.082 2.297 -0.367
JAPK 7.520 13.536 5.182
*‫ߚ = ܫܣ‬ா௉ ா௉ ௗ௧
+ ߚே
ே ௗ௧
+ ߚ௄
௄ ௗ௧
Unlike previous studies which found a non-negligible gap between rates

of output growth and rates of aggregate input growth (i.e. the Solow
residual), these results show that growth in U.S., German and Japanese
manufacturing value-added is nearly fully explained by growth in the
relevant fixed-weight factor input growth indexes (Beaudreau 1995, 1998).
For the complete period 1950-1984, U.S. manufacturing value-added
increased at an average annual rate of 2.684%, while the aggregate input
increased at 2.655%. For the first sub-period 1950-1973, it increased at
3.469, while the aggregate input increased at 3.466%. In the second sub-
period, 1974-1984, it increased at an average annual rate of 0.121, while the
aggregate input increased at 0.310.
The results show that foremost among the causes of growth in U.S.,
German and Japanese manufacturing value-added is electric power
consumption. In the U.S. case, electric power consumption increased at an
average annual rate of 4.052 percent over the period 1950–1984. Per worker
consumption of electric power in U.S. manufacturing in this period goes
from 12, 534kwh in 1950 to 41, 688kwh in 1984, a total increase of 232
percent (U.S. Department of Commerce various years). When multiplied by
the relevant output elasticity (i.e. 0.537244), the growth in electric power
consumption in U.S. manufacturing accounts for 82% of overall output
growth (i.e. 2.17/2.68), which corroborates the predictions of energy
organization-based production analysis. Prior to the energy crisis (i.e.
1973), electric power consumption in manufacturing increased at an average
annual rate of 5.371%. Output increased at an average annual rate of
3.466%. Electric power consumption growth, it then follows, accounts for
83 percent of output growth (i.e. 2.88/3.46). In the post-energy-crisis period,
electric power consumption increased at an average annual rate of 0.246,
while output increased at an average annual rate of 0.121.
5.4 The Role of Capital and Labor in Growth Accounting

Traditional growth accounting, best associated with the work of Edward
(Denison 1962), and energy organization-based growth accounting (Beaudreau
1998) differ in their view of the causes of growth. The former attributes
growth to capital, labor and a residual, while the latter attributes growth to
energy consumption, and variations in second-law efficiency. The former
views energy as an intermediate input, and, consequently, not productive in
the physical sense; the latter views capital and labor as organizational
inputs, and, consequently, as not productive in the physical sense.
Both cannot be right. Capital cannot be viewed as physically productive
by economists, and not physically productive by engineers. Labor cannot be
viewed as physically productive by economists, and not physically
productive by engineers. Clearly, more work needs to be done on what we
consider to be the fundamental questions in production theory.
5.5 Conclusions
While this chapter has been somewhat critical of growth theory in its
current state, especially the lack of diversity in research programs, it
findings and conclusions should nonetheless be seen as prospective in
nature, providing an additional—we hesitate to use the word novel—
approach to thinking about and modeling growth. In other words, there is
room for competing and complementary approaches. For example, the
energy-organization approach to growth, used here, can be viewed as
complementary to general purpose technology (GPT) analysis (Bresnahan
and Trajtenberg 1992, Helpman 1998) where the emphasis, empirically

speaking, has been on energy-related innovations (the steam engine, the
electric motor).
The theory of economic growth presented here is not only consilient
with the theory of material processes as found in process engineering, but is
also consilient with work on entropic processes found elsewhere in nature
(Wilson 1998). As such, it is consistent with the principles of classical
mechanics and thermodynamics. It is felt that this constitutes a step in the
right direction, in so far as growth theory, and indeed, in so far as production
theory in general is concerned. Much, however, remains to be done,
especially in so far as modeling the organization factor input is concerned.
5.6 References
Aghion, P. and P. Howitt. 1998. Endogenous Growth Theory. Cambridge,
MA: MIT Press.
Alting, L., 1994. Manufacturing Engineering Processes. New York, NY:
Marcel Decker.
Beaudreau, B.C. 1995. The Impact of Electric Power on Productivity: The
Case of U.S. Manufacturing 1958-1984. Energy Economics. 17: 231-
236.
Beaudreau, B.C., 1996. “R&D: to Compete or to Cooperate.” Economics of
Innovation and New Technology. 4, 173–186.
Beaudreau, B.C. 1998. Energy and Organization: Growth and Distribution
Reexamined. Westport, CT: Greenwood Press.
Beaudreau, B.C. 1999. Energy and the Rise and Fall of Political Economy.
Beiser, A. 1983. Modern Technical Physics. Menlo Park, CA: The
Benjamin-Cummings Publishing Co.
Berndt, E., Wood, D.O. 1975. “Technology, Prices and the Derived Demand
for Energy.” The Review of Economics and Statistics: 259–268.
Betts, J.E. 1989. Essentials of Applied Physics. Englewood Cliffs, NJ.
Blanchflower, D.G., Oswald, A.J. and S. Peter. 1996. “Wages, Profits, and
Rent Sharing.” Quarterly Journal of Economics. 60: 227–251.
Bresnahan, T. and Trajtenberg M. 1992. General Purpose Technologies:
Engines of Growth? National Bureau of Economic Research Working
Paper No. 4148. Cambridge, MA.
Denison, E.F. 1962. The Sources of Economic Growth in the United States
and the Alternatives Before Us. Committee for Economic Development,
New York, NY.
Gullickson, W. and Harper, M.J. 1988. Multifactor Productivity in U.S.

Manufacturing1949–1983. Monthly Labor Review:18–28.
Helpman, E. 1998. General Purpose Technologies and Economic Growth.
Hisnanick, J.J. and Kymm, K. 1992. “The Impact of Disaggregated Energy
on Productivity.” Energy Economics: 274–278.
Jones, C. 1995. “R&D-based Models of Economic Growth.” Journal of
Political Economy. 103: 759–784.
Jorgenson, D.W. 2001. “Information Technology and the U.S. Economy.”
American Economic Review. 91: 1–32.
Kummel, R. Lindenberger, D. and W. Eichhorn. 1998. “The Productive
Power of Energy and Economic Evolution.” Indian Journal of Applied
Economics (special issue on macro and microeconomics).
Maddison, A. 1987. “Growth and Slowdown in Advanced Capitalist
Economic Literature. 25: 649–698.
Mansfield, E., J. Rapoport, J. Schnee, S. Wagner and M. Hamburger. 1971.
Research and Innovation in the Modern Corporation. New York, NY:
W.W. Norton.
Romer, P.M. 1986. “Increasing Returns and Long-Run Growth.” Journal of
Political Economy. 94:1002–1037.
Romer, P.M. 1994. “The Origins of Endogenous Growth.” Journal of
Economic Perspectives. 8: 3–22.
Sonenblum, S. 1990. Electrification and Productivity Growth in Manufacturing.
in: Shurr S., et al. (eds.), Electricity in the American Economy. Westport,
CT: Greenwood Publishing.
Van Reenen, J. 1996. “The Creation and Capture of Rents: Wages and
Innovation in a Panel of U.K. Companies.” Quarterly Journal of
Economics. Vol. 61:195–226.
Wilson, E.O. 1998. Consilience, the Unity of Knowledge. New York, NY:
Vintage Books.
6
THE DYNAMO AND THE COMPUTER:

AN ENGINEERING PERSPECTIVE ON THE
MODERN PRODUCTIVITY PARADOX
Abstract
Purpose–The purpose of this chapter is to provide an engineering

perspective on the modern productivity paradox. Specifically, to shed new
light on the failure of information and communication technology (ICT) to
increase overall factor productivity.
Design/methodology/approach–To this end, alternative approaches to
modeling material processes are presented and discussed. Empirical
evidence is brought to bear on the question of ICT productivity. Finally, the
implication of the findings for production and management technology are
presented and discussed.
Findings–The principal finding is theoretical in nature, namely that,
according to classical mechanics and applied physics, ICT is not physically
productive. Rather, information is an organizational input.
Practical implications–By identifying the role of ICT in material
processes, the chapter provides a framework to better understand and
evaluate ICT investment, both at the firm and industry level. While ICT does
not contribute to increasing physical output, it does nonetheless increase
profitability. On a broader level, the chapter provides a framework to
evaluate ICT-related public policy measures.
Originality/value–Among the contributions of the chapter are the use of
basic engineering principles to shed light on the modern productivity
paradox and the conclusion that information, unlike energy, is not physically
productive and as such cannot be counted upon to increase output.
6.1 Introduction
The information and communications technology (ICT) revolution
remains an enigma. Three decades after the introduction of the microchip,
and 15 years after the birth of the internet, productivity, as measured by
output per worker, has increased, especially in the USA, and in the period
90 6. The Dynamo and the Computer
1995-2001, but by less than expected (Gordon 2000, 2006, Oliner and
Sichel 2000, Gust and Marquez 2000). In the 1980s and 1990s, it was
generally believed that ICTs would usher in a third industrial revolution,
and more importantly, a period of uninterrupted productivity growth, not
unlike the first two industrial revolutions, the first resulting from the steam
engine, and the second, from the dynamo. Both were associated with
extended periods of above-average growth, lasting decades:1
The advent of microcomputers since the beginning of the 1970s has brought
a new dimension to power electronics and drive technology. The impact of
this evolution is as significant as the advent of power semiconductor devices
in the 1950s. In the forthcoming industrial revolution, that is the
“computerized automation of factories,” microcomputers will not only
provide intelligence to higher-level factory automation, but will play a vital
role in the control of lower-level power electronics and motion control
systems. Microcomputers have now been accepted universally for the
control of power electronic and drive systems. It is interesting to note that
both the ends of the power-electronics spectrum are digital: one provides the
brain, and the other provides the muscle (Bose 1987).
That the ICT revolution has not, at least yet, produced anything
comparable to the growth rates in the 1920s (the roaring 1920s), has raised
a number of questions (Maddison 1987, David 1990, Gordon 2000, 2006,
Gust and Marquez 2000). Foremost among these has, without a doubt, been
why? Why have ICTs not, as was generally expected, increased overall
productivity, and, consequently, wealth? Related questions include, how
does the computer stack up against the steam engine, against the dynamo?
David (1990), in an article entitled The Dynamo and the Computer: An
Historical Perspective on the Modern Productivity Paradox, raised the
possibility of lengthy diffusion lags, not unlike those experienced by U.S.
manufacturers with the dynamo (electrical motor) in the late 19th/early 20th
century:
My aim here simply is to convince modern economic analysts (whether

perplexed by the productivity slowdown, or not) of the immediate relevance
of historical studies that trace the evolution of techno-economic regimes
formed around the general-purpose engines. The latter, typically, are key
functional components embodied in hardware that can be applied as
elements or modular units of the engineering designs developed for a wide
variety of specific operations or processes (David 1990, 355).
1 See Jorgenson (2001) for more on the ICT revolution and productivity in the USA.
Implicit in David’s work and, indeed, in the current growth literature

(Bresnahan and Trajtenberg 1994, Oliner and Sichel 2000) is the assumption
that the computer and the dynamo are, technologically speaking, comparable.
That is, output is assumed to be increasing – by a non-negligible amount–in
both. This raises the obvious counterfactual question, namely, what if they
are not? What if information (computer) and energy (dynamo) differ
fundamentally–that is, as far as productivity goes? If so, then the issue of
diffusion lags becomes a second-order consideration. Unfortunately, the
role of ICT in production, specifically, its contribution to productivity
growth, is not well understood. Little theoretical work exists (Oliner and
Sichel 1994, Jorgenson and Stiroh 1995, Jorgenson 2001). For example, in
Jorgenson’s work, ICT enters via the capital input, specifically via “capital
services provided by computer equipment capital” (i.e. ICT-related capital).
Just how more and better information contributes to increasing physical
output is not well-specified2. Similarly, the role of energy in production is
not well understood (Berndt and Wood 1975). Is it a physically productive
factor input, or is it an intermediate input?
This raises another important theoretical question, namely, are energy
and information “technologies” in the traditional sense, or are they factor
inputs? If they are technologies, then their effect will be relatively short-
lived. If they are factor inputs, then it stands to reason that their effect (on
output) will be felt over time, say through information-deepening and
energy-deepening, the process whereby information consumption and
electric consumption per unit of capital increase over time.
Theoretically, at least, it stands to reason that for ICT to contribute to
sustained, long-term economic growth, information has to be considered as
a factor input. That is, more information will contribute to greater wealth.
Ideally, what is needed is a more complete theory of production, one general
enough to incorporate energy and information as factor inputs. Engineering,
specifically material process engineering, provides an alternative to
Jorgenson-style attempts at including energy and information in the production
function.
6.2 Production Processes as Seen by Engineers

Consider, first, the following description of production processes found
in Alting’s (1994) Manufacturing Engineering Processes: “the term process
can, in general, be defined as a change in the properties of an object,
2 Oliner and Sichel also enlarge the definition of computers to encompass all of
information processing equipment and also computing software and computer-using
labor. The end result, however, remains unchanged.
including geometry, hardness, state, information content (form data), and

so.” To produce any change in property (value-added VA), according to
Alting (1994), three essential agents must be available: (1) material, (2)
energy and (3) information.
Figure 6.1: Manufacturing Production Processes
Source: Alting (1994).
Figure 6.1, taken from Alting (1994), is the corresponding flow diagram.
In short, production processes are energy and information based. Capital
and labor, the workhorses of production theory in economics, are, at least
on the surface, not mentioned, at least not explicitly. However, as it turns
out, they play a crucial role in organizing energy and information. We will
return to this later.
Unlike standard production theory where energy and information enter
parametrically via the technology scaler (A in the Cobb-Douglas specification),
in engineering they constitute the two key factor inputs (Alting 1994, Beiser
1983). According to Figure 6.1, energy and information are what transform
materials. This raises a number of fundamental questions, notably, what is
the exact role of each? Are they both physically productive? Are they
comparable as Paul David and Dale Jorgenson maintain? Consider classical
mechanics and thermodynamics. According to classical mechanics, energy
being the source of force, is the basis of work. For example, work is defined
as the product of force and distance (W=fd). This principle carries over to
thermodynamics where energy, work and heat behave according to the
equation ܷ݀ = ߜܳ െ ߜܹ, where dU is the net increase in the internal
energy of a thermodynamic system, GQ is the heat entering the system and
GW is the work done by the system.3 The amount of useful work that can be
extracted from a thermodynamic system is determined by the second law of
thermodynamics. In essence, it states that not all energy can be transferred
into useful work. The better is the system, the more work can be extracted
3 This is known as the first law of thermodynamics.

from energy; however, second-law efficiency is always and everywhere less

than unity4.
According to classical mechanics, information is embodied in
machinery, directing and applying force/energy:
Machinery is used to change the magnitude, direction and point of

application of required forces in order to make tasks easier. The output of
useful work from any machine, however, can never exceed the total input of
work and energy (Betts 1989, 172).
This was understood in economics as far back in time as 1832 when

Babbage (1832) in The Economy of Machinery and Manufactures pointed
out:
Of the class of mechanical agents by which motion is transmitted – the lever,

the pulley, the wedge, and many others–it has been demonstrated, that no
power is gained by their use, however combined. Whatever force is applied
at one point can only be exerted at some other, diminished by friction and
other incidental causes [...]. Of those machines by which we produce power,
it may be observed that although they are to us immense acquisitions, yet in
regard to two of the sources of this power–the force of wind and of water–
we merely make use of bodies in a state of motion by nature; we change the
directions of their movement in order to render them subservient to our
purposes, but we neither add to nor diminish the quantity of motion in
existence. The force of vapour is another fertile source of moving power;
but even in this case it cannot be maintained that power is created. Water is
converted into elastic vapor by the combustion of fuel. The chemical
changes which thus take place are constantly increasing the atmosphere by
large quantities of carbonic acid and other gases noxious to animal life.
Information plays a similar role in thermodynamics where it defines the

relevant thermodynamic system (microscopic). Production processes
(thermodynamic systems) differ with regard to their overall efficiency,
defined as overall work divided by available energy. Accordingly, better
systems (better information) increases productivity, ceteris paribus; however,
more of the same information has no such effect. What’s more, the increase
in productivity associated with “better information” is of a one-shot nature
(punctual). Given the overall stability of second-law efficiency for any
given thermodynamic system, it cannot be counted on as a source of growth
in the long run.
4Useful energy (i.e. that which can be extracted from a given supply of energy) is
sometimes referred to as exergy or energy availability.
This brings us to conventionally defined labor which, surprisingly, is

ignored in process engineering and descriptions of thermodynamic systems
(production processes). Labor is conspicuous by its absence, raising the
obvious question of why? Why is the labor input ignored? The reason, we
posit, has something to do with the role of labor in modern production
processes, namely as machine operatives, a point made by Alfred Marshall
over a century ago. Labor oversees machinery (thermodynamic system):
of more than half the working men even in such a country as England in
with much less toil for themselves. Nothing could be more narrow or
many times as much work in the course of a day as the old hand loom did;
than his did (Marshall 1890, 218).
Broken down by tasks, machine operatives, like modern day, computer-

based manufacturing control devices collect information on material
processes and respond to such information in predetermined algorithms. Put
differently, workers are overseers in the sense of overseeing material
processes (Ludema et al. 1987, Lutters et al. 1997). This is also known as
the human-machine interface defined as operator interface terminals with
which users (in this case, labor) interact in order to control other devices.
The point here is that the modern-day labor input is no longer a source of
energy/heat but rather a human-based control device, collecting and
processing production process-related information. This, we argue, provides
new avenues for formalizing the role of computers and control devices in
manufacturing processes, namely as a low-cost substitute for human control
devices. Ludema et al. (1987, 71), in Manufacturing Engineering: Economics
and Processes, describes automation and control similarly:
The next step is to relieve human of the task of guiding or controlling the
tool or machine. Tools or machines that control their own motions within
prescribed limits are said to be automatic; or by virtue of having control
devices connected to them, tools or machines are automated [.. .]. In many
instances, a tool or machine is automated to maintain product quality, A
hamburger grill may become too hot to produce quality hamburgers, unless
controlled. Likewise, the chemistry of an uncontrolled electroplating bath
may vary from the acceptable range, necessitation the stopping of production
while adjustments are made[. . .]. Automatic machines can often produce
parts with greater accuracy, uniformity, and speed than can manually-
operated machines; they also relieve humans of some tedious, dangerous,
and hard jobs.
As such, replacing a worker by a control device (computer, etc.) will, in

most instances, be profit-increasing; however, it will not increase physical
productivity the latter being measured as the ratio of output of VA (work)
per unit of energy (second-law efficiency). Firms will simply reduce their
demand for labor, or put differently, human-based control devices. Ceteris
paribus, conventionally defined labor productivity (ratio of output to labor
input) will increase owing to a lower denominator (i.e. labor input).5
To summarize, we have shown that labor and capital, the bread and
butter of standard neoclassical economics, can be deconstructed in terms of
each’s information content. The modern-day labor input is, for the most
part, a supervisory input, collecting and processing information on specific
material processes. Capital, on the other hand, embodies the information
that defines the material process in question. What is important to note,
however, is the fact that neither is physically productive, not being sources
of force/heat.
6.3 Implications for the Modern Productivity Paradox

As pointed out in the Introduction, the purpose of the chapter is to
explore the possible contribution of the fields of engineering and classical
mechanics to our understanding of the role of energy and information, as
well as capital and labor in modern production processes, and, in the
process, provide new leads in the still unresolved debate surrounding the
modern productivity paradox (Solow Paradox). Here, using the elements of
the engineering approach to understanding material processes presented
above, we reexamine the modern productivity paradox. To begin with, we
examine the question of ICT-based investment and growth. According to
our results, energy (force and heat) is the only physically productive factor
input, while information is organizationally productive (machinery and
labor). It therefore stands to reason that the reported absence of a significant
positive correlation between ICT-based investment and growth (Gust and
Marquez, 2000) is not only not surprising, but also in fact predicted by the
theory. Information, by its very nature, cannot increase physical output,
except indirectly, via second-law efficiency.
5Perhaps this explains why workers who use computers have higher earnings than
workers who do not (Kreuger 1993, Black and Lynch 1996).
Evidence presented by Gust and Marguez (2000) shows that foreign

countries did not experience the same acceleration in productivity growth
from 1995 to 2001 as the USA, despite “significant proliferation of (ICT)
technologies abroad” (Gust and Marquez, 2000, 14). For example, they
show that despite high-ICT goods to gross domestic product ratios, Ireland,
Japan, Finland, and Sweden did not experience high rates of growth of multi-factor
productivity. Again, this is not unexpected from the applied-physics point of view.
In more recent work using industry data for four European countries
(France, Germany, The Netherlands, and the UK) and the USA, Inklaar et
al. (2005) show that the ICT-based productivity gains at the aggregate level
reported by Jorgenson and Stiroh (1995) and others do not generalize across
industries but rather are concentrated in the financial services/communications
industries. Table 6.1 shows the extent to which the effect of ICT on labor
productivity from 1995 onwards was limited to the service sector. The
greatest gains occurred in the financial services and business sectors (both
in the four European countries and in the USA). The contribution of ICT–
using manufacturing and non-ICT industries in the four European countries
combined (0.03 and 0.11, respectively) represents roughly one-quarter of
the overall gain (0.54). Similar results were found for the USA, raising
further doubt regarding the ICT-productivity nexus and, more importantly,
raising questions regarding its status as a general purpose technology. In
both periods (1979-1995) and (1995-2000), ICT had a negligible effect on
aggregate labor productivity. And moreover, what effect it did have
(principally in the service sector) has recently been called into question by
the findings of Triplett and Bosworth (2008) who show that shift from the
old US Standard Industrial Classification system to the new North American
Industry Classification System introduced a number of distortions which,
when accounted for, reduce considerably the reported labor productivity
gains.
That ICT has not produced the anticipated increases in physically
productivity (i.e. increase labor or capital productivity) however does not
imply that it has not been and is not profit-increasing. ICT investment has
grown rapidly throughout the 1990s and 2000s which raises the question,
why? The answer, we believe, lies in the resulting cost savings. Being
substitutes for human being-based control devices, ICT-based control
devices, are on average cheaper, and, in many cases, more reliable, lowering
overall costs and raising second-law efficiency by minimizing machine
Table 6.1: Contributions to Aggregate Productivity Growth of Industry

ICT Capital Deepening, EU-4 and USA
1979-1995 1996-2000
EU-4 USA USA-EU EU-4 USA USA-EU
Total economy 0.33 0.46 0.12 0.53 0.86 0.33
ICT-producing 0.04 0.06 0.02 0.07 0.11 0.04
industries
Electrical and electronic 0.01 0.04 0.02 0.02 0.05 0.04
Equipment and 0.03 0.02 0.00 0.05 0.05 0.00
instruments
communications
ICT-using industries 0.21 0.28 0.07 0.35 0.57 0.22
ICT-using 0.02 0.02 0.01 0.03 0.03 0.01
manufacturing
Wholesale Trade 0.03 0.08 0.05 0.07 0.13 0.06
Retail trade 0.01 0.04 0.03 0.03 0.05 0.02
Financial intermediation 0.08 0.11 0.03 0.10 0.27 0.17
Business services 0.07 0.04 20.03 0.12 0.09 20.04
Non-ICT industries 0.08 0.11 0.03 0.11 0.18 0.07
Agriculture, forestry 0.00 0.00 0.00 0.00 0.00 0.00
and fishing
Mining and quarrying 0.00 0.01 0.01 0.00 0.00 0.00
Non-ICT manufacturing 0.04 0.04 0.00 0.04 0.05 0.01
Transport and storage 0.00 0.01 0.00 0.01 0.02 0.01
Social and personal 0.01 0.01 0.01 0.01 0.03 0.02
services
Non-market services 0.01 0.03 0.01 0.02 0.04 0.03
Other non-ICT 0.02 0.02 0.00 0.03 0.03 0.00
Source: Inklaar et al. (2005, 519)
breakdowns and downtime.6 The resulting robotization and rationalization,

both of which have resulted in massive job destruction, are as such profit-
increasing strategies. As labor costs on average represent 50-70 percent of
a firm’s overall costs, it stands to reason that by replacing workers with
6According to ABB, automation assets typically include hard measurement assets

such as pressure transmitters, flow meters, and control valves. Production assets
include all other equipment, which varies greatly and is vertical industry specific.
Asset classes include vessels, pumps, heat exchangers, rotating equipment, etc. 3PM
contracts vary greatly with each service supplier, with the scope dependent upon the
resource knowledge of the provider. ARC research has confirmed that the fastest
growing segment of 3PM services is provided by automation suppliers. These
suppliers have gained the trust of many users and have won 3PM contracts for all
assets, regardless of manufacturer.
control devices, firms reduce their costs, and increase their profits. The
automation and robotization literature contains numerous references to such
windfalls. It therefore stands to reason that the only place computers are
likely to show up is in labor (production and non-production worker) and
capital statistics, the result of firms substituting inanimate forms of
supervision (computers and control devices) in the place of animate (read:
human) forms. Labor input per unit output will decrease, while capital input
per unit output will increase, which is precisely what we observe in the data.
Output will not increase as a result; however, output per remaining labor–
predominantly, non-production labor–will increase.7
Figure 6.2: U.S. Net Stocks of Fixed Assets
Evidence is provided by the output and input data for the U.S., German,
and Japanese manufacturing presented in Table 6.2 taken from Beaudreau
(1999). Output is measured in terms of real VA, while inputs include electric
power consumption, broadly defined labor (L), and capital (K). We see that
in all three cases, broadly defined labor has actually decreased from 1974 to
1988. Broadly defined capital, on the other hand, has increased at rates equal
to or greater than the rate of growth of VA. For example, in US
manufacturing, VA increased at an average annual rate of 0.121 percent
7 Cast in terms of the U.S. Department of Commerce’s production versus non-

production workers nomenclature, production workers per unit of output will
decrease, while non-production workers per unit output will increase. Animate
supervision has given way to inanimate supervision. Similar arguments have been
made in the popular press, most notably by Jeremy Rifkin in The End of Work.
from 1974 to 1984. Broadly defined labor decreased at an average annual

rate of 0.091 percent, while broadly defined capital increased at an average
annual rate of 3.4008 percent. This stands in marked contrast with 1950-
1973 when capital and labor were increasing in VA.

Manufacturing

1950-1984 1962–1988 1965-1988
EP 0.533046 0.728269 0.609444
(10.079) (4.493) (2.067)
L 0.418822 0.249191 0.193766
(18.231) (2.332) (1.847)
K 0.064250 0.0339365 0.193766
(2.768) (0.543) (1.608)
R2 0.984 0.968 0.981
F 1032.52 367.85 265.98
Disaggregating total U.S. current-cost net stock of fixed capital as

reported by the U.S. Department of Commerce, Bureau of Economic
Analysis (2004) (Nonresidential Equipment and Software) into total
equipment (Industrial Equipment), and an information processing and
related equipment (Information Processing Equipment and Software)
component, we see (Figure 6.2) that from roughly 1975 on, the greatest
gains have been in information processing and related equipment. Further
evidence of this is provided by Jorgenson (2001) where (Figure 6.3) the
average annual growth rate of IT capital services has increased markedly
from 1973 onwards. As a result, Alfred Marshall’s “machine operatives”
have been, are, and continue to be replaced by computer-based control
devices. As U.S. manufacturing VA has not increased at comparable rates,
it stands to reason that investment in ICT has, as pointed out above, come
at the expense of labor–that is, human-based supervision.
Figure 6.3: IT and non-IT Capital Services
6.4 Conclusions
The purpose of this chapter was exploratory: draw from the laws of
classical mechanics and thermodynamics in search of new leads, new
perspectives in the debate over the modern productivity paradox. As pointed
out, historical comparisons such as Paul David’s and indeed those found in
the current growth literature, while extremely useful, have to rest on solid
theoretical grounds. Alikes have to be compared to alikes; apples cannot
and should not be compared to oranges. As we have shown in this note,
according to classical mechanics and thermodynamics, the dynamo and the
computer are fundamentally different, adding credence to Joel Mokyr’s
warning that: “Historical analogies often mislead as much as they instruct
and in technological progress, where change is unpredictable, cumulative,
and irreversible, the analogies [are] more dangerous than anywhere”
(Mokyr 1997, 35). One is, according to classical mechanics, the source of
all transformations in the universe, including value-added (wealth), while
the other is an organizational input.
Information, like tools, has never been, is not, and will never be
physically productive per se. This, we argue, provides an interesting
perspective on the modern productivity paradox. Computers have not shown
up in the productivity data for the simple reason that information is not
physically productive. Further, it corroborates Gordon’s (2000, 2006) claim
to the effect that the ICT revolution and the second industrial revolution
appear to be, qualitatively and quantitatively, incomparable. Clearly, more
work on the role of information and indeed, on all other non-traditional
factor inputs (e.g. energy) in wealth-generating material processes is needed
if only to reassess and reevaluate growth-related public policy. If
information is not productive as originally thought, then investments in ICT
has to be reassessed as does growth policy in general. For the past three
decades, growth policy has been based on massive investment in ICT via
direct investment and various policy incentives (tax credits, etc.). Lastly,
our findings have important implications for the question of productivity at
the firm level. For years, ICT has been front and center in the productivity
debate. Firms have invested billions in ICT in the hope of increasing profits
and dividends. As we have argued, profits and dividends have increased but
for entirely different reasons, namely by reducing costs.
It is our view that more work along the lines suggested by our model and
findings is needed. Specifically, micro-models of the role of information in
production and management technology are needed. Such models are
needed to better evaluate the various ICT innovations at the firm, industry
and national level. In sum, the time has come to integrate ICT in a
meaningful way into production and management technology.
6.4 References
Alting, L. 1994. Manufacturing Engineering Processes. New York, NY:
Marcel Decker.
Babbage, C. 1832. The Economy of Machinery and Manufacturing.
London: C. Knight.
Beaudreau, B.C. 1999. Energy and the Rise and Fall of Political Economy.
Beiser, A. 1983. Modern Technical Physics. Menlo Park, CA: Cummings
Publishing Company.
Berndt, E. and Wood, D.O. 1975. “Technology, Prices and the Derived
Demand for Energy.” The Review of Economics and Statistics. Vol. 57:
259-68.
Betts, J.E. 1989. Essentials of Applied Physics. Englewood Cliffs, NJ:
Prentice-Hall.
Black, S.E. and Lynch, L.M. 1996. “Human Capital Investments and
Productivity.” American Economic Review: Papers and Proceedings.
May.
Bose, B.K. 1987. “Introduction to Microcomputer Control.” in Bose, B.K.

ed. Microcomputer Control of Power Electronics and Drives. New
York, NY: IEEE Press.
Bresnahan, T. and M. Trajtenberg. 1994. “General Purpose Technologies:
Engines of Growth?” Working Paper No. 4148, National Bureau of
Economic Research, Cambridge, MA, August.
David, P.A. 1990. “The Dynamo and the Computer: an Historical
Perspective on the Modern Productivity Paradox.” American Economic
Review, Papers and Proceedings. May: 355-61.
Gordon, R. 2000. “Does the New Economy Measure Up to the Great
Inventions of the Past?” Journal of Economic Perspectives. No. 4: 49-
74.
Gordon, R. 2006. “The 1920s and the 1990s in Mutual Reflection.” in
Rhode, P. and G. Toniolo. eds. The Global Economy in the 1990s: A
Long-run Perspective. Cambridge: Cambridge University Press.
Gust, C. and J. Marquez, 2000. “Productivity Developments Abroad.”
Federal Reserve Bulletin. Washington, DC: Federal Reserve System:
665-81.
Inklaar, R. M. O’Mahony and M.P. Timmer. 2005. “ICT and Europe’s
Productivity Performance: Industry-Level Growth Account
Comparisons with the United States.” Review of Income and Wealth. 51,
no. 4: 505-36.
Jorgenson, D.W. 2001, “Information Technology and the US Economy.”
American Economic Review. 91, no 1: 1-32.
Jorgenson, D.W. and K. Stiroh. 1995. “Computers and Growth”, Economics
of Innovation and New Technology. 3, no. 3: 295-316.
Kreuger, A. 1993. “How Computers Have Changed the Wage Structure:
Evidence from Micro Data 1984-1989.” Quarterly Journal of
Economics. 108, no. 1: 33-60.
Ludema, K.C., R.M. Caddell and A. Atkins. 1987. Manufacturing Engineering,
Economics and Processes. Englewood Cliffs, NJ: Prentice-Hall.
Lutters, D., A.H. Streppel and H.J.J. Kals. 1997. “Product Information
Structure as the Basis for the Control of Design and Engineering
Processes.” Proceedings of the 29th CIRP International Conference on
Manufacturing Systems: 167-86.
Maddison, A. 1987. “Growth and Slowdown in Advanced Capitalist
Economies.” Journal of Economic Literature. 25: 649-98.
Marshall, A. 1890. Principles of Economics 8th ed. London: Macmillan.
Mokyr, J. 1997. “Are We Living in the Middle of an Industrial Revolution.”
Federal Reserve Bank of Kansas City Economic Review. 82, no. 2: 31-
43.
Oliner, S.D. and D.E. Sichel. 1994. “Computers and Output Growth
Revisited: How Big is the Puzzle?” Brookings Papers on Economic
Activity. 2: 273-317.
Oliner, S.D. and D.E. Sichel. 2000. “The Resurgence of Growth in the Late
1990s: Is Information Technology the Story.” Journal of Economic
Perspectives.14, no. 4: 3-22.
Triplett, J. and B. Bosworth. 2008. “The State of Data for Services
Productivity Measurement in the United States.” International
Productivity Monitor. 16: 53-71.
U.S. Department of Commerce. 2004. Bureau of Economic Analysis U.S.
Net Stock of Private Fixed Assets. Washington, DC.
7
THE PHYSICAL LIMITS TO ECONOMIC

GROWTH BY R&D FUNDED INNOVATION
Abstract
For over three decades, worldwide R&D expenditure has risen steadily,
reaching $1.3 trillion in 2011. Underlying this unprecedented growth is a
deeply-held belief that R&D is a prime mover of economic growth.
Ironically, despite three decades of massive R&D expenditure, growth levels
have remained substantially lower than that of the immediate post-World
War II period. This raises important theoretical questions regarding R&D
and its impact on growth per se. For example, R&D-growth has been
modeled and continues to be modeled as an unbounded set. This has not
been inconsequential because it has introduced an upward bias in growth
projections as evidenced in the literature. More importantly, are there
physically-determined upper limits to R&D-based growth and, if so, what
are they? This chapter uses the physical sciences to map the physical limits
to R&D-based innovation. A consilient model of economic growth is
presented and upper bounds for energy efficiency-based growth rates are
provided, both for individual energy sectors and globally. We find that with
economic growth by innovation being limited by physical conditions,
increasing the rate of economic growth can only come through increasing
the rate of energy consumption
7.1 Introduction
Virtually ignored in economics for most of the 20th century, research and
development (R&D) literally took off in the 1980s. R&D and innovation
took center stage in the debate over economic growth; theoretically,
empirically and policy-wise. Building on the pioneering work of Denison
(1962) and Griliches (1995) and the writings of Moses Abramovitz1 the
1Moses Abramovitz was a 20th-century American economist and a professor who

wrote books such as Thinking About Growth and Other Essays on Economic Growth
and Welfare, Cambridge University Press, 1989. 406.
106 7. The Physical Limits to Economic Growth by R&D Funded Innovation
growth literature virtually exploded with articles about the myriad aspects
of innovation. Governments heeded the call and re-examined incentives,
industry structure and corporate taxation.
Now, some three decades later, massive spending on R&D (Table 7.1)
growth has not returned GDP2 growth to the levels in the 1950-1970s. Why
have record levels of R&D failed to jump-start growth? This chapter is an
attempt to provide answers to this and other R&D-growth related questions.
Our starting point is the physical R&D-growth boundary, defined as the
upper limits of R&D-based economic growth, which is defined by the
underlying physical laws. By the latter, it should be understood as the
fundamental laws of classical mechanics and thermodynamics as applied to
material processes. Among the problems associated with the dominant R&D
growth model is its unbounded nature. For example, the technology scalar
in the Solow-Swan growth model (Solow 1956, Guerrini 2005) is
theoretically unbounded from above, which is in clear violation of the first
and second laws of thermodynamics and of matter. In other words, R&D
cannot create matter or energy and cannot violate the laws of physics
(Beaudreau 1998, 1999).
Table 7.1: Global R&D Spending in 2011-billion U.S.$
Americas 491.8
US 427.2
Other 64.6
Asia 473.5
Japan 152.1
China 174.9
India 38
Europe 326.7
Rest of world 41.4
Total 1, 333.0
We reexamine the physical R&D-growth boundary using the energy-

organization (EO) approach to modeling material processes. According to
this approach, wealth is defined as an increasing function of broadly-defined
energy and organization. Included in organization are machinery and
equipment, supervision, information, and management. Linking the two is
the concept of second-law efficiency, or simply put, energy efficiency.
Wealth, or output, growth is bounded from above by maximum (i.e. 100%)
2GDP (gross domestic product) is the sum of all of the goods and services produced
and delivered. GDP is a measure of wealth and can be expressed as GDP/capita
when comparing the wealth of populations.
efficiency as work output cannot exceed energy input. This is followed by

estimated sector-specific and aggregate levels of energy efficiency, as well
as projections for future growth over extended time horizons. For purposes
of this chapter we assume maximum energy efficiency is achieved by 2100.
In fact, today, many processes and industries are already very close to
maximum energy efficiency.
We use increases in energy efficiency because it is a well-known and
understood terminology. However, we immediately convert an energy
efficiency increase to a decrease in energy intensity. This is because energy
efficiency has no defined units whereas energy intensity is measured as
energy per unit of output, which can be compared to initial energy intensity.3
For example, in this chapter the maximum reduction in energy intensity is
used to compare world energy intensities in 1990 and 2100.
7.2 The Productivity Slowdown, Policy Responses

and R&D Expenditure
The mid-1970s witnessed a marked decline in GDP gross domestic
product growth rates in Western industrialized economies as well as in the
world economy. For most of the post-World War II (WWII) period up to
the mid-1970s, economic growth was not only high, but robust. G6
countries could expect growth in the 4-6% range. After the mid-1970s,
growth plummeted to just over one percent (Maddison 1987). These lower
growth rates altered the social fabric, a fabric that had been based on rising
per-capita wealth. Governments floundered under the weight of growing
deficits and ballooning debt loads.
The economics profession was at a loss to explain this sharp drop in
growth. In the face of what Moses Abramovitz referred to as “our measure
of ignorance,” the profession embarked on a research program unequaled in
scope and breadth since the Keynesian revolution. The question of growth
soon came to dominate the research agenda as researchers shifted their
3 Adding energy-efficiency increases gives misleading results. For example, suppose
there were two consecutive 50% increases in energy efficiency for the fuel
consumption of an automobile that was initially 10 l/km. The first 50% increase
allows the automobile to travel 150 km on 10 l of fuel and energy intensity would
be (10x(100/150))1/4 6.67 l/km. The second 50% increase in energy efficiency
allows the automobile to travel 150 km on 6.67 l of fuel. The energy intensity is now
(6.67x(100/150)) 1/4 4.44 l/km. The first 50% increase in energy efficiency reduced
fuel consumption by 3.3 l/km and the second by 2.2 l/km. Thus, the second 50%
increase in energy efficiency gives less actual reduction in energy intensity.
attention to try to determine the underlying causes of economic growth

because R&D as a cause of growth was a hypothesis, not a proven theory.
Nevertheless, as government revenues declined and deficits and debt
increased, attention shifted to policies that might restore growth levels to
post-WWII levels. The amount of growth literature was increasing and
according to this literature, roughly half of post-WWII growth came from
technological change. On this basis, speculative at best, most Western
industrialized countries tried to ameliorate the situation through tax
incentives, subsidies, and changes to intellectual property laws.
7.3 Residual Growth and R&D

Capital and labor had been front and center in the science of wealth
formation throughout the 19th century and information, knowledge and
R&D were absent from economic theory (Smith 1776). This changed in the
post-WWII period with the creation of government statistical agencies. If
Keynesian-style governments were to fine-tune the economy, they would
need detailed knowledge and information about GDP, employment and
investment. Typically, in growth accounting exercises today, capital and
labor growth rates are factored out of the corresponding GDP growth rates,
leaving a residual that is attributed to knowledge, information and
ultimately R&D.
There were attempts to explain the difference as the “efficiency
residual,” “technical change,” or most accurately a “measure of our ignorance”,
as much of observed economic growth remained un-explained. Then, the
focus became more formalized with estimates of multifactor productivity
growth, often referred to as TFP (total factor productivity), when Solow and
Swan developed a growth model that was widely accepted. However, it
could only explain half of the observed growth. The other half was termed
a residualthe Solow residual. Finally, the focus switched to trying to
understand the unexplained portion of economic growth.
In the 20th century industry and government-sponsored research on
growth became specialized. Today, the world allocates roughly 1.5 trillion
dollars annually to R&D expenditure. According to UNESCO, there are
roughly 7.2 million scientists and others engaged in R&D activity
(UNESCO 2010). Has growth increased as a result? Is there evidence of an
R&D-growth nexus? We now examine the empirical evidence.
7.4 The Evidence

As evidenced by the empirical literature, researchers focused on the
relationship between measures of research activity and either TFP or long
run patterns of economic growth (Ladu and Meleddu 2014). TFP is the
growth of output that cannot be explained by the growth of traditional inputs
such as labor and capital and is termed a “residual,” and often referred to as
the Solow residual. Typically, research activity is measured by the level of
technical knowledge as by Equation 7.1 where qt is the measure of new
knowledge in period t. The stock of knowledge, Tt, is then assumed to
evolve according to:
ܶ‫ݐ‬+1 = (1 െ ߜܶ )ܶ‫ ݐ‬+ ‫ݐݍ‬ 7.1
where GT is the rate at which ideas “depreciate.”

Output as TFP is modeled in terms of two functions. First, yt, aggregate
output, is a function of Tt, in addition to aggregate capital, kt, and labor, nt,
in Equation 7.2:
ߙ (1െߙ݇ )
‫ݐ݊ ݇ ݐ݇ ݐߠܶ ݐݖ = ݐݕ‬ 7.2
where Tt is a factor of TFP, and zt is a productivity residual. Second, the

supply of new knowledge is modeled in terms of an idea “production
function,” Equation 7.3:
߮ ట
‫ݐݔ ݐܶ ݐݏ = ݐݍ‬ 7.3
where xt is R&D expenditure, and st is a residual analogous to zt of Equation

7.2. Empirically, st captures the fact that measures of R&D activity display
far less variation than measures of output. Output, or GDP, grows on
average by a factor Ȗs.
Estimates of the parameters of Equations 7.2 and 7.3 have yielded
disappointing results. Using patent applications as an indicator of
knowledge Tt (Porter and Stern 2000, Abdih and Joutz 2006) the
contribution of ideas to TFP, was found to be small. For example, estimates
of ș in Equation 7.2 lie in the range 0.05-0.2, and often lack statistical
significance. Estimates of the ideas production function, Equation 7.3
generally detect a downward trend in st, a decline that is robust to a diversity
of approaches. For example, a selection of citation-weighted patent grants
(Caballero and Jaffe 1993), aggregate patent application data (UNESCO
2010) including international patenting (Ladu and Meleddu 2014) and
simultaneously estimating the aggregate production of ideas all give very

similar results. This downward trend lacks a theoretical basis. For example,
in the case where “easy” ideas are discovered first, so that ideas become
progressively harder to uncover, the value of j should be negative. It should
depend on the number of ideas that have already been uncovered, not on the
date.
A value of ȥ > 0 implies that past ideas are useful for the production of
new ones, whereas a value of ȥ < 0 suggests that research uncovers the ideas
that are easiest to find. These effects are known respectively as “standing
on shoulders” and “fishing-out.” Energy efficiency is a good example of the
latter because as it approaches its theoretical limit, gains are increasingly
difficult to achieve.
Endogenous growth models of this class are typically constructed so that
ȥ=1 in which increases in capital variety drive growth (Romer 1990), and
ideas lead to growth through investment-specific technical change (Krussel
1998). However, if ȥ=1, then the growth rate depends on the population size
or, more generally, on the size of the economy (Jones 1995). Recent
evidence suggests the rate is, in fact, close to constant or even increasing
over the post-War period d in other words ȥ=1 (UNESCO 2011, Ladu and
Meleddu 2014).
To summarize, the evidence is unequivocal: any evidence of a
relationship between the output of ideas from R&D and economic growth
is either weak or non-existent. Clearly, there is a need for a theory that can
discriminate between productivity-increasing and non-productivity-
increasing ideas and innovations. The debate must move away from the
current practice of “residual-based inferences,” i.e., away from the ztTș in
Equation 7.2, and to testable predictions. In the next section, we present just
such a theory.
7.5 Wealth: The Underlying Physics

This section presents an alternative approach to understanding wealth as
measured by GDP, one that imposes more structure and limits on the
relationship between inputs and outputs. In doing so, we can shed new light
on the role of innovation and R&D on material processes and hence
aggregate economic growth.
We use the EO (energy-organization) approach to modeling material
processes (Beaudreau 1998). It models wealth in terms of two universal
inputs, namely broadly-defined energy and broadly-defined organization.
Both are necessary conditions in all material processes whether it be in
biology, chemistry, engineering or economics. The model is formalized in
terms of Equation 7.4 where Wt, Et, Lt, and St refer to wealth, energy, tools
and supervision at time t, respectively and K refers to second-law efficiency,
which, as shown, is a function of Lt and St.
ܹ௧ = K(‫ܮ‬௧ , ܵ௧ ) ‫ܧ‬௧ 7.4
K(‫ܮ‬௧ , ܵ௧ ) corresponds to defined organization input, while Et corresponds to

energy input.
While Et is sometimes referred to as energy consumption per se,
technically it refers to available work or negentropy. As energy cannot be
created or destroyed, it follows that energy is not consumed, but rather
overall entropy is increased. Second-law efficiency (i.e. K) is assumed to be
increasing because of the quality of tools and supervision. For the sake of
discussion, it will be assumed that the latter are qualitative and not
quantitative variables. A good example of an improved tool is James Watt's
external condenser that increased steam engine efficiency, i.e.K, by 200%.
All material processes are powered by energy. Energy and energy alone
is productive in the physical sense. Tools and supervision are necessary
inputs, but not productive in the physical sense, but only affect output via
their effect on second-law efficiency. Lastly, this model and its predictions
are consistent with the laws of classical mechanics and thermodynamics,
making for a truly consilient model of material processes. Importantly, it
predicts that increasing energy use will cause increased growth (Beaudreau
1998).
ௗௐ ଵ ௗK ଵ ௗா ଵ
= + 7.5
ௗ௧ ௐ ௗ௧ K ௗ௧ ா
Unlike conventional models where the effects of R&D on growth are

unbounded, the EO model imposes upper bounds. For example, in
conventional models, technological change could theoretically double,
triple or quadruple output as there are no stated limits. Such is not the case
here. According to Equation 7.5, growth in the EO model is defined as the
ௗா ଵ
sum of energy consumption growth (i.e. ) and second-law efficiency
ௗ௧ ா
ௗK ଵ
growth (i.e. ). Because R&D affects the latter which is bounded, then
ௗ௧ K
so too is R&D-based growth bounded. Examples of K growth include such
as more efficient electrical motors, more efficient production processes, and
more efficient lighting. In short, anything that increases the amount of work
from a given supply of energy.
Any innovation that raises company, industry, or aggregate efficiency

will increase output. Because K is bounded from above and is extremely
stable over time, process-based R&D will have a limited effect on the
growth of material wealth. Once energy efficiency, K, has reached its
maximum, R&D can no longer increase output. Put differently, if qt in
Equation 7.1 or ztTș in Equation 7.2 has no effect on K, then it cannot
increase output and wealth. Then, Energy consumption growth is the key
source of growth, a result that is universal in its scope.
7.6. Estimates of Potential Increases in Energy Efficiency

Currently, there is much interest in improving energy efficiency in
applications such as industries (Morefeldt and Silveira 2014), specific
products (Armstrong et al. 2014, Burman et al. 2014) and energy security.
We examine the aggregate energy efficiency increase from of all sources of
energy efficiency increase (Giacone and Manco 2012). We present
estimates of h(t) levels and potential rates of growth by sector and for the
world economy as a whole. These are based on industry and sector
engineering data, and represent “best estimates” of the potential gains in
energy efficiency (Johansson 2013). As such, they define the physical limits
of growth and, hence, are important in the debate over R&D and growth.
Specifically, they define the upper limits of productivity growth in a world
in which the only physically productive factor input is energy.
Table 7.2: World Energy Consumption by End-Use Sector, 1990
Energy consumption sector Primary EJ/yr Percentage

Electricity generation 137 37.5
Transportation 68 18.6
Residential 44 12.1
Industrial 80 21.9
Commercial 36 9.9
Total World energy 365 100
In keeping with the energy-organization approach, Table 7.2 presents

world energy consumption in 1990 by end-use sector measured in exajoules
(1EJ=1018J). By combining these sectors with estimates of K (t), we can
approximate the level of world GDP in 1990 via Equation 7.4. Using
industry and sector estimates of possible potential increases in energy
efficiency over the course of the next century until 2100, we calculate the
corresponding decreases in energy intensity efficiency. These are then
aggregated at the sector level, and finally, at the world level.
7.6.1. Electricity Generation

World electricity demand is continuing to grow because electricity can
often do more in less space. Electricity generation is powered by the primary
energies which are mainly coal, petroleum, natural gas, uranium, wind and
solar. All require extraction processes and sometimes minimal conditioning
before being consumed to generate electricity. Each of the four non-energy-
producing sub-sectors in Table 7.3 mainly uses secondary energy such as
electricity and refined petroleum derivatives, such as gasoline.
7.6.2 Fossil Fuel Energy

Average world energy efficiency for fossil fuel power plants is 34% with
some up to 40% for natural gas (International Energy Agency 2008).
Supercritical designs involving new materials would allow higher steam
temperatures and pressures and raise energy efficiencies up to about 50%
(U.S. Environmental Protection Agency 2010). We use this as our
achievable level of energy efficiency, which corresponds to a 47% increase
in efficiency over current levels for coal. The use of oil for generating
electricity is small and also has a 47% increase in energy efficiency.
CCGTs (combined cycle gas turbines), where very hot combustion gases
expand through a gas turbine and the waste heat in the exhaust gases
generates steam for a steam turbine, could eventually reach efficiencies of
50%-68% (Higher Heating Value basis) (Nakicenovic et al. 2000). For
natural gas, we estimate the corresponding increases in average energy
efficiency for CCGTs at 100%, or thermal energy efficiency of 68%.
Gasification of coal and biomass make them good fuels for a CCGT.
However, syngas has no energy efficiency advantage over natural gas
because of the energy required in the gasification process. For example,
with a gasification efficiency of 70% (U.S. Environmental Protection
Agency 2008) and CCGT efficiency of 68%, the overall efficiency to
electricity would be 47.6%, or below the lower end of the range of 50%-
68% for natural gas CCGT power plants. There is a distinct advantage in
gasifying solid biomass, such as wood, because the overall efficiency of the
combined cycle system is almost double the efficiency of burning solid
biomass in a conventional power plant (U.S. Department of Energy 2000).
7.6.3 Nuclear Energy

The EIA (Energy Information Administration) scale for measuring
nuclear primary energy for generating electricity uses the heat content of
the actual amount of uranium consumed. Current energy efficiency of

generating nuclear electricity ranges from 31% to 38% (World Nuclear
Association 2014). There is potential for improved energy efficiency of
nuclear electricity generation, possibly into the range of 48%-59% for high
temperature molten salt reactors with helium for heat transfer (International
Energy Agency 2008). In our analysis, the corresponding nuclear energy
efficiency increase is set at 50%, for an average thermal energy efficiency
range of 46%-57%. Nuclear fusion, if and when it becomes viable, is likely
to have about the same energy efficiency as nuclear fission when used as a
source of heat for steam/electric power plants.
7.6.4 Renewable Energies including Solid Biomass Fuels

The EIA scale for measuring primary energy specifies that primary
energy used for electricity generated from renewable energies is defined
using the 1990 heat rate of 10, 400 btu/kWh, or 3.05 times gross electricity
output, which is the fossil fuel equivalent (Lightfoot 2007). This ratio
stabilizes or fixes the efficiency of generating electricity from hydro and
other renewable sources, such as wind and solar. Stabilizing these
conversion rates implies that energy efficiency does not change. Thus, there
is no change in energy efficiency, or in energy intensity, for these energy
sources in Table 563, hence the zeroes on lines 5 and 6 in columns C and D.
Even if the efficiency were free to change, there would be little
difference in the result. For example, there is little possibility of significant
increases in hydro and wind because efficiency is already 85% for hydro
plants, and 76% of the theoretical Betz limit of 59% for a 45% efficient wind
turbine. The amount of sunlight falling on a given area of land is fixed and,
therefore, solar efficiency relates to land area. For example, solar land
efficiency is 7.5% using 15% efficient photovoltaic cells when the average
land to collector ratio is approximately 2:1 (National Renewable Energy
Laboratory 2011) for non-tracking arrays. Tracking collectors capture more
of the sunlight during a day, but the extra land required to prevent one
collector from shading another can reduce the solar land efficiency by
approximately 3:1. Potential photovoltaic conversion efficiency of silicon
cells to electricity might be raised from the current range of 11%-15% up
to, possibly, close to a maximum of 30% in normal sunlight (Green et al.
2007).
7.4.5 Proportion of Fuels for Electricity Production in 2100

Renewable energy sources are currently approximately 14% of world
energy supply. Physical limitations of these energies are not likely to allow
a significant increase, if any, in this share. Therefore, in Table 7.3, the sum
of hydro and other renewable sources is 14%. The actual percentage of
renewables is of little significance because the EIA arbitrarily defines their
efficiency (U.S. Energy Information Administration 2013). The best coal is
currently being mined and only lower grades will remain in 2100. Oil and
natural gas might well be scarce by 2100 and/or too valuable for providing
essential petrochemicals to burn them to generate electricity because viable
alternatives are available. Thus, in 2100 only 16% of electricity production
is by coal and none by oil or natural gas. The bulk of electricity generation
is by nuclear fission energy, i.e. 70% (Lightfoot 2006).
Table 7.3: World Electricity Generation: Primary Energy Intensity

Relative to 1990a
A B C D E F
Energy Energy Potential Potential Energy Energy
source consumed increase decrease consumed intensity
in 1990 in 1990 in energy in energy in 2100 decrease
EJ/yr (%) efficiency intensity (%) 1990 to
to 2100 1990- 2100 D x
(%) 2100 (%) E (%)
Oil 14.9 10.9 47 34.2 0.0 0.0
Natural gas 21.1 15.4 100 50.0 0.0 0.0
Coal 51.8 37.9 47 34.2 16.0 5.5
Nuclear 21.5 15.7 59 37.1 70.0 26.0
Hydro 23.8 17.4 0 0.0 9.0 0.0
Other renew 3.5 2.6 0 0.0 5.0 0.0
Total 136.6 100.0 - - 100.0 31.4
Energy 31.4
intensity
decline from
1990 to 2100
Residual 68.8
energy intensity
in 2100
Corresponding 45.0
overall increase
in energy
efficiency
aThe following is a short description of the construction of this table, which is similar
to that of Tables 6.4-6.7. Columns A and B are 1990 numbers for each sector.
Column C is the potential increase in energy efficiency as previously described in
the text. Column D is the decrease in energy intensity that corresponds to the
increase in energy efficiency. Column E is the percentage of each energy source in
the mix in 2100. Multiplying the values in Column D with those in Column E results
in Column F, which is the decrease in energy intensity for each source in the mix.
The individual decreases in energy intensity are added to obtain the total of 31.4%.
The residual energy intensity is 68.6% of that in 1990 and the equivalent average
increase in energy efficiency is 45.9%
Table 7.4: World Transportation: Primary Energy Intensity Relative

to 1990
A B C D E
U.S. trans. Potential Potential World trans. Energy
Energy increase in decrease in energy intensity
consumed in energy energy consumed in decrease
2012 (%) efficiency to intensity 2100 (%) 1990-
2100 (%) 1990-2100 2100C x D
(%) (%)
Cars and light 58.6 270 73.0 58.6 42.8
trucks
Trucks, ships, 29.6 50 33.3 29.6 9.9
trains
Air 8.0 100 50.0 8.0 4.0
Other 3.8 50 33.3 3.8 1.3
Totals 100.0 - - 100 57.9
Energy 57.9
intensity
decline from
1990 to 2100
Residual 42.1
energy
intensity in
2100
Corresponding 137.8
overall increase
in energy
efficiency
7.4.6 Co-Generation
Co-generation is defined as the recovery of thermal energy that is
normally lost or wasted. It is often difficult to obtain large and/or consistent
benefits from co-generation because normally waste heat cannot be stored
until needed. Thus, it is necessary to balance the amount and timing of the
loads between electricity generation and heat utilization.
Currently, approximately 8% of world electricity capacity is capable of

using cogeneration (World Watch Institute 2010). Cogeneration is used in
fossil fuel electricity generating stations where waste heat preheats
combustion air and boiler feed water. District heating of buildings using low
temperature heat is a form of cogeneration that is applicable in areas of
relatively dense population. Co-generation is likely to remain a small
contributor to improved energy efficiency and is not a significant
consideration in this analysis.
Column C in Table 7.3 sets the upper bound for the projected gain in
energy efficiency in the world electricity generation sector over the next
century at 45.9%.
7.4.7 Transportation
Transportation in Table 7.2 is estimated to consume 18.6% of world
energy in 1990. Currently, 95% of U.S. transportation energy is supplied by
oil. The figures in Table 7.4 are for the U.S. in 2012 (Center for Climate and
Energy Solutions 2012, U.S. Department of Energy 2014) because
comparable figures for the world are not available. As such, they overstate
the contribution of cars and light trucks. However, it could be argued that
the rest of the world will tend towards a similar breakdown over the course
of this century.
7.4.7.1 Cars and Light Trucks
As a starting point to estimate an overall increase in energy efficiency

of cars and light trucks we use the gasoline consumption for U.S. light-duty
vehicles of wheelbase less than 3.07 m which was 9.9 L/100 km in 2012
(Metz et al. 2001). Increased energy efficiency comes from a change to the
concept, such as hybrid vehicles, which have gains in energy efficiency
ranging from very little to 140%, depending on the driving conditions (Metz
et al, 2001). Additional efficiency increases of up to 156% may be possible
by improvements to current four-stroke engines, continuously variable
transmissions, lightweight materials, reduced rolling resistance and
improved aerodynamics (Transportation Research Board and National
Research Council 2002). Together these efficiency increases reduce the
overall fuel rate to 3.6 L/100 km. For purposes of this analysis, and to err
on the high side, the final average fuel rate for cars and light trucks will be
2.35 L/100 km, which is a decrease in energy intensity of 73%, and

equivalent to an increase in energy efficiency of 270%.4
7.4.7.2 Trucks, Trains and Ships
Highway weight and size restrictions limit the size and weight of heavy
trucks, and relaxation of these restrictions is not likely. Improvements in
energy efficiency must come, therefore, from propulsion systems and
reductions in air and rolling resistance. For purposes of our analysis, the
maximum increase in energy efficiency for large trucks was estimated at
50%, which raises current efficiency from 45% to 67%. Trains have the
advantage of the low rolling resistance of steel wheels on steel rails. The
maximum increase in efficiency here is set at 50% because all of the
components of the propulsion system are well developed. Doubling the size
of a ship increases the carrying capacity by eight times, but the power
required increases by only four times. Thus, there is an energy advantage
for large ships as evidenced by the emergence of large supertankers, which
require special port facilities. The size of port and canal facilities, as it turns
out, limits the size of most ships. It is unlikely that this will change
significantly and, therefore, the average size of ships cannot increase
significantly. The efficiency of large ship propulsion systems from fuel to
propeller is currently at 42% (U.S. Department of Energy 2001). Large
increases in efficiency do not appear likely from improved hull shapes and
anti-fouling methods. Therefore, the maximum increase in energy
efficiency is set at 50% for purposes of this analysis.
7.4.7.3 Aircraft
Similarly, large aircraft are more efficient than smaller ones, which
explains the trend towards larger aircraft. However, limiting the size of
aircraft are route traffic patterns, runway construction and airport facilities.
With these constraints, it is unlikely that the average size of aircraft will
4 At this point, it is necessary to discuss hydrogen gas as an automobile fuel. The

most efficient and widely used process for manufacturing hydrogen is steam
reforming of natural gas, which is approximately 70% efficient. PEM (proton
exchange membrane) fuel cells have efficiencies in the range of 40%-60% Thus;
overall efficiency of hydrogen to electricity is in the range of 28%-42%. The
efficiency of automobile gasoline engines is in the lower half of this range and diesel
engines are in the upper half. Thus, the use of hydrogen made from fossil fuels
provides little, if any, gain in energy efficiency, and a substantial loss of efficiency
when made by electrolysis.
increase substantially. Most increases in energy efficiency will continue to

come from lightweight materials and engine efficiency (U.S. Department of
Energy 2001). The projected energy efficiency increase is 100%. The
“Other” category appears to be about half heavy fuel oils for marine use and
half unidentified uses. Because of the associated uncertainty, the energy
efficiency increase is set at 50%.
To summarize, according to our estimates, overall energy intensity in
world transportation can, at the most, fall by 57.9% resulting in an energy
intensity of 42.1%. This is equivalent to an increase in energy efficiency of
137.8% over that in 1990.
Table 7.5: World Primary Energy Intensity Relative to 1990
A B C D E
U.S. Potential Potential World Energy
residential increase in decrease in residential intensity
energy energy energy energy decrease
consumed in efficiency to intensity consumed in 1990-
2010 (%) 2100 (%) 1990-2100 2100 (%) 2100C x D
(%) (%)
Space heating 44.7 300 75 45 33.8
Water heating 16.4 50 33 17 5.7
Space cooling 9.2 200 12 8.0
Lighting 5.9 400 80 4 3.2
Refrigeration 3.9 300 75 4 3.0
Electronics 4.7 300 75 5 3.8
Wet cleaning, 3.3 100 50 3 1.5
drying
Cooking 3.7 300 75 4 3.0
Computers 1.5 100 50 2 1.0
Other 6.8 100 50 4 2.0
Totals 100.0 - - 100 64.9
Energy 64.9
intensity
decline from
1990 to 2100
Residual 35.1
energy
intensity in
2100
Corresponding 184.9
overall
increase in
energy
efficiency
7.6.5 Residential
Residential energy consumption in the U.S. is given in Table 7.5 (U.S.
Department of Energy 2012). The U.S. data is used as a base because it is
readily available and over the next century it is highly probable that world
residential energy consumption patterns will approach U.S. levels. The
values in Column D tend to reflect this.
The projected increases in energy efficiency are listed in Column B and
range from 50 to 300% (Letschert et al. 2013). Only a minor part of the
projected increase in energy efficiency for space heating and cooling derives
from increased efficiency of furnaces, heaters and air conditioners as energy
efficiency of these items is already well developed (Wada et al. 2012). For
example, the efficiency of older heating systems is 56-70% and the current
annual fuel utilization efficiency recommendation is 78-83%. The bulk of
the increase is from better windows, reduced air infiltration, better
insulation, and the replacement of less energy-efficient heating and cooling
systems. Energy efficiency increases of 50% for water heating will come
mainly from better storage tank and piping insulation and, possibly, from
combined space and water heating systems. The energy efficiency of
refrigerators has increased 300% since 1974 and most of the improvement
in energy efficiency is likely to come from replacement of old, less energy
efficient models. Lighting holds the potential for large increases in energy-
efficiency mainly through wider use of more efficient light bulbs.
Overall energy intensity for the residential sector is projected to decrease
by 64.9% to a level of 35.1%, which represents an increase in energy
efficiency of 184.9% from that in 1990.
7.6.6. Industrial
Table 7.6, Column A, is a breakdown of industrial energy consumption
for the U.S. in 2004. Once again, as similar information is not available for
the world, estimates were made of how energy might be used by all world
industry in 2100 (Column D). Based on historical experience and the fact
that the use of robots and automation is expected to increase, the estimates
for percentage of boiler fuel and process heating and cooling are reduced in
2100 and machine drives and facilities are higher.
The energy efficiencies of industrial boilers, process heating and
cooling, and machine drives are generally significantly higher than 50%
today. The projected increase in energy efficiency for these three categories
is 25%. Facilities include buildings, which may have a potential for energy
efficiency increase as high as 200%. “Other” is unknown and has been
assigned 200% energy efficiency increase. For purposes of our analysis,

electro-chemical energy consumption is negligible. There is always a
continual incentive to improve efficiency of electrochemical applications
because electricity is a major part of production cost. Thus, there is little
potential for additional increases in energy efficiency. As such, overall
energy intensity for the world industrial sector is projected to decrease by
36.3% to a level of 63.7% of that in 1990, which is equivalent to an increase
in energy efficiency of 57.1%.
7.7 Commercial
Column A in Table 7.7 gives the breakdown for commercial primary
energy consumption for the U.S. in 2010 (U.S. Department of Energy 2010)
which, given the nature of the technology, appears to be reasonable for the
world. Estimated overall energy intensity for commercial can decrease by
43.7% to a residual of 56.3%, which is equivalent to an increase in energy
efficiency of 77.6% over that of 1990.
Table 7.6: World Industrial: Primary Energy Intensity Relative to 1990
A B C D E
U.S. Potential efficiency to Potential World
industrial increase in 2100 (%) decrease in industrial
energy energy energy energy
consumed in intensity consumed in
2004 (%) 1990-2100 2100 (%)
(%)
Boiler fuel 35 25 20 20 4
Process heat 39 25 20 20 4
and cool
Machine drive 12 25 20 25 5
Facilities 8 200 66.7 15 10
Electro- 2 25 20 0 0
chemical
Other 4 200 66.7 20 13.3
Totals 100 - - 100 36.3
Energy 36.3
intensity
decline from
1990 to 2100
Residual 63.7
energy
intensity in
2100
Corresponding 57.1
overall
increase in
energy
efficiency
Table 7.7: World Commercial: Primary Energy Intensity Relative to

1990
A B C D E
U.S. Potential efficiency to Potential World
commercial increase in 2100 (%) decrease in commercial
energy energy energy energy
consumed intensity consumed in
in 2010 (%) 1990-2100 2100 (%)
(%)
Lighting 13.6 100 50 15 7.5
Space heating 26.6 100 50 25 12.5
Space cooling 10.1 50 33.3 10 3.3
Ventilation 6.1 50 33.3 6 2.0
Refrigeration 4.5 200 66.7 5 3.3
Water heating 6.7 50 33.3 7 2.3
Electronics/co 5.4 200 66.7 6 4.0
mputers
Cooking 2.3 50 33.3 5 1.7
Other 24.6 50 33.3 21 0.7.0
Totals 100 - - 100 43.7
Energy 43.7
intensity
decline from
1990 to 2100
Residual 56.3
energy
intensity in
2100
Corresponding 77.6
overall
increase in
energy
efficiency
Table 7.8: World primary energy intensity relative to 1990
A B C
Portion of world Decrease in energy Contribution to
energy consumption in intensity by sector to decrease of each
1990 (%) 2100 (%) energy sector to world
energy intensity A x B
(%)
Electricity 37.5 31.4 11.8
generation
Transportation 18.6 57.9 10.8
Residential 12.1 64.9 7.9
Commercial 21.9 36.3 7.9
Industrial 9.9 43.7 4.3
Totals 100 - 42.7
World primary 57.3
energy intensity
as a percent of
1990
Equivalent 74.5
increase in world
energy efficiency
7.6. Results: The Upper Bound of R&D Energy Efficiency

The purpose of this chapter is to define and bound the physical R&D-
energy efficiency-growth to 2100. Aggregating across all five energy
sectors we construct a composite index by weighting the reported sectorial
decreases in energy intensity (Column B) by their portion of world primary
energy consumption in 1990 (Column A), yielding the values reported in
Column C of Table 8, the sum of which is 42.7%. Thus, world energy
intensity in 2100 would be 57.3% of that in 1990, which corresponds to
74.5% rise in overall energy efficiency.
R&D-based growth will, at best, increase output over the course of the
next 110 years at the equivalent to an energy efficiency increase of 74.8%.
After that, and even much before in some sectors, growth comes from
increased energy consumption resulting from additional facilities for
producing goods and services.
In terms of confidence levels, these results of energy intensity decline
are robust. In other words, relatively large changes in the input make small
changes to the output. For example, if the share of electricity generation
were to grow from 37.5% to 50%, i.e., by 33.3%, then world energy
intensity would increase from 57.3% to 59.5% of that in 1990, a change of
3.8%. Suppose transportation grew from 18.7% to 25% of world energy
consumption, an increase of 33.7%, then world energy intensity would

decrease from 57.3% to 56.0% of that in 1990, a decrease 2.2%.
The same is true of the individual components of each energy sector. For
example, if increases in residential space heating and cooling were 150 and
100% respectively, rather than 300 and 200%, decreases of 50%, then world
energy intensity would increase from 57.3% to 58.3% of that in 1990, and
increase of 1.7%.
Thus, it is evident there is very low probability of world minimum
energy intensity being significantly different from the estimated limit of
57.3% of that in 1990.
7.7 Conclusions
In this chapter, we examined the question of the physical R&D-Energy
Efficiency-Growth upper boundary. Until now, there have been no physical
constraints imposed on economic growth based on R&D as R&D-based
growth was, for the most part, assumed to be unbounded. Using basic
physics, we examined the physical limits to non-energy consumption-based
growth, i.e., the physical limits to energy efficiency. Our conclusions are:
1. The output of current R&D models of economic growth is not

unbounded, but is bounded by the underlying physics of material
processes.
2. Material processes, of which production processes are a subset, are
powered by energy and defined by organization. Better organization
increases energy efficiency, defined as the work accomplished by a
given, fixed amount of energy.
3. Energy efficiency varies across material processes within an industry
and across industries according to the laws of thermodynamics.
4. Energy efficiency imposes an upper-bound on the gains from
process-based innovation and hence from R&D expenditures.
5. Energy efficiency is the only scientifically-valid definition of TFP
(total factor productivity).
6. In most processes and sub-processes, energy efficiency is near or at
its theoretical limit, thus imposing a de facto limit on future R&D
expenditures.
7. When increasing energy efficiency is no longer productive,
economic growth comes from increased energy consumption
through additional facilities for producing goods and services.
8. Directing R&D expenditures towards those projects that are farthest

from the upper bound of energy efficiency have the most potential
for success.
9. Minimum average world energy intensity projected for 2100 is
approximately 57.3% of that in 1990, which is equivalent to an
energy efficiency increase of 74.5%.
These findings are consistent with the laws of physics that underlie the
material processes that constitute the world economy. They clearly indicate
that R&D alone cannot restore the high levels of growth experienced in the
post-WWII era up to the mid-1970s.
7.8 References
Abdih Y, F. Joutz. 2006. Relating the Knowledge Production Function to
Total Factor Productivity. International Monetary Fund Staff Papers.
Armstrong P, D. Agir, I. Thompson, and M. McCulloch. 2014. “Improving
the Energy Storage Capability of Hot Water Tanks Through Wall
Material Specification.” Energy. 78:128-40.
Ayres R, H. H. Turton H, and T. Casten. 2007. “Energy Efficiency,
Sustainability and Economic Growth.” Energy 32:634-48.
Beaudreau B.C. 1998. Energy and Organization: Growth and Distribution
Reexamined. Westport, CT: Greenwood Press; 1998.
Beaudreau B.C. 1999. Energy and the Rise and Fall of Political Economy.
Westport, CT: Greenwood Press; 1999.
Burman E, D. Mumovic, and J. Kimpian. 2014. Towards Measurement and
Verification of Energy Performance under the Framework of the
European Directive for Energy Performance of Buildings.” Energy.
77:153-63.
Caballero R. and A. Jaffe. 1993. “How High are the Giant's Shoulders: An
empirical Assessment of Knowledge Spillovers and Creative Destruction
in a Model of Economic Growth.” NBER Macroeconomics Annual
Center for Climate and Energy Solutions. 2012. Annual Report.
Chang M.C. 2014. “Energy Intensity, Target Level of Energy Intensity, and
Room for Improvement in Energy Intensity: An Application to the Study
of Regions in the EU.” Energy. 67:648-55.
Choi J.K, D. Morrison, K. Hallinan, and R. Brecha. 2014. “Economic and
Environmental Impacts of Community-Based Residential Building
Energy Efficiency Investment.” Energy.78:877-86.
Denison E. 1962. The Sources of Economic Growth in the United States and
the Alternatives Before Us. Committee for Economic Development.
Washington, DC.
Giacone E. and S. Manco. 2012. “Energy Efficiency Measurement in
Industrial Processes.” Energy. 38:331-45.
Green, M.A., K. Emery, Y. Hisikawa and W. Warta. 2007. Solar Cell
Efficiency Tables (Version 30) Progress in Photovoltaics. 15:425-30.
Griliches Z. 1995. “The Discovery of the Residual: An Historical Note.”
National Bureau of Economic Research Working Paper 5348.
Guerrini L. 2005. The Solow-Swan with Abounded Population Growth
Rate. Journal of Mathematical Economics. 42, no. 1:14-21.
Hall C and K. Klitgaard. 2011. Energy and the Wealth of Nations:
Understanding the Biophysical Economy. New York, NY: Springer
Press; 2011.
International Energy Agency. 2008. Report OECD/IEA.
Johansson B. 2013. “A Broadened Typology on Energy and Security.”
Energy. 53: 199-205.
Jones C. 1995 “R&D-Based Models of Economic Growth.” Journal of
Political Economy.103: 759-84.
Krusell P. 1998 “Investment-Specific R&D and the Decline in the Relative
Price of Capital.” Journal of Economic Growth. 3:131-41.
Ladu M, and M. Meleddu. 2014. “Is There Any Relationship Between
Energy and TFP (Total Factor Productivity)? A Panel Cointegration
Approach for Italian Regions.” Energy 75:560-7.
Letschert V. L.B. Derroches KeJ, M. McNeil. 2013. “Energy Efficiency:
How Far Can We Raise the Bar? Revealing the Potential of Best
Available Technologies.” Energy. 59:72-82.
Lightfoot H.D. 2006. “A Strategy for Adequate Future World Energy
Supply and Carbon Emission Control.” in: Climate Change Technology
Conference: Engineering Challenges and Solutions in the 21st Century.
Engineering Institute of Canada, Ottawa, Canada.
Lightfoot H.D. 2007. “Understand the Three Scales for Measuring Primary
Energy and Avoid Errors.” Energy. 32:1478-83.
Lightfoot H.D. W. Manheimer D.A. Meneley, D. Pendergast and G.
Stanford. 2006. “Nuclear Fission Fuel is Inexhaustible.” in: Climate
Change Technology Conference: Engineering Challenges and Solutions
in the 21st Century. Engineering Institute of Canada, Ottawa, Canada.
Maddison A. 1987. “Growth and Slowdown in Advanced Capitalist
Economic Literature. 25, no.2:649-98.
Metz B. O. Davidson E. Swart and J. Pan. 2001. Climate Change 2001:

Mitigation. Contribution of Working Group III to the Third Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge
UK: Cambridge University Press.
Morefeldt J. and S. Silveira. 2014. “Capturing Energy Efficiency in
European Iron and Steel Production-Comparing Specific Energy
Consumption and Malmquist productivity Index.” Energy Efficiency.
7:955-72.
Nakicenovic Nebojsa, Joseph Alcamo Joseph, Gerald Davis, Bert de Vries,
Joergen Fenhann, Stuart Gaffin et al. 2000. Special Report on Emissions
Scenarios, a Special Report of Working Group III of the Intergovernmental
Panel on Climate Change. Cambridge: Cambridge University Press.
National Renewable Energy Laboratory. 2011. “Land use requirements for
solar power plants in the United States.”
Porter ME, S. Stern S. 2000. “Measuring the Ideas Production Function:
Evidence from International Patent Output.” NBER Working Paper
w7891.
Romer P.M. 1990. “Endogenous Technological Change.” Journal of
Political Economy. 98: 71-102.
Smith A. 1776(1990) An Inquiry into the Nature and Causes of the Wealth
of Nations. Chicago, IL: Encyclopaedia Britannica.
Solow R.M. 1956. A Contribution to the Theory of Economic Growth.
Quarterly Journal of Economics 70, no.1.
Transportation Research Board and National Research Council. 2002.
Effectiveness and Impact of Corporate Average Fuel Economy (CAFE)
Standards. Washington, DC: The National Academies Press.
U.S. Department of Energy. 2000. Project Update: the Vermont Gasifier.
June. Report DOE/GO¬102000-1051.
U.S. Department of Energy. 2001. International Energy Outlook.
DOE/EIA-0484. Washington, DC.
U.S. Department of Energy. 2012. Building Energy Data Book.
Washington, DC.
U.S. Department of Energy. 2014. Transportation Energy Data Book.
Vehicle Technologies Office, Office of Energy Efficiency and
Renewable Energy. Washington, D.C.
U.S. Department of Transportation, 2013. Bureau of Transportation
Statistics. Washington, D.C.
U.S. Energy Information Administration. 2013. Independent Statistics and
Analysis. International Energy Outlook. Washington, DC.
U.S. Energy Information Administration. 2014. Monthly Energy Review.
U.S. Environmental Protection Agency (EPA). 2010. Office of Air and

Radiation, Available and Emerging Technologies for Reducing
Greenhouse Gas Emissions from Coal-Fired Electric Generating Units.
U.S. Environmental Protection Agency. 2008. Combined Heat and Power
Partnership, Biomass CHP Catalog, Chapter 5. Biomass Conversion
Technologies. Table 5-16. Biomass Gasification Cost and Performance.
Washington, DC.
UNESCO. 2010. Science Report. The Current Status of Science Around the
World, Paris.
Wada K. K. Akimoto F.Sano, J. Oda, and T. Homma. 2012. Energy
Efficiency Opportunities in the Residential Sector and Their Feasibility.”
Energy. 48:5-10.
World Nuclear Association, 2014. Advanced Nuclear Power Reactors.
World Watch Institute. 2010. One Twelfth of Global Electricity Comes from
Combined Heat and Power Systems. Product number: VST121.
Xie B.C, L.F. Shang, S.B. Yang, and B.W. Yi. 2014. “Dynamic
Environmental Efficiency Evaluation of Electric Power Industries:
Evidence from OECD and BRIC (Brazil, Russia, India and China)
Countries.” Energy. 74:147-57.
8
THE ECONOMIES OF SPEED, KE = 1/2MV2

AND THE PRODUCTIVITY SLOWDOWN
Abstract
Drawing on basic physics, Kummel (1982) and Beaudreau (1995, 1998)

attributed the productivity slowdown to the OPEC price-hike-led decrease
in the rate of growth of energy consumption in the mid-1970s. The high
post-WWII energy use growth rates observed in most OECD countries fell
drastically, decreasing productivity and GDP growth. However since,
considerable doubt has been cast on this view. For example, why did the rate
of growth of energy use in manufacturing, specifically electricity use, fall
when and where the price of electricity was either unaffected or increased
only slightly afterwards? Second, why did it fall instantaneouslythat is,
without the usual lag? Third, why did energy consumption growth rates not
return to their pre-1973 level once real energy prices had returned to their
pre-1973 levels? Drawing from kinetics, this chapter presents an alternative
hypothesis, namely that energy demand-related factors, notably the physical
limits to energy-based speed-ups, not energy supply-related factors, may
have been behind this sudden decrease in productivity growth and hence
behind the productivity slowdown. Specifically, in many industries and
sectors, maximum machine speed/velocity may have beenor was near to
beingreached in the late 1960s/early 1970s, making further increases
physically impossible or not economically viable.
8.1. Introduction
Energy-based theories/explications of the productivity slowdown
attribute the marked decrease in output growth in the 1970s to the OPEC
crisis and the ensuing energy price increases (Baily 1982, Jorgenson 1981,
1983). There are however a number of problems with these theories. First,
there is the question of timing. Oil prices increased in 1973, but electricity
prices remained stable well into the late 1970s. So, why did electricity use
growth rates fall in 1973? Second, there is the question of expectations.
Were higher energy prices perceived of by firms as temporary or permanent? If
130 8. The Economies of Speed, KE=1/2mv2 and the Productivity Slowdown
they were perceived of as temporary, then energy consumption growth should

have continued unaffected, for a while that is. Third, if higher real energy
prices were indeed to blame for the slowdown in the rate of growth of
energy use, then why did growth rates not return to their former levels or
higher once the real price of energy returned to its previous level? The data
show that despite lower energy prices, the demand for energy failed to return
to its previous levels in the 1980s and 1990s. Last, there is considerable
evidence that shows that productivity growth had begun to decline in the
late 1960s, well in advance of the OPEC oil embargo (Fisher 1988), which
raises the possibility that the decrease in the rate of growth of energy use
may have been the result of demand factors or what we shall refer to as
technological factors.
According to basic kinetics, energy use increases by the square of
velocity, or conversely, velocity increases at the square root of energy use.
Successive increases in energy use will result in smaller and smaller
increases in machine speed/velocity and ultimately in productivity.1 Second,
machine speed is not, in most cases, unbounded/unlimited. Machine speeds
will typically reach levels beyond which various technology-related
constraints (i.e. heat, vibration) will limit further increases/improvements.
In other words, machine speed is bounded from above. This raises the
possibility that the observed slowdown in the growth of energy use in the
1960s, 1970s, and 1980s may have been the result of purely technical factors
rather than supply factors as has been suggested. This chapter examines the
evidence, focusing particular attention on the role of kinetics in the
manufacturing sector and the implications for the demand for energy, output
and growth, and comes to the conclusion that technical factorsspecifically,
the laws of physicscould quite possibly have been at the root of the
decrease in the rate of growth of energy use and the ensuing decrease in the
rate of productivity growth. By kinetics, it should be understood the actual
speed at which machines operate, and second the control devices that
oversee their utilization over time.
The chapter is organized as follows. To begin with, we examine the role
of speed-ups and control technologies in various sectors of the economy.
We then turn and model the role of speed in production processes. This is
achieved by incorporating the laws of kinetics into a simple model of output
involving only capital/machines. The results of previous and new
econometric tests are presented. We then present our main hypothesis,
namely that the well-documented decrease in the rate of growth of energy
consumption in the manufacturing sector and indeed throughout Western
1This is not to be confused with lower second-law efficiency (Thermodynamics).
Rather, it is a kinetics-based decrease in productivity.
industrialized economies in the 1970s may have been demand, not supply,
based. Put differently, physical constraints on machine speeds may have put
an end to energy deepening, resulting in a lower rate of growth of output.
We end with a discussion of the possible implications of our findings.
8.2. Speed, Energy Deepening and Control Technologies

The integral role of machine speed and the associated electric power use
in industrial productivity has been recognized by engineers, by
entrepreneurs, by academics and by governments alike throughout the
industrial era. Take, for example, William Longston's testimony in 1832 on
working conditions in the U.K. textiles industry before the Committee on
the Factories Bill.
9397. Is the intensity of application and of labour altered, either against or

in favour of the operative and of the children employed in mills and
factories?I was a great number of years out of any factory, but those who
were my acquaintances during my boyhood have often conversed with me,
and they very frequently say that it cannot be less than double in intensity
and exertion of physical application.
9398. State why you believe that the labour of those employed has doubled
since the first introduction and use of cotton machinery, or at least since you
first knew it?The reason why I believe so is from some calculations which
I have been obliged to make, and by my own observation during the time I
was manager of a mill in 1830 and 1831, when I had some of the same
operations under my own observation.
9399. Have you any objection to put in those calculations Certainly not. [The
following document was then put in and read.]
9400. It appears by this document that the work done is very greatly
increased between the years 1810 and 1832; has the machinery been so
altered as to produce that amazing difference, or does it result from
accelerating the speed of the machinery?—It is from accelerating the speed
generally; and another cause is, that more and more exertion is required from
the individual working at the machine; these are the two causes ([33]; 430).
The General Electric Company, as early as 1937, pointed to increased

machine speedcontrolled machine speedas one of the defining features of
modernity and of productivity growth. Under the title of Today is the Day
of Speed, it maintained that:
Our transportation systems, our industrial processes, our factory machineryall

these have felt the magic hand of speedcontrolled speed that has given us
more things to enjoy and more time in which to enjoy them; that has
produced more goods for more people at less cost and that has created a
better standard of living for the average man. These are the benefits of ever-
increasing speed and accurate control.2
Henry Ford, in his description of mass production, pointed to machine

speed as the key element of his new technique:
Mass production is not merely quantity production for this may be had with
none of the requisites of mass production. Nor is it merely machine
production, which also may exist without a resemblance to mass production.
Mass production is the focusing upon a manufacturing project of the
principles of power, accuracy, economy, system, continuity and speed (Ford
1926)
Graham Laing, in a book entitled Towards Technocracy, pointed to

machine speed as a key factor in the unprecedented productivity gains in the
1920s and 1930s.
Industrial processes have been speeded up, new inventions are being added
to manufacturing, new economies of personnel and of management have
been made in industry. The 1929 production can undoubtedly be achieved
with thousands, and probably millions, fewer workers (Laing 1933).
Alfred Chandler, in his history of 20th-century business, echoed this

view, generalizing it to the U.S. economy as a whole.3
In modern mass production, as in modern mass distribution and modern

transportation and communications, economies resulted more from speed
than from size. It was not the size of the manufacturing establishment in
terms of the number of workers and the amount and value of productive
equipment, but the velocity of throughput and the resulting increase in
volume that permitted economies that lowered costs and increased output
per worker and per machine (Chandler 1977, 244).
2 Source: The General Electric Machine Tool Speed Show, http://www.youtube.
com/watch?v=CUYajEF7X-U.
3 Alfred Chandler's economies of speed are consistent with Dale Jorgenson and
Nathan Rosenberg's views of the role of energy in productivity growth in general

and with productivity growth in 20th century U.S. manufacturing (Jorgenson 1981,
1983, Rosenberg 1983). In fact, one could argue that speed provides the missing link
in their work, connecting energy use directly with productivity.
Sidney Sonenblum, in his work on electrification and productivity

growth in manufacturing, also pointed to speed, or the accelerating the rate
of throughput as a key element in productivity growth:
During these years, the focus of managerial attention shifted from enlarging
the scale of operations to increasing operating efficiency by speeding up the
rate of throughput in the plant. High priority was assigned to modifications
of factory design and layout in order to better integrate worker and machine
tasks. Advances in the electrification of machine drive were indispensable
to the realization of these new objectives and may, indeed, have served to
stimulate the new managerial perspectives that emerged (Sonenblum 1990,
291)
Running through each of these accounts of speed and its many

advantages is the notion of control. General Electric referred to “accurate
control,” while Ford referred to “accuracy, system and continuity.” Hence,
speed and control are to be understood as complementary inputs. Theoretically,
control can be defined in terms of four functions (i) constancy of speed (ii)
minimal breakdown-related downtime, (iii) sub-process coordination and
(iv) machine programming. Put differently, the better able is the
firm/engineer at maintaining a constant speed, the greater the output. The
same holds for machine breakdown. As not all machines/sub-processes
operate at the same speed, it is essentialin order to avoid bottlenecksto
coordinate speeds throughout the plant. Lastly, because firms typically
produce many different goods/models of goods with the same machinery, it
stands to reason that more efficient machine programming will reduce
downtime and hence, increase the average operating speed per period of
time.
According to Warren T. Devine, control technologies underwent a series
of innovations in the 20th century that were instrumental in increasing
machine speed. Specifically, hydraulic drive and control mechanisms gave
way to servomechanisms, which in the 1950s, gave way to numerical
control, which reduced machine downtime considerably. Not only would
productivity rise as the result of greater machine speeds (i.e. owing to
greater energy use), the machines themselves would be more fully utilized
in any given time period. He noted:
Numerically-controlled machinery had a number of advantages over

conventional manually controlled machinery. The time required to get a
newly designed part into productionthe machine setup timewas
sometimes as much as 65 to 75 percent less with numerical control (Devine
1983, 50).
In a 1966 report entitled “Technological Trends in Major American

Industries,” the U.S. Department of Labor pointed to control technologies
in the form of the computerization of data processing and increased
mechanization (read: faster speeds) as the leading innovations of the post-
WWII period. Under the heading of “Trend Toward Increased Mechanization,”
it pointed out that:
Improvements in machinery that do not involve drastic departure from

conventional design will continue to be an important factor in raising
productivity in many industries. Faster operation, larger size, automatic
loading and unloading devices and automatic lubrification significantly
reduce the amount of labor required per unit of output. The integration of a
number of separate operations into one large specialized machine which
performs a long cycle of operations with a minimum of intervention by the
machine tender constitutes a more advanced type of mechanization.
Examples of greater mechanization are found in many industries: faster
textile machine speeds with larger packages of stock; continuous steel
casting machines that require one-half the number of steps of traditional
ingot casting; machinery in meatpacking for continuous production of
frankfurters; tape controlled line casting machines in printing; faster, larger
capacity machines in tire and tube manufacture. Other examples are
mechanical “lumberjacks” to cut trees in the lumber industry; larger capacity
stripping equipment in copper mining; greater use of continuous coal mining
machines; and a machine that combines a number of operations in shirt
making.
As fabricating operations become highly mechanized, new ways are
sought to achieve labor savings in moving goods and materials from one
plant operation to the next. Mechanized material handling often is
introduced or improved to utilize more fully the high speed and large
capacity of modernizing fabricating equipment. (U.S. Department of
Commerce 1966, 5)
This literature is characteristic of the 1960s. No such literature exists for

the 1970s where speed-up- and machine kinetics-based control technology-
based productivity gains appear to have run their course.4 This, we maintain,
was only to be expected given the law of kinetics.
4 The nature of control technologies changed in the 1980s and 1990s. Instead of
focusing on minimizing downtime and increasing average machine speed, they
focused on automating supervision, rendering labor expendable. In others words,
they ushered in the era of the intelligent plant/factorythat is, a factory without
workers.
8.3 Model
Surprisingly, little is known of the role of energy in material processes
in general, and in productivity in general. Typically, it has been modeled as
another input in a multi-input Cobb-Douglas, or constant elasticity of
substitution (CES) production function (Beaudreau 1998, 1999, Kummel
1982). Accordingly, firms are free to either increase of decrease the amount
of energy deployed within a given plant (or material process) at will—that
is, without limit. Left unspecified are the underlying mechanics (i.e.
microfoundations) of say a higher energy-capital ratio, other than the
standard reference to the laws of diminishing marginal productivity. This
raises a number of questions regarding energy use in manufacturing
processes. Can it be increased at will, and what the physical laws that govern
its use in? In this section, we attempt to answer these questions by
introducing the laws of physics, specifically, the laws of machine kinetics
into a standard model of production.5
8.3.1. Firm level

We begin by defining the firm as a series of n sub-processes, each of
which is governed by the laws of basic machine kinetics, according to which
output yi(t) is an increasing function of Chandlerian machine speed(s) si(t)
and the machines themselves, denoted by ki(t).6 The individual n machine
speeds, defined as the machine rates of output per unit time, are governed
by two factors, namely translational kinetics (KE = 1/2mv2, where KE =
kinetic energy, m = mass, and v = velocity), and by machine utilization rates
which are a function of a number of factors including machine downtime
and market conditions (i.e. the business cycle).7
Translational kinetics defines machine speed (e.g. revolutions per
minute, feet per minute) for a given mass m and kinetic energy KE.8
5 In so doing, we are attempting to shed light on the black box of not only
neoclassical economics, but of energy-augmented models of production.
6 The quantity theory of money can be seen as a special case, where V, the velocity
of money, is defined as the speed at which money is exchanged. M, the money stock,
can be seen as the capital in this case. Hence, the greater is V, the more productive
is the money stock.
7 This equation is derived from special relativity, when v, the velocity, is much less
than c, the speed of light. The Lorentz factor allows us to derive KE = 1/2mv2 from
E = mc2.
8 Machine speed can be defined either in time or in number of product (i.e. maximum
speed is 10s per product, maximum speed is six products per minute).
Machine downtime in turn can be broken down into (i) maintenance (ii)
idleness due to lack of coordination between sub¬processes (iii) retooling
for a new product and (iv) market conditions. To capture these effects, we
define machine average speed in terms of Equation 8.2 where ߛ௜ (‫ )ݐ‬is a
downtime scaler. If machines are never idled owing to any one or a
combination of factors (e.g. maintenance, market conditions), then ߛ௜ (‫ )ݐ‬is
zero. However, the more a machine is shut down as opposed to slowed down
the greater will ߛ௜ (‫ )ݐ‬be [0< ߛ௜ (‫ < )ݐ‬0.5].9 Equation 8.3 defines maximum
sub-process i machine speed.
‫ݏ[݊݅݉ = )ݐ(ݕ‬ଵ (‫݇)ݐ‬ଵ , ‫ݏ‬ଶ (‫݇)ݐ‬ଶ (‫)ݐ‬, ‫ݏ‬ଷ (‫݇)ݐ‬ଷ (‫)ݐ‬, … , ‫ݏ‬௡ (‫݇)ݐ‬௡ (‫ ])ݐ‬8.1
‫ݏ‬௜ (‫݁ = )ݐ‬௜ (‫)ݐ‬଴.ହାఊ೔ (௧) 8.2
Assume that ‫ݏ‬௜ (‫ )ݐ‬is bounded from above. In other words, for each sub-
process, there exists a maximum speed beyond which it is impossible to
go.10 It will be defined as a combination of (i) translational/rotational
kinetics (ii) material tolerances and (iii) minimal downtime (owing to
maintenance, retooling, etc).
‫ݏ = )ݐ( ݅ݏ‬ҧ ݅ (‫)ݐ‬ 8.3

‫)ݐ( ݅݁ = )ݐ( ݅ݕ‬0.5+ߛ݅ (‫)ݐ( ݅݇ )ݐ‬ 8.4
Equations 8.1-8.4 define the process technology available to the firm at

time t. It consists of the menu of set of sub-process speeds and
corresponding energy use levels available to the firm as well as the levels
of sub-process capital at time t. What is immediately obvious is that the
overall speed of production and hence, overall productivity will depend on
individual sub-process machine speeds, more specifically on the lowest sub-
process machine speed.11 Further, it illustrates a number of important
9 Consider the following example. Suppose that a firm has two machines, both of
which operate according to the law of kinetics. Suppose that demand is halved in
which case it can choose to either slow both down by 50 percent, or shut one of the
two machines down. In the former case, will be zero and the law of kinetics will be
in force. However, in the latter case, as one of the two machines will be shut down,
output will be linear in the energy input (i.e. homogeneous of degree one). When it
restarts the second machine (i.e. increases energy consumption), output will double.
Hence, in this case ߛ௜ (‫ )ݐ‬will be non-zero
10 ‫ݏ = )ݐ( ݏ‬ҧ (‫ )ݐ‬can be viewed as a combination of the asymptote of ‫= )ݐ( ݏ‬
௜ ௜ ௜
݁௜ (‫)ݐ‬଴.ହାఊ೔ (௧) , and the physical upper limits of machine speed.
11 In this case, machine speed defines total factor productivity or capital productivity.
This would continue to hold if labor was included. It is important to keep in mind
phenomena, namely the relationship between energy use ݁௜ (‫ )ݐ‬and speed,

‫ݏ‬௜ (‫ )ݐ‬as well as the increasing incremental cost of speed¬ups. Specifically,
energy costs increase exponentially with machine speed.12 It therefore
stands to reason that there will come a time when speed-related productivity
gains will have (i) been exhausted physically (i.e. the upper bound will have
been reached) and/or (ii) be highly uneconomical that is, the cost of speed-
ups will be prohibitive.13
8.3.2. Aggregate Level

Aggregating across firms, we obtain Equations 8.5-8.7, which describe
aggregate output in terms of the average speed of production processes S(t),
and the capital requirements of the latter K(t). Ceteris paribus, the greater
is aggregate energy use per unit K(t); the greater is average machine speed
S(t), and the greater is overall average machine productivity. Furthermore,
the greater is the aggregate level of machine downtime, the greater is *(t),
and hence the more productive is energy given the greater presence of
machine restarts.14
Y(t) = S(t) K(t) 8.5
S(t)=E(t)0.5+*(t) 8.6
Y(t)=E(t)0.5+*(t) K(t) 8.7
8.4. Empirical Evidence

This approach to modeling material processes differs from conventional
approaches in that capital (and labor) is assumed to be an organizational
input (Beaudreau 1998), providing the setting in which energy transforms
material inputs, creating wealth.15 Hence, in keeping with classical
that in modern material processes, labor is a supervisory input. Hence, faster

machines will raise output per unit labor (supervisory) input, despite no more effort
on its part.
12 This is consistent with the notion of diminishing marginal factor productivity.
13 A good example of this is the Anglo-French Concorde supersonic air travel
consortium which was ended owing to prohibitive energy costs.

14 It is our view that while ߛ (‫ )ݐ‬can be an important factor at the industry level, it
௜
will be less of a factor at the aggregate level. Further, by definition, it can never be
negative.
15 This is consistent with basic process engineering where capital is seen in terms of
tools, providing mechanical advantage, but not being a source of energy per se.
Interestingly, labor as a factor input is ignored altogether. See Alting (1994) and
Beiser (1983).
mechanics, only energy is assumed to be physically productive. As such,

output is an increasing function of machine speed, machines, and control
devices the latter affecting the utilization of machine utilization. The
underlying idea is that most machinery is used to accomplish numerous
tasks per period t, requiring retooling/downtime. Control devices, especially
reprogrammable control devices, reduce downtime, thus increasing
utilization rates and average machine speed per period t. Average, aggregate
machine speed is defined by Equation 8.6, where it is a function of energy
use and ߛ௜ (‫ )ݐ‬the rate of growth of machine utilization due to the increasing
use of control technologies.
In this section, we present existing estimates of the energy use/ electric
power output elasticity (0.5+ߛ௜ (‫ ))ݐ‬as well as new results. Equation 8.4
maintains that the output elasticity for energy use should be in the 0.5+ߛ௜ (‫)ݐ‬
range. Given that ߛ௜ (‫ )ݐ‬is variable and positive, it stands to reason that the
energy output elasticity should vary across firms, industries and countries.
However, at the very least (i.e. when ߛ௜ (‫=)ݐ‬0), a ten percent increase in
energy use should result in a five percent increase in aggregate machine
speed and hence, a five percent increase in output per period of time,
assuming that the corresponding maximum machine speed has not been
reached.
Beaudreau (1998, 1999); Kummel, Henn and Lindenberger (2002)16 and
Stresing, Kummel and Lindenberger (2008) provided direct as opposed to
indirect estimates of the energy in this case, electricity used output elasticity
in U.S., German and Japanese manufacturing.17 Recently, Giraud and
Kahraman (2014) provided estimates of the primary energy output elasticity
for 50 countries, reporting elasticities between 0.6 and 0.7. Referring to
Table 8.1 which presents these output elasticities, we see estimates in the
range of 0.30-0.7474, with the average at 0.5303.18 These estimates were
obtained using a number of techniques, ranging from OLS to cointegration
models.
16 No standard errors or t-statistics are provided in Kummel et al. (2002) and Stresing
et al. (2008).
17 For indirect estimates, see Berndt and Woods (1975). Indirect estimates subsume
perfect competition in all factor markets. Direct estimates, on the other hand, make
no such assumption.
18 Beaudreau (1999) conducted a simple growth accounting exercise using these
output elasticities as well as input-output growth rates before and after 1973. He was
able to show that factor input growth (energy, capital and labor) accounted for
almost all of the variation in manufacturing output growth. See the results in the
Appendix.
Table 8.1: Estimates of the Electricity-Use Output Elasticity-

Manufacturing
Method Source Country and Period Estimate (t-stat)

OLS Beaudreau (1995) U.S. (1950-1984) 0.5330 (10.791)
OLS Beaudreau (1998) U.S. (1958-1984) 0.4483 (12.469)
Germany (1962-1988) 0.7474 (3.135)
Japan (1962-1988) 0.6055 (3.017)
LINEX Kummel et al Germany 0.64
(2002) U.S. (1960-1993) 0.51
Japan (1965-1992) 0.61
U.S.-Total (1960- 0.30
1993)
Germany-Total (1960- 0.44
1989)
Cointegration Stresing et al. Germany (1960-1989) 0.517
(2008) Japan (1965-1992) 0.350
U.S. (1960-1978) 0.663
Table 8.2: OLS-AR(1) Estimates-Log Linear Specification
Country Constant EP K L R2
U.S. 1.1678 0.4902 0.7826 0.1034 0.9990
(1958-1984) (0.5342) (3.169) (0.9005) (5.136) ȡ=0.688
Germany -0.2128 0.6124 0.5587 -0.1086 0.9494
(1963-1988) (0.1961) (3.710) (0.5520) (3.2171) ȡ=0.557
Japan -3.1852 0.6970 0.5175 0.3789 0.9818
(1965-1988) (1.328) (2.514) (2.383) (0.9804) ȡ=0.885
Canada -3.6435 0.2451 0.2655 1.2673 0.9844
(1962-1988) (2.421) (0.8631) (1.216) (3.070) ȡ=0.660
Britain 1.2630 0.9141 -0.0972 -0.0870 0.9232
(1963-1988) (0.936) (5.762) (4.000) (0.7156) ȡ=0.502
Finland -0.4482 0.7081 0.3040 0.0583 0.9856
(1963-1988) (0.6075) (4.772) (1.702) (0.3133) ȡ=0.364
Table 8.3: OLS-AR(1) Estimates-Linear Specification
U.S. -64.7637 0.4554 0.0963 1.1141 0.9891
(1958-1984) (3.186) (4.009) (1.084) (5.417) ȡ=0.663
Germany -0.2128 0.6124 -0.1086 0.5587 0.9994
(1963-1988) (0.1961) (3.710) (0.5222) (3.217) ȡ=0.557
Japan -164.97 0.5618 0.2374 1.3221 0.9820
(1965-1988) (1.814) (2.983) (3.683) (1.643) ȡ=0.827
Canada -130.442 0.3426 0.1582 1.6709 0.9821
(1962-1988) (2.347) (1.210) (6.2092) (2.836) ȡ=0.651
Britain 17.1535 0.8570 -0.0289 -0.0459 0.911
(1963-1988) (0.3206) (5.605) (0.1008) (0.1720) ȡ=0.583
Finland -17.9512 0.57475 0.3122 0.1854 0.9918
(1963-1988) (0.8936) (6.931) (2.553) (0.9723) ȡ=0.294
Table 8.4: OLS-AR(1) Estimates-Log Differences Specification
U.S. 0.0447 0.3824 0.9839 0.9646 0.9179
(1958-1984) (3.168) (2.498) (2.9561) (6.499)
Germany -0.0016 0.6093 0.0650 0.4675 0.6682
(1963-1988) (0.1469) (3.682) (0.2197) (1.872)
Japan 0.0360 0.7176 0.3052 0.6291 0.6710
(1965-1988) (1.930) (2.942) (1.128) (1.228)
Canada 0.0572 0.0528 -0.9138 1.4123 0.715
(1962-1988) (2.391) (0.2072) (1.933) (4.244)
Britain -0.0118 0.7898 -0.2098 0.0891 0.6432
(1963-1988) (0.9165) (5.122) (0.4099) (0.5597)
Finland 0.0139 0.7066 -0.1171 -0.3518 0.48
(1963-1988) (0.371) (4.364) (0.1352) (0.8735)
Using comparable data, we estimated a simple three-factor input

production function for U.S., German, Japanese, Canadian, British, and
Finnish manufacturing.19 Three econometric specifications were employed,
namely (i) a linear with AR(1) error structure, (ii) a log-linear with AR(1)
error structure and (iii) a differences-in-logs specification.20Ordinary least
19 Output, labor and electricity use data were obtained from United Nations
Industrial Statistics Yearbook (1960-1988) (United Nations 1984); capital data were
obtained from OCED, Flows and Stocks of Fixed Capital 1989 (OECD 1989). These
data are available from the author upon request.
20 The variables used were rva, real value added, ep electric power use in kwh, k
capital stock, and l, hours workers. The estimated equations were: in the linear case,
rva(t) = a0 + a1ep(t) + a2k(t) + a3l(t); in the log-linear case, lnrva(t) = a0 + a1lnep(t)
squares with an autoregressive error of order one was used in the case of the
first and second, while simple OLS was used in the third. The results are
presented in Tables 8.2-8.4, where we see estimates of the EP (electric
power use) coefficients in the predicted range of 0.50+Ȗ(t). For example, in
the per unit of capital on electric power per unit capital and labor per unit
capital. The results were however not significantly different.21
We then proceeded to test the model using 2-digit SIC data for U.S.
manufacturing from 1947 to 1984. The data in this case were taken from the
U.S. Annual Surveys of Manufactures (1947-1984) as well as the U.S.
Census of Manufactures (1947-1984). The results are presented in Table
8.5, where two sets of electric power output elasticities are reported.
Column 1 contains the electric power consumption output elasticity when
both electric power consumption (purchased and generated) and production
workers were included as independent variables, while Column 2 contains
the output elasticity when only electric power consumption was used as the
independent variable. As was the case with aggregate data, the output
elasticities are once again centered around the predicted-by-the-law-of-
kinetics value of 0.50. Column 3 reports the R2 for the latter case. What is
noteworthy is the extent to which variations in electric power consumption
explain roughly 80 percent of the variation in value-added by 2-digit SIC
industry.
+ a2lnk(t) + a3lnl(t), and the differences in logs equation, dlnrva(t) = a0 + a1dlnep(t)

+ a2dlnk(t) + a3dlnl(t). ln and dln denote log and differences in logs, respectively.
21 One of the difficulties encountered was the fact that from 1973 onwards, capital
literally exploded with massive investments in control technologies, which we

believe thwarted our attempt at accounting for scale effects.
Table 8.5: 2-Digit SIC industry Electric Power Output Elasticities

1947-1984
SIC Industry Elasticity-I (t- Elasticity-II (t- R2

statistic) statistic)
20 Food and Kindred Products 0.610 (30.873) 0.552 (13.074) 0.867
21 Tobacco Products 0.386 (12.604) 0.399 (13.315) 0.872
22 Textile Mill Products 0.401 (5.243) 0.301 (5.154) 0.505
23 Apparel and Other Textile 0.142 (7.800) 0.200 (7.489) 0.718
Products
24 Lumber and Wood 0.518 (17.614) 0.474 (10.923) 0.821
Products
25 Furniture and Fixtures 0.227 (5.785) 0.490 (13.566) 0.884
26 Paper and Allied Products 0.5937 (17.558) 0.673 (32.139) 0.975
27 Printing and Publishing 0.156 (3.3158) 0.458 (14.599) 0.895
28 Chemicals and Allied 0.519 (9.651) 0.595 (9.475) 0.775
Products
29 Petroleum and Allied 1.153 (8.486) 0.658 (9.161) 0.763
Products
30 Rubber and Misc. Products 0.247 (1.822) 0.804 (24.828) 0.959
31 Leather and Leather 0.382 (7.373) 0.048 (0.8439) 0.020
Products
32 Stone, Clay and Glass 0.627 (11.702) 0.769 (22.983) 0.953
Products
33 Primary Metal Industries 0.607 (16.826) 0.552 (13.074) 0.867
34 Fabricated Metal Products 0.480 (11.841) 0.603 (28.612) 0.970
35 Machinery, Except 0.624 (18.007) 0.624 (18.007) 0.967
Electrical
36 Electric and Electronic 0.403 (6.740 0.744 (23.012) 0.953
Equipment
37 Transportation Equipment 0.737 (20.826) 0.791 (19.149) 0.933
38 Instruments and Related 0.662 (15.220) 0.812 (37.508 0.981
Products
39 Miscellaneous 0.400 (4.1755) 0.337 (3.134) 0.308
Manufacturing Industries
*rva = Real Value Added; ep = Electric Power Consumption; pw = Production
Workers. Model-I: lnrva = a + blnep + clnpw; Model-II: lnrva = a + blnpw.
8.5. Further Evidence of the Law of Kinetics

in Manufacturing
The Law of Kinetics referred to above is straightforward in its
implications. Successive increases in energy use will lead to smaller and
smaller increases in speed and productivity-in short, the returns to
increasing energy intensity, while still positive, will be decreasing. It
therefore stands to reason that industries that had experienced high growth
rates prior to 1973 should have experienced lower ones afterwards (vertical
axis) and vice-versa. Figure 8.1, taken from Baily (1982); presents a
breakdown of pre- and post-1973 productivity growth rates for U.S. 2-digit
SIC industries. We see that industries that experienced high growth rates
prior to 1973 (horizontal axis) experienced larger decreases in adjusted KLP
growth afterwards (vertical axis).
Data on multifactor productivity growth in major U.S. industries show
a similar pattern. Referring to Table 8.6, we see that post-WWII
productivity growth rates in general fall after 1965, but remain positive
(1965-1973), before falling significantly in the post-1973 period. However,
it is important to point out that some industries continue to experience high
growth rates. For example, multifactor productivity growth remained high
in Textiles, Apparel, and Electrical Machinery after 1973. Figure 8.2 plots
these data points for the four sub-periods. What is particularly noteworthy
is the overriding downward trend in productivity growth across all
industries in each of the four sub-periods. Again, what is immediately
obvious is the fact that productivity growth had begun to decrease well
before 1973.
Aggregate productivity data show the same pattern. Table 8.7 taken The
Economic Report of the President as cited in Fisher (1988) shows the U.S.
productivity growth rate declining throughout the post-WWII period, going
from 2.9 percent in 1947-1955 to 2.5 percent in 1955-1968 and 1.5 percent
in 1968-1973.
Figure 8.1: Productivity Growth Slowdown and Past Productivity

Growth, by Manufacturing Industry
Our model predicts that output per unit of energy input (i.e. in this case,
electric power), will be decreasing in the use of energy per unit of capital.
Hence, greater electric power use per unit of capital (machinery and
equipment) will result in lower incremental output. Once machinery and
equipment reaches its maximum speed, then output per unit energy input
should stabilize. Figure 8.3 presents the ratio of value-added per kwh for 20
U.S. manufacturing industries from 1947 to 1984. What is immediately
obvious is (i) the similarity of these results and those of Baily and
Chakrabarti (1988) reported earlier, and (ii) the fact that electric power
productivity appears to stabilize in virtually all industries after 1970. This
leads us to conclude that the energy deepening that had characterized the
post-WWII period appeared to have come to an end in the latter half of the
60s, well ahead of the OPEC-induced energy crises.
Table 8.6: Multifactor Productivity Growth in Major U.S. Industries,

1948-1985
Inductry 1948-1985 1948-1965 1965-1973 1973-1979 1979-1985

Major Aggregates
Farming 3.4 3.4 2.8 1.8 5.6
Nonfarm 1.1 2.1 0.9 -0.2 -0.4
nonmanufacturing
Manufacturing except 1.9 2.6 1.9 0.6 1.1
nonelec. mach.
Manufacturing Industries
Food 2.6 2.9 3.6 0.1 2.9
Tobacco 0.2 2.7 2.0 -0.2 -8.0
Textiles 3.9 4.6 2.0 5.7 2.6
Apparel 2.1 1.9 2.7 2.8 1.5
Lumber 2.4 3.8 1.0 1.5 1.3
Furniture 1.7 2.0 1.6 2.1 0.5
Paper 2.0 2.0 3.7 0.1 1.8
Printing and Publish. 1.0 2.1 0.8 -0.2 -0.9
Chemicals 3.2 4.1 4.0 1.1 2.0
Petroleum 0.6 2.9 0.8 -1.4 -4.0
Rubber 1.9 2.2 1.7 -0.8 4.0
Leather 0.8 1.1 1.8 0.2 -0.6
Stone, Clay and Glass 1.4 2.2 1.0 0.5 0.9
Primary Metals 0.0 1.0 0.7 -2.8 -0.9
Fabricated Metals 1.3 1.8 1.3 0.0 1.1
Nonelectrical Machinery 2.5 1.4 1.5 0.4 9.1
Electrical Machinery 3.4 4.2 2.7 3.5 1.9
Transportation 2.0 3.5 1.2 0.1 0.6
EEquipment
Instruments 2.5 3.6 2.4 2.7 -0.5
Misc. Manufacturing 2.0 2.6 2.9 0.1 1.2
Nonmanuf. Industries
Mining 0.6 3.1 2.0 -6.0 -1.5
Construction -0.2 2.9 -3.9 -2.2 -2.0
Transportation 1.6 2.1 2.7 1.3 -1.3
Communications 3.9 5.6 3.4 2.4 1.3
Public Utilities 3.5 5.8 3.0 -0.6 1.7
Trade 1.9 2.6 2.4 0.4 0.8
Finance and Insurance 0.3 1.3 0.6 -0.7 -2.0
Real Estate 0.6 1.8 0.3 1.4 -3.2
Services 0.7 0.4 1.4 0.4 0.8
Source: Baily and Chakrabarti (1988), 6.
8.6. The Limits to Speed, Control Technologies

and the Productivity Slowdown
The post-WWII period was characterized by speed-ups both in general
and in industry (i.e. industrial processes). Throughout this period, speed
records were being set. For example, the fastest airplane was the Lockheed
SR-71 Blackbird, capable of speeds over 2200 miles per hour. The two-seat
SR-71 was developed in the early 1960s by the U.S. Air Force as a strategic
reconnaissance aircraft. The first flight of an SR-71 was in 1964 at a
classified location in Nevada. The aircraft's first operational sortie was
flown out of Okinawa, Japan in 1968.22 The World Water Boat Speed
record, like the airspeed record, dates back to 1978. Australian Ken Warby
set the record in 1978 when he averaged 317.60 mph in a 27-foot jet-
powered hydroplane called "Spirit of Australia."
In this chapter, we have argued that while greater machine speeds were
instrumental in raising productivity, the laws of physics imposed and
continue to impose upper limits/bounds on speed-ups, and hence on
productivity growth. Theoretically, as energy use increases exponentially
with speed/velocity, there will come a time when further speed-ups will be
either physically impossible (owing to material tolerances), or not
economically viable. We maintain that this point in time was reached in the
late 1960s/early 1970s, bringing to an end a century of speed-up-based
productivity growth. Put differently, by the late 1960s/early 1970s, most
machines had reached their maximum speed. The evidence is there for all
to see. Automobile, airplane, power tool, household appliance (washers,
dryers) speeds have not increased since the 1960s. Clearly, there are some
exceptions as some speeds have increased.23 However, on the whole,
process speeds have not.
22 Most of the SR-71 fleet has now been retired, except for two Blackbirds currently
on loan to NASA's Dryden Flight Research Center where the aircraft are being used
as "test beds" for high altitude research.
23 A good illustrative example is provided by supersonic air travel, more specifically,
the joint British/French Concorde that flew at Mach II. Energy costs were
astronomical, which in addition to the problems associated with the associated sonic
boom, did the venture in. Today, commercial jet aircraft travel at speeds which in
many cases are lower than those in the 1970s, owing principally to the associated
energy costs.
Figure. 8.2: Productivity Growth by Manufacturing Industry 1948-

1985.*
*The four data points plotted here refer to the four periods Industry 1948-1965,
1965-1973, 1973-1979, 1979-1985.
Table 8.7: U.S. Annual Productivity Growth 1947-1980
1947-1955 1955-1968 1968-1973 1973-1980

2.9% 2.5% 1.5% 0.4%
Source: Economic Report of the President, 1987, Table B-43 quoted in Fisher
(1988).
8.6.1 Asymptotic s and the Productivity Slowdown

We have shown that the post-WWII era was characterized by what
Alfred Chandler referred to as the economies of speed and consisted of the
wholesale application of speed-ups and numerical control technologies to
industrial processes, increasing average machine speed and conventionally-
measured productivity. Machines turned faster and downtime (which
reduced statistical/measured machine speed) fell drastically. However, we
maintain that by the late 1960s, the large majority of these machines had
reachedor were near to reachingtheir theoretical maximum speed (Equation
8.3), resulting in a tapering off of productivity growth, which coincided with

a decrease in the rate of growth of energy usein this case, electricity
consumption. By the early 1970s, it had fallen to near zero, bringing with it
the rate of growth of productivity. The OPEC-induced price increases, while
contemporaneous events, were largely inconsequential.24
Referring to Figure 8.4 which presents the evolution of the ratio of
electric power consumption to capital for five OECD countries in the pre-
1973 period, we see that the energy intensity of manufacturing production
processes had leveled off by the 1960s.25 The marked drop in the rate of
growth of electric power in 1973 and after, we maintain, may have been the
culmination of this process. Production processes, in general, could no longer
be either speeded-up, or more fully utilized (via reprogrammable control
devices).
It is important to point out that this is an aggregate result. It does not
imply that machine speed-ups in general were/are a thing of the past. For
example, in certain industries, machine speeds continued to rise. A good
example is the pulp and paper industry where newsprint machine speeds
have continued to increase into the 21st century despite higher energy prices.
This may explain why the electric power to capital ratio (EPK) in Finland,
a large paper producer, increased in the late 1960s (see Figure 8.4). To better
understand the role of speed in material processes and indeed throughout
the economy, we now turn and examine important developments in various
industries. We will focus on two developments, namely increases in
machine speed and the introduction of new control technologies which as
pointed out above were and are complementary phenomena.
8.6.1 Continuous Flow and Batch Material Processes

There are, in general, two types of material processes, continuous flow
and batch, both of which are energy-based. For example, a newsprint paper
machine is an example of a continuous-flow material process where wood
pulp is transformed into paper. The key distinction is whether the process is
interrupted. Batch processes are, by definition, interrupted, while continuous
flow ones are not. All processes are governed by the laws of kinetics
(machine or chemical). Increasing machine speed and electricity use will
increase output per period of time. Likewise, increasing the temperature of
24 It was suggested that perhaps the OPEC oil embargo precipitated the decline in
energy use. We remain skeptical as, if nothing else, higher energy prices will,
assuming that machine speed can be increased, result in further profit-increasing
speed ups.
25 These ratios were obtained using the same data.
Figure 8.3: Value-Added per kwh 1948-1980 by 2-Digit SIC Industry.
a batch process-based chemical reaction will increase output per period of

time.26 Whether it be batch or continuous, the speed at which the various
transformations take place is not unlimited, owing to a number of factors,
not the least of which are the laws of kinetics, specifically KE = 1/2mv2.27
26 This is known as Arrenius’ Law, according to which the rate of a chemical reaction
depends on, among other things, the absolute temperature.

27 The rotational energy or angular kinetic energy is the kinetic energy due to the
rotation of an object and is part of its total kinetic energy. Looking at rotational
energy separately around an object's axis of rotation, one gets the following
dependence on the object's moment of inertia: Erotational=1/2iȝ2 , where ȝ is the
angular velocity i is the moment of inertia around the axis of rotation E is the kinetic
energy. The mechanical work required for/applied during rotation is the torque times
the rotation angle. The instantaneous power of an angularly accelerating body is the
torque times the angular velocity. For free-floating (unattached) objects, the axis of
rotation is commonly around its center of mass. Note the close relationship between
the result for rotational energy and the energy held by linear (or translational)
motion: Etranslational=1/2mv2. In the rotating system, the moment of inertia, i, takes
the role of the mass, m, and the angular velocity, ȝ, takes the role of the linear
velocity, v. The rotational energy of a rolling cylinder varies from one half of the
In short, kinetic energy increases at the square of velocity. Hence, the cost
of continuous Àow process speed-ups will increase exponentially over time.
Other factors limiting speed-ups are factors like material tolerances, bearing
speeds, bottlenecks, etc.
8.7 Case studies

8.7.1. Manufacturing
It is generally argued that the manufacturing sector, especially in the
U.S., benefited from two waves of electrification (Devine 1990). The first
wave occurred in the early 20th century when electric power largely replaced
steam power. This is typically referred to as the second industrial revolution.
The second wave occurred in the post-WWI period when production was
reorganized around the principle of unit drive—that is, individual electric
motors. Production processes, consisting of myriad sub-processes, were
reorganized. According to Warren Devine,”Reorganization led to the
integration of productionthe coordination of related mechanized processes
and the adjustment of the tempo of these processes with one another so that
they could be carried out in a natural or continuous sequence (Devine 1990,
201).” Also increasing productivity per machine in this period was the
introduction of reprogrammable control devices that reduced machine
downtime considerably. Together, these developments set the stage for
record output and productivity growth as the newly coordinated production
processes were speeded up. In this section, we examine the role of speed-
ups and control-technology developments in a handful of manufacturing
industries, including pulp and paper, petroleum refining and information
technology. Unfortunately, data on machine speed and/ or control
technologies by industry are extremely hard to come by, forcing us to resort
to anecdotal evidence.28
The pulp and paper industry, like the transportation sector, provides a
good example of speed-up related productivity gains, as the underlying
technology (Fourdrinier Process) has remained virtually unchanged for over
the past two centuries. What has changed, however, is the speed at which
the various sub-processes (e.g. forming, pressing and drying) operate. Until
the 1990s, most newsprint paper machines in North America dated back to
the 1930s. Surprisingly, these machines were not unprofitable as they had
translational energy (if it is massive) to the same as the translational energy (if it is
hollow). Source: http://en.wikipedia.org/wiki/Rotational-energy.
28 This is surprising given the important role of machine speed in productivity
growth.
undergone, over the course of the intervening period, successive speed-ups,

placing them at par with new machines.29
Figure 8.4: Electric Power to Capital Ratios-Manufacturing Pre-1973
According to Hardman and Cole (1960): “further increases in the

efficiency of the paper machine depend on changes in the technology of
29 What eventually did them in was their sizemore specifically, their width.
paper production, on the perfection of the design of machines and separate

units, and on an increase in productivity in terms of speed and width.”
Cohen (1984) showed that capacity in the U.S. Pulp and Paper industry
more than doubled from 1914 to 1940 despite the fact that the number of
machines remained constant. Specifically, average machine capacity increased
from 4, 182 to 10, 550 tons. These increases were made possible “largely
by faster (and wider) paper machines. Between 1915 and 1940, maximum
Fourdrinier machine speeds more than doubled. (Cohen 1984, 780)”
Referring to Table 8.8, we see that machine speeds increased by 4.17
percent on average from 1919 to 1931, but only by 2.51 percent from 1931
to 1940 in keeping with the law of kinetics.
Output in the petroleum industry increased rapidly from 1947 to 1951,
a little more slowly from 1951 to 1956 and still more slowly after 1956.
Output per man-hour increased at an average annual rate of 5.2 percent,
higher than the overall average for manufacturing. In a report prepared for
United States Select Committee on Small Business by Leon Greenberg,
Assistant Commissioner for Productivity and Technological Developments,
this increase was attributed to (i) increased use of computer control (ii)
improvements in cracking (iii) product mix and petrochemical Products (iv)
the increased size and complexity of plants (v) increasing expenditures for
research and development. As he made clear, improvements in control
technology affected what we shall refer to as “refining speed” in two ways,
namely by reducing downtime, and increasing throughput rates. In his
words,”Direct benefits of electronic data processing include increased
production with the same equipment, reduced operating costs and improved
quality control (Greenberg 1964, 7).” Catalytical cracking, he reported,
increased refining capacity of existing plants by more than 20 percent. He
went on to point out that between 1947 and 1962, the number of operating
refineries in the industry decreased from 361 to 287, but that in the same
period, crude oil throughput capacity of operating refineries rose 5.3 million
barrels per day to 9.8 million barrels, and total refinery capacity rose from
5.5 million to 10.1 million barrels per day.30
8.7.2 Transportation
The transportation sector provides what we feel is the best example of
the principles at play in this chapter and indeed throughout the post-WWII
period. A motor vehicle provides transportation services. The greater the
speed at which it travels, the greater its productivity. Doubling the speed of
30 No mention is made of speed-ups which in this case consist of increasing the
rate/speed at which petroleum is refined.

a truck, for example, will double the number of miles traveled, and hence,
will double its supply of transportation servicesits productivity. Second,
the fewer hours the truck in question is idle (e.g. sitting in the yard), the
greater is its productivity. The post-WWII period, we maintain, witnessed
substantial increases in vehicle productivity, increases that owed primarily
to increased speed and more efficient utilization of rolling stock. By the late
1960s/early 1970s, most of these gains had been achieved. Maximum
speeds had been reached, and downtime (idle time) had been reduced
considerably. A good example of a speed-increasing technological change
in the post-WWII period was the introduction of container shipping or
containerization. Containerization increased productivity markedly. It did
so in two ways, first by increasing the productivity of ports. Less time was
needed to unload large cargo ships (i.e. lower Ȗ(t)). Second, owing to the
latter, ship downtime was reduced considerably so that instead of lying idle
in port, they could be providing services, which in this case consists of
“transporting” cargo. Conceptually, this corresponds to an increase in the
relevant si(t) (see Equation 8.1). It is important to note that this was a one-
shot occurrence. Once minimal port time is reached, no further gains can be
achieved.31
8.7.3 Coal Mining

According to Warren Devine, productivity improvement in the underground
coal industry came in two stages. Steady but comparatively slow
productivity growth occurred prior to the 1950s as coal cutting and hauling
were mechanized via electricity. The second stage (after 1950) was marked
by mechanically integrated production, whereby coal could be cut, loaded,
and removed in a continuous sequence. Surface coal mining also benefited
from electricity.
Large-scale surface mining was not possible before mechanization.
Specialized machinery began to be introduced around World War I, but it
was mid-century before appropriate machines and techniques were widely
employed. Electric power was not integral to the mechanization of surface
mining, but electricity has always played a supporting role. Mechanical
power transmission proved to be cumbersome and costly on very large
mobile machinery; electric power was often advantageous even though it
31 Port productivity, like that of the retail sector, is increasing in the speed at which
merchandise moves through the port itself. The greater the turnover, the greater the
productivity. In the retail sector, this is measured in terms of average shelf life and
sales per square foot. The lower is shelf life, the greater is store productivity, and the
greater are sales per square foot, the greater is productivity.
had to be delivered to machines via cable of generated on board. Larger and

larger machines were deployed (particularly after 1950) and many of them
used electricity provided in these ways. This larger machinery facilitated
the mining of coal seams of thickness and extent far exceeding those
previously mined and productivity grew rapidly. (Devine 1990, 202)
8.7.4. The Construction Sector

The construction sector has been and continues to be labor-intensive.
Specifically, physical labor has been used to add value to building materials.
Among the tasks performed include materials handling, their transformation
(cutting, drilling), and the assembly. For our purposes, let us define productivity
in this sector as the average time needed to complete a standard building
(e.g. a 25’x40’, one storey bungalow). Historically, the most significant
change in this process consisted of the introduction of electric-powered
tools, including electric-powered cranes, winches, and power tools (saws,
drills, nailers) which increased individual sub-process speeds, thus
increasing the overall speed at which the “standard” building could be
completed. As it turns out, most of these tools were introduced in the 1950s
and 1960s, increasing productivity significantly. However, like all other
machines, the maximum speed at which these tools operate was reached
early on. Electric-powered drill, saw, and nailing machine speeds have, for
the most part, not increased in the last 40 years.
8.7.5 Information Technology Revolution

The Information Technology (IT) Revolution provides another example
of the law of kinetics at work. Computer/processor productivity is measured
in terms of CPU speed. The faster the CPU, the more operations can be
carried out on a given processorhence, the greater the productivity. Over
the course of the last 30 years, CPU speed has increased exponentially,
going from 12.5 Hz in 1982 to 3.6 GHz in 2004 where according to many it
has become “stuck.” (see Figure 8.5). Among the reasons invoked is rising
energy costs which increase cubically with the clock rate.
The power-performance relationship has been formalized in terms of
what is known as Dennard scaling.
Table 8.5: Chronology of Maximum Fourdrinier Newsprint Paper

Machine Speeds
Year Maximum Speed Period Average Annual

(feet per minute)
1887 250
1896 500 1887-1919 3.43
1911 700
1919 735
1921 1000 1919-1931 4.17
1931 1200 1931-1940 2.51
1940 1500
Source: Cohen (1984), 787.
Another important observation is the concept of Dennard scaling

according to which the amount of power required to run the transistors in a
specific unit volume stays constant despite increasing their number, such
that the voltage and current scale with length. Yet, this observation is no
longer becoming valid as transistors are growing very small. The scaling of
voltage and current with length is reaching its limits, since transistor gates
have become too thin, affecting their structural integrity, and currents are
starting to leak.
Furthermore, thermal losses occur when you are putting several billions
of transistors together on a small area and switching them on and off again
several billion times per second. The faster we switch the transistors on and
off, the more heat will be generated. Without proper cooling, they might fail
and be destroyed. One implication of this is that a lower operating clock
speed will generate less heat and ensure the longevity of the processor.
Another severe drawback is that an increase in clock speed implies a voltage
increase and there is a cubic dependency between this and the power
consumption. Power costs are an important factor to consider when
operating computing centers. (http://www.comsol.com/blogs/havent-cpu-
clock-speeds¬increased-last-years/)
We maintain that the microprocessor, like all other material processes,
obeys the laws of kinetics, and furthermore that the CPU slowdown mimics
the productivity slowdown in the 1960s and 1970s. Both are manifestations
of the limits imposed by these same laws.
8.8 Implications
There are limits in the physical world. And non-linearities. One such
non-linearity is the relationship between speed and energy use. In an ideal
world, the relationship between the two would be linear. Doubling energy
would double speed, and in so doing double output. Unfortunately, this is
not the case. As argued, most production processes have presently reached
their maximum operating speed (both physically-viable and economically
viable). Large speed-based productivity gains were achieved in the post-
WWII period, owing to greater machine speeds and to new management
techniques (e.g. reprogrammable control technologies) and exhausted by the
mid-to-late 1960s. In this section, we examine some of the implications of
our findings.
Figure 8.9: History of Clock Rate and Power for Intel x86
Microprocessors
Source: Patterson and Hennessy (2005), 342.
8.8.1. Knowledge-Based Growth is Bounded

Our findings show that when the laws of physics are taken into
consideration, knowledge-based growth is, in general, bounded and hence,
limited. Productivity growth in the post-WWII period was based on rising
energy consumption, governed by the laws of kinetics. These laws are
immutable and knowledge-independent. Research and development or
knowledge cannot alter them. Once a machine or sub-process has reached
its theoretical maximum speed, nothing can be done to increase it. Or at

least, not at a reasonable cost.32
This is not to say that process-based research and development or
knowledge is completely unproductive, but rather that it should be viewed
as an enabling factor, allowing firms to reach the upper limits implied by
the laws of physics. However, the point we wish to make is that it cannot
restore growth rates to post-WWII levels. By increasing material tolerances
(e.g. bearing tolerances) or by removing bottlenecks research and development
can increase subprocess speeds. However, at the aggregate level, machine
speed will be bounded by the laws of kinetics. In many ways, our findings
corroborate the findings of Jones (1995) which cast serious doubt on the
“productivity” of research and development expenditure.
8.8.2. The Singularity of the Post-WWII Period in History

The high-productivity growth in the post-WWII period, we maintain,
may have been a “singularity” in Western industrial and economic history.
It represented a period of time in which speed up-based productivity gains
at the aggregate level were still physically and economically viable. That
period may now be over with all that this would imply. This is not to say
that speed-up-based productivity gains are impossible, but rather that, in the
advanced industrial economies (G-6 countries), maximum aggregate (i.e.
economy-wide) operating speeds have for the most part been reached. Our
model predicts that nations that upgrade their manufacturing sectors will
experience high growth rates, but will, eventually, return to the low growth rates
of the post-1973 period. This is consistent with Gordon (2012) who argued that
the second industrial revolution was exceptional in its effects on wealth, and
moreover, appears to have been a singular event in human history.33
8.8.3. Conventional Labor Productivity

Despite not being modeled explicitly, our model predicts that while
conventionally-measured labor productivity increased throughout the post-
WWII period, largely the result of speed-ups, that from 1973 onwards it
32 A good illustration is provided again by the Concorde which doubled the speed
of air travel to Mach II, but at an exorbitant cost ($5000 versus $800), one that
eventually did the project in.
33 Our results also rationalize Olson's (1988) findings to the effect that post-WWII
productivity growth was greater in Europe than in North America. Specifically,

operating speeds in North America were greater than those in Europe, making for a
situation in which greater gains could be achieved in Europe.
could/can only grow by raising the capital-labor ratio, either by investing in

more capital per worker, or reducing labor for a given amount of capital.34
The widespread application of new, computer-based control technologies
in the 1980s and 1990s is an example of the latter. By decreasing the
demand/need for human supervision, the capital-labor ratio increased,
increasing conventionally-measured labor productivity. In other words, less
labor was needed to supervise a given level of capital. It bears noting,
however, that the capital in question was not more productive, physically
speaking.
8.8.4. The Inelastic Demand for Speed-Up Based Energy Use

We maintain that speed-ups rank among the most profitable process-
related investments a firm can make. The reason is simple: Most if not all
of the firm’s costs remain the same (capital and labor), while its revenues
increase substantially. Take for example the case of a speed-up that doubles
the firm’s output. Assuming that electricity prices account for five percent
of total costs, its stands to reason that its revenues will double, but its costs
will only rise by 20 percent. Hence, it stands to reason that if the firm in
question has not reached its maximum speed (Equation 8.3), then speed-ups
will be profit increasing. This allows us to conclude that higher energy
prices may not have been the cause of the decrease in the rate of growth of
energy use in the 1970s. Rather, physical factors may have been at play.
This is further borne out by the fact that machine speeds have continued to
increase in certain industries (e.g. pulp and paper).
8.8.5. ICT and Productivity: General and Sector Specific

As we have shown, developments in control technologies have increased
overall productivity by reducing downtime and hence increasing capital and
labor utilization. Hence, it stands to reason that ICT can affect productivity.
Better control technologies such as reprogrammable machine tools will
increase average machine speeds and output. The point however is that these
gains are bounded from above. Once machine downtime has been altogether
eliminated (or near-eliminated), then ICT-based productivity gains will cease.
34 Conventionally-measured labor productivity can be modeled as:

௬೔ (௧) ௘೔ (௧)బ.ఱశം೔ (೟) ௞೔ (௧)
= . Speed-ups will increase labor productivity by increasing the
௡೔ (௧) ௡೔ (௧)
numerator on the right-hand side of this equation. It bears reminding that according
to our approach, labor is not physically productive, but rather is an organizational
input, which explains its absence from the production function.
Oliner and Sichel (2002) and others have argued that the higher
productivity rates observed in the late 1990s and early 2000s owed in large
measure to productivity gains in the ICT sector. This, we maintain, is
consistent with our findings. As with all new industries/sectors, the
associated sub-processes and processes benefit from important machine
speed-ups and better control technologies, which has the effect of increasing
overall productivity. However, like all other industries, these gains diminish
over time, and eventually disappearthat is, once maximum machine speeds
have been reached.
8.8.6. The Universal Nature of the Productivity Slowdown

Perhaps one of the most vexing aspects of the productivity slowdown
has been its universal character. In short, productivity growth in all
countries plummeted at roughly the same time, which given the presence of
significant heterogeneity across countries (i.e. in the price of energy), raised
a number of questions. Of all the explanations, the only one that was
consistent with this fact was the OPEC-led energy crisis hypothesis. This
owed in large measure to the fact that energy markets were and are highly
integrated. Our work, we argue, adds to the list as it predicts that machine
upper-speed limits, universal in nature, would have been reached at roughly
the same time across all countries. A good example is airplane speed which
increased throughout the post-WWII period, but which attained its upper
maximum in the late 1960s/early 1970s.
8.9. Summary and Conclusions

To many observers, the OPEC-induced oil price hikes in the 1970s and
the resulting decrease in the rate of growth of energy use were among the
leading causes of the productivity slowdown. After all, energy lies at the
core of each and every material process. Everything fit: higher energy price
put an end to decades-long energy deepening, bringing productivity with it.
But there were problems. One was the question of “contemporaneity”.
Energy demand decreased swiftly and universally, despite considerable
heterogeneity in prices. Second, the demand for speed-ups are, as we have
shown, generally price inelastic. Third, by the 1990s, the rate of growth of
the demand for energy use did not return to its pre-1973 levels despite the
fact that the real price of energy had fallen to its pre-1973 level. Last, there
is considerable evidence that shows that the productivity slowdown started
well before the 1973 oil price shock.
This chapter offered an alternative explanation, namely that the fall in

the rate of growth of energy use may have had little if nothing to do with
the OPEC price hikes, and everything to do with basic physics. Speed-up-
based productivity growth, the staple of the post-WWII period, may have
reached its systems-wide physical limit. Machine speeds, in general, could
not have been increased any further. The “Age of Speed” may have been
coming to an end. This is not to say that speed-ups were no longer possible,
but rather that maximum system-wide (i.e. aggregate) machine speed may
have been reached. In some industries, speed continued to rise, if only
marginally. In others, new control technologies reduced machine downtime,
thus increasing average speed per period of time. Alfred Chandler's
“economies of speed,” had run their course. While this possibility may be
sobering, it is nonetheless reassuring. However, maybe a newly industrial
revolution will change the limits of machine speed again.
8.10 Appendix
U.S. 1950–1984 1950–1973 1974–1984

USVA 2.684 3.469 0.121
USAI* 2.674 3.472 0.310
USEP 4.052 5.371 0.246
USN 0.662 0.900 -0.091
USK 3.694 3.614 3.4008
Germany 1963–1988 1962–1973 1974–1988
GERVA 2.462 6.522 1.486
GERAI* 2.433 5.190 1.080
GEREP 2.894 5.883 1.366
GERN -0.785 0.592 -0.938
GERK 2.945 5.620 1.406
Japan 1965-1988 1965-1973 1974-1988
JAPVA 3.826 8.844 3.099
JAPAI* 3.566 9.856 1.538
JAPEP 3.559 11.320 0.965
JAPN -0.082 2.297 -0.367
JAPK 7.520 13.536 5.182
*‫ߚ = ܫܣ‬ா௉ ா௉ ௗ௧
+ ߚே
ே ௗ௧
+ ߚ௄
௄ ௗ௧
8.11 References
Alting Leo. 1994. Manufacturing Engineering Processes. New York, NY:
Marcel Decker Inc.
Baily N, Chakrabarti A.K. 1988. Innovation and the Productivity Crisis.
The Brookings Institution.
Baily Martin Neil. 1982. The productivity slowdown by industry. Brookings
Papers on Economic Activity. 2:423-59.
Beaudreau Bernard C. 1995. “The Impact of Electric Power on Productivity:
the Case of U.S. Manufacturing 1958-1984.” Energy Economics. 17,
no.3:231-6.
Beaudreau Bernard C. 1998. Energy and Organization: Growth and
Beiser Arthur. 1983. Modern Technical Physics. Menlo Park, CA: The
Benjamin/ Cummings Publishing Company.
Berndt Ernst, Wood David. 1975. “Technology, Prices and the Derived
Demand for Energy.” Review of Economics and Statistics: 259-69.
Chandler Jr Alfred D. 1977. The Visible Hand, the Managerial Revolution
Cohen Avi J. 1984. “Technological Change as Historical Process: the case
of the U.S. Pulp and Paper Industry, 1915-1940.” Journal of Economic
History. 44, no.3:775-99.
Devine Warren D. 1990. “Electricity in Information Management: the
Evolution of Electronic Control.” in: Schurr Sam H. et al., eds.
Feyrer J. 2011. “The U.S. Productivity Slowdown, the Baby Boom and the
Management Quality.” Journal of Population Economics. 24, no. 1:267-
84.
Fisher Stanley. 1988. “Symposium on the Slowdown in Productivity
Growth.” Journal of Economic Perspectives. 2, no. 4:37.
Ford, Henry. 1926. Mass Production. Encyclopaedia Britannica.
Giraud Gael and Kahraman Zeynep. 2014. “How Dependent is Growth from
Primary Energy? Output Energy Elasticity in 50 Countries.” Paris
School of Economics. Working Paper.
Gordon Robert J. 2012. “Is Economic Growth Over? Faltering Innovation
Confronts the Six Headwinds.” NBER Working Paper No. 18315.
Greenberg Leon. 1964. Productivity and Technological Change in the
Petroleum Refining Industry. U.S. Senate, Washington: Select
Committee on Small Business.
Griliches Zvi. 1994. “Productivity, R and D, and the Data Constraint.”
American Economic Review. 84, no. 1:1-23.
Hardman H, E.I. Cole EI. 1960. Papermaking Practice. Manchester.

Jones Charles. 1995. “R & D-Based Models of Economic Growth.” Journal
of Political Economy. 103, no. 4:759-84.
Jorgenson D.W. 1981. “The Role of Energy in Productivity Growth.” in:
Kendrick J.W, ed. International Comparisons of Productivity and
Causes of the Slowdown. Cambridge MA: MIT Press.
Jorgenson D.W. 1983. “Energy Prices and Productivity Growth.” in: Schurr
S, et al., eds. Energy, Productivity, and Economic Growth. Cambridge,
MA: Oelgeschlager, Gunn, and Hain.
Kummel R. J. Henn J, and D. Lindenberger. 2002. “Capital, Labor, Energy
and Creativity: Modeling Innovation Diffusion.” Structural Change and
Economic Dynamics. 13:415-33.
Kummel Reiner. 1982. “The Impact of Energy on Industrial Growth.”
Energy. 7, no.2: 189-201.
Laing Graham A. 1933. Towards Technocracy. New York: Angelus Press.
OECD. 1989. Flows and Stocks of Fixed Capital 1964-1989. Paris: OECD
Statistics Directorate.
Oliner Stephen D. and Daniel E. Sichel. 2002. “Information Technology
and Productivity: Where are We Now and Where Are We Going?”
Federal Reserve Board.
Olson Mancur. 1988. “The Productivity Slowdown, The Oil Shock and the
Real Cycle.” Journal of Economic Perspectives. 2, no. 4:43-69.
Patterson David A, and John Hennessy John. 2005. Computer Organization
and Design: The Hardware/Software Interface. Morgan Kaufmann.
Rosenberg Nathan. 1972. Technology and American Economic Growth.
Armonk, NY: M.E. Sharpe.
Rosenberg N. 1983. “The Effects of Energy Supply Characteristics on
Technology and Economic Growth.” in: Schurr S, et al., eds. Energy,
Productivity and Economic Growth. Cambridge, MA: Oelgeschager,
Gunn and Hain.
Sonenblum Sidney. 1990. “Electricity and Productivity Growth in
Manufacturing.” In: Schur S.H. et al., eds. Electricity in the American
Economy, Agents of Technological Progress. Westport, CT: Greenwood
Press.
Stresing R., D. Lindenberger and R. Kümmel. 2008. “Cointegration of
Output, Capital, Labor, and Energy.” European Physics Journal.
U.K. Committee on Factories Bill. 1832. “Factories Bill”. Hansard House
of Commons Debates. 1832.
U.S. Department of Commerce, Annual Survey of Manufactures, (Washington,
D.C.: Bureau of the Census, various years).
U.S. Department of Commerce. 1975. Historical Statistics of the U.S.:

Colonial Times to 1970. Washington, DC.: Bureau of the Census;
Bicentennial Edition.
Washington, DC.: Bureau of Economic Analysis.
U.S. Department of Labor. 1966. Technical Trends in Major North
American Industries. Department of Labor Statistics, No. 1474,
Washington, DC.
U.S. Department of Labor. 1967. Productivity and Technological Change
in the Petroleum Refining Industry. Washington, DC: Department of
Labor Statistics.
United Nations. 1984. Industrial Statistics Yearbook. New York, United
Nations.
9
A KINETICS-BASED APPROACH
TO PRODUCTION:
THEORY AND EVIDENCE*
Abstract
Production theory has, over the course of the past two centuries, been
besieged by criticism, ranging from its weak fundamentals to its lack of
coherence with the physical sciences. Yet, no real alternatives have emerged
with the result that neoclassical production theory stands today, much as it
did over a century ago. This chapter presents a consilient theory of
production, one that is grounded in classical mechanics, is empirically
validated, and sheds light on myriad productivity-related phenomena.
Specifically, a two-tiered approach to modeling production is proposed. In
the first tier, a kinetics-based theory is developed where output is an
increasing function of energy consumption, in keeping with basic physics.
In the second tier, the organization of energy-based material processes (the
first tier) is modeled. The resulting model is estimated using U.S.
manufacturing two-digit SIC data from 1947 to 1989.
9.1 Introduction
Economics is, by definition, the science of wealth, a fact that places
wealth at the center of the analysis. Understanding wealth is, as such, the
key to, and basis of economics as an intellectual endeavor. Getting wealth
right, it therefore follows, is a sine qua non condition for success in all other
sub-fields (e.g. labor economics, macroeconomics, industrial organization).
Unfortunately, despite over two centuries of effort, wealth and its creation
remain a challenge, as evidenced by the many puzzles and paradoxes that
characterize the literature, including the decades-old productivity slowdown
and the current information paradox.
This raises the question of scientific validity, namely when is a hypothetical
model of wealth valid or correct? Clearly, it can be internally valid (i.e. to
the profession) if it is accepted by the majority. Take, for example, classical
166 9. A Kinetics-Based Approach to Production
production theory according to which output is an increasing function of

labor, which for decades was internally validated/valid/correct. It wasn’t
however until its shortcomings, specifically regarding the role of capital in
the creation of wealth (mid-19th century), were identified that it was
abandoned and replaced by the new-classical or neoclassical approach.
There is also external validity. That is, to what extent is production
theory valid outside of economics? After all, economists are not the only
scholars/intellectuals that study the wealth-creating process. As it turns out,
the question of external validity came to a head in the early 20th century
when a growing chorus of scholars openly doubted the validity of
neoclassical production theory. Among these were Nobel prize laureate
Frederick Soddy, Thorstein Veblen, Howard Scott, Walter Rautenstrauch,
Frederick G. Tryon, and Woodlief Thomas. In general, they were either
engineers or scientists, schooled in basic mechanics and thermodynamics,
who felt that production theory was woefully inadequate—not to mention
incompleteas it abstracted from the key input in all known material
processes, namely energy/force.
This oversight came back to haunt the profession in the 1970s when the
price of petroleum and other forms of energy quadrupled in price. Not
having heeded earlier advice, the profession found itself unable to analyze
the effect of higher prices on output, employment and productivity. In 1975,
Ernst Berndt and David Wood responded with the KLEMS (capital, labor,
energy, materials, services) framework, which consisted of increasing the
dimensionality of the basic neoclassical model, the result of which was three
new output elasticities, namely the energy, the materials, and the services
output elasticities. Assuming competitive factor markets, they went on to
show, surprisingly, that energy, the cornerstone of all material processes,
and the only source of work (mechanics), was relatively unimportant in
generating wealth. Rather, machines and equipment and labor were the key
productive inputs.
This highlights the current dilemma in economics, namely that while
being internally valid, formalizations like KLEMS are increasingly under
attack externally. Take, for example, the challenge posed by the workerless
factory. According to neoclassical production theory, no output should
result as the labor input is essentially zero. Yet, output is produced,
oftentimes more than previously, thus putting the onus on capital, or
machinery and equipment. However, according to classical mechanics,
tools, not being sources of energy, are not physically productive, leaving us
with a conundrum, namely that the workless factory should not be
productiveyet it is.
Which brings us to the purpose of this chapter, namely to provide a

model of wealth that is internally and externally valid. To this end, we begin
by examining non-mainstream contributions, from Frederick Soddy to
Nicholas Georgescu-Roegen. This will be followed by a series of engineering-
inspired models that capture the material aspects of the wealth-creating
process, culminating in what we refer to as the kinetics-based theory of
production in which the laws of kinetics are integrated into functional
representations of production processes. As internal and external validity is,
in large measure, based on the predictive power of the model, we estimate
our model using aggregate and sectoral U.S. manufacturing data.
9.2 Non-Mainstream Contributions

It is often thought or believed that of all the non-mainstream economists,
there was none greater than German political economist Karl Marx. After
all, he is credited with single-handedly changing the course of political
economy, especially distribution theory. A careful reading of the first seven
chapters of Das Kapital, published in 1867, reveals what is a characteristically
classical approach to wealth, putting labor at the core of production and
hence of wealth. Like his classical forebearers, he held labor to be at the
center of all wealth creation. However, if one takes the time to read Das
Kapital from cover to cover (which few do) in Chapter 15 one discovers a
more compelling description of wealth creation in the age of machinery, one
that is based on classical mechanics.
Mathematicians and mechanicians, and in this they are followed by a few

English economists, call a tool a simple machine, and a machine a complex
tool. They see no essential difference between them, and even give the name
of machine to the simple mechanical powers, the lever, the inclined plane,
the screw, the wedge, etc.
As a matter of fact, every machine is a combination of those simple
powers, no matter how they may be disguised. From the economic
standpoint this explanation is worth nothing, because the historical element
is wanting. Another explanation of the difference between tool and machine
is that in the case of a tool, man is the motive power, while the motive power
of a machine is something different from man, as, for instance, an animal,
water, wind, and so on. According to this, a plough drawn by oxen, which
is a contrivance common to the most different epochs, would be a machine,
while Claussens circular loom, which, worked by a single labourer, weaves
96, 000 picks per minute, would be a mere tool. Nay, this very loom, though
a tool when worked by hand, would, if worked by steam, be a machine. And
since the application of animal power is one of man’s earliest inventions,
production by machinery would have preceded production by handicrafts.
When in 1735, John Wyatt brought out his spinning machine, and began the
industrial revolution of the 18th century, not a word did he say about an ass
driving it instead of a man, and yet this part fell to the ass. He described it
as a machine to spin without fingers.
All fully developed machinery consists of three essentially different
parts, the motor mechanism, the transmitting mechanism, and finally the tool
or working machine. The motor mechanism is that which puts the whole in
motion. It either generates its own motive power, like the steam-engine, the
caloric engine, the electromagnetic machine, etc., or it receives its impulse
from some already existing natural force, like the water-wheel from a head
of water, the wind-mill from wind, etc. The transmitting mechanism,
composed of fly-wheels, shafting, toothed wheels, pullies, straps, ropes,
bands, pinions, and gearing of the most varied kinds, regulates the motion,
changes its form where necessary, as for instance, from linear to circular,
and divides and distributes it among the working machines. These two first
parts of the whole mechanism are there, solely for putting the working
machines in motion, by means of which motion the subject of labour is
seized upon and modified as desired. The tool or working machine is that
part of the machinery with which the industrial revolution of the 18th
century started. And to this day it constantly serves as such a starting-point,
whenever a handicraft, or a manufacture, is turned into an industry carried
on by machinery. (Marx 1867, Chapter 15).
Clearly, there was more to Marx’s thought than the simple labor theory
of value. In fact, one could argue that he was well aware of classical
mechanics and the role of force in material processes, not to mention the
role of tools in material processes.
Perhaps the most influential of 19th-century iconoclastsin large part,
much in spite of himselfwas William Stanley Jevons, the father of
neoclassical production theory. In the The Theory of Political Economy
published in 1874, he outlined what was to become neoclassical production
theory, namely that wealth is an increasing, continuous, twice-differentiable
function of homogenous labor and capital. A lesser known, but equally
important contribution, of his was The Coal Question: An Inquiry Concerning
the Progress of the Nation, and the Probable Exhaustion of Our Coal-
Mines, published in 1865 in which he addressed the question of Great
Britain’s dwindling coal reserves. In the opening salvo, he declared:
Day by day it becomes more evident that the Coal we happily possess in
excellent quality and abundance is the mainspring of modern material
civilization. As the source of fire, it is the source at once of mechanical
motion and of chemical change. Accordingly, it is the chief agent in almost
every improvement or discovery in the arts which the present age brings
forth. It is to us indispensable for domestic purposes, and it has of late years
been found to yield a series of organic substances, which puzzle us by their

complexity, please us by their beautiful colours, and serve us by their
various utility.
And as the source especially of steam and iron, coal is all powerful. This
age has been called the Iron Age, and it is true that iron is the material of
most great novelties. By its strength, endurance, and wide range of qualities,
this metal is fitted to be the fulcrum and lever of great works, while steam is
the motive power. But coal alone can command in sufficient abundance
either the iron or the steam; and coal, therefore, commands this agethe Age
of Coal.
Coal in truth stands not beside but entirely above all other commodities.
It is the material energy of the countrythe universal aidthe factor in
everything we do. With coal almost any feat is possible or easy; without it
we are thrown back into the laborious poverty of early times. (Jevons 1865,
xi)
Paradoxically, some nine years later (i.e. in 1874), coal or the energy
input had disappeared completely from what is largely considered to be his
magnum opus, namely The Theory of Political Economy, where capital is
included in the production function and, more importantly, is assumed to be
physically productive. In short, both labor and capital were assumed to by
physically production and more importantly, were substitutable. One could
argue that internal validity (i.e. vis-à-vis the debate over the role of capital
in wealth) is what prevented Jevons from incorporating energy into the
corpus of neoclassical analysis.
Perhaps the greatest of British iconoclasts was Nobel-prize laureate
chemist Frederick Soddy, who after his pioneering work with Ernest
Rutherford on atomic transmutation turned his attention to economics,
largely in response to the alleged “misspecification” of production theory,
more to the point, to the absence of energy from the analysis. The gist of his
critique can be found in the following allegory:
At the risk of being redundant, let me illustrate what I mean by the question,
How do men live? by asking what makes a railway train go. In one sense or
another the credit for the achievement may be claimed by the so-called
engine-driver, the guard, the signalman, the manager, the capitalist, or share-
holder, or, again, by the scientific pioneers who discovered the nature of fire,
by the inventors who harnessed it, by labour which built the railway and the
train. The fact remains than all of them by their united efforts could not drive
the train. The real engine-driver is the coal. So, in the present state of
science, the answer to the question how men live, or how anything lives, or
how inanimate nature lives, in the sense in which we speak of the life of a
waterfall or of any other manifestation of continued liveliness, is, with few
and unimportant exceptions, By sunshine. Switch off the sun and a world
of by far the greater part of the life of inanimate nature. The volcanoes, as
now, might occasionally erupt, the tides would ebb and flow on an otherwise
stagnant ocean, and the newly discovered phenomena of radioactivity would
persist. But it is sunshine which provides the power not only of the winds
and waters but also of every form of life yet known. The starting point of
Cartesian economics is thus the well-known laws of the conservation and
transformation of energy, usually referred to as the first and second laws of
thermodynamics. (Soddy 1924, xi)
In short, according to Soddy, energy was the cornerstone of all human

activity, including production. Labor, capital, information, technology etc.
are all accessory inputs, necessary for but not the actual source of wealth.
Despite much promise, the proposed Cartesian economics, based on the
laws of basic physics (mechanics and thermodynamics) failed to make
inroads into mainstream economics.
As it turned out, the torch of “production theory” iconoclasticism would
soon cross the ocean, surfacing in North America. However, the resulting
strain would be less ideological and more practical. Specifically, the late
19th-century conversion from belting and shafting to electric unit drive (i.e.
individual electrical motors integrated in machinery) had witnessed a non-
negligible increase in energy use in U.S. manufacturing, one that produced
a sizeable increase in output against a backdrop of lower capital
expenditure/stock. The latter owed to the fact that electric generators and
motors were less costly than the elaborate shafting and belting they
replaced.
With the passage of time, it became increasingly obvious that energy use
in general, and electric power in particular, had become the driving force,
increasing output and wealth.
F.G. Tryon of the Institute of Economics (Brookings Institution) was
among the first to point to the incongruity between production processes as
modeled in economics and those he observed in early 20th century America.

than it has yet received by economists. The industrial position of a nation
may be gauged by its use of power. The great advance in material standards
of life in the last century was made possible by an enormous increase in the
next century turns perhaps more than on anything else on making energy
energy.
These considerations have prompted the Institute of Economics to
undertake a reconnaissance in the field of power as a factor of production.
One of the first problems uncovered has been the need of a long-time index
of power, comparable with the indices of employment, of the volume of
production and trade, of monetary phenomena, that will trace the growth of
the factor of power in our national development (Tryon 1927, 281).
One year later (i.e. in 1928), Woodlief Thomas of the Division of

Research and Statistics of the Federal Reserve Board, published an article
in the American Economic Review entitled “The Economic Significance of
the Increased Efficiency of American Industry, in which he attributed the
“striking changes in American industry to power-related developments:
Large-scale production is dependent upon the machine process, and the

increasing use of machinery and power and labor-saving devices has
accompanied the growth in size of productive units. The growing use of
power in manufacturing, for example, is reflected in the increase in
horsepower of installed prime movers. This does not tell the whole story,
moreover, for owing to increased use of electricity, the type of power used
is now more efficient—requiring less fuel and labor for its production. Out
of a total installed horsepower in factories of thirty-six million in 1925,
twenty-six million or 72 per cent was transmitted to machines by means of
electric motors, as compared with 55 percent in 1919, 30 per cent in 1909,
and only 2 per cent in 1899. Between 1899 and 1925 horsepower per person
employed in factories increased by 90 percent and horsepower per unit of
product increased by 30 percent. Power has been substituted for labor not
only through machines of production but also in the form of automatic
conveying and loading devices. (Thomas 1928, 130)
In little time, this incongruity reached academia, specifically Columbia

University where a group of engineers, known as the Technocracy Alliance
outrightly rejected mainstream approaches to understanding wealth (essentially
neoclassical production theory), arguing that they ignored mechanics,
thermodynamics, process engineering and with the then state of the art
regarding material processes in general.
With the passage of time, it became increasingly obvious that while
neoclassical production theory continued to hold sway within the profession
(i.e. internally valid), it came under increasing scrutiny outside (i.e. external
validity). However, as energy consumption and output increased monotonically
throughout the post-WWII period, there was little internal or external
pressure on production theory. Growth theorists (e.g. Robert Solow, Moses
Abramovitz, Edward Denison, Zvi Griliches) contented themselves with the
view that all non-labor and non-capital sources of growth (e.g. energy)
could be included in a portmanteau variable, namely the Solow Residual. In
essence, all that could not be explained by labor and capital, two non-
physically productive inputs, would be included in the residual.
The problem with residuals in growth theory, as it turned out, was not
with their existence, but rather with their disappearance. And this is what
happened in the 1970s when the Solow residual suddenly disappeared,
ushering in the productivity and growth slowdown, a slowdown that has not
been since been reversed. Among the alleged causes was the 1973 OPEC
oil embargo. Unfortunately, because energy was absent from production
theory, it was unclear if and how higher oil prices could impact GDP. The
response was not long in coming. In 1975, Ernst Berndt and David Wood
proposed the KLEMS approach to study the effects of oil prices shocksand
energy price shocks in generalon the economy. The upshot was damning
of the energy input: only 4-5 percent of output could be attributed to it.
Hence, the OPEC oil shock could only lead to a downturn in the presence
of factor input complementarities, specifically the capital-energy
complementarityor substitutability.
At roughly the same time, the internal validity of standard neoclassical
production theory came under fire from a Romanian economist, Nicholas
Georgescu-Roegen, who argued that like all other material processes in the
universe, production can and indeed should be seen as entropy increasing,
hence as an irreversible process. His principal target was the standard
neoclassical approach to wealth creation which, in its simple version, was
reversible.
9.3 The Critics: Big on Principles, Small on Specifics

As has been shown, the majority of what we refer to as the non-
mainstream critics lamented the absence of energy from production theory
and from economics in general. However, unfortunately, they stopped short
of providing workable/credible alternatives. For example, no one proposed
an alternative theory of material processes based on energyessentially on
the principles of classical mechanics. Similarly, no one took the time to
analyze or examine the underlying mechanics, namely of the exact way in
which energy is combined with tools and conventionally-defined workers to
produce wealth. At the very best, we have the KLEMS approach where energy
is simply added, along with materials and services, to capital and labor
without much fore- or after-thought to the list of substitutable/complementary
factor inputs. Just how energy interacts with capital and labor was left
unspecified, other than invoking a sense of complementarity.
In the next section, we present a physics-based theory of material
processes, where this interaction takes the form of either machine kinetics
and/or chemical kinetics. That is, in displacement-based material processes
(translational and rotational), more energy per unit of capital will affect
output via the law of kinetics in general, and via machine speed in particular.
In non-displacement-based material processes, more energy per unit of
capital will lead to greater operating temperatures, greater material
breakdown and higher costs (via Arrhenius’ Law). In the remainder of this
section, we examine the literature—however scant—on the role of
speed/kinetics in production and in productivity growth, starting with
William Longston’s testimony on working conditions in early 19th-century
textiles industry before the Committee on the Factories Bill.
9397.Is the intensity of application and of labour altered, either against or in

favour of the operative and of the children employed in mills and
factories?I was a great number of years out of any factory, but those who
were my acquaintances during my boyhood have often conversed with me,
and they very frequently say that it cannot be less than double in intensity
and exertion of physical application.
9398. State why you believe that the labour of those employed has doubled
since the first introduction and use of cotton machinery, or at least since you
first knew it? The reason why I believe so is from some calculations which
I have been obliged to make, and by my own observation during the time I
was manager of a mill in 1830 and 1831, when I had some of the same
operations under my own observation.
9399. Have you any objection to put in those calculations Certainly not. [The
following document was then put in and read.]
9400. It appears by this document that the work done is very greatly
increased between the years 1810 and 1832; has the machinery been so
altered as to produce that amazing difference, or does it result from
accelerating the speed of the machinery?It is from accelerating the speed
generally; and another cause is, that more and more exertion is required from
the individual working at the machine; these are the two causes. (Committee
on Factories Bill 1832, 430)
The General Electric Company, as early as 1937, pointed to increased

machine speedcontrolled machine speedas one of the defining features of
modernity and of productivity growth. Under the title of Today is the Day
of Speed, it maintained that:
Our transportation systems, our industrial processes, our factory machineryall

these have felt the magic hand of speed, controlled speed that has given us
more things to enjoy and more time in which to enjoy them; that has
produced more goods for more people at less cost and that has created a
better standard of living for the average man. These are the benefits of ever-
increasing speed and accurate control.
Henry Ford, in his description of mass production, pointed to machine

speed as the key element in his new technique:
Mass production is not merely quantity production for this may be had with
none of the requisites of mass production. Nor is it merely machine
production, which also may exist without a resemblance to mass production.
Mass production is the focusing upon a manufacturing project of the
principles of power, accuracy, economy, system, continuity and speed.
(Ford 1926, 821).
Graham Laing, in a book entitled Towards Technocracy, pointed to

machine speed as a key factor in the unprecedented productivity gains in the
1920s and 1930s.
Industrial processes have been speeded up, new inventions are being added
to manufacturing, new economies of personnel and of management have
been made in industry. The 1929 production can undoubtedly be achieved
with thousands, and probably millions, fewer workers. (Laing 133, 23)
Alfred Chandler, in his definitive work on the early 20th century, echoed
this view, generalizing it to the U.S. economy as a whole.
In modern mass production, as in modern mass distribution and modern

transportation and communications, economies resulted more from speed
than from size. It was not the size of a manufacturing establishment in terms
of the number of workers and the amount and value of productive
equipment, but the velocity of throughput and the resulting increase in
volume that permitted economies that lowered costs and increased output
per worker and per machine. (Chandler 1977, 244)
Sidney Sonenblum, in his work on electrification and productivity

growth in manufacturing, also pointed to speed, or the accelerating the rate
of throughput as a key element in productivity growth:
During these years, the focus of managerial attention shifted from enlarging
the scale of operations to increasing operating efficiency by speeding up the
rate of throughput in the plant. High priority was assigned to modifications
of factory design and layout in order to better integrate worker and machine
tasks. Advances in the electrification of machine drive were indispensable
to the realization of these new objectives and may, indeed, have served to
stimulate the new managerial perspectives that emerged. (Sonenblum 1990,
291)
Running through each of these accounts of speed and its role in

productivity is the notion of control. General Electric referred to “accurate
control,” while Ford referred to “accuracy, system and continuity.” Hence,
speed and control are to be understood as complementary inputs. Theoretically,
control can be defined in terms of four functions (i) constancy of speed (ii)
minimal breakdown-related downtime, (iii) sub-process coordination and
(iv) machine programming. Put differently, the better able is the
firm/engineer at maintaining a constant speed, the greater the output. The
same holds for machine breakdown. As not all machines/sub-processes
operate at the same speed, it is essentialin order to avoid bottlenecksto
coordinate speeds throughout the plant. Lastly, because firms typically
produce many different goods/models of goods with the same machinery, it
stands to reason that more efficient machine programming will reduce
downtime and hence, increase the average operating speed per period of
time.
According to Warren T. Devine, control technologies underwent a series
of innovations in the 20th century that were instrumental in increasing
machine speed. In a nutshell, hydraulic drive and control mechanisms gave
way to servomechanisms, which in the 1950s, gave way to numerical
control, which reduced machine downtime considerably. Not only would
productivity rise as the result of greater machine speeds (i.e. owing to
greater energy use), the machines themselves would be more fully utilized
in any given time period. He noted:
Numerically-controlled machinery had a number of advantages over

conventional manually controlled machinery. The time required to get a
newly designed part into productionthe machine setup timewas
sometimes as much as 65 to 75 percent less with numerical control. (Devine
1990, 50)
In a 1966 report entitled “Technological Trends in Major American

Industries,” the U.S. Department of Labor pointed to control technologies
in the form of the computerization of data processing and increased
mechanization (read: faster speeds) as the leading innovations of the post-
WWII period. Under the heading of “Trend Toward Increased
Mechanization,” it pointed out that:
Improvements in machinery that do not involve a drastic departure from

conventional design will continue to be an important factor in raising
productivity in many industries. Faster operation, larger size, automatic
loading and unloading devices and automatic lubrification significantly
reduce the amount of labor required per unit of output. The integration of a
number of separate operations into one large specialized machine which
performs a long cycle of operations with a minimum of intervention by the

machine tender constitutes a more advanced type of mechanization.
Examples of greater mechanization are found in many industries: faster
textile machine speeds with larger packages of stock; continuous steel
casting machines that require one-half the number of steps of traditional
ingot casting; machinery in meatpacking for continuous production of
frankfurters; tape controlled line casting machines in printing; faster, larger
capacity machines in tire and tube manufacture. Other examples are
mechanical “lumberjacks” to cut trees in the lumber industry; larger capacity
stripping equipment in copper mining; greater use of continuous coal mining
machines; and a machine that combines a number of operations in shirt
making.
As fabricating operations become highly mechanized, new ways are
sought to achieve labor savings in moving goods and materials from one
plant operation to the next. Mechanized material handling often is
introduced or improved to utilize more fully the high speed and large
capacity of modernizing fabricating equipment. (U.S. Department of Labor
1966, 5)
Clearly, machine speed and its relationship to energy use, specifically,

to electric power use appear to have been key elements in raising overall
productivity. In the next section, we attempt to formalize this both in terms
of a theoretically-consilient and empirically consistent model of output. By
theoretically consilient, it should be understood the property of being
consistent or in-keeping with the principles of related fields such as classical
mechanics, thermodynamics and process engineering. By empirically
consistent, it should be understood the property of being consistent with the
datathat is, is confirmed by the data. Put differently, models of production
should at the very least be able to confirm the relevant underlying laws of
physics in material processes.
9.4 Model
In keeping with Beaudreau (1998) who classified inputs in terms of two
main categories, namely broadly-defined energy and organization, we
propose a two-tiered approach to understanding production. The first tier is
purely physical and is governed by the laws of physics, specifically, the laws
of machine and chemical kinetics. Neither tools and/or equipment nor
conventionally-defined labor (supervisors) is physically productive, and
hence is parametric to this tier. Tier I is universal in its application and
reach, accessible to industrial engineers, to physicists as well as to
economists and production specialists, thus ensuring both internal and
external validity. The second tier, Tier II, is the organization tier which
focuses on the definition and supervision of first tier material processes. It

focuses on defining and the overseeingin short, the organizationof
machines. It focuses on what Alfred Marshall referred to, in 1890, as
machine operatives. It is important to point out that such operatives are not
a source of power/energy and hence are not physically productive; hence
they cannot be substituted for primary power. Traditional factor substitution
is, as such, rendered unfeasible and theoretically impossible across our two
tiers. Machinery and equipment, not being sources of power, cannot be
substituted for energy.
Table 9.1: Manufacturing Processes and Corresponding Kinetics
Type Kinetic Law Examples Acceleration

Mechanical- Translational Material Higher Speed
Translational Kinetics Handling
KE = 1/2mv2 Transportation
Mechanical- Rotational Grinding, Higher Speed
Rotational Kinetics Shaping,
KE = 1/2mi2 Assembling,
Reducing
Chemical/Thermal Chemical Refining, Higher
Kinetics Electrolysis Temperature
k(t) = AeíEa/RT Cracking Higher Voltage
9.4.1 Tier I: Machine and Chemical Kinetics at the Sub-Process

and Plant Levels
We begin by defining the firm as a series of n sub-processes which can
be of two types, namely mechanical (nm) and chemical (nc) (Equation 9.1).
Mechanical sub-processes are governed by the laws of translational and
rotational kinetics, while chemical sub-processes are governed by the laws
of chemical kinetics. Included in the mechanical sub-processes category are
the material-handling processes (pumps, conveyor belts, etc.) between the
various sub-processes. That is, in sequential production processes, the
output of one sub-process becomes the input for the next sub-process.
‫ݕڿ݊݅݉ = )ݐ(ݕ‬௜௠ (‫ = ݅׊)ݐ‬1, 2, 3 … . ݊௠ ; ‫ݕ‬௜௖ (‫ = ݅׊)ݐ‬1, 2, 3 … . ݊௖ ‫ ۀ‬9.1
Mechanical sub-processes are assumed to be governed by the laws of
basic machine kinetics (translational and rotational), according to which
output yi(t) is an increasing function of operating speed(s) si(t) and the
machines themselves, denoted by ki(t). The measured individual n machine
speeds, defined as the machine rates of output per unit time, is governed by
translational/rotational kinetics (KE = 1/2mv2, where KE=kinetic energy,
m=mass, and v=velocity). As a result, abstracting from the question of

machine downtime, the effect of an increase in energy use on operating
speed and hence on output, will be determined by translational/rotational
kinetics. However, in the presence of machine downtime, the statistical
relationship between energy (KE) and output (y(t)) will be described by
Equation 9.2 where ߛ௜௠ (‫ = ݅׊)ݐ‬1, 2, 3 … . , ݊௠ , where [0 ൑ ߛ௜௠ (‫ )ݐ‬൑ 0.5] is
an adjustment factor. Translational kinetics defines operating speed (e.g.
revolutions per minute, feet per minute) for a given mass (m) and kinetic
energy (KE). Machine downtime in turn can be the result of (i) maintenance
(ii) idleness due to lack of coordination between sub-processes and (iii)
retooling for a new product. To capture both effects, we define
measured/statistical operating speed in terms of Equation 9.2 where ߛ௜௠ (‫)ݐ‬
captures sub-process i downtime in period t. The closer ߛ௜௠ (‫ )ݐ‬is to 0.5, the
higher is average downtime per t, and hence the higher will be the resulting
output elasticity.
೘ (௧)
‫ݏ‬௜௠ (‫)ݐ‬଴.ହାఊ೔ ‫ ݅׊‬െ 1, 2, 3, … . , ݊௠ 9.2
Thermal and chemical sub-processes can be described by the laws of

chemical kinetics (e.g. Arrhenius Equation) according to which chemical
reaction rates are an increasing function of temperature ܶ௜௖ which in turn is
a function of energy consumption ݁௜௖ , ‫ܧ‬௔௜ ௖
being activation energy and R
being the activation energy constant (8.314). As in the previous case,
increasing T by increasing energy use, will result in a higher reaction rate,
which in our analysis is akin to an increase in machine speed.
೎
‫ݏ‬௜௖ (‫ = )ݐ‬ൣ‫ܣ‬௖௜ ݁‫ି݌ݔ‬ாೌ/ோ்(௘೔ ) ‫ = ݅׊‬1, 2, 3, … . , ݊௖ ൧ 9.3
As was the case with mechanical sub-processes, the effect of increased

energy use on operating speed (reaction rates) will be determined by two
factors, namely chemical kinetics (Arrhenius’ Law), and by machine
downtime. The more downtime there is, the more linear will be the
relationship, and vice-versa. Equations 9.2 and 9.3 describe the two types
of sub-process production functions, the first for rotational and
translational-based processes, and the second for chemical and thermal-
based processes.
We assume that ‫ݏ‬௜௠ (‫ )ݐ‬and ‫ݏ‬௜௖ (‫ )ݐ‬are bounded from above. In other
words, for each subprocess, there exists a maximum speed beyond which it
is impossible to go. It can be defined as a combination of (i)
translational/rotational chemical kinetics (ii) material tolerances and (iii)
average downtime (owing to maintenance, retooling, etc.).
Equations 9.2 and 9.3 define the sub-process/process technologies

available to firms at time t. What is immediately obvious is that the overall
speed of productionand hence, overall productivity of the processwill
depend on individual sub-process machine speeds, more specifically on the
lowest sub-process speed. Further, they illustrate a number of important
phenomena, notably the relationship between energy use (machine and
chemical kinetics), and secondly the corresponding process speeds/reaction
rates. In the case of translational and rotational mechanical processes, it
increases quadraticallythat is, doubling speed will quadruple energy use.
For example, doubling the speed of a conveyor belt will quadruple energy
consumption.
It is our view that these equations provide a formalization of the West’s
experience with energy-based technologies in the 19th and 20th centuries
(i.e. the 1st and 2nd Industrial Revolutions). As indicated by the quotations
provided in Section 9.2, rising productivity was associated with greater
energy use per unit of capital, the latter operating on the former via
increased operating speeds. As Alfred Chandler and numerous others
pointed out, the productivity gains registered in the late 19th/early 20th
centuries owed in large measure to rising machine speeds (Jerome 1934,
Chandler 1977, Devine 1983). As the experience at Ford Motor showed, the
use of electric-powered conveyor belts was the key components in the
success of the assembly line.
9.4.2 Tier II: Organization and the Demand for Supervision

and Tools
While early models of production (i.e. classic) viewed labor as being
physically productive (e.g. classical theory of value), by the end of the
century, most mainstream economists, including neoclassical writers,
viewed labor for what it had become, namely what Alfred Marshall referred
to as “machine operatives.” Take, for example, the following excerpt from
Alfred Marshall’s Principles of Economics where he refers to labor as
“managers”:
of more than half the working men even in such a country as England . . . in
with much less toil for themselves . . . . Nothing could be more narrow or
many times as much work in the course of a day as the old hand loom did;
than his did. (Marshall 1890, 218)
This change was echoed in official statistics. For example, the U.K.
Board of Trade, in its Censuses of Production, no longer referred to workers
or production workers, but rather to operatives. The concept of labor
productivity, it therefore follows, took on a new meaning, specifically as a
measure of output per machine manager or operative. Implicitly, labor was
not physically responsible for/involved in generating wealth, but rather was
responsible for overseeing/managing the corresponding machines. To
capture this, we model the demand for supervision/machine operatives as a
function of output, specifically desired output. The greater the desired or
targeted level of output on the part of firms, the greater the demand for
supervision.
As machine operatives are involved in all nm and nc subprocesses defined
by Equation 9.1, it stands to reason that the demand for supervision will
depend on a number of factors, from the individual sub-process supervision
technology, to average overall machine speed, to the overall scale of
operation. For example, if the firm automates a given sub-process, then it
would stand to reason that the demand for supervision per unit of output
would fall as a result. The same would hold for an increase in machine
speed. Only with an increase in the overall scale of operations (i.e. all sub-
processes are increased by the same factor) will the demand for supervision
per unit of output stay the same.
݊(‫)ݐ(ݏ[ߙ = )ݐ‬, ߤ(‫)ݐ(ݕ])ݐ‬ 9.4
The firm-level demand for supervision can be formalized as Equation 4,

where nt refers to the number of overall machine operatives or supervisors,
and Į refers to the demand for supervision per unit y(t), which is a function
of s(t), average overall machine speed at time t, and ȝ(t), the average overall
level of automation at time t. As such, an increase in s(t), machine speed,
will result in an increase in y(t) per unit n(t), thus contributing to lowering
D. However, it bears reminding that such an increase owes in no measure to
labor’s intrinsic properties or productivity, but rather to greater machine
speed and more output per unit capital (i.e. machinery) and supervision.
Similarly, innovations in ICT-based technological change will affect
labor demand. Specifically, innovations in machine control technology will,
in general, reduce Į, thus reducing the demand for supervision per unit
outputin some cases, reducing it to zero (i.e. the case of total factory
automation). Measured output per machine operative will, consequently,
rise, again in no part due to the intrinsic value or contribution of

conventionally-defined labor.
݇(‫)ݐ(ݏ[ߚ = )ݐ‬, K(‫)ݐ(ݕ])ݐ‬ 9.5
Like the supervisory input, we view tools (capital) as an organizational

input, one that defines a given material process, but not one that is physically
productive. According to classical mechanics, tools provide mechanical
advantage, which is defined as the advantage gained by the use of a
mechanism in transmitting force; specifically, the ratio of the force that
performs the useful work of a machine to the force that is applied to the
machine. The demand for tools can, as such, be modeled analogously to that
of supervisionthat is, as a function of projected output, y(t). According to
Equation 9.5, the demand for tools is an increasing function of the latter
variable, with ȕ[s(t), Ș(t)] being the corresponding scaler. As can be seen,
the per unit output demand for tools (capital) is a decreasing function of
machine speed, and of Ș(t), the level of second-law efficiency (in short, the
productivity of energy). The more efficient is the energy input, the less
capital is required per unit output.
9.4.3 Aggregating Across Firms within an Industry/Sector

Aggregating across sub-processes within firms, and between firms
within a given industry/sector, we obtain Equations 9.6-9.10, which
describe aggregate industry/sector output in terms of the average speed of
the overall production processes S(t), as well as the supervisory N(t) and
capital requirements K(t) of the latter. Ceteris paribus, the greater is
aggregate energy use per unit K(t), the greater is the aggregate speed S(t),
and the greater is overall output (i.e. Y(t)). Furthermore, the greater is the
rate at which firms in general are able to reduce average machine downtime
(i.e. ī(t)), the greater is machine speed per period t, and hence, the greater
is aggregate productivity and output. However, this makes for a lower
energy output elasticity as successive increases in energy will serve to
increase machine speed and not start up idle machines.
9.3.1 Tier I
Y(t) = S(t)K(t) 9.6
S(t) = E(t)0.5+ī(t) 9.7
Y(t) = E(t)0.5+ī(t)K(t) 9.8
Table 9.2: Production per Factor Indexes-PFI
Index Definition Parameters

Energy PFI y(t)/e(t) ei(t)0.5+Ȗi(t)ki(t)
Labor PFI y(t)/n(t) 1/Į[s(t), ȝ(t)]
Capital PFI y(t)/k(t) 1/ȕ[s(t), Ș(t)]
9.4.2 Tier II
N(t) = A[S(t), ȝ(t)]Y(t) 9.9
K(t) = B[S(t), K(t)]Y(t) 9.10
9.5 Production Activity Indicators

This approach to production has important implications for conventional
productivity indexes or indicators. Specifically, the conventional concepts
of labor and capital productivity are deemed to be theoretically invalid as
they assume (incorrectly) that labor and capital are physically productive,
when in fact they are organizational (Tier II) variables and, as argued earlier,
not physically productive. Theoretically, the only physically productive
factor is energy, implying that the only scientifically-legitimate “productivity
index” is that of energy. In light of this and the need for production
indicators, we present a new indicator, namely the production per factor
input index, or production per factor index for short (PFI).
It is important to point out what these indexes are and are not. First, they
are production per factor measures and not productivity measures, the
exception being the Energy PFI, which de facto measures the average
physical productivity of energy. The Labor PFI and Capital PFI are simple
measures of output per factor input, which vary over time according to
underlying technology parameters such as machine speed, information
technology and second-law efficiency.
For example, an increase in machine speed will, ceteris paribus, increase
the labor and capital PFIs. However, neither factor will have contributed to
the increase. As such, it cannot be maintained that either of the factors is
more productive. Rather, both are witnesses, of sorts, of greater output,
without being responsible for it.
9.6 Empirical Evidence

The tiered approach to understanding material processes has important
implications for the associated empirics. For example, traditionally, output
has been regressed against all factor inputs (e.g. the KLEMS method). This
approach is abandoned on the grounds that it is theoretically unjustifiable
(i.e. physically-productive versus organizational inputs) and serves to
confuse rather than illuminate. Labor is not and has not been physically
productive for over two centuries (starting with the introduction of the steam
engine). Capital has never been, nor will never be physically productive.
Neither can be substituted for each other as each fulfills an entirely different
function. Moreover, neither can be substituted for energy as neither is a
source of energy. Hence, for these and innumerable other reasons, we
proceed by (i) testing the predictions of machine and chemical kinetics in
economics and (ii) testing the derived input demand for organization
specifically conventionally-defined labor.
9.6.1 Tier I: Testing the Theory of Industrial Kinetics

This approach to modeling material processes differs from conventional
approaches in that capital (and labor) is assumed to be an organizational
input (Beaudreau 1998), providing the setting for what we consider to be
the most fundamental relationship in all material processes, namely of
energy transforming material inputs, creating wealth. Hence, in keeping
with classical mechanics, only energy is physically productive. As such,
output is an increasing function of machine speed, machines, and control
devices the latter affecting the utilization of machine utilization. The
underlying idea is that most machinery is used to accomplish numerous
tasks per period t, requiring retooling/downtime. Control devices, especially
reprogrammable control devices, reduce downtime, thus increasing
utilization rates and average machine speed per period t. Average, aggregate
machine speed is defined by Equation 9.6, where it is a function of energy
use and Ȗ(t), the rate of growth of machine utilization due to increasing use
of/advances in control technologies. It is important to point out that small
changes in utilization rates will have important effects on Ȗ(t) at time t for
the simple reason that these changes will be applied to all k(t), the entire
capital stock.
Table 9.3: Estimates of the Electricity-Use Output Elasticity-

Manufacturing
Method Source Country and Period Estimate (t-stat)

OLS Beaudreau U.S. (1950-1984) 0.5330 (10.791)
(1995)
OLS Beaudreau U.S. (1958-1984) 0.4483 (12.469)
(1998)
Germany (1962- 0.7474 (3.135)
1988)
Japan (1962-1988) 0.6055 (3.017)
LINEX Kummel, Henn Germany 0.64
and Lindenberger U.S. (1960-1993) 0.51
(2002)
Japan (1965-1992) 0.61
U.S.-Total (1960- 0.30
1993)
Germany-Total 0.44
(1960-1989)
Cointegration Stresing, Germany (1960- 0.517
Kummel 1989)
and Lindenberger Japan (1965-1992) 0.350
(2008)
U.S. (1960-1978) 0.663
In this section, we present previous aggregate estimates of the energy

use/electric power output elasticity (0.5+ī(t)) as well as new disaggregated
results. Equation 9.6 maintains that the output elasticity for energy use
should be in the 0.5 +Ȗ(t) range. Given that Ȗ(t) is variable, it stands to reason
that the energy output elasticity should vary across firms, industries and
countries. However, at the very least, a ten percent increase in energy use
should result in a five percent increase in aggregate machine speed and
hence, a five percent increase in output per period of time, assuming that
the corresponding maximum machine speed has not been reached.
Beaudreau (1995, 1998), Kummel, Lindenberger and Eichorn (1998)
provided directas opposed to indirect estimates of the energy output
elasticity in U.S., German and Japanese manufacturing. Recently, Giraud
and Kahraman (2014) provided estimates of the primary energy output
elasticity for 50 countries, reporting elasticities between 0.6 and 0.7.
Referring to Table 9.3 which presents electricity use output elasticities, we
see estimates in the range of 0.30 to 0.7474, with the average at 0.5303.
These estimates were obtained using a number of techniques, ranging from
OLS to cointegration models.
Using similar data, we estimated a simple three-factor input production

function for U.S., German, Japanese, Canadian, British, and Finnish
manufacturing.1 Three econometric specifications were employed, namely
(i) linear, (ii) log-linear and (iii) log differences.2 Ordinary least squares was
used in the case of the latter, while an OLS-AR(1) approach was used in the
case of the former. The results are presented in Tables 9.4-9.6, where we
see estimates of the EP (electric power use) coefficients in the predicted
range of 0.50+Ȗ. For example, in the linear case, estimates ranging from
0.4902 in the case of the U.S. to 0.9141 in the case of Britain were obtained.
The Canadian elasticity was an anomaly at 0.2451. As collinearity was
suspected, the tests were repeated without labor. The resulting electric
power use elasticities are 0.7379, 0.7889, and 0.8412, respectively for the
log-linear, differences in logs, and linear cases.
These results are consistent with the predictions of our model. More to
the point, they are consistent with the predictions of mechanical and
chemical kinetics. For the most part, the estimates of the energy output
elasticity are greater than the theoretical value of 0.50. The difference, we
believe, owes to, among other things, ī(t) which is country-specific,
continuous increases in energy efficiency, measurement errors, and the very
nature of the estimates. Specifically, in addition to capturing kinetics (per
unit of capital), they capture scale effects as the stock of capital was
increasing over time. An attempt was made at eliminating these “scale”
effects by regressing output per unit of capital on electric power per unit
capital and labor per unit capital. The results were however not significantly
different.
1 Output, labor and electricity use data were obtained from United Nations Industrial
Statistics Yearbook (1960-1988); capital data were obtained from OCED, Flows and
Stocks of Fixed Capital 1989. These data are available from the author upon request.
2We opted to estimate the output elasticities directly as opposed to indirectly. This
owed to a number of factors, including the nature of our work (i.e. estimating the
production function itself) and the belief that factor markets are not competitive,
especially the electricity market, making indirect estimation techniques inappropriate.
Table 9.4: OLS-AR(1) Estimates-Log Linear Specification
U.S. 1.1678 0.4902 0.7826 0.1034 0.9990
(1958-1984) (0.5342) (3.169) (0.9005) (5.136) ȡ=0.688
Germany -0.2128 0.6124 0.5587 -0.1086 0.9494
(1963-1988) (0.1961) (3.710) (0.5520) (3.2171) ȡ=0.557
Japan -3.1852 0.6970 0.5175 0.3789 0.9818
(1965-1988) (1.328) (2.514) (2.383) (0.9804) ȡ=0.885
Canada -3.6435 0.2451 0.2655 1.2673 0.9844
(1962-1988) (2.421) (0.8631) (1.216) (3.070) ȡ=0.660
Britain 1.2630 0.9141 -0.0972 -0.0870 0.9232
(1963-1988) (0.936) (5.762) (4.000) (0.7156) ȡ=0.502
Finland -0.4482 0.7081 0.3040 0.0583 0.9856
(1963-1988) (0.6075) (4.772) (1.702) (0.3133) ȡ=0.364
Table 9.5: OLS-AR(1) Estimates-Linear Specification
U.S. -64.7637 0.4554 0.0963 1.1141 0.9891
(1958-1984) (3.186) (4.009) (1.084) (5.417) ȡ=0.663
Germany -0.2128 0.6124 -0.1086 0.5587 0.9994
(1963-1988) (0.1961) (3.710) (0.5222) (3.217) ȡ=0.557
Japan -164.97 0.5618 0.2374 1.3221 0.9820
(1965-1988) (1.814) (2.983) (3.683) (1.643) ȡ=0.827
Canada -130.442 0.3426 0.1582 1.6709 0.9821
(1962-1988) (2.347) (1.210) (6.2092) (2.836) ȡ=0.651
Britain 17.1535 0.8570 -0.0289 -0.0459 0.911
(1963-1988) (0.3206) (5.605) (0.1008) (0.1720) ȡ=0.583
Finland -17.9512 0.57475 0.3122 0.1854 0.9918
(1963-1988) (0.8936) (6.931) (2.553) (0.9723) ȡ=0.294
Table 9.6: OLS-AR(1) Estimates-Log Differences Specification
U.S. 0.0447 0.3824 0.9839 0.9646 0.9179
(1958-1984) (3.168) (2.498) (2.9561) (6.499)
Germany -0.0016 0.6093 0.0650 0.4675 0.6682
(1963-1988) (0.1469) (3.682) (0.2197) (1.872)
Japan 0.0360 0.7176 0.3052 0.6291 0.6710
(1965-1988) (1.930) (2.942) (1.128) (1.228)
Canada 0.0572 0.0528 -0.9138 1.4123 0.715
(1962-1988) (2.391) (0.2072) (1.933) (4.244)
Britain -0.0118 0.7898 -0.2098 0.0891 0.6432
(1963-1988) (0.9165) (5.122) (0.4099) (0.5597)
Finland 0.0139 0.7066 -0.1171 -0.3518 0.48
(1963-1988) (0.371) (4.364) (0.1352) (0.8735)
We then tested the model using 2-digit SIC data for U.S. manufacturing
from 1947 to 1984. The data, in this case, were taken from the Annual
Surveys of Manufactures as well as the Census of Manufactures. The results
are presented in Table 9.7, where three sets of electric power output
elasticities are reported. Column 1 presents the relevant output elasticity
when both electric power consumption (purchased and generated) and
production workers were included as independent variables, while Column
2 presents the output elasticity when only electric power consumption was
used as the independent variable. Column 3 presents the output elasticity
when the dependent and independent variables were measured relative to
the level of production workers, the idea being that this would eliminate
cyclical biases/effects. As was the case with aggregate data, the output
elasticities were centered around the predicted-by-the-law-of-kinetics value
of 0.50. In fact, in the first case, the average output elasticity was 0.493.
9.6.2 2-Digit U.S. Manufacturing Sectoral Estimates of the Input

Demand Elasticity for Supervision
In this section, we present estimates of two demand input elasticities for

supervision, namely the demand input elasticity for supervision with regard
to output (y(t)) and secondly, that with regard to the energy inputin this
case, electric power consumption. These are defined as the percentage
increase in the demand for supervision (i.e. machine operatives) divided by
either the percentage increase in the level of output or the percentage
increase in the energy input at the 2-digit sectoral level (see Table 9.8).
Given the persistence of energy deepening in the form of a rising electric
power to machinery/equipment ratio throughout the period under study, it
would stand to reason that the elasticity with regard to energy would be
systematically less than that with regard to output. The estimates presented
in Table 9.6 confirm this. The demand for supervision per kwh was less than
the demand for supervision per unit of output in virtually all industries.
Table 9.7: 2-Digit SIC Industry Supervision Input Elasticities 1947-

1984
SIC Industry Elasticity-EP (t- Elasticity- R2

stat.) RVA (t-stat.)
20 Food and Kindred Products -0.049 (4.558) -0.071 (3.36) 0.244
21 Tobacco Products -0.700 (2.621) -1.204 (2.88) 0.191
22 Textile Mill Products -0.476 (7.796) -0.336 (1.25) 0.042
23 Apparel and Other Textile Products -0.025 (1.084) 0.065 (0.580) 0.010
24 Lumber and Wood Products -0.043 (1.405) 0.027 (0.401) 0.004
25 Furniture and Fixtures 0.0186 (13.691) 0.153 (3.726) 0.284
26 Paper and Allied Products 0.085 (5.113) 0.134 (5.795) 0.489
27 Printing and Publishing 0.388 (35.772) 0.097 (6.846) 0.572
28 Chemicals and Allied Products 0.021 (1.103) 0.037 (1.519) 0.061
29 Petroleum and Allied Products -0.387 (14.896) -0.414 (5.16) 0.439
30 Rubber and Misc. Products 0.495 (30.930) 0.676 (43.02) 0.981
31 Leather and Leather Products -0.444 (2.324) 1.478 (12.13) 0.807
32 Stone, Clay and Glass Products 0.0179 (0.605) 0.087 (2.166) 0.118
33 Primary Metal Industries -0.254 (2.884) 0.307 (2.017) 0.104
34 Fabricated Metal Products 0.080 (4.075) 0.365 (15.89) 0.881
35 Machinery, Except Electrical 0.055 (2.147) 0.208 (4.763) 0.393
36 Electric and Electronic Equipment 0.286 (11.006) 0.406 (14.86) 0.863
37 Transportation Equipment 0.028 (0.839) 0.68 (1.63) 0.068
38 Instruments and Related Products 0.334 (19.402) 0.415 (22.35) 0.934
39 Misc. Manufacturing Industries -0.204 (3.501) 0.549 (2.115) 0.122
*EP=Electric Power Consumption; PW=Production Workers; RVA=Real Value
Added. Model-I: ln(PW)=Į+ȕln(EP); Model-II: ln(PW) = Į+Ȗln(RVA).
What is interesting is the fact that in many industries, the input

elasticities were negative, which confirms the well-documented decrease in
supervisor demand per unit output in manufacturing (due to increased speed
as well as AI-based automation of supervisory activity) as a whole in this
period (Rifkin 1995). In other words, supervisors/machine operatives were
being called upon to oversee machines and processes that were turning out
more and more output. Industries such as SIC 21 Tobacco Products and SIC
22 Textile Mill Products witnessed the lowest input elasticities, indicating
that supervisory technology would have undergone important
modifications/change. What is also interesting to note is the fact that the R2’s
are all considerably lower than those reported in Tier I output elasticities
(Table 9.5). In other words, output varies more closely with energy use than
with the demand for supervision, which is understandable given the
indivisible nature of supervisory inputs (i.e. conventionally-defined

workers).
9.7 Applications and Implications

In this section, we consider the various applications and implications of
the kinetic-based theory of production presented here. It bears noting that
unlike conventional production theory which consists largely of stylized
correlations (Cobb-Douglas, CES, Trans-log), the model presented here is
a bona fide theory, grounded in translational, rotational and chemical
kineticsin short, in the laws of basic physics.
9.7.1 Factor Substitution: Separating Fact from Fiction

The conventional, mainstream view regarding factor substitution is
founded on the (erroneous) view that all factor inputs are essentially alike,
and hence substitutable (Solow 1974). Put differently, it assumes that all
factor inputs are physically productive (i.e. sources of energy), making
substitution possible. For example, tools and energy are seen as substitutable,
as are materials and energy, or labor and materials (Berndt and Wood 1975).
It is our view that this somewhat simplistic (and erroneous) formalization
had its roots in the 18th-century transition from an artisanal to an industrial
economy, where brawn and muscles (e.g. human being-based energy) were
replaced by steam power. Hence, human force/energy/work was replaced
by btus and hps from Boulton-Watt steam engines (referred to as “capital”).
While this form of substitution was (or can be viewed as) legitimate,
substitution as found in the current context is not, owing in large measure
to the very nature of the labor (and capital) input—that is, not being sources
of energy.
Fast forward to the late 19th/early 20th century where labor had, as
Marshall put it, become a supervisory input, overseeing machinery. Clearly,
in this case the capital-labor substitution of the early 19th century was no
longer physically possible, as was the labor-energy substitution. This
follows from the fact that labor was no longer powering material processes,
and capital was neither a source of energy, nor a source of supervision.
The tiered framework presented above puts these questions in what we
feel is the proper perspective. Standard production analysis combines all
three in a single, multivariate production function. Doing so connotes the
notion that factor inputs are comparable and hence substitutable. The tiered
approach, based on kinetics and supervisory technology, highlights the
basic fundamental difference between the two universal factor inputs
(Beaudreau 1998), namely broadly-defined energy and broadly-defined

organization. In so doing, statements like those of Robert Solow to the effect
that “the world can get along without natural resources so exhaustion is just
an event, not a catastrophe” would be dismissed outright as it confuses
energy with organization. In fact, such statements would be dismissed as
they would be seen as violating the basic laws of physics.
9.7.2 The End of Human Supervision and Not

the “End of Work”
Our analysis provides important insights into the nature of material
processes and the contribution of the various “factor inputs.” One such
insight has to do with Jeremy Rifkin’s notion of the “end of work,”
according to which innovations in control technology have rendered and
continue to render conventional labor redundant, thus the title of the book
(Rifkin 1995). Specifically, our results show that semantically, this is an
inaccurate description of the underlying forces. According to basic physics,
work is what is accomplished using force/energy. Moreover, as pointed out,
conventionally-defined workers or labor have not “worked” in over two
centuries. Rather, they have provided and continue to provide supervisory
services, or supervision.
This leads us to argue that what Rifkin and others have been describing
is not the end of work, but rather the end of human supervision. Work (i.e.
the physical definition) in the economy has, since 1995, increased by a
factor of two as evidenced by a doubling of GDP. The point is that this has
been achieved with less human supervision. Again, this highlights the
importance and relevance of having definitions and concepts in economics
that are both internally and externally valid. Rifkin’s prophecy of the “end
of work” makes no sense to an engineer or to a physicist. In short, a more
appropriate title for his book (although one that would be less catchy,
marketing-wise) would have been “The End of Human Supervision.”
Material processes will always be supervised, whether by man or machine.
And they will always work.
9.7.3 The Labor PFI in the post-WWII Era

The last application concerns what is typically referred to as labor
productivity, but which we refer to as labor product per factor index,
oftentimes used as a basis for establishing remunerationin short, it is often
argued that wages should track labor “productivity.” In this section, we
maintain that the post-WWII labor PFI increased in two distinct phases,
namely the speed phase and the automation phase. The former refers to the
increase in labor PFI in the immediate post-WWII period owing to greater
machine speeds, which increased the amount of product per labor or
supervisory input. Again, it is imperative to point out that labor was not
responsible for the increase, but rather was simply a witness to greater
machine speeds.
The second phase, which began in the 1980s and continues to this day is
the automation phase which witnessed the increasing use of inanimate
supervision technologies, commonly referred to as factory automation.
Here, product per factor input increased via a decrease in the denominatoras
opposed to an increase in the numerator in the first phase. As the remaining
supervisors (i.e. labor) were not responsible for the increase in the Labor
PFI, it stands to reason that their remuneration would not, in any noticeable
way, be affected. Perhaps this explains the wage-productivity gap that has
been identified in the literature.

This chapter set out to present an internally and externally-valid theory
of wealth, one that is consistent with the basic laws of physics, and one that
is consistent with both the overall goals and objectives of economics as the
science of wealth as a whole. Finding most non-mainstream critiques to be
short on specifics, we developed a model that is dualistic in nature, focusing
on the physical underpinnings of output (Tier I) as well as the supervisory
and tool-related aspects (Tier II). Delineating the two was the physics-based
construct of physical productivity. Supervisory activity was deemed to be a
necessary part of production activity, without contributing physically to
production.
Our model is the first ever to explicitly invoke the laws of kinetics in
production theory, and the first-ever to confirm empirically what is a key
law of mechanicsand physicsin economics. Moreover, it provided a
theoretical support for the observed 20th-century increase in productivity.
Specifically, the productivity gains in the early 1900s identified by Alfred
Chandler can be attributed in large measure to higher machine
speeds/throughput rates and not to an increase in physical capacity.
It is our view that such models are not only a welcomed alternative to
what are archaic approaches to understanding material processes, they are
necessary to resolving a number of the puzzles and paradoxes in economics.
For example, there is the question of the “information paradox” according
to which “we see computers everywhere except in the productivity data.”
The kinetics-based approach allows us to conclude that it is not a paradox
at all, given that information is not physically productive, and can only
contribute marginallyif at allto productivity via second-law efficiency.
As the latter is bounded from above and highly stable, it stands to reason
that ICT has not, cannot, and will not increase productivity. This stands in
contrast to the two other GPTs, namely the steam engine and the electric
motor, both of which resulted in greater energy consumption per machine,
and hence, greater productivity and output.
Lastly, they provide a long, overdue bridge between classical mechanics,
basic physics, process engineering and economics. While the economics
profession has paid and continues to pay lip service to the fact that its
formalizations of production are grounded in engineering data, the resulting
models have been and continue to be orthogonal to material processes as
seen in the physical sciences. It was shown that this bridge provides
valuable insights into such things as productivity and product indexes, the
most telling example being the theoretically-correct measure of Labor’s
Product Per Factor Input, which measures output per unit labor, without
connoting of physical productivity. Such insights are immensely important
in moving the debate over output, wages and profits (i.e. the debate
instigated by Thomas Piketty’s Capital in the 21st Century) along.
9.9 References
Alting, Leo. 1994. “A Morphological Process Model.” in Manufacturing
Engineering Processes. New York: Marcel Decker.
Beaudreau, Bernard C. 1995. “The Impact of Electric Power on
Productivity: The Case of U.S. Manufacturing 19581984.” Energy
Economics. 17, no. 3: 231–236.
Beaudreau, Bernard C. 1998 Energy and Organization: Growth and
Beiser, Arthur. 1983. Modern Technical Physics. Menlo Park, CA: The
Benjamin/Cummings Publishing Company.
Cobb, Charles and Douglas, Paul. 1928. “A Theory of Production.”
American Economic Review. 18, 1928: 139–165.
Devine, Warren D. 1990. “Electricity in Information Management: The
Evolution of Electronic Control.”, in Schurr, Sam. H. et al. (eds.),
Giraud, Gael and Zeynep Kahraman. 2014. “How Dependent is Growth
from Primary Energy-Output Energy Elasticity in 50 Countries.”
Working Paper, Paris School of Economics.
Jevons, W.S. 1865. The Coal Question. London: MacMillan and Co.
Kummel, Reiner, 1982. “The Impact of Energy on Industrial Growth.”
Energy 7, no.2: 189-201
Kummel, Reiner, Dietmar Lindenberger, and Wolfgang Eichhorn. 1998.
“The Productive Power of Energy and Economic Evolution.” Indian
Journal of Applied Economics, Special Issue on Macro and Micro
Economics.
Laing, Graham. 1933. Towards Technocracy. New York, NY: Angelus
Press.
Marshall, Alfred. 1890. Principles of Economics. London, MacMillan.
Marx, Karl. 1867. “Machinery and Modern Industry.” in Marx, Karl, Das
Kapital. New York, NY: The Modern Library.
Rifkin, Jeremy. 1995. The End of Work. New York, NY: G.P. Putnam’s
Sons.
Samuelson, Paul. 1976. “A Comprehensive Restatement of the Theory of
Cost and Production,” in Samuelson, Paul A. Foundations of Economic
Analysis. New York, NY: Atheneum: 57-89.
Sciences upon State Stewardship, London: Henderson
Solow, Robert. 1974. “The Economics of Resources or the Resources of
Economics.” The American Economic Review, 64(2), Papers and
Proceedings of the Eighty-sixth Annual Meeting of the American
Economic Association:1-14.
Thomas, Woodlief. 1928. “The Economic Significance of the Increased
Efficiency of American Industry.” American Economic Review. 18:
122–138.
Tryon, F.G. 1927. “An Index of Consumption of Fuels and Water Power.”
Journal of the American Statistical Association. 22: 271–282.
U.S. Department of Commerce. Annual Survey of Manufactures.
Washington, DC: Bureau of the Census.
U.S. Department of Labor. 1966. Technical Trends in Major North
American Industries. Department of Labor Statistics, No. 1474,
Washington, DC.
Wibe, Soren. 1984. “Engineering Production Functions: A Survey,”
Economica: 401-411.
10
IDEAS THAT DO AND DON’T MATTER….

FOR GROWTH
Abstract
All existing forms of man-made technology, including material process

technology, started with an idea and thus, are idea-based. However, not all
goods and services-related ideas affect material processes. Ergo, some ideas
matter for growth, while others don’t. In this chapter, we document the
former using the science of material processes as our guide. Specifically, the
model of production developed in Chapters 2, 8, and 9 will be used to
separate the proverbial wheat from the chaff.
10.1 Introduction
The gist of this chapter is quite straightforward, namely that all existing
forms of man-made technology, including material process technology,
started with an idea and thus, is idea-based. However, not all goods and
services ideas affect material processes. Ergo, some ideas matter for growth,
while others don’t. In this chapter, we document the former using the
science of material processes as our guide. Specifically, the model of
production developed in Chapters 3, 4, 8, and 9 will be used to separate the
proverbial wheat from the chaff.
10.1.1 Innovations/Ideas that Matter

As argued in the previous chapters, the material processes that lie at the
very core of economics, like those in all other scientific fields, have toin
fact, mustobey the laws of physics. First and foremost among these is the
fact that the growth of a nation’s wealth is inextricably tied to growth in its
energy use (read: work) as well as improvements in energy efficiency,
known formally as second-law efficiency. As shown, the rise of Western
material civilization was achieved as the direct result of a manifold increase
196 10. Ideas That Do and Don’t Matter….For Growth
in energy consumption in general as well as numerous improvements in

second-law efficiency, both in general and on a per capita basis.
This has been achieved largely as the result of innovations in both our
ability to extract energy (negentropy) from the environment as well as our
ability to apply more and more energy to a given set of tools (simple and
complex) that together define existing and new material processes. For
example, mechanized weaving resulted from the application of a continuous
and in most cases, increasing amounts of inanimate force to the looms that
had up until then been powered by weavers. And, as is true in the case of all
innovations, this started with an idea–the idea of mechanizing (read: apply
inanimate power) what was an artisanal process. In this chapter, we examine
the set of ideas that over the centuries led to increasing energy use, thus
contributing to the vast increase in wealth that collectively defines the
industrial era. The discussion will be structured around the energy-
organization approach presented in Chapter 5 and organized around three
themes, namely the innovations affecting material processes, the innovations
affecting the availability of energy, and the innovations/developments that
prompted the demand for greater speed (material process speed in general,
and machine speed in particular).
Table 10.1: Second-Law Efficiency Increasing (increasing work per

unit energy)
Idea/Innovation Period Effect

Bronze Age 2, 700-900 BCE Increased second-law efficiency;
extended tool life, increased
available energy
Iron Age 1, 200-550 BCE Increased second-law efficiency;
extended tool life, increased
available energy
Watt External James Watt 1765, Patent Increased second-law efficiency,

Condenser Steam granted in 1769 rendered steam economically
Engine viable, ushered in a new era
Reciprocating Steam Matthew Boulton and James Increased second-law efficiency,
Engine Watt 1788 increased speed, engine scaling
from large to small
High-Pressure Richard Threvithick 1800 Increased second-law efficiency,
Steam Engine George Corliss 1849 allowed for greater speeds
Steam Turbine Charles Parsons 1877, Reduced wear and tear, led to
Charles Curtiss 1896 greater machine speeds
Gas Turbine General Electric 1949
Electric Unit Drive 1910s, 1920s Led to greater machine speeds
10.1.2 Innovations/Ideas That Affected Material Processes

Directly
Looking back, there are two types of innovations/ideas that affected
material processes directly, namely second-law efficiency-based (i.e.K(t))
and energy use per unit of tools/capital-based (i.e. e(t)). As a general rule,
the latter were/are achieved by increasing operating speeds (see Chapters 6
and 7). In this section, we examine both historically.
Table 10.2: Innovations That Increased Energy Use Per Unit Capital
Per Unit Time
Idea/Innovation Inventor/Period Effect

Shafting, Gearing and Led to application of
Belting force/energy to
material processes
(manual)
Rotary Drive Matthew Boulton and Transformed batch
James Watt 1788 material processes
into continuous ones
Mechanized Metal Increased amount of
Stamping work per period of
time
High -Speed Electric Unit
Drive
Mechanized Assembly Henry Ford 1913 Increased amount of
Line work per period of
time
Mechanized Material Permitted the more
Handling (Conveyor intensive use of
Belts, High-Speed energy in material
Pumps) displacement
Jet Engine
Automatic Control Massachusetts Institute
Technology (servos, of Technology, Control
sensors, etc) Engineering Laboratory
1945
Improved Lubrication Reduced heat, wear
Technology and tear, allowed for
faster speeds and
more work per period
of time
Table 10.3: Innovations/Ideas That Increased Energy Availability
Innovation/Idea Inventor Effect

Windmill Technology Allowed for the
application of wind
energy to material
processes
Waterwheel Technology Allowed for the

application of hydraulic
energy to material
processes
Denis Papin, Thomas Allowed for the
Savary, Thomas application of chemical
Newcommen, James energy to material
Watt processes
Steam Engine-Rotary James Watt and Made
Matthew Boulton 1788
High-Pressure Boilers Richard Trevithick Made for greater

energy availability per
period of time
Made for more energy
available per period of
time
Large-Scale Thermal Thomas Edison, Samuel Economies of scale,
Generating Stations Insull lower cost per kwh
Direct Current Michael Faraday and Allowed for power
Joseph Henry 1820 transmission
Alternate Current Nicolai Tesla Improved power
transmission
Electrical Grid Thomas Edison, Samuel Increased load factor

Insull
Nuclear Power Unlocked nuclear

energy
Solar Power Unlocked solar energy
10.2 Innovations/Developments That Increased

the Demand for Speed
In this section, a Schmooklerian approach to innovation is adopted, the
idea being that many if not most of the innovations listed above were not
spontaneous occurrences, but rather were the result of an event or events
that served to put a premium on process speed. Perhaps the best metaphor
comes by way of war in general, and artillery in particular. More specifically,
as the history of firearm and projectile technology makes clear, the greater
the rate at which a firearm can deliver its projectile, the greater the damage
it can inflict on its target. The best example of this is the introduction of the
machine gun which increased greatly the rate of fire, conferring a marked
advantage over conventional technology to the shooter. In short, war has
been the single most important source of the demand for speed-increasing
ideas/innovations. A related example is the development of the Garand M-
1 at the Springfield Armory in Springfield, Massachusetts which increased
not only the speed but the reliability of semi-automatic rifles.
It could be argued that Adam Smith’s proverb that “specialization is
determined by the extent of the market” is a variation on this theme. By
specialization, Smith understood the mechanization of what were, at the
time, artisanal material processes. In short, it was about adopting what he
naively referred to as “fire power” that is, the Watt external condenser steam
engine.
In fact, his argument in favor of free trade was, in large measure, a
corollary of this proverb. Specifically, by expanding one’s trading network,
a country increases the “extent of the market” and in so doing, fosters
conditions that are conducive to specialization–in short, adopting a more
energy-intensive, speed-increasing material process technology.
10.2.1 The Pent-Up Demand for Iron and Copper Bar,

and Textiles and Clothing in 18th Century England
The 18th century witnessed an important change in England’s trading
patterns (Thomas and McCloskey 1984). In short, its trade which had until
then been concentrated in the Mediterranean shifted to the Atlantic Basin –
in short, triangular, mercantile trade that saw goods (textiles, iron and
copper bar) being traded for slaves in Africa, the slaves then being traded
for staples in the West Indies and America, and staples being traded
ultimately for goods, with the process repeating itself.
Initially, the iron and copper bar that was used by English merchants
was imported from Sweden and Russia. However, as trade increased, so too
did the demand for iron, which in turn, put a premium on cheap iron and
copper bar. This led to a number of developments including the use of the
atmospheric steam engine as a source of power to pump water out of English
iron mines, the puddling technique, and the development of an indigenous
copper industry.
10.2.2 The Pent-Up Demand for Consumer Goods

in Post-Bellum America
As described in Alfred D. Chandler’s 1977 magnum opus, the The
Visible Hand: the Revolution in American Business, the opening up of the
U.S. West, the result of (i) the Louisiana Purchase and (ii) the massive
investments in railroads, prompted a series of innovations which increased
productivity in U.S. industry, notably by increasing process speeds. This
was achieved in large measure by the electrification of processes that until
then had been powered by either water power or steam power. This was
achieved by replacing the cumbersome belting and shafting power-driving
technology with electric unit drive.
The creation of a mass-market, according to Chandler, witnessed the
creation of nation-wide distribution channels, which in turn led to greater
scale and scope, in so far as industry was concerned. The result was the
second industrial revolution.
In many regards, the opening up of the West can be considered as being
analogous to the creation of the British empire in the 17th and 18th century.
In other words, an expanded market prompted merchants to seek out, speed-
increasing innovations in order to adequately serve the various sub-
segments of the market.
10.2.3 The Demand for Armaments/Military Supplies in WWII

War, it could be argued, exerts a similar effect on the armament industry
in that it puts a premium on productivity on the one hand, and efficacy on
the other. The better able is a country to produce armaments, the greater the
probability of being successful (being victorious). Further, the greater is the
productivity of such armaments, the greater the chances of being successful.
Such was the case in World War II when a series of speed-related
innovations improved the Allies’ chances of victory. One such innovation
was the production of aircraft at Ford’s Willow Run factory. At its peak
monthly production (August 1944), Willow Run produced 428 B-24s with
highest production listed as 100 completed Bombers.
A second innovation was the development of the Garand M-1 semi-

automatic rifle at the Springfield Armory, which increased the speed (i.e.
the rate of fire) and efficacy of conventional hand-held rifles. According to
General Patton, it was the one single innovation that won the War for the
Allies.
10.3 Innovations/Ideas that Don’t Matter

The gist of this volume is that growth is a material process and since
ideas are not material, they don’t matter. Put differently, ideas are abstractions
and as such cannot affect growth except in so far as they affect either energy
use/intensity or energy efficiency. This holds regardless of the set of
products and services being produced. The following is a sample of ideas
which, according to this criterion, don’t matter in so far as growth is
concerned.
10.3.1 Automation and Robotization

A vestige of an earlier age when human muscle/brawn powered material
processes, labor productivity has been used and continues to be used as a
measure of productivity. As pointed out earlier, the very concept of labor
productivity is fundamentally flawed as labor has not been a source of
force/energy for over two centuries. Instead, modern-day labor is akin to
machine operators/supervisors. That is, whose task it is to oversee the
workings of one or a series of material processes, without however powered
them.
Automation and robotization are similar and different. The former refers
to the automatic control of material processes, usually resulting in a
decrease in the number of machine operatives. The second refers to either
(i) replacing brawn and muscles with a source of inanimate energy and/or
(ii) the use of automatic control technologies in lieu of human supervision.
Automation, by decreasing the number of supervisors/operators, will
invariably increase the aforementioned labor productivity as it reduces the
denominator without affecting the numerator. While not in any way
responsible for the increase, their “productivity” nonetheless increases as a
result. Capital productivity, defined as output divided by a measure of the
capital input will, on the other hand, decrease as more tools (information
technology) will now be required for a fixed amount of output.
This finding, we believe, goes a long way to explain Robert Solow’s
1987 Information Paradox, according to which “we see computers
everywhere, but in the productivity data.” Other than labor productivity
growth, the vertiginous rise in the use of computers in the 1990s, 2000s, and
2010s had little to no effect on overall productivity growth. While a paradox
to most, this is consistent, indeed even predicted, by the energy-organization
approach outlined in Chapter 5. In short, information is not physically
productive. As such, the information and communications technology
revolution, by increasing the rate of transmission of information and
reducing its cost of storage, had no effect on growth.
There is, however, one exception, namely the capacity of ICT to increase
second-law efficiency. Specifically, improvements in control technology
made possible by ICT could, conceivably, increase growth, but only on a
punctual basis. In other words, these gains are made at a point in time and
not necessarily over time. Hence, while they could potentially affect growth
in the year they are applied, they will have little to no bearing on the overall
rate of growth.
10.3.2 New Products

Products (goods and services) are, by definition, that which results from
the large-scale application of force in well-defined material processes, the
result of which is wealth. They are typically defined as a collection of
characteristics which together define goods and services. The point of the
matter is that varying the composition of a or many products will have no
bearing whatsoever on growth, as to pretend so is to confuse process from
result. In short, new products are not a form of energy and as such, cannot
increase work, nor growth.
It should be noted that roughly 85 percent of all research and
development expenditure in the world is product-based, a fact which we
believe explains the failure of forty years of research and development
expenditure to restore growth rates to their pre-productivity slowdown
levels – that is, to levels in the post-WWII period.
10.4 Prognosis for Future

These results raise the obvious question, namely can Western
industrialized economies grow at post-WWII rates–that is, like before? After
all, four decades of policy, from free trade, to deregulation, to tax policy, to
education, to R&D policy have been devoted to restoring the high growth
rates that shaped our modern world.
The answer, we believe, is as unequivocal as is the record: the high
growth rates of the post-WWII period are a thing of the past, a singularity
that will in all likelihood not be repeated soon. The reason is simple, namely
that machine/process speeds, the source of productivity growth throughout

the industrial age, have for all intents and purposes plateaued. In short, the
age of speed has come to an end, and with it, any hope of a return to the
past. In other words, because most processes have reached their kinetic
limit, they cannot be speeded up, making for the current situation in which
growth rates have and continue to hover around 1.5 percent. Table 10.4
presents a non-exhaustive list of processes that have, for all intents and
purposes, reached their kinetic limit, and as such, cannot be speeded up.
Figure 1 shows the evolution of CPU clock speed over the course of the past
three decades. As shown, it has, for all intents and purposes, plateaued, as
have numerous other process speeds.
While there can and will be exceptions, we are of the view that growth
rates cannot be ratcheted up to previous levels. In other words, the post-
WWII period was a singular event, the likes of which cannot and will not
be reproduced. To argue the contrary would be to violate the basic laws of
physics, a practice all too common in economics.
Table 10.4: Processes That Have Reached Their Kinetic Limit
CPU Clock Speed

Airplane Speed
Power Tool Speed
Automobile Speed
Farm Tractor Speed
Automobile Assembly Line
Material Handling Speeds
10.5 Exponential R&D Growth and Anemic Productivity

Growth
Global R&D expenditure has exploded in the post-productivity
slowdown era as governments have embarked on ambitious plans to
increase growth with the goal of ultimately restoring 4-5 percent annual
growth. This has been achieved via a panoply of policy measures, ranging
from increased government R&D spending, subsidies, recruiting institutions
of higher learning to do their part, fiscal incentives for both individuals and
corporations, free-trade, and deregulation. The result has been what we refer
to as the R&D full monty, namely that societies, in the name of wanting to
restore growth to post-WWII levels, have put innovation and R&D at the
center of the policy agenda.
Consider, for example, the case of the United States. The National
Science Foundation estimates that U.S. R&D funding reached an all-time
high of $499 billion in 2015. Of that total, the federally sponsored share fell
to a record-low 23 percent while the business sector’s share rose to a record-
high 69 percent. According to new estimates from the National Science
Foundation’s National Center for Science and Engineering, total spending
on R&D reached $499 billion last year, buoyed by record levels of business
spending. If confirmed, this will represent the largest amount the U.S. or
any nation has ever spent on R&D in a single year. It is one of a number of
indicators that the U.S. remains the international leader in science and
technology even as China poses a challenge to U.S. dominance in the field.
Of total U.S. R&D in 2015, businesses funded $355 billion, or 69 percent,
continuing a long-term trend of private enterprise financing an increasingly
large majority of R&D nationwide. The federal government, the second-
largest funder of U.S. R&D, sponsored an estimated $113 billion, or 23
percent of the total.
Figure 10.1: U.S. R&D by Performing Sectors and Source of Funding

1953-2015
Source: Henry (2016), 2.
Business-funded R&D has grown over the decades and now comprises
about two-thirds of U.S. R&D spending, while federally funded R&D has
gradually declined to its lowest point as share of the economy since records
began in 1953 (see Figures 10.1 and 10.2). Public Domain Federal funding
for R&D has declined for four years in a row, reaching its lowest level since
2007 in 2015. The federal share of total U.S. R&D is also now the lowest it
has been dating back to 1953, when NSF first began recording data on the
subject.
Figure 10.2: Ratio of U.S. R&D to Gross Domestic Product by Source

of Funding 1953-2015
Source: Henry (2016), 4
The U.S. case is highly representative and indicative of R&D expenditure

in virtually all industrialized and industrializing countries. Specifically, there
was an important upswing in R&D activity in the 1980s, largely in response
to the productivity or growth slowdown. Like all policy measures, this has
come at a cost. Over two percent of the world’s overall wealth is now
allocated to R&D. The resulting fiscal incentives have increased inequality,
the most telling example being the “trickle-down principle” whereby tax
cuts to the rich will kick start productivity growth, thus raising both wages
and profits. Institutions of higher learning have been solicited via an
increased emphasis on the STEM (science, technology, engineering and
math) curricula, at the expense of traditional/classical curricula. In short,
universities the world over have been transformed into institutes of
technology, where value is measured in terms of a researcher’s ability to

deliver marketable goods and services.
That all of these measures, over a period of forty years, have failed to
resuscitate post-WWII growth rates should, of course, be a red flag. Clearly,
something is wrong. Never before have we as a species devoted as many
resources to innovation, yet growth remains patently anemic. Call it the
R&D Paradox, namely “we see R&D everywhere except in the growth
statistics.”
This raises the question: how have the leading figures in New Growth
Theory responded? After all, the record speaks for itself: forty years of
massive R&D spending/activity have failed to make a dent. The pat
response is the “low-lying fruit allegory” namely that the returns to R&D
are decreasing as all of the goods ideas (low-lying fruit) have been
discovered, leaving the proverbial high-lying, high-cost fruit to be discovered.
In other words, they are not on the verge of abdicating, of declaring defeat.
This, we maintain, owes in whole or in part to the patent lack of alternatives.
As pointed out in the Introduction, the forty years of New Growth Theory
have failed to generate new ideas about production, ideas that could shed
light not only on the productivity slowdown, but the efficacy of the trillions
of dollars spent over this period in the hope of restoring growth. In many
regards, this is consistent with Paul Krugman’s indictment of New Growth
Theory on the grounds that it is too open-ended and hence, lacking any
structure.

In this chapter, we have argued that while ideas, in general, don’t matter,
some ideas do, notably those that pertain to material processes–that is, those
that either increase second-law efficiency or increase the amount of energy
per unit tools/capital. In both cases, the result is more work per unit
tools/capital.
All other ideas, while of value, cannot and will not increase growth.
While this may appear to be counterintuitive in this, the age of the Solow
residual, the Ak Model and New Growth Theory, it is not only consistent
with, but predicted by the laws of physics. Put differently, the last four
decades have, for lack of a more complete theory of production, left the
question of what matters, specifically what ideas matters and don’t, wide
open. Clearly, this was and remains a weakness in the growth literature and,
more importantly, a reflection of the poor state of production theory in
general.
On another note, the findings of this chapter are consistent with the
overall literature, scholarly and popular, on the innovations that changed the
world. For example, the steam engine is widely seen as one of the defining
innovations that marked the start of the industrial revolution and the
dawning of the modern world. Another is electricity which survey after
survey show to be another defining innovation in the rise of modernity. At
the turn of the current century, a local newspaper surveyed its readers on the
question of 20th-century innovations, specifically of the most far-reaching
innovation. The results put electricity and electric power on the top of the
list. Other energy-related innovations include the internal combustion
engine which gave us the automobile and other forms of powered
transportation, and the jet engine which did likewise for flight. Both
increased the speed with which we as a species can displace ourselves in
space and time. That the general public would come to these conclusions is
not, in our view, surprising given the underlying science, namely the laws
of physics.
10.6 References
Henry, M. 2016. US R&D Spending at All-Time High, Federal Share
Reaches Record Low (aip.org/fyi/2016/us-rd-spending-all-time-high-
federal-share-reaches-record-low).
Schmookler, J. 1962 Economic Sources of Inventive Activity* The Journal
of Economic History, 1962, 22, vol.1:1-20
Schmookler, J, 1966 Invention and Economic Growth, Cambridge, MA:
Harvard University Press, 1966.
Thomas, R.P. and D. McCloskey” Overseas Trade and Empire, 1700-1820,”
Chapter 4 in Floud, R. and D. McCloskey, The Economic History of
Britain, 1700-Present (1984), vol. 1, 87-102.
SUMMARY AND CONCLUSIONS
This volume set out to do a number of things, including expanding the

set of ideas regarding the process of wealth creation with the explicit
purpose of recasting the debate over the sources of economic growth. Until
now, the literature has been dominated to the point of being limited to the
neoclassical theory of production and growth which as we have argued is
largely impressionistic in nature, not to mention a theory whose underlying
axioms sit in violation of the laws and principles of basic physics.
Unfortunately, over the past century, new approaches and new ideas
regarding production and growth have been–and to a certain extent,
continue to bedenigrated and outrightly dismissed. This volume can be
viewed as a first step, as the first step to render the economic growth
literature consilient with the literature in all related, material process fields,
and, as such, as the first step to a new, scientific era of work on economic
growth. The articles/chapters presented here have one thing in common,
namely being grounded in the laws of physics, be they classical,
thermodynamic, or kinetic. Further, they can be and should be viewed as
the logical conclusion/extension of a view of production that had its origins
in the early 20th century and included contributions by the likes of Frederick
Soddy and Howard Scott, two non-economists.
Taken collectively, they provide a consilient approach to production and
growth, one that parallels the findings of other material sciences, and thus,
one whose predictions are consistent with those of the science of material
processes in general. As mentioned, the study of production and growth in
economics has ignored the basic laws of physics, both in their suppositions
and their omissions. The energy-organization and kinetics approaches
presented here go a long way to correct this.
However, in the end, the proof of the pudding is in the eating. Do these
approaches provide new and valuable insights? Do they solve heretofore
unresolved issues? It is our view that they do. Among others, they are
capable of rationalizing a number of unresolved issues, including the
information paradox, the productivity slowdown and indeed the failure of
government policy in the post-productivity slowdown era to restore growth
to post-WWII levels. In all of these cases, the reason is simple, namely that
the limits to machine/process speed have been reached, making further
speed-up-based productivity growth improbable if not impossible. No
210 Summary and Conclusions
amount of research and development, of information technology, of

deregulation, of free trade, of tax cuts, and of rhetoric can change this one,
immutable fact, namely that the age of speed is over, never to return.
It goes without saying that this constitutes a wholesale condemnation of
growth theory and indeed of production theory from the get-go. In other
words, this volume implicitly rejects all of production theory from Smith to
Romer. Clearly, this is a bold assertion on our part.1 After all, because the
question of wealth lies at the very core of all of economics, it follows that it
amounts to a wholesale condemnation of the science in general.
However, the important thing is that it does not stop at that. Rather, it
should be seen as a first step in the long, overdue need to render the science
of wealth consilient with all other sciences of material processes. Put simple,
the laws of nature (read: physics) cannot be overlooked or discarded, but
rather must be incorporated into the corpus of economic thought.
And on this note, one thing is clear: ideas as defined by such notables as
Robert Solow and Paul Romer haven’t mattered, don’t matter and will never
matter to growth. What matters is the ability to do work and, as we have
argued repeatedly throughout this volume, this depends on energy.
Moreover, they imply that Moses Abramovitz’s “measure of ignorance”
pertained not to the shocks (i.e. to A), but rather to our understanding of the
very material processes that together make up a modern industrialized
economy.
We are not the first to make such an assertion. In fact, it was the
cornerstone of the Technocracy movement, not to mention Frederick
Soddy’s Cartesian Economics which its name implies was an attempt at a
rapprochement between physics and economics. So too was Technocracy
which focused almost exclusively on energy.
The science-based approach to growth presented here, we feel, is a first
step in the colossal task ahead of integrating the science of material
processes into the many sub-fields of economics, notably in income
distribution and of international trade, not to mention labor economics (or
human resources economics).
Lastly, there is the question of public policy, specifically the public
policy implications of our results/findings. First and foremost is the question
of the limits of productivity growth. For forty years, governments the world
over have struggled with what was and continues to be anemic growth rates,
enacting myriad policies geared toward restoring growth. These have been
largely unsuccessful and a source of bewilderment on the part of scholars
1However, it is our view that what has been and continues to be bold are the myriad
engimas regarding production and growth that today constitute the science of
economics.
and elected officials. Our results/findings shed light on this perceived-of

anomaly, and in so doing, provide insight into such policies. For example,
as growth is a material process, it stands to reason that the panoply of
policies enacted over the last four decades (R&D subsidies, deregulation,
free-trade, supply-side economics, STEM education, fiscal policy) was
futile, as none affected the underlying material processes.
Second, with the end of the “Age of Speed” came the end of material
process-based “speed-ups” and consequently, the end of productivity
growth. As such, the objective of restoring post-WWII productivity growth
levels is an illusion, fed by and maintained by poor science-notably the
open-ended nature of the A-scaler in neoclassical production theory. As
such, governments and indeed the World will have to live with low
productivity growth for the foreseeable future, if not forever. Perhaps the
best metaphor is the speed of jet airplane travel which has actually decreased
relative to the 1970s. Or computer CPU clock speed which has also
plateaued.
This brings us to the question of R&D spending, specifically the value
of devoting over two percent of world GDP to R&D. As we have shown,
the bulk of this came on the heels of the productivity/growth slowdown. For
the past four decades, projects have and continue to be justified by their
growth-enhancing possibilities à la Romer. Does this mean that R&D
should be halted and the monies spent elsewhere?
There are two answers to this question. First, while R&D has not
contributed to restoring growth rates, it has contributed to improving goods
and services-in short, it has affected quality as opposed to quantity. And, of
course, this has value. The point here is that R&D will have to be justified
on the basis of its ability to increase consumers’ willingness-to-pay as
opposed to increasing growth. On the other hand, one could argue that the
nearly $2 trillion currently spent on R&D could be better spent on providing
basic needs to the World’s poor, and that this would constitute a better
outcome.

Bernard C. Beaudreau - Science and The Wealth of Nations-Cambridge Scholars Publishing (2021)

Uploaded by

Copyright:

Available Formats

You might also like

Bernard C. Beaudreau - Science and The Wealth of Nations-Cambridge Scholars Publishing (2021)

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

Bernard C. Beaudreau - Science and The Wealth of Nations-Cambridge Scholars Publishing (2021)

Uploaded by

Copyright:

Available Formats

Science and the Wealth

The Physics of Economic Growth

This book first published 2021

Cambridge Scholars Publishing

Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK

British Library Cataloguing in Publication Data

Copyright © 2021 by Bernard C. Beaudreau

ISBN (10): 1-5275-6581-5

List of Tables and Figures ......................................................................... xi

Introduction ............................................................................................ xvii

1 The Enigmatization of Economic Growth ............................................... 1

2 Thinking About Production: From Smith to Romer .............................. 19

3 The Impact of Electric Power on Productivity: A Study of U.S.

4 The Energy-Organization Framework ................................................... 51

5 Engineering and Economic Growth....................................................... 77

6 The Dynamo and the Computer: An Engineering Perspective

7 The Physical Limits to Economic Growth by R&D Funded

7.7 Conclusions ................................................................................. 124

8 The Economies of Speed, KE=1/2mv2 and the Productivity

9 A Kinetics-Based Approach to Production: Theory and Evidence...... 165

10 Ideas That Do and Don’t Matter….For Growth ................................ 195

Summary and Conclusions ..................................................................... 209

This volume is a response of sorts to the papers presented in a session at

The ultimate irony of idea-based growth theory is the total

This quotation captures the purpose of the present volume, namely of

referred to as the A parameter, or technology as the leading cause. And so

It would be an understatement to point out that this whole episode, from

productivity slowdown and the information paradox. Both are shown to

1-The Enigmatization of Economic Growth. Real World Economic

2-Thinking About Production: From Smith to Romer

3-The Impact of Electric Power on Productivity: The Case of U.S.

elasticities were estimated directly, as opposed to indirectly. These were

4-The Energy-Organization Approach to Modeling Material Processes

5-Engineering and Economic Growth. Structural Change and

6-The Dynamo and the Computer: An Engineering Perspective on the

7-The Physical Limits to Economic Growth by R&D Funded

8-The Economies of Speed, KE=1/2mv2, and the Productivity

9-The Kinetics Approach to Production: Theory and Evidence, 2018

10-Ideas That Do and Don’t Matter…..for Growth.

THE ENIGMATIZATION OF ECONOMIC

Table 1.1: Material Processes, Energy and Organization

Material Process Energy Input Organization Inputs

1.2 Consilient Formalizations of Growth

Beaudreau (1998) provided a consilient approach to understanding

ܹ (‫)ݐ( ܶ[ߟ = )ݐ‬, ܵ(‫)ݐ‬, ‫)ݐ(ܧ])ݐ(ܫ‬ 1.1

Frederick Soddy captured the essence of material processes–animate and

1.3 Literature Review: Steady-State Growth Models

Table 1.2: Neoclassical and Other Enigmas

by simply decreeing capital to be physically productive. No justification

1.3.1 Endogenous or New Growth Theory

directions. For example, instead of focusing on the Solow residual which is

1.3.2 Non-Steady State Models (shocks)

1.3.4 Policy Implications

1.3.5 Consilient Approaches

such as the ecological approach and the biophysical approach, while

Figure 1.1: Actual and Predicted Growth of U.S. GDP 1960-1978

Source: Kummel (1982), 196.

Let us begin with the question of post-WWII growth. A consilient