Elite Networks: The Political Economy of Inequality 0197774237, 9780197774236

Table of contents :
Cover
Elite Networks
Copyright
Dedication
Contents
Preface
Acknowledgments
Introduction: Connecting Elite Power to Inequality
I.1. Main Contributions
I.2. Why Join Elite Networks? The Three Principles
I.3. Why Should Elite Networks Have Any Effect on Inequality?
I.4. The Cyclicality of Inequality
I.5. What This Book Does Not Do: External Validity Limitations and Potential Extensions
I.6. Structure of the Book
1. Why Study Elite Networks?
1.1. Defining Elite Networks: Informal Mutually Beneficial Interactions between Political and Corporate Elites
1.1.1. Does It Pay Off to Be Part of the Network?
1.1.2. The Role of Institutions
1.1.3. Elite Theory in Sociology
1.1.4. The Spontaneous Nature of Elite Networks
1.2. Visualizing Elite Networks
1.3. Putting Elite Networks on the Map of Political Interactions
1.4. Quantifying Elite Network Outcomes
Part I: The Impact of Elite Networks on Inequality
2. Evolution of Elite Networks and Inequality: Inequality in the Very Long Run
2.1. The Agricultural Revolution: From Hunter-​Gatherers to Settlers
2.2. How Exactly Did Farming Translate to Inequality?
2.3. Proto-​States and the Formation of Hierarchical Orders
2.4. Concentration of Wealth in Ancient Civilizations: Rise of the Malthusian Economic Model
2.5. The Long Malthusian Trap: Inequality and Elite Concentration during the Middle Ages until the Industrial Revolution
2.6. Wealth Concentration and Elite Power in the 19th and 20th Centuries: From a “Society of Rentiers” to a “Society of Managers”
3. Democracy and Inequality in the Short Run
3.1. The Switch toward Democracy
3.2. The Relationship between Democracy and Inequality
3.3. The Rise of Government Spending in Democracies
3.4. The Rise of Inequality in Democracies
3.5. Economic Explanations of the Rise of Inequality
3.6. Political Factors That Could Help Explain the Rise of Inequality
4. Political Networks and Wages of Top Corporate Income Earners
4.1. Data and Variables
4.1.1. Measuring Networks and Connections of Individuals and Firms
4.1.2. Correlation between Political Connections and Network Size
4.2. Inequality in Top Executive Incomes
4.3. Empirical Strategy and Results
4.3.1. Results: United States
4.3.2. Results: United Kingdom
4.4. The Impact of Political Connections on Executive Salaries in Other Countries
Part II: Inside the Logic of an Elite Network
5. The Internal Logic of an Elite Network
5.1. Network Theory of Elites
5.1.1. The Link between Clustering and Size of Networks
5.1.2. The Centrality of Elite Network Superhubs
5.1.3. Historical Examples of Elite Networks
5.1.4. Incentives for Firms and the Impact on Inequality
5.2. Economic Theory: Costs and Benefits of Elite Network Membership
5.3. Joining Elite Networks to Mitigate Risk and Achieve Status: The Three Principles
6. Motivation for Politicians: Extracting Rents and Staying in Power
6.1. Corruption, Clientelism, and Rent-​Extraction
6.1.1. Political Agency
6.2. Do Voters Punish Corruption?
6.3. Elite Networks and the Selectorate Theory
6.3.1. The Nonlinearity between Corruption and Political Survival
6.4. Case Study: Using Corruption and Minimum Winning Coalitions to Stay in Power
6.4.1. The Impact of Minimum Winning Coalitions
7. The Role of the Firm
7.1. Rent-​Seeking vs. Customer-​Seeking Firms
7.2. How Firms Benefit from Collusion with Politics
7.3. The Revolving Door between Politics and Finance
7.4. Case Study: Political Allocation of TARP Funds during 2008 and 2009
Part III: Reducing Political Power, the Root Cause of Inequality
8. Capitalism and Democracy
8.1. How Far Have We Come?
8.2. Persistent Incentive to Change
8.3. The Gradual Shift to Democracy
8.4. Democracies as a Trial-​and-​Error Process
8.5. Socialism’s Rapid Industrialization with a Lack of Trial-​and-​Error in the Underdeveloped Periphery
8.5.1. Condensing 200 Years of Gradual Progress into 30 Years of Rapid Growth
9. The Pitfalls of Political Power: Expanding the Scope of Government to Reduce Inequality
9.1. Progressive Taxation
9.1.1. The Unprecedented 30-​Year Period That Brought Down Inequality
9.1.2. The Artificial Decrease of Inequality
9.2. Adverse Selection into Politics
9.2.1. Personalization of Governing Institutions
9.3. The Importance of Trust
10. The Three Levers
10.1. Moving beyond Standard Approaches: Introducing the Three Levers
10.2. First Lever: Reducing Political Power
10.2.1. Full Transparency of Budgetary Spending and Decision-​Making
10.2.2. Encourage and Promote Free, Independent Media
10.2.3. Term Limits and Strict Conditions for Holding Office
10.2.4. Reduce Discretionary Decision-​Making Power and Introduce Rule-​Based Politics
10.2.5. Reduce the Scope of Centralized Government
10.2.6. Minor Reforms That Could Strengthen the First Lever
10.3. Second Lever: Re-​Empowering the Citizens
10.3.1. Give Citizens the Ability to Take Control of Their Tax Payments
10.3.2. Mandatory Participatory Budgeting for Community Public Good Provision
10.3.3. Granting Impeachment Powers to the Citizens for Direct Punishment of Officials
10.4. Third Lever: Re-​Empowering the Community
10.4.1. Decentralize and Depersonalize Decision-​Making over Allocation of Public Funds
10.5. Second-​ and Third-​Order Effects of the Three Levers
Afterword
Notes
References
Index

Citation preview

Advance Praise for Elite Networks “Vuković’s sweeping reimagining of the drivers of economic inequality introduces the notion of elite networks—​informal social networks between political and economic leaders—​as a key mechanism behind the steady accumulation of wealth by small elites in almost all societies. Moving beyond explanation, he identifies concrete steps to advance economic equality through decentralizing reforms. A powerful, compelling combination of theory, history, and data analysis that should reshape how we understand the links between economic and political power.” —​Jacob N. Shapiro, Princeton University “Elite Networks is a powerful and groundbreaking analysis of how privileged groups exploit their positions to secure economic resources. In this wide-​ranging tour de force, Vuković shows how political influence pervades every aspect of our economy, underpinning the global rise in inequality.” —​Ben Ansell, University of Oxford, and author of Why Politics Fails “This is a supremely ambitious and provocative book. It offers a sweeping account of the sources of inequality across the world and attempts to uncover and document universal laws about capitalism, social and political organization, and the distribution of resources akin to Thomas Pikketty’s Capital in the 21st Century. Deemphasizing the idea that market power or enduring competitive advantages are a symptom of the search for profits in the crucible of creative destruction, Vuković stresses that, even in the most innovative countries, corporations secure barriers to entry that engender entrenched class stratification. It therefore follows that sealed off elite networks are at the heart of modern, post Malthusian inequality everywhere, both in the developed and developing world. Whether you agree or not, this is a must read!” —​Victor Menaldo, Professor, University of Washington, and Co-​founder of the UW Political Economy Forum “Elite Networks is a groundbreaking book that delves deep into the roots of income inequality. Empirically and descriptively rich, the book elucidates how extractive political power drives this pervasive problem, unveiling the intricate web of informal relationships between politicians and influential corporate figures. The book's strength lies in its ability to connect the dots between intra-​elite trust, loyalty, and the alarming income disparities we witness today. What sets the book apart is its bold assertion that income inequality is not an inherent feature of economic systems but a result of deliberate actions stemming from the quest for centralized political power. In a world grappling with inequality, this book challenges us to rethink our approach and offers a path toward a fairer society. Elite Networks is a must-​read for anyone seeking a deeper understanding of income inequality and a roadmap for meaningful change.” —​Josip Glaurdic, University of Luxembourg

Elite Networks The Political Economy of Inequality V U K V U KOV IĆ

Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2024 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. Library of Congress Cataloging-in-Publication Data Names: Vuković, Vuk, 1988– author. Title: Elite networks / Vuk Vuković. Description: New York, NY : Oxford University Press, [2024] | Includes bibliographical references and index. Identifiers: LCCN 2023056791 (print) | LCCN 2023056792 (ebook) | ISBN 9780197774236 (paperback) | ISBN 9780197774229 (hardback) | ISBN 9780197774250 (epub) Subjects: LCSH: Strategic alliances (Business)—Political aspects. | Elite (Social sciences)—Economic aspects. | Business and politics. | Income distribution—Political aspects. Classification: LCC HD69. S8 V85 2024 (print) | LCC HD69. S8 (ebook) | DDC 322/.3—dc23/eng/20240221 LC record available at https://lccn.loc.gov/2023056791 LC ebook record available at https://lccn.loc.gov/2023056792 DOI: 10.1093/​oso/​9780197774229.001.0001 Paperback printed by Integrated Books International, United States of America Hardback printed by Bridgeport National Bindery, Inc., United States of America

To my loved ones, Barbara, Nikola, Josip, and Ema

Contents Preface  Acknowledgments 

xi xv

Introduction: Connecting Elite Power to Inequality  I.1. Main Contributions  I.2. Why Join Elite Networks? The Three Principles  I.3. Why Should Elite Networks Have Any Effect on Inequality?  I.4. The Cyclicality of Inequality  I.5. What This Book Does Not Do: External Validity Limitations and Potential Extensions  I.6. Structure of the Book 

1 7 11 17 21

1. Why Study Elite Networks?  1.1. Defining Elite Networks: Informal Mutually Beneficial Interactions between Political and Corporate Elites 

34



46 59 64



1.1.1. Does It Pay Off to Be Part of the Network?  1.1.2. The Role of Institutions  1.1.3. Elite Theory in Sociology  1.1.4. The Spontaneous Nature of Elite Networks 

1.2. Visualizing Elite Networks  1.3. Putting Elite Networks on the Map of Political Interactions  1.4. Quantifying Elite Network Outcomes 

26 29

34 37 39 41 42

PA RT I :   T H E I M PAC T O F E L I T E N E T WO R K S O N I N E Q UA L I T Y 2. Evolution of Elite Networks and Inequality: Inequality in the Very Long Run  2.1. The Agricultural Revolution: From Hunter-​Gatherers to Settlers  2.2. How Exactly Did Farming Translate to Inequality?  2.3. Proto-​States and the Formation of Hierarchical Orders  2.4. Concentration of Wealth in Ancient Civilizations: Rise of the Malthusian Economic Model  2.5. The Long Malthusian Trap: Inequality and Elite Concentration during the Middle Ages until the Industrial Revolution  2.6. Wealth Concentration and Elite Power in the 19th and 20th Centuries: From a “Society of Rentiers” to a “Society of Managers” 

71 72 75 77 81 85 90

viii Contents

3. Democracy and Inequality in the Short Run  3.1. The Switch toward Democracy  3.2. The Relationship between Democracy and Inequality  3.3. The Rise of Government Spending in Democracies  3.4. The Rise of Inequality in Democracies  3.5. Economic Explanations of the Rise of Inequality  3.6. Political Factors That Could Help Explain the Rise of Inequality 

101 101 104 108 115 120 126

4. Political Networks and Wages of Top Corporate Income Earners  132 4.1. Data and Variables  133

4.1.1. Measuring Networks and Connections of Individuals and Firms  4.1.2. Correlation between Political Connections and Network Size 



4.3.1. Results: United States  4.3.2. Results: United Kingdom 

134 139



4.2. Inequality in Top Executive Incomes  4.3. Empirical Strategy and Results 

139 152



4.4. The Impact of Political Connections on Executive Salaries in Other Countries 

164

154 159

PA RT I I :   I N SI D E T H E L O G IC O F A N E L I T E N E T WO R K 5. The Internal Logic of an Elite Network  5.1. Network Theory of Elites 



5.1.1. The Link between Clustering and Size of Networks  5.1.2. The Centrality of Elite Network Superhubs  5.1.3. Historical Examples of Elite Networks  5.1.4. Incentives for Firms and the Impact on Inequality 

5.2. Economic Theory: Costs and Benefits of Elite Network Membership  5.3. Joining Elite Networks to Mitigate Risk and Achieve Status: The Three Principles 

171 172 174 177 179 184

186 189

6. Motivation for Politicians: Extracting Rents and Staying in Power  6.1. Corruption, Clientelism, and Rent-​Extraction 

195 196



6.2. Do Voters Punish Corruption?  6.3. Elite Networks and the Selectorate Theory 

199 201



6.4. Case Study: Using Corruption and Minimum Winning Coalitions to Stay in Power 

205



6.1.1. Political Agency 



6.3.1. The Nonlinearity between Corruption and Political Survival 



6.4.1. The Impact of Minimum Winning Coalitions 

197

204

210

Contents  ix

7.  The Role of the Firm  7.1. Rent-​Seeking vs. Customer-​Seeking Firms  7.2. How Firms Benefit from Collusion with Politics  7.3. The Revolving Door between Politics and Finance  7.4. Case Study: Political Allocation of TARP Funds during 2008 and 2009 

213 216 218 221 226

PA RT I I I :   R E DU C I N G P O L I T IC A L P OW E R , T H E R O O T C AU SE O F I N E QUA L I T Y 8.  Capitalism and Democracy  8.1. How Far Have We Come?  8.2. Persistent Incentive to Change  8.3. The Gradual Shift to Democracy  8.4. Democracies as a Trial-​and-​Error Process  8.5. Socialism’s Rapid Industrialization with a Lack of Trial-​and-​Error in the Underdeveloped Periphery 

8.5.1. Condensing 200 Years of Gradual Progress into 30 Years of Rapid Growth 

235 235 241 243 245 246 248

9.  The Pitfalls of Political Power: Expanding the Scope of Government to Reduce Inequality  9.1. Progressive Taxation 

260 260



263 265



9.1.1. The Unprecedented 30-​Year Period That Brought Down Inequality  9.1.2. The Artificial Decrease of Inequality 



9.2. Adverse Selection into Politics 

270



9.3. The Importance of Trust 

275



9.2.1. Personalization of Governing Institutions 

10. The Three Levers  10.1. Moving beyond Standard Approaches: Introducing the Three Levers  10.2. First Lever: Reducing Political Power 



10.2.1. Full Transparency of Budgetary Spending and Decision-​Making  10.2.2. Encourage and Promote Free, Independent Media  10.2.3. Term Limits and Strict Conditions for Holding Office  10.2.4. Reduce Discretionary Decision-​Making Power and Introduce Rule-​Based Politics  10.2.5. Reduce the Scope of Centralized Government  10.2.6. Minor Reforms That Could Strengthen the First Lever 

10.3. Second Lever: Re-​Empowering the Citizens 

272

278 278 281 282 284 287 290 295 297

300

x Contents

10.3.1. Give Citizens the Ability to Take Control of Their Tax Payments  10.3.2. Mandatory Participatory Budgeting for Community Public Good Provision  10.3.3. Granting Impeachment Powers to the Citizens for Direct Punishment of Officials 

300 303 305



10.4. Third Lever: Re-​Empowering the Community 

308



10.5. Second-​and Third-​Order Effects of the Three Levers 

316



10.4.1. Decentralize and Depersonalize Decision-​Making over Allocation of Public Funds 

Afterword  Notes  References  Index 

313

321 323 341 357

Preface This is a book about political power and inequality. It shows how the collusion between political and corporate power affects the distribution of incomes. Its most important goal is to offer a novel perspective on the long-​run origins of inequality by introducing the concept of elite networks and examining their effects on the distribution of power and incomes. Inequality is a man-​made, yet persistent condition of human history ever since the Agricultural Revolution. It is by no means a modern phenomenon, nor is it linked to one economic system. It has been present, to a varying level of intensity, during prehistoric times, in ancient civilizations, during feudalism, and in modern times in both capitalist and socialist economic systems. In premodern, Malthusian, societies poverty was a widespread constant. Living standards for the vast majority of the population were abysmal, while a small ruling class owned 90% of all societal wealth. In order to achieve above-​subsistence income during those times people could either engage in wars, generating wealth via violent extraction, or by being lucky enough to be born into the ruling class, which typically gained its status through violence at some point in their history. Inequality was high, but it also fluctuated and typically decreased only as a consequence of adverse events like wars, pandemics, or societal upheavals. In modern societies, particularly since the mid-​20th century, we are witnessing a decline of poverty and an unprecedent increase in living standards for the first time in human history. Inequality, however, still persists given that forces determining inequality are different from forces that used to keep populations stranded in poverty and dismal living standards. The argument presented in this book is that there exists a constant force among humans, an incentive rooted deeply within our behavioral patterns, that is responsible for persistently pushing inequality up. This does not imply that we have to justify it or endure it. Quite the contrary. But in order to solve it, we first need to understand its root causes. Which brings us to the second goal of the book: defining elite networks. Elite networks are informal social networks between politicians in power and top corporate executives or owners of politically connected firms where personal ties and long-​term interactions build trust and loyalty between involved actors. Both groups draw benefits from these interactions. Politicians stay in power and may extract bribes and other favors, while firms are rewarded with exclusive government contracts, favorable regulation, or direct subsidies. Top corporate executives successfully acquire rents for their firms as a result of their

xii Preface elite network interactions, and are rewarded by their firms with higher salaries. This consequentially widens the dispersion of earnings between the top 1% or top 0.1% of income earners (most of whom are corporate executives) and everyone else. This is what we observe in the modern era, where elite networks exist through interactions between corporate executives, owners of capital, and politicians. In premodern times their structure was different—​a network formed between the nobility, clergy, and the rulers—​but their effect on society was the same. Throughout human history an incentive to join elite networks has been driven by the same reason, a force I define as the violence power principle. The use of violence based on holding power motivated human behavior for ages and served as a perpetual justification for wealth extraction and entrenched inequality. Wealth and opportunities were constrained only to those with access or proximity to power—​the early ruling elites, which had the monopoly over the use of violence and coercion. Religion served an important role as well. Rulers used religion to grant themselves divine rights to govern, thus ensuring the loyalty of their subordinates. No one questions the divine legitimacy of a ruler, especially when the ruler upholds that ideological advantage through a network of preachers. The early elites thus had every incentive to preserve their position and make it permanent and hereditary. Only a descendent of nobility can claim divine right to a throne. Political and social institutions were designed as hierarchical orders with a goal to strengthen the elites’ legitimacy over the hold of power and use of violence. The army became a necessary institution to preserve that legitimacy. It was the dominant political-​economic system for centuries, even millennia, and it is the legacy of every country today. For thousands of years, wealth was condensed in the hands of a small elite with legitimacy to use violence in their quest for power. With the rise of Enlightenment and subsequently the Industrial Revolution, the emphasis on individual freedom started gradually overtaking the dominant coercion narrative. However, one does not simply erase the long shadow of history in how these hierarchical orders were established and what their effect was. The elites always had strong instincts for self-​preservation, so much so that even during revolutionary times, the majority of elite networks kept their fortunes and positions in the social hierarchy. They managed to permanently preserve the violence power mechanism, in some countries dominantly, in others latently. Only recently have our societies learned how to become more inclusive, giving impetus for the voluntary exchange principle—​coercion-​free exchange that delivers the greatest benefit to all involved parties, protected by inclusive political institutions delivering the necessary public goods. However, despite the enormous progress our societies have made because of the voluntary exchange

Preface  xiii mechanism, it is difficult to overcome a principle that has guided human behavior for such a long time. The violence power principle implies concentrating political and economic power in hierarchical governing institutions with little sense of accountability. This is true for governments and corporations, both of which seek to eliminate competing powers. Corporations today, similar to nobility in the past, seek protection of their interests via proximity to political power. This is what drives modern-​day inequality and is guided by rent-​extraction in the form of misuse and usurpation of political power. Obviously not all corporations exist under the realms of political power, but those that do are the drivers of income inequality in the same fashion as in the early days of capitalism or in premodern times: lowering risk-​exposure via seeking political protection through forming elite networks. Big Tech firms are a good example here. They achieved market dominance based on their proprietary innovations and the network externality effects arising from those innovations. However, today, when voters and policymakers want to limit their power by changing the regulatory framework—​such as limiting the use of private data or the scope of algorithms that anticipate our behavior better than ourselves—​the response from the tech industry is to infiltrate into elite networks. They hire former in-​office politicians to serve as de facto lobbyists, thus exploiting their deep network of connections to their benefit. This can be fully legitimate, but it is exactly the type of behavior that firms use to protect their status. The consequence is a growing dispersion of income driven by between-​firm differences. Whenever an innovator turns into a rent-​seeker, their high relative incomes are no longer purely a result of innovation or productivity growth. Joining elite networks is therefore a trade-​off between coercion-​free earnings on the market, and coercion-​driven earnings generated through proximity to power. The violence power principle explains the main motivation of elite network formation and how it is used to perpetuate economic inequality. Inequality, however, has also been cyclical, subject to ups and downs, which might seem contradictory to the argument just presented. If the origin of inequality is rooted in the quest for political power, which is a persistent incentive in any society, then we should always observe a consistently high level of inequality across every society. Inequality is subject to two additional forces that determine its long-​run cyclicality: one that pushes inequality up constantly, and the other that lowers it occasionally. I call the first a force of wealth concentration, and the second a force of occasional destruction. The force of occasional destruction consists of four types of events that have been recognized as powerful levelers of inequality throughout history: wars, plagues, state failures, and revolutions. In the 20th century, when inequality

xiv Preface declined faster than ever before, it did so because societies underwent 30 unprecedented years in the first half of the 20th century that included two major wars, a postwar pandemic, a major economic depression, and hyperinflation. All reforms enacted after these events were designed to prevent such outcomes from occurring ever again, and societies finally flourished. Therefore, only powerful, unpredictable events can seriously undermine inequality—​both directly, and through policy solutions in its aftermath—​simply because they tend to undermine the existing order and lead to a dissolution of existing elite networks. Over time, the elites adapt, meaning that inequality eventually resurfaces. This is driven by the forces of wealth concentration, where uninterrupted periods of economic progress often lead to misuse of political power to once again concentrate wealth within the elites. In preindustrial times this was obvious given that economic development itself was contingent on political power. Wealth was expropriated by the ruling elites. During the early stages of the Industrial Revolution inequality was even higher due to privileges granted by proximity to political power. In modern societies, the role of political power has been subdued and progress is no longer driven purely by the ruling elites, but rent-​seeking corporations still benefit from its proximity, further building momentum for inequality. The goal of this book is to show how long-​run forces behind inequality were always rooted in centralized and extractive political power. Centralization of power enables the network effect that creates elite networks, which enable rent-​ seeking opportunities to its members, consequentially increasing inequality. Inequality is not an artifact of a particular economic system, it is not a modern nor a natural phenomenon, but a man-​made phenomenon rooted deeply within the, often violent, quest for political power. Having redefined the sources of inequality, we also need a redefinition of its solutions. Inequality cannot be lowered or curbed by giving more power to politicians and regulators to deliver just outcomes. Greater centralized political power only increases incentives for elite network formation. Progressive taxation on the other hand, is merely dealing with the consequences instead of the true causes of inequality. A top executive will benefit from proximity to political power regardless of their top marginal tax rate. An owner of a firm with political protection avoids any punishment for regulatory noncompliance. If we wish to truly lower inequality and prevent incentives of elite network formation we must first and foremost lower centralized political power and re-​empower the citizens and the community, and we must do so by rebuilding trust and relying on the democratic trial-​and-​error mechanism. The final goal of the book is to deliver at least some answers as to how we can achieve this.

Acknowledgments Like every book, mine too was a product of a joint effort containing many revisions, different versions, helpful comments, and useful additions, however with me holding full accountability over all errors and omissions. The origin of the book was my Oxford DPhil thesis, written as a three-​paper dissertation from 2016 to 2019 under the supervision of my extraordinary mentor Professor Ben Ansell, to whom I own considerable gratitude. Not only for his academic prowess but also for his genuine guidance as a mentor during the DPhil and later in my career. His invaluable advice, comments, criticisms, and many suggestions helped make the thesis, and later the book, into what it is today. Of the three papers, two were published and their results are summarized and presented in the book (in Chapters 6 and 7), while the third paper has been adapted into the crucial empirical chapter of the book (Chapter 4). The rest of it has been completed in the years after my time at Oxford. Therefore, the empirical findings developed within the thesis represent the cornerstone argument of the book, but everything else, the entire theoretical and historical background developed to support the findings, was written later. In this process I had great support from my editor at Oxford University Press, James Cook, two excellent reviewers who greatly improved the quality of the arguments, and the entire OUP editorial team that handled my book. To all of them I also express genuine gratitude. During my time at Oxford, I benefited from conversations and comments on my work from a number of individuals, all of whom I owe a special thank you: Andrew Eggers, Elias Dinas, Robin Hardin, Jacob Nyrup, Josip Glaurdić, Valentino Larcinese, the late Alan Krueger, Joshua Angrist, Jacob N. Shapiro, Dejan Kovač, Boris Podobnik, Vedrana Pribičević, Jurica Zrnc, and Victor Menaldo. I would also like to extend a thank you to the participants at various conferences where I presented the empirical findings of my papers, which are crucial parts of the book: the 2017 and 2018 American Political Science Association Annual Meetings in San Francisco and Boston, the 2018 PIIRS Conference at Princeton University, the 2017 Nuffield Politics Graduate Student Conference at Oxford, etc. The findings on corruption and local government in Croatia (presented in Chapter 6) have been presented at the Croatian Parliament, in front of the National Council for Monitoring the Anti-​Corruption Strategy in November 2017. I thank the members of the Council for expressing interest in the research that came as a result of this thesis.

xvi Acknowledgments For data extraction, I would like to thank Mile Šikić and Marko Rakar for extracting parts of the original procurement database I used, and to Dejan Vinković for helping me assemble the network datasets. I would also like to thank my research assistants from Oraclum, Luči Krnić and Sanja Deur, for help on drawing all the network graphs used in Chapter 1. I am thankful to the Center for Responsive Politics for providing the data on lobbying and campaign contributions, and to the Wharton Research Data Service (WRDS) for access to the BoardEx database on individual-​level corporate and political connections. During my research phase for the thesis, I held numerous interviews with individuals involved in tracking corruption in the public procurement process in Croatia, as well as individuals involved in oversight or decision-​making regarding the allocation of TARP funds in the United States, all of whom wished to remain anonymous. A special thank you goes to Damon Silver from the TARP Oversight Subcommittee for providing important guidelines over the TARP distribution process. I am indebted to all of these individuals and institutions for their valuable and indispensable help in various parts of the thesis. While doing my DPhil research at Oxford, I was a member of Pembroke College, from which I received incredible support, both academic and financial. Pembroke provided me with an academic scholarship and gave me a travel grant for every single conference I wanted to attend and present my results. In addition to my college, I also received numerous outside scholarships: the Sir Richard Stapley Educational Trust grant, the IHS fellowship, the Koch Foundation dissertation grant (awarded to me by my later associate, Scott Alford), the John Blundell Scholarship from the Adam Smith Institute, in addition to numerous other grants. A special thank you to that end goes to Jessica and Peter Frankopan for the Frankopan Fund Scholarship I have received each year while at Oxford. Every single one of these scholarships were of immense importance to me, and I am thankful to everyone who supported my research. The one person, however, who has made the greatest effort in diligently reading all my work during my entire academic career was my father, Nikola. I am forever indebted to him for his continuous intellectual support, his immense personal library that was at my disposal since I was a child, and the numerous long conversations we still have about economics, politics, and history. Finally, the book would be impossible without the ongoing love and support from my wife, Barbara, and our three kids, Nikola, Josip, and Ema. To them I dedicate my work.

Introduction Connecting Elite Power to Inequality

Citizens of “developed” nations find the expropriative antics of foreign despots exotic, sometimes even laughable. But the shakedowns practiced by their own politicians are different only in the complexity, camouflage, and rhetoric surrounding the expropriation threats and ultimate extraction schemes. —​Fred S. McChesney (1997) Money for Nothing, p. 68

Can you buy a Big Mac in Tunisia? Back in 2008 McDonald’s tried to award their franchise to a local company to open a restaurant in the capital city. It was the “wrong” company. The government of the all-​powerful President Zine El Abidine Ben Ali stopped the deal, citing disagreements over the purchases of domestic Tunisian-​made meat in order to ensure that the meat was Halal. McDonald’s operates in many Muslim countries and certainly does produce meat that is Halal, even when it is not locally bought. The problem was that by doing so the franchise would avoid signing a big contract with the local meat company under direct protection of the president. According to a classified report1 from the American Embassy in Tunisia from 2008, “Corruption in Tunisia: What’s Yours Is Mine,” the real reason was obvious: there are still several examples of foreign companies or investors being pressured into joining with the “right” partner. The prime example remains McDonald’s failed entry into Tunisia. When McDonald’s chose to limit Tunisia to one franchisee not of the GOT’s [Government of Tunisia’s] choosing, the whole deal was scuttled by the GOT’s refusal to grant the necessary authorization and McDonald’s unwillingness to play the game by granting a license to a franchisee with Family connections.

This is a typical story surrounding Ben Ali’s 24-​year reign over Tunisia. The reign started in 1987 and ended abruptly in 2011 after mass protests that successfully overthrew him and later sentenced him to life imprisonment—​however in absentia, as he fled the country—​on charges of theft, money laundering, violence, Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0001

2 Introduction and murder. After he was ousted, his personal wealth was estimated to around $13bn, about a quarter of the country’s GDP. According to a 2014 World Bank working paper2 that reconstructed the story behind Ben Ali’s fortune, by the time he was ousted he and his family had ownership in almost 220 firms, which were responsible for 3.2% of the country’s output and 21.3% of its profits in 2010. These were the “right” companies to do business with in Tunisia. Ben Ali had a simple strategy. He manipulated the country’s investment laws to enact entry barriers for any competition and thus enabled his companies to charge monopoly prices. His family was present in all the major industries: telecoms, banking, transport, distribution, real estate, hotels, restaurants; altogether he restricted access in 45 different sectors where his interest was present. What were the consequences of such restricted competition and monopoly pricing on Tunisia’s population? Primarily, higher prices of goods and services (telephone prices on international calls, for example, were 20 times higher than the market price). More importantly, there were lack of employment opportunities to anyone not connected to the family’s clan, not to mention failures of many entrepreneurs or investors who were not connected to the government. Higher levels of income and wealth inequality were a natural consequence. The connected few in the political and corporate world benefited at the expense of the unconnected majority. The situation has not improved much in the decade after Ben Ali’s departure. Well-​connected firms still profit from privileged access to government. The dictator is gone, but the crony system remains. What does collusion between politicians and corporate executives look like? It looks exactly like Ben Ali’s Tunisia. It looks like the Siemens scandal of bribing public officials across the world to get lucrative projects, thus earning a hefty profit for both the company and its chief executives. It looks like the 1MDB scandal in Malaysia, a development fund set up by a businessman-​turned-​fugitive, Jho Low, who, using an elaborate scheme and a huge personal power network spanning from Middle Eastern royal families to Hollywood to Goldman Sachs, funneled billions of dollars both to himself and the then-​prime minister Najib Razak. Goldman’s executives earned $600m in fees, but their former Southeast Asia chairman Tim Leissner faced embezzlement charges, while Goldman is seeking a settlement with the US Justice Department. It looks like the Gürtel corruption scandal in Spain involving a businessman, Francisco Correa, who was funneling money to ruling party politicians in exchange for exclusive procurement contracts. It looks like the infamous Tangetopoli affair in Italy in the 1990s, uncovered by a special investigation “Manu pulite” (clean hands) that led to indictments of over half the Italian parliament and local government, effectively destroying the dominant Christian Democracy party in 1994.3 The system was so corrupt that a bribe or a “kickback” (tangente) was considered a standard

Introduction  3 commitment every firm who wanted to do a job with the government had to pay. Italy had its elite network institutionalized. It looks exactly like what was uncovered during “Operation Car Wash” (Lava Jato) in Brazil in 2014.4 One of the biggest “corruption machines” ever built involved 20 multinational corporations and a whopping $800 million paid in bribes. At the center of it was a Brazilian construction conglomerate Odebrecht, whose senior executives engineered massive frauds across Latin America and beyond. They were signing large construction contracts (building airports, power plants, dams, refineries) and were extracting money by creating shell firms with secret bank accounts to pay for fake invoices from fake customers. To cover it all up they heavily bribed whichever government official they needed, from local mayors to presidents. Their bribes helped fund political campaigns of several presidents across the continent, and pay off hundreds of legislators in each country where they did business. Needless to say, they built their own elite network, as efficient and lucrative as the one built by Ben Ali in Tunisia, and as institutionalized as the one in Italy. Lava Jato exposed the extent of Brazil’s (and Latin America’s) deeply corrupt system in which politics and business virtually hijacked the country’s economic growth and kept inequality among the highest in the world. Ben Ali’s Tunisia is a case in point of when a ruling politician captures the state and uses it for personal enrichment of himself, his family, his close cronies, and all the firms connected to them. Suharto’s Indonesia, Fujimori’s Peru, or Putin’s Russia are almost identical examples, to mention only a few. Sometimes the pull factor comes directly from politicians (like in Italy, Tunisia, or Russia), sometimes from corporate executives (as with Odebrecht, Siemens, or 1MDB), but in each case their collusion reaps negative effects for the vast majority of the population. Heavily protected cartels and monopolies raise prices thus lowering disposable incomes and living standards. The political and corporate collusion restricts climbing the social ladder only to those connected with the elites. In the extreme cases, because huge parts of public budgets get expropriated for rent-​ extraction, it reduces the supply of necessary public goods like health care, education, or basic infrastructure (sewage, water, electricity), leaving major parts of the population stranded in abject poverty. As exemplified by McChesney’s quote from the beginning of this chapter, there is not much of a difference—​in the logic of obtaining and consolidating power—​between a despotic ruler in a kleptocracy and the connections he has with local corporations and local cronies, and an elected official in an institutionally stable democracy and the connections he has with a plethora of powerful interest groups. Almost all despotic leaders in history have one thing in common: creating a powerful coalition with the nation’s elites that helped them preserve their position of power, very often at the expense of the people who

4 Introduction were condemned to poverty and dismal living standards. Whether these were the nobles during the ages of medieval kings, the conquistadors, and other mercantilist colonial elites during the Age of Discovery, the military juntas in postcolonial kleptocracies, the oligarchs in postcommunist kleptocracies, or even the modern corporate elites that finance political campaigns in exchange for preferential treatment—​in each case the collusion between business or military power with political power generated negative socioeconomic effects on the rest of the population. Modern democracies are hardly comparable even to their historical antecedents let alone monarchies or present-​day autocracies, but their elites are guided by similar principals. Particularly when the positions of power are subject to less scrutiny and lower accountability. Consider another set of examples limited to the United States, the beacon of democracy and a role model of a well-​defined constitutional order. During the second half of the 19th century New York politics was dominated by the infamous Democratic Party political machine Tammany Hall. This group of politicians did not just collude with business interests—​they merged politics and business into one, benefiting vastly from their accumulated position of political power. Tammany Hall was an organization notorious for its patronage systems and outright corruption. It built a massive organization of loyal supporters, making sure that its members were well-​placed within the system to obtain full control. Their actions included everything from direct bribery, extortion, budgetary misappropriation, vote-​buying, nepotism, even directly buying judges and thwarting any sign of reform of its deeply corrupt system of governance.5 The most famous leader of this organization was William “Boss” Tweed. Although he was only a mere member of the New York State Senate, his tenure as Tammany boss landed him immense political power, giving him direct control over the New York job market, as well as the voting process itself. Tweed was a large landowner in the city, director of a railroad company, a bank, a printing company, and a hotel. The law did catch up to him eventually. In 1877 he was convicted for stealing between $25 million and $45 million (which in today’s inflation-​adjusted terms would be over $1 billion6). Other notorious examples of Tammany Hall political-​made millionaires were John Kelly and his successor Richard Croker, who were running the machine like a “well-​oiled criminal organization,” or the case of George W. Plunkitt, who admittedly bought land he knew the city was planning to buy, and sold it back to the city for an inflated price. It was, in his own words, “honest graft.”7 There is hardly any difference between the modus operandi of Tammany bosses and any robber baron of the Gilded Age. Or Ben Ali in modern-​day Tunisia. Political corruption and collusion between politics and business did not end with the Gilded Age in America. It evolved. Consider the recent case of Robert Rizzo, a city manager of the small town Bell in California.8 Rizzo was in power for

Introduction  5 17 years in this poor, mainly Hispanic community with a population of slightly over 35,000 people. During this time, he managed to consistently misappropriate public resources for his own benefit and for the benefit of his small group of business and political cronies. He paid himself an annual salary of $787,637, almost twice as high as that of the US president, and four times as much as the state’s governor. The law caught up with him in 2014 and sentenced him to 12 years in prison on various corruption charges.9 A similar case of political bribery and misuse of power to benefit a small group of people close to him was that of former Illinois governor Rod Blagojevich, who ended up with a 14-​year prison sentence. His fellow party member and longest-​serving Chicago mayor Richard Daley avoided prison, but his reign was marred with significant controversies of allocating city contracting jobs to personal friends and political allies. National-​level politics in the United States is certainly not immune to collusion. However, the monetary amounts uncovered by investigations are quite modest compared to even US local politics, let alone their foreign counterparts. The House, for example, faced two big corruption scandals in the 1990s, the House banking scandal and the Congressional Post Office scandal, both of which led to embezzlement, bribery, and money laundering charges against several Congressmen. Around the same time DC politics was dominated by a powerful lobbyist Jack Abramoff, dubbed the “Man Who Bought Washington.”10 Abramoff was the embodiment of a powerful network of interest, a true superhub, the key person connecting business and politics. He bribed Congressmen with money, golf courses, tickets to events, and expensive meals to get whatever was necessary for his clients. House Majority Leader Tom DeLay was one of the victims of Abramoff ’s web of deceit, indicted over alleged money laundering. These cases are not few and far between. They are massive and reoccurring. Very often, however, political capture in the United States is not driven by politicians, but by the corporate world. Connections between Wall Street executives and office-​ holding politicians (presented in greater detail in Chapter 7), or between politicians and the tech community are much more subtle today than they were during the Gilded Age, but that does not make them any less benign than a relationship between a corrupt mayor and his cronies, or between a kleptocratic ruler and his oligarchs in any postcommunist society. If anything, the money involved is greater and their consequences much worse. Consider the frauds involving Elizabeth Holmes’s Theranos in 2016 or Sam Bankman-​Fried’s FTX in 2022. Each is a case in point of how unicorn start-​ups, offering fanciful technological solutions to big issues, used political connections to boost their companies’ valuations, only to create an exuberant self-​defeating spiral that ended up hurting their investors and defrauding their customers. The meteoritic rise of both companies was made possible primarily due to their

6 Introduction access and infiltration into the biggest elite networks in corporate and political America. Holmes filled her board with top-​tier former politicians, including two former secretaries of state, George Shultz and Henry Kissinger, in addition to other former senators, generals, and cabinet members. It was the connection with Shultz in particular that convinced a host of other former politicians and corporate executives from his network to join in and support the company. Bankman-​Fried went even further, becoming one of the biggest donors to the Democratic Party, using his parents’ political connections to navigate the venture capital investor field very quickly and successfully, while building a fraudulent product that lost billions for its customers. In these cases, politicians were among the many victims of the fraudsters, but they trusted them simply because they were part of the same elite network. Another good example of (mis)using elite network status to commit fraud was Bernie Madoff ’s four-​decade long fraud. Unlike Holmes and Bankman-​Fried who most likely did not initiate their ventures with an idea of committing fraud, and later became victims of their own fake-​it-​till-​you-​make-​it mindset, Madoff ’s business was a pyramid scheme by design practically from the onset. And yet, even when his fraud was becoming obvious and exposed in the media, he repeatedly used his influence to avoid prosecution for years. On several occasions he used his ties to the SEC to prevent ongoing investigations. He didn’t lobby or bribe anyone, nor was there any direct quid pro quo involved. He called the chairperson of the SEC, his friend, and politely questioned why they were even investigating him, which ended the investigation. For members of elite networks there is no need to buy influence. Personal relationships are more than enough to sway decisions. Eventually he was brought down, not by the SEC, but by the financial crisis of 2008, as more and more of his clients wanted to withdraw their money due to nationwide financial difficulties, and the pyramid scheme was fully exposed. The examples mentioned thus far were all, conveniently, fraudsters and criminals using political influence for personal enrichment. However, there are many more examples of nonfraudsters who use political connections to advance their interests—​the highly connected corporate CEOs, board members, and owners of politically connected firms which are the focus of this book. The vast majority of them are not engaged in any illicit or even unethical activity. They become illegal when fraud is the object, but much more often than not, none of these people are committing fraud. They are merely advancing their interests, and this causes unintended social consequences. Needless to say, for all those who use political protection to advance their interest, there are many more who do not and who succeed based on value added and innovation. These individuals and their firms are not the issue. The issue is a systemic fault of allowing certain elite networks to capture political power.

Introduction  7 Traditionally robust US institutions are failing to prevent such outcomes because they get captured by elite networks, who have little constraint in bending the rules to their will. The law does react, but only ex post, after a serious fraud has already been exposed (as with Madoff or Enron). It is becoming increasingly difficult to prevent political capture from elite networks ex ante. Fukuyama’s seminal contribution on US political decay presents a detailed historical overview of how strong US institutions, developed in the aftermath of the Gilded Age and as a direct response to cronyism of robber barons and political machines like Tammany Hall, started to subside and decay precisely because of interest group state capture and lack of institutional adaptation in the late 20th century.11 Lobbying activities, super PACs, or the revolving door practice of former staffers working for industry giants are all completely legal and even fully legitimate ways for various individuals and organized groups to pursue their interest. In this pursuit some groups will fare better than others. According to Olson, because of their inability to solve the collective action problem, the latent (unorganized, dispersed) groups will lose out at the expense of the privileged (organized, narrow) ones, thus skewing the distribution of resources toward groups with more power and with a more narrowly defined set of interests.12 The question is does the pursuit of one’s interests hamper market outcomes, benefiting the few at the expense of the many? The aforementioned cases, and many others occurring across the globe—​ from the least developed, institutionally deficient countries with very low democratic capital to the most developed, constitutionally well-​defined countries endowed with democratic capital—​suggest that even rich and advanced democracies are not immune to the logic of power and behavior of elite networks.

I.1.  Main Contributions The main contribution of the book is to show how elite networks—​informal social networks between politicians and politically connected firms—​generate a negative effect on inequality. The inequality literature mostly overlooks this very important mechanism of how the misuse of power translates into inequality. The mechanism is rather simple, it has been with us for centuries (with elected politicians and corporate executives replacing kings and nobility), and mostly works in the following way: politicians and senior corporate executives or owners of connected firms draw benefits from continuous mutual interactions. Politicians are given campaign donations and other forms of support which enable them to win elections and stay in office, while firms are rewarded with government procurement contracts, favorable regulation and legislation, monopoly

8 Introduction rights over a certain good or service, or direct subsidies in times of need. Top-​ level corporate executives responsible for securing rents for their firms get rewarded with higher compensation. This is where interactions of elite network members yield a direct negative impact on the distribution of incomes. Senior corporate executives (especially for publicly listed firms) are all in the top 1% and even top 0.1% of income earners in a country. The inequality literature is consistent in showing that the growth of incomes of the highest earners is the main cause of rising inequality in contemporary Western democracies. The broader implication of this book is that the high income inequality we observe today could also have been influenced by political connections of top income earners. In fact, the book shows that inequality typically is a result of extractive political power, both today and historically. Individual-​level empirical findings presented in the book validate the many theoretical and case-​study arguments which suggest that usurpation of power by rent-​extracting executives carries a large weight in explaining the causes of inequality. The book does not aspire to become a theory of everything. Its long-​run historical analyses, theoretical explanations, and detailed empirical results are necessary to justify the inclusion of political capture by elite networks as one important and crucially overlooked factor of explaining inequality. However, while contemplating the theory and empirical results presented in the book, it is important not to neglect all the standard economic theories that explain the rise and persistence of inequality, nor to neglect economic theories that explain corporate success and failures based on competitive advantages, differences in productivity, and adaptation of innovation. Throughout the book I reference the vast and instructive inequality literature13 according to which the leading cause of rising inequality worldwide over the past four decades has been concentration of wealth and incomes into a small minority of the population, the top 1% or 0.1% of income earners. Globalization, technological progress generating creative destruction, global competition for talent and skills, particularly among top-​tier corporate bosses leading to the emergence of superstars in the corporate sphere (the “supermanagers”14), concentration of firm power, tax policy reforms, and a number of other factors all have merit in explaining the rising trends of the past decades. However, these theories, with a few exceptions,15 only explain the short-​term trend, mostly of the 20th century, and do not engage in providing a long-​run theory of inequality persistence. Inequality was present throughout recorded human history and across different political systems and forms of governance, meaning that its origination has to have deeper underlying root causes. In this book I link the persistence and cyclicality of inequality to political power, examining how the misuse of power by those with proximity to it presents a yet unexplored factor in observing the inequality phenomenon.

Introduction  9 The mechanism of elite network political capture does not, in any case, suggest that success of firms is contingent on securing political support, especially not in post–​Industrial Revolution times. The way firms adapt innovation and how they add value to their customers has been and always will be the crucial determinant of their success. The majority of between-​firm differences will arise as a consequence of specific and enduring competitive advantages. However, in that competitive race, some firms will make a decision to secure privileges contingent on the personal relationship their management has with key government officials. At a certain stage of the firm’s life cycle, in order to preserve their status, some firms can and will engage in political capture. Incumbents use political capture to influence policies and prevent challengers from disrupting them. Obviously, this cannot last forever. All incumbent firms that were large and powerful for a very long time eventually succumbed to their challengers when time was ripe. But this creative destruction process takes time, and before being completed it renders economic inefficiencies as the market mechanism is prevented from performing its optimal allocation process. Industrial economic theory, theory of the firm, or other microeconomic toolkits that are useful in understanding between-​firm differences fail to explain what happens when political power intervenes in the allocation process. Firms thrive on attracting top human capital and securing a technological edge in order to maintain their competitive advantage. However, human capital and technology are nonuniformly distributed, and they tend to flow to very specific geographical clusters. These same clusters attract not only capital but also power, and it is in the intersection of money and power where elite networks tend to form. The clustering is driven by motives of securing competitive advantage and attracting human capital, but its consequence is even greater concentration of wealth and power within a small group of people. This is not done by some intelligent design, but as a consequence of a network homophily effect that wants to optimize efficiency and performance. Out of that network effect arises the spontaneous incentive for greater assortativity up to a point where a network becomes highly topocratic—​it matters more who you know than what you know. At this phase transition point a firm will cross the line from being a customer-​seeker (focusing on competing for customers in the marketplace via its competitive advantage and innovative solutions) into a rent-​seeker (competing in the political arena for favors and privileges). Historically, this phase transition was only possible after the Industrial Revolution era. Industrialization brought new opportunities for small innovators and merchants to generate above-​subsistence wealth, as they gradually, over more than a century, started to replace the landed aristocracy and nobility. But at the same time, early capitalism was full of industrial behemoths, led by individuals who could hardly have built their empires without political support

10 Introduction (from 19th-​century US robber barons like Rockefeller, Carnegie, Vanderbilt, Morgan, to Tammany Hall businessmen, to the 18th-​century British East India Company, etc.). Those who had initial backing of political power made corporate conglomerates, others were forced to compete on the marketplace, innovate, and grow through the value added they offered to their customers. The transition from one type of firm to another is still a latent motive present in firms when faced with declining market share and nimble challengers. The self-​preservation instinct is strong and it will lead to elite network formation if necessary. Many firms, particularly small and midsized ones, never get to that transition point, but these firms are not where inequality originates. Inequality originates among the top 0.1%, meaning that we must look at the structure of those within that group to understand which factor, in addition to skills, talent, or luck, drives their abnormal earnings, and ask, can this factor be attributed to market forces alone, or is there a missing link? *** Before we dive more deeply into the relationship between elite networks and inequality, it is important to settle the definitions of a few concepts used throughout the book. First and foremost, the definition and classification of rents and the concept of rent-​seeking. Rents are typically defined as a source of income from exclusive ownership of a limited resource, without adding any value or wealth to society. Shiller uses a classic example of a feudal lord who places a chain across a river on his land and then charges a fee to any boat that wants to pass through.16 There is no value added to society, a cost is imposed to users (for something that used to be free), and the only person to benefit is the lord himself. This simple example illustrates that the opportunity to extract rents must be given by some source of power and derived from ownership over a limited resource. This is where the political process kicks in, as rents could only be obtained if the rights to extract them were given to a market participant by holders of political power. The state can generate rents in a variety of ways. It can enact barriers to entry for competitors through things like patents, intellectual property protections, or licenses to specific occupations, it can directly grant monopoly rights over a resource, it can artificially reduce supply via things like tariffs to protect certain industries, it can give out subsidies, or it can engage in illicit activities and corruption where a monetary concession is exchanged for some form of privilege. Then there are modern rents, arising from digital technology firms and their network externality effects. The modern tech giants didn’t generate their market dominance by obtaining political protection through barriers to entry or monopoly rights. They’ve achieved it through their innovative products and demand-​side economies of scale where adaptability of a product depends on how

Introduction  11 many other people use the same product. After becoming too big and realizing their dominance through pure disruptive innovation, they are now seeking to preserve their position through lobbying the regulatory process over the usage of personal data and algorithms to target users. They are only now resorting to what the political economy literature defines as classic rent-​seeking—​misusing the political process to their benefit and preservation of status. The usage of the concept of rents in this book is deliberately broad, but also in line with how the political economy literature tends to examine rent extraction. Rents are just one way of describing the outcome of what political influence brings to an elite network member. They are an outcome of misusing political power in order to generate a nonmarket source of income or other nonmarket benefit. And it is through the misuse of power that rent-​seeking tends to occur. When top executives no longer focus on their disruptive innovation but shift their activities to elite networking in order to affect regulation and legislation, this is when they are rent-​seeking: trying to gain a privileged position on which they prove their value to their company. Rent-​seeking is a wealth-​reducing activity that generates a host of negative outcomes by impairing the allocative efficiency of the market mechanism. As far as the literature is concerned, rents are by definition a by-​product of political capture. This is also how they are seen in the book; an outcome of elite networking influence over political power. Rents don’t get translated into inequality per se; the consequence of elite network interactions do. The definition of inequality also comes with a clarification. In the book I interchangeably use the concepts of “income” and “wealth” inequality, even though the two are not the same nor do they necessarily have the same causes. In the initial historical overview in Chapter 2, I focus the majority of my attention on wealth inequality as the main source of inequality during the preindustrial age, up until the 20th century. Income from capital—​owning land, means of production, or inheritance—​was the key source of differentiation in wealth. In the 20th century this changed; income from capital lost its relative importance, and income from labor became the crucial determinant of differences in personal wealth.17 Hence the necessity to, at least in principle, equate wealth and income inequality; one was important throughout human history until the 20th century, and is the subject of the theoretical part of the book, while the other became important as of the 20th century, and is the subject of the empirical part of the book.

I.2.  Why Join Elite Networks? The Three Principles There is a deep, structural connection between elite power and the consequences it projects on the distribution of incomes and wealth in a society. In order to

12 Introduction theoretically justify the origin of such a connection the book must answer three basic questions: (1) why is there an incentive to join an elite network?; (2) why should elite networks have any effect on inequality, or better yet how do their actions translate into inequality?; and (3) if there is a relationship, how does it explain the cyclical nature of inequality? The following three subsections offer a brief answer to each of these questions which are later tackled, theoretically and empirically, throughout the book. The first theoretical question concerns the motivation for elite membership. Why do people have incentives to be connected to those in power, and why do those in power need a crucial group of key supporters around them? To answer this, we must first develop a theory of how people mitigate risk, ensure their survival (how they realize the Lockean necessity of subsisting18), and most importantly preserve their status. Within a society, from ancient times until today, there are three basic principles of securing one’s existence and status: (1) the voluntary market exchange principle, (2) the principle of gift exchange, and (3) the violence power principle. Only after understanding these three principles can we provide an answer as to why people have incentives to form and join elite networks. The voluntary market exchange principle is present when people satisfy their needs by exchanging goods and services voluntarily without any coercion, realized within a set of institutions whose rules all actors consensually accept. These institutions can arise spontaneously (based on social contracts or trust) or can be designed as public goods which offer a number of institutional protections to engage in a transaction or relationship with another individual: protecting basic human rights; ensuring equality before the law and equality of opportunity; protecting and enabling free exchange and entrepreneurship; protecting freedom of speech, freedom to assemble, and freedom to promote ones interest; and protecting private property. Societies have not been able to achieve any significant form of progress in the long run without abiding to the voluntary exchange principle and the institutional support behind it. Historical examples of this are ample. Whenever there was even a remote possibility to gain wealth based on voluntary exchange (or free trade) societies prospered.19 Whether this was in ancient times of the Roman Republic, the Persian Empire, or Chinese dynasties, or during the postindustrial era with a gradual adaptation of democratic institutions, societies prospered when exchanges between individuals were not bounded by any form of coercion and when people entered into exchange on equal terms. When institutions were designed to minimize incentives to engage in coercion during the process of exchange, economies prospered. When institutions were designed to do the opposite, economies stagnated.20 The greater the freedom to engage in exchange, the better the economic outcome of a society.

Introduction  13 The second principle is the gift exchange principle. It operates in many forms. Anthropological research has found that a number of small communities fostered the gift exchange culture as a predecessor to what we know today as the market exchange. There were no monetary transactions involved, just the necessity to repay one’s gift with another or with an act of kindness. Even today many occupations still depend on gift exchanges. There are various forms of charities and foundations which are financing universities, think tanks, and cultural events, not to mention tackling a host of social issues, from poverty to fighting diseases to basic education. The most widespread form of the gift exchange principle are gifts received within a family, with particular emphasis on inheritance. Monetary transactions within a family are purely altruistic and are never driven by market exchange principles nor do they necessarily rely on outside institutions to enforce them (with a few exceptions, inheritance and divorce being the most prominent). They serve the purpose of solidarity through which parents take care of their children and when they get old, their children take care of them. No contract is necessary to enforce this relationship. The principle of inheritance is therefore a natural tendency of any person with children to leave whatever he or she created during their lives to the ones they care about the most. However, inheritance can also be the source of inequality, particularly if a large fortune is being transferred to the benefactor. The issue of its legitimacy is not in the act of inheritance itself, but in the source of wealth: was it created and maintained according to the free exchange principle, which means that the benefactor needs to operate under the same principle to ensure the wealth is not lost, or was it created and maintained based on proximity to political power, in which case the inheritance is rightly considered unjust. This brings us to the third principle of securing one’s existence: the violence power principle—​the negation of the free exchange and solidarity principles. The ability to wield violence and secure power can only be used by a small minority within a society: the ruling elite and everyone in its proximity. Only the ruling elite has the legitimacy to use violence and coercion granted by its hold on power. Violent expropriation of property and coercion of individuals into slaves dominated human societies until the modern age and has existed throughout humanity as a permanent, latent pattern of behavior. When the costs and risks of violence are lower than the anticipated benefits, a person has an incentive to secure their existence through coercion and expropriation. In premodern times the cost of theft was minimal and the risk was low, particularly if one had no competition in theft and could rely on weapons and soldiers as a legitimate threat. Wars in premodern times were seen primarily as a method of enrichment of all men who went into battle through pillaging the conquered territories. The latent desire for violent conquest was in the human DNA ever since the Homo sapiens started their spread from Africa onto other continents some 70,000 years ago leaving

14 Introduction every other human species they’ve encountered extinct. It was the dominant form of behavior throughout antiquity, the Middle Ages, the ages of discovery and colonization, and even during the aftermath of the Industrial Revolution. The key benefactors of violent conflict had strong incentives to preserve their position, to make it permanent and hereditary. They molded political institutions to strengthen their extractive power with the help of religion as a justification of their divine right to hold power. The result was several millennia of slavery and feudal systems and a society confined within the Malthusian trap of long-​term relative economic stagnation, a relative demographic stagnation, and Hobbesian living standards. Applying the violence power principle inherently led to rising inequality where wealth and opportunities were constrained to anyone with access to power or proximity to power: the rulers and the nobility that supported them. Over time, with the rise of Enlightenment and gradual development of western European societies during colonial and postindustrial times, the emphasis on individual freedom started replacing the dominant coercive collectivism. Deidre McCloskey claims that the change of rhetoric and ethics and ideology, in other words new and better ideas, were the key motivational factors behind the Great Enrichment achieved via the Bourgeois Deal.21 The change in rhetoric too was a gradual process, to a large extent driven by the spread of books after the invention of the printing press in the 15th century. Spreading knowledge via books was much more efficient than ever before, but it took 200 years to create a generation of intellectuals and scholars—​itself a minority elite that led the Enlightenment movement.22 Nevertheless, despite these positive changes where the voluntary exchange principle was very slowly overtaking the violence power principle, societies are still consumed with the latter. Centuries of philosophical and political economic discussions on the limits of state power—​that carry on until this very day—​are best proof of this. History shows many examples of attempts of special interests to misuse the voluntary exchange principle and establish a clear dominance hierarchy with the legitimacy over the use of violence and coercion. In Europe this had last happened in the 20th century with the rise of Nazism and fascism. In such cases predatory behavior, lawlessness, and destruction become the modus operandi of political action. Even with their defeat it is impossible to exterminate incentives for coercion based on power. The violent power mechanism is therefore permanently present, either dominantly or latently, in every society. The rise of the voluntary market exchange principle in the battle of ideas ever since the Enlightenment and particularly after the Industrial Revolution generated extreme levels of wealth accumulation, but it also perpetuated extreme disparities in wealth. The question is are these disparities in wealth purely a consequence of the voluntary exchange principle, or are there different, nonmarket

Introduction  15 factors that can be attributed to the rise of inequality? The argument delivered in this book is that the usage of violence based on power has been a persistent incentive in a quest to extract wealth throughout human history. The rise of the voluntary exchange principle, despite its enormous ability for the first time in history to generate inclusive wealth cannot simply overrule a principle that has guided human behavior for more than eight millennia. The violence power principle is achieved through institutionalized political and economic power. Power concentrated in political parties and government institutions whose internal structure is highly hierarchical. Political power knows no competition. It is monopolistic and depends first and foremost on how much the citizens hold it accountable. Without accountability any type of government descends into the basic instinct of autocratic decision-​making and misuse of power, regardless if it operates within a democratic context or not. Governments facing zero accountability are free to redistribute resources however they see fit, just as any autocratic government in history did. Corporations are similar hierarchical institutions. Internally, their governance structures are not democratic but autocratic. When they depend on market principles and free competition, they are forced to adapt internally to maintain their position on the market. However, corporations often attempt to avoid the costs of adaptation and achieve market dominance based on gaining privilege from access to political power. Then they are free to transfer their costs and risk to the consumers, worsening their position. Corporations will therefore have a stronger incentive to reduce risk and costs by achieving a politically privileged position. They will enter into elite network relationships in order to achieve a nonmarket way of capturing a market position. They are encouraged by the prize in terms of nonmarket rents. Their profits may arise from their market power, but their market power has been granted to them by political power. Many individuals seek to use access to power to secure a rent-​seeking position with a much lower risk and lower cost than when having to engage in a free exchange mechanism. Access to political power is gained through personal connections and a system of favors that build trust. Whoever enters into such relationships has incentives to achieve a nonmarket way of capturing a market position. Despite the fact that there is a risk of getting uncovered or losing power, individuals enter elite networks to, essentially, lower risk. They seek protection from politics in terms of legal or regulatory acts, or they seek direct rents from government contracts. From the perspective of the firm this interaction is economically rational as it involves trading off low cost (of effort and engagement) for a great potential return that involves reduced risk exposure. The acts of lobbying and campaign spending for example are completely legal and legitimate ways of influencing government decision-​making. Exchanging personal favors between connected individuals in positions of power is not necessarily always

16 Introduction legal but it is, in most cases, completely legitimate. Using these connections implies trading off costs of engagement with lucrative prizes—​rents.23 The incentive to join an elite network can therefore be summarized by trading off the voluntary exchange principle with the violence power principle. This happens whenever the costs and risks for building a position of proximity to power are lower than the expected benefits from engaging in voluntary market exchange. Chapter 5 develops this further. Elite network theory is very much aligned with the North, Wallis, and Weingast violence and social order (VSO) framework24 but also the narrow corridor (NC) framework developed by Acemoglu and Robinson.25 We can think of the violence power principle as a typical outcome of the natural state, a limited access order (VSO) run by a despotic leviathan (NC), where wealth and power are concentrated in the hands of those who have a monopoly over the use of violence. This was the dominant form of social organization throughout human history. Limited access orders are by necessity dependent on kinship and close personal relationships. Governing institutions themselves become personalized, and economic outcomes are manipulated by the ruling elites which cater to privileged interests. All political and economic outcomes are merely a consequence of interpersonal relationships among agents within the ruling elite networks. Liberty is constrained and wealth allocation is determined by the despotic leviathan. On the other hand, the voluntary exchange principle is the foundation of open access orders (VSO) or shackled leviathans (NC). In it, institutions are depersonalized and subject to promoting equal opportunity and free exchange. This makes it much harder to erect dominance-​based hierarchical orders and develop clientelistic relationships, making it much more difficult to use violence to distribute wealth. However, maintaining such an order is not easy nor straightforward. Institutions constantly evolve, but even when a society enters the narrow corridor there is no guarantee it will succeed in solving the violence power incentive that propagates the concentration of wealth and power within elite networks. What is necessary is to impose and maintain the Red Queen Effect26—​always running just to stay in the same place—​a metaphor for persistent societal mobilization to preserve the foundations of liberty and keep in check the power structures running the country. An important addition of elite network theory to the VSO and NC frameworks is an explanation of how individuals within different political orders adapt to the principles of securing survival, and what serves as motivation for economic and political activity in the first place. When the institutional order of a society is a limited access one (a despotic leviathan), in which the violence power principle guides social interaction, elites will form around whoever wields power. When an institutional order is an open access one (shackled leviathan), dominated by

Introduction  17 the voluntary exchange principle, incentives to use violence are curbed and elites adapt by engaging in the political process in legitimate ways. But the struggle for power remains. Given that societies have spent most of their history as limited access orders guided by the violence power principle of survival, any transition into voluntary exchange takes time and considerable effort in order to curb elite network power incentives. However, it is only within open access orders that elite network incentives can be curbed and placed under control. Under closed access orders, this isn’t even possible. Once the transition from a limited into an open access order is achieved, the issue still becomes how to curb power concentration among the elites. A society needs a strong state and ruling elites to ignite development. But a society also needs strong social mobilization to keep both the state and the elites in check. Problems arise when elites become too powerful and hidden so that society is unable to constrain their power. Hence the necessity for the community (the “third pillar” of every society27) to involve itself into the political process to maintain the Red Queen Effect and preserve the balance of power between society, the state, and its elites. This argument is presented in the final chapter of the book, Chapter 10, examining how to curb the violence power principle by reducing centralized political power, and thus reducing incentives for elite network usurpation of power.

I.3.  Why Should Elite Networks Have Any Effect on Inequality? One of the central arguments of this book is that elite networks are and always were responsible for the accumulation of wealth at the top. In order to understand why elite networks carry such an effect on the distribution of wealth and income, we need a brief historical overview of the origination of wealth and the creation of first power structures. This argument, developed in greater detail in Chapter 2, reinforces the logic behind the creation of elite networks and helps understand how they generate an effect on inequality today. A broad consensus among economic historians and anthropologists is that the creation of wealth came as a result of plant and animal domestication during the Agricultural Revolution some 10,000 years ago (8,000 years B.C.).28 Domestication created food surpluses and consequently a possibility of food storage. Having surplus food meant that there was no need for everyone to work in agriculture anymore, encouraging people to specialize and do other things such as to become bureaucrats, soldiers, and rulers. These nonproducers could then be fed from the surpluses created for the whole society. This crucial factor

18 Introduction of greater food production enabled the growth of early populations and further encouraged sedentary lifestyle as well as larger and denser societies. It was the starting point for the origination of political and religious organizations, both of which were crucial in enabling cooperation between unrelated individuals in large societies. One of the underlying tasks of the political organization was to protect the food surpluses by organizing armies. Thus, the first soldiers emerged, loyal to the ruler who further strengthened his rule by assigning himself divine characteristics (which is where religion comes in as the first form of elite propaganda). Another consequence was a gradual development of the technology for warfare (guns and swords), which was useful in later conquests to acquire new land and amass new wealth. Therefore, the creation of initial wealth in terms of surplus food almost instantly enabled the creation of a hierarchical power structure that included a ruler, a military to support him, and a religious leader that justifies his divine existence. All of these formed the first ruling elites. In addition to protecting food surpluses, the first proto-​states also had a role in protecting domestic population from outside violence. Olson29 describes this as a process of replacing a social disorder called roving bandit with a stationary bandit model. People preferred the certainty of having one ruler protecting them to the uncertainty of having a series of roving bandits randomly pillaging them. The demand for protection under one ruler implied granting unlimited power to that ruler, accepting his authoritarian leadership, and accepting his use of violence to both protect but also expropriate. As the proto-​states were gradually expanding into the first empires (Sumerian, Egyptian, Akkadian, Babylonian, Assyrian, Persian, Chinese dynasties), and later into more powerful and organized empires (Roman), they were at all times characterized by a hierarchical and autocratic social order with a ruling elite at the top of the wealth distribution—​in other words a stationary bandit model. The most important artifact of the stationary bandit model was the role of the state’s ruling elites in defining early ownership rights over land and all types of property (including slave labor) and enabling the transmission of wealth within a society. Scheidel argues that the early premodern states and empires enabled the concentration of income and wealth within the ruling elite through basic rent-​ extraction. He refers to these elites as the “original 1 percent” whose purpose was to uphold power in order to seize wealth.30 The very distribution of income and wealth in early societies and early empires was based on proximity to power. Any income generated from labor or capital by the nonelites was quickly expropriated by those with proximity to political power, leaving the vast majority of the population to live on subsistence income. The stationary bandit model, based on extractive political institutions—​to borrow the term from Acemoglu and Robinson31—​could explain why the world had been stuck in a Malthusian trap for 10,000 years, ever since evolving from hunter-​gatherer societies. This

Introduction  19 natural economic system where technological progress only led to population growth instead of better living standards,32 was supported by the holders of political power. Having or being close to political power was the only way to become rich (in relative terms) in a Malthusian economy. This argument is further supported by Milanovic, Lindert, and Williamson,33 who find that preindustrial societies (ranging from ancient Rome and Byzantium to medieval England, Holland or France) were characterized by a very high extraction ratio—​the rate at which potential inequality is turned into actual inequality. The higher the extraction ratio for a given society, the greater the wealth extracted by the top income groups, and hence the more unequal the society. Essentially, they confirm the idea that extractive and repressive ruling elites were responsible for creating and maintaining unequal societies. To sum up, political inequality, or the inequality in the distribution of power, was the primary source of income and wealth inequality ever since our species evolved from hunter-​gatherer societies. This artifact of societal development was, and still is, a key characteristic of many states and nations—​all, in fact, which have had or still have some type of authoritarian rule. After the fall of the Roman Empire, medieval kingdoms replaced the earlier structures and enabled the extractive stationary bandit model to spread out, giving rise to the first nobles as the supporting elite group of the ruler. Their political power gave them the right to own land, (slave) labor, and ultimately amass large wealth. The medieval nobility and the elites that were created at the onset of the Middle Ages, as well as the fortunes that they’ve amassed are still in the upper tail of the income distribution until this very day.34 Clark’s natural law of social mobility predicts a regression to the mean over time for elites and impoverished groups, however this regression is very slow; it can take up to five hundred years for an elite family to regress to the mean. The seeds of today’s inequality were therefore planted during times of stationary bandit extraction-​based politics. This is the legacy that elite networks have had on the distribution of income and wealth. Piketty’s seminal contribution testifies to this argument.35 His long-​run verdict is that European inequality of the 18th and 19th century was driven primarily by rentiers and income from capital. This was, in fact, a characteristic of all traditional agrarian societies. The pre-​20th-​century elites lived on rents they extracted from their ownership of inherited capital and land. The upper decile of income earners owned over 90% of total wealth in a country, where the top 1% (kings and nobility) owned about 50%–​60% of it while the other 9% (landed gentry and clergy) owned the other 30%–​40%. A vast majority of people owned almost no wealth at all. Gaining wealth—​land or capital—​was possible either through expropriation, carried out by the rulers themselves, or by inheritance, which is how wealth got transmitted among the nobility and gentry. The income hierarchy of pre-​20th-​century societies was therefore dominated by capital as the

20 Introduction main source of income. This means that a person was much better off to inherit wealth rather than to earn it. A top 1% inheritor had about 2.5 times more income than a top 1% labor earner. No job could be as lucrative as owning land or capital. Piketty’s fundamental force for divergence—​ where the return on capital (r) is bigger than the economic growth rate (g)—​was a defining force of all agrarian and pre-​20th-​century societies. These were societies which had a high concentration of wealth among the elites in addition to a high level of persistence of wealth in the hands of the elites. Even after revolutionary periods this artifact did not change. Clark shows this perfectly on the example of China’s Communist revolution, which despite being determined to purge the country of its “class enemies”—​the landlords and businessmen—​still has these “class enemies,” originating from the same elite surname groups of the late 19th century, overrepresented in the top earner groups in modern-​day China, and are even present in the highest structures of its Communist Party. In general, Clark’s conclusion for a number of countries (Sweden, United States, United Kingdom, China, India, Japan, Korea, and Chile) is that social mobility across all societies tends to be very low regardless of any societal changes, including revolutions, major policy reforms like the introduction of the welfare state, or even ground-​ breaking events like the Reformation, the Enlightenment, or the Industrial Revolution.36 It was not until the beginning of the 20th century when the structure of capital started to change and its value relative to national income started to decline. The two world wars and the Great Depression between them marked the ultimate downfall of the rentier and made the shift from what Piketty calls a “society of rentiers” to a “society of managers,” or a “society of superstars.” In other words, the top decile of the income distribution has ever since consisted mostly of individuals who live on wages instead of rents, i.e., from labor instead of capital. Today the top 1%—​the new elite—​are the top managers (CEOs, presidents) of global companies and entrepreneurs who created entirely new industries. Their income is mainly determined by labor, however they do also benefit from exposure to financial markets through capital income and returns on investments (e.g., equity positions and dividends). The structure of the elites had changed in the 20th century. However, their resulting effects remain the same. Atkinson, Piketty, and Saez attribute the rise of inequality since the 1970s in the United States (but also in other Anglo-​Saxon countries) to the rise in incomes of the top 1% and top 0.1% of income earners, where salaries and capital gains from stock investments represented the largest contributor to this increase.37 The so-​called rise of the supersalaries of top managers, a consequence of a multitude of factors that include an increasing demand for top talent (the superstar effect), significantly advanced not only their

Introduction  21 position in the income hierarchy but also, more importantly, their position in the social hierarchy. Furthermore, Bakija, Cole, and Heim confirm that 60% to 70% of individuals in the top 0.1% of the income distribution are top managers of global companies, most of them in nonfinancial industries.38 By all means of measurement this group is the new elite. Supermanagers signal their importance to their organizations by having access to a wide network of influence. Their degree and betweenness centrality (the positioning of an individual in a key space of a network where he or she is highly connected to all other nodes) enables them considerable power in justifying their own abnormal earnings (abnormal in a sense that their total compensations are higher than their productivity). Chapter 4 will uncover which effect is more important—​being connected to a lot of people, or being connected to the key people in politics. Historically elite persistence was always closely correlated with political power. One could not maintain its dominant position in the social hierarchy without being somewhat dependent on the ruler. Today, this relationship, although different in the scope of political power the ruler can use to expropriate its subordinates, is essentially the same as before. The power of a ruler to take away human lives has vanished in democracies (at least in domestic affairs), however the ruler still has enough power in terms of budgetary redistribution so that his role attracts anyone dependent on these resources, be it various interest groups seeking to gain public resources or special treatment for its members, or firms seeking to gain regulatory, legislative, or other form of advantage over its competitors. Furthermore, it has been shown that supermanagers often use their political power to lower top marginal income tax rates or finance various NGOs so as to indirectly promote their cause.39 This too can affect the rise of top incomes and widen income inequality. Arguably this political effect—​lobbying and using campaign donations to persuade politicians to make legal changes in the tax codes—​is part of a much wider network effect. The way a supermanager gains access to a politician is not through some crude and well-​defined institutional mechanism where more campaign funding automatically transfers into favorable decisions for the giver. No, influence is gradually constructed through a series of informal meetings and gatherings, where favors are exchanged between powerful individuals. This is the foundation of an elite network.

I.4.  The Cyclicality of Inequality Inequality is a cyclical phenomenon. Its cyclical nature implies that the elite network framework used in this book needs to be able to explain why inequality

22 Introduction sometimes goes up and why it sometimes goes down. If inequality and concentration of wealth within a select narrow elite are indeed consequences of political power, and if hierarchical political orders led by autocratic rulers were indeed the dominant sociopolitical orders ever since the Agricultural Revolution, well into the 19th and 20th centuries, then the logical conclusion is that inequality should have been very high and relatively stagnant during the entire history of mankind. Elites always existed and they always successfully extracted and concentrated wealth due to their access or proximity to political power. However, even if inequality observed during preindustrial times tends to be really high in certain time periods and certain areas (wealth inequality Gini coefficients used to be as high as 0.8 and 0.9)—​it is never relatively stagnant. On the contrary, it is quite dynamic. The Kuznets theory had a very simple inverted U-​shaped explanation where inequality tends to be low for lower levels of economic development, then it significantly increases as society undergoes a transition to an industrial society, after which it starts to decline due to forces of modernization.40 This theory was limited to its historical context: Kuznets was observing the pre-​and postindustrial trends in inequality until the 1960s when he made his seminal contribution. However, the Kuznets theory fails to explain the U-​shaped inequality trends that occurred by the end of the 20th century. It should have never happened in developed societies. This is why Milanović suggests that inequality moves in cycles, so-​called Kuznets waves, thus explaining the peak after the Industrial Revolution, the trough after the wars in the 20th century, and the new peak at the end of the 20th century. Milanović argues, rectifying Kuznets, that the 20th-​ century upswing in inequality that falls beyond the scope of Kuznets’s original model is, in fact, a second Kuznets curve appearing in modern times and is being driven by the same forces that drove the first Kuznets curve—​advancements and innovations in technology driving a shift from manufacturing into services (just as the first Kuznets wave was driven by a shift from agriculture into manufacturing), and the effect of globalization which increased concentration of returns from both labor and capital at the top. The decline of inequality in preindustrial times, according to Milanović, was driven by random events like wars, epidemics, or new discoveries given that there was no relationship between relatively stagnant incomes and inequality in preindustrial times.41 The only thing missing from the Kuznets waves theory is an explanation of inequality trends in the very long run. It is an excellent toolkit for understanding modern cycles of inequality since the 19th century, but it does not deliver a theoretical justification of why inequality moved upward since the first settler societies, throughout antiquity and the Middle Ages, and during modern times, both pre–​and post–​Industrial Revolution. Random exogenous events might have exerted pressures to lower inequality, but the reason for its persistent

Introduction  23 upswings is deeply rooted somewhere else. This book provides an answer to the crucial question: What is the main driving force behind inequality throughout human history? The elite network theory recognizes two forces that affect the cyclicality of inequality over time. One force pushes inequality up constantly, while the other force pushes inequality down occasionally. The occasional force, which I will call the force of occasional destruction, is composed of four types of events, some endogenous, some exogenous, all perfectly encapsulated in Scheidel’s theory. These are the Four Horsemen: war, plague, state failure (dissolution of empire), and revolution. Scheidel makes a powerful argument that whenever there was a significant leveler of inequality in history it was driven by or was a consequence of one of the aforementioned catastrophic events.42 Both Piketty and Milanović agree that 20th-​century leveling of inequality was to a large extent due to the impact of the two world wars, although they do emphasize the role of social reforms and the rise of the welfare state that added to the gradual leveling of inequality until the 1970s (Milanović calls them benign forces that were only possible in postindustrial societies with rising mean incomes43). Franchise extensions, welfare state reforms, and demands for redistribution all arose internally in the 20th century, potentially as a consequence of social rebellions against high levels of inequality. These too can be considered as occasional forces, arising endogenously. Scheidel on the other hand dismisses the benign force theory by arguing that the demand for these reforms (like the extension of the franchise after World War I or the expansion of the welfare state after World War II) came as a direct consequence of the wars. Even the rise of socialism, first in Russia in 1917 and later across Europe and the world can be attributed as a consequence of the wars.44 In other words, there was no natural force—​like modernization or economic development—​that led inequality down, at least not significantly. There were only large-​scale unpredictable events that happen randomly and eventually served the purpose of leveling the inequality trend. Whenever such an event happens the powerful elites and their wealth are affected. They are either ousted and removed from access to power (in a revolution or during state failures), killed (during plagues or epidemics), or have their wealth destroyed (in wars). In each case the source of power is altered and elite members shift their relative positions within a society. This does not mean they become extinguished—​Clark shows that elites can be quite resilient during long periods of time.45 The elites merely adapt, but this adaptation takes time, so there is a prolonged period before inequality starts going up again. Which brings us to the first force, one that pushes inequality up constantly. This force is an artifact of elite entrenchment and political power, and a consequence of elite existence throughout history. I call it the force of wealth concentration. During uninterrupted periods of economic development political

24 Introduction power is used to expropriate newly created wealth and this has a tendency to concentrate within the elites. Throughout history, particularly during preindustrial times, whenever there was rapid economic development, it was contingent on political power. Autocratic political orders always featured extractive ruling elites. The ruling elites and their networks had incentives to expropriate wealth and were driving the concentration of wealth during periods of economic growth and expansions. This was particularly prominent during the Industrial Revolution when inequality reached significantly higher levels on the wave of rapid economic development driven primarily by those with access to political power and privilege. It is also true in any modern society which has economic growth contingent on political power; and in any preindustrial society where economic growth was limited within the constraints of the Malthusian trap. In each case, whether we are observing a premodern society captured in the Malthusian trap or a modern society captured by entrenched political and corporate elites, their mere existence was what made wealth concentration high. The only way for inequality to subside was a war, an epidemic, or a similar major disruptive event. Only when the elites were disrupted was it possible for inequality to go down. Figure I.1 depicts the theory of cyclical inequality when faced with the two major opposing forces. Note that the force of occasional destruction is always stronger but given that the leveling from this force happens as a consequence of a short-​run event, its effects are deep but they soon subside and societies can once again move in the direction given by the force of wealth concentration. Wealth inequality

Elite concentration of wealth

Elite concentration of wealth

Elite concentration of wealth State collapse

War Plague

Time

Figure I.1  The cyclical nature of inequality is driven by two forces, one pushing up constantly, and the other pushing down occasionally.

Introduction  25 A further theoretical justification of the force of wealth concentration can be found in Olson’s 1982 classic, The Rise and Decline of Nations. Olson argued that stable democratic societies have in-​built mechanisms that encourage the rise of special interest power. According to Olson’s logic, prolonged political and economic stability will gradually lead to an accumulation of power of special interest groups. Derived from his central theory on collective action, Olson suggests that many individuals will not actively seek to form groups even when it is in their best interest to do so (when the gains of organizing would outweigh the losses). On the other hand, smaller groups with a narrower focus will form because they have a strong incentive to solve the free rider problem and impose collective action among their members.46 From this he derives several propositions on how the logic of collective action should impact economic outcomes (economic growth in particular). The first one is that special interest groups accumulate gradually over a long period of time, and its accumulation of power is limited to societies with a “continued freedom of organization.” He predicts that societies which did not have any major disruption in the formation of its interest groups (like the United States or the United Kingdom) will see their interest groups growing stronger over time. Their accumulation of power in stable societies will, over time, reduce economic efficiency and result in what Olson calls an “institutional sclerosis.” The longer the time of stability in which such groups operate, the stronger their negative effect on economic growth. If, however, a society develops only “encompassing” groups which include the majority of members of a society within them, only then will they render a positive effect on economic growth as the benefits achieved by such groups will be distributed across a wide range of its members thus making most people in a society better off.47 Following the same reasoning, a prolonged period of stability will gradually lead to an accumulation of power in the hands of elite networks, where powerful interests will successfully fight to preserve their status quo. Therefore, a necessary outcome of prolonged stability will be an accumulation of elite network power that could retard economic growth, as anticipated by Olson, but also curb institutional rules to their benefit—​like the use of violence, theft, and expropriation—​at the expense of the rest of society. Long periods of uninterrupted stability, because of the accumulation of power within well-​organized elite networks, can be attributed to wealth concentration among the elites, and hence serve as a potent motivational factor behind inequality. This is a powerful force that is realized automatically, whenever there are no disruptive events that halt its progress. To sum up, inequality is a cyclical phenomenon because random events affect the allocation of wealth within societies. Absent from violent random shocks societies would have persistent wealth accumulation in the hands of elite

26 Introduction networks and very high and stagnant levels of wealth and income inequality, led by the forces of wealth concentration. However, because of violent shocks—​the forces of occasional destruction—​elite networks will occasionally be disrupted and inequality trends will be leveled. The long period of (relative) stability will once again give rise to the forces of wealth concentration and the cycle is completed. None of this suggests that elite-​driven inequality is imminent nor that it is a fact of life that can be justified. There are ways to constrain the forces of wealth concentration outside causing violent disruptive shocks. The problem is that throughout history societies have seldom recognized its true causes, obviously failing to ever truly address them. As this book shows, the root cause of the forces of wealth concentration and by extension the upward trends in inequality, is political power. The final part of the book will address the specific sets of reforms necessary to redistribute political power from centralized office-​holders to citizens and communities. It is envisioned as a long and continuous process, resting on the democratic trial-​and-​error mechanism gradually rebuilding trust in social and political institutions. Inequality is a trend that accumulates over time. Its remedies also need time to accumulate enough social capital to deliver the expected effect. Simple redistribution of unequal incomes will not suffice. Time has come to change the way we think about the causes of inequality.

I.5.  What This Book Does Not Do: External Validity Limitations and Potential Extensions The book delivers theoretical and empirical validation of its hypotheses. The chapters presenting empirical results primarily address the internal validity constraints. However, they do not touch on any external validity issues. Furthermore, the measurement of an elite network presented in the empirical chapters is a mere approximation, as interactions between its members are difficult to uncover, yet alone measure. The closest I’ve come in accurately defining an elite network is in Chapter 4, where I directly implicate top executives who associate with politicians on an informal basis (via memberships in the same clubs, societies, and associations) or who used to work together. A similar, although less restrictive approximation of an elite network is given in Chapter 7, where the key was to implicate firms, looking at the benefits firms get if they donate money to political campaigns, lobby politicians, or have former executives who used to work in government. The most lenient approximation is given in Chapter 6, where I discuss research linking political survival to suspicious procurement contracts given to firms. Each of these approximations offers the best definition given the available data.

Introduction  27 Furthermore, the scope of the book is limited only to the countries that were examined. The impact of elite networks on inequality was shown only for the United States and the United Kingdom. Second, each case provides an entirely different and very specific context, meaning that I could not show on a single dataset for a single country how both politicians and firms benefit, nor how their mutual benefits carry an indirect effect on inequality. The overall conclusion is drawn from specific cases, which raises a slight concern over the generalization of the hypothesis. Third, the estimated empirical effects are in most cases local average treatment effects (constrained within a specific as-​if random context), making it even more difficult to generalize the conclusion to all countries and all cases. To make the argument externally valid I would need to directly approximate elite networks across different societies in the same way as presented in parts of the book, and observe how it differently affects inequality in each of them. If greater inequality is indeed rooted in politics, as the book aims to show, I should observe cross-​country differences in inequality based on the differences in how connected the corporate sector is to politics. In particular, I would need to show that in countries where top executives are not as closely connected to politics as they are in the United Kingdom or the United States, earnings are distributed more equally. If low-​inequality countries are indeed mostly deprived of these types of connections at the highest levels (i.e., if they have institutions that prevent elite networks from abusing power), then we could make the conclusion that politics is an important determinant of higher inequality. If, however, this is not the case, the conclusions of the book would only suffice for a limited sample and perhaps even a limited time period. Furthermore, in Chapter 4 the findings are limited with how long it can follow the origins of an elite network. The data can only be traced back until the year 2000, meaning that I present a rather static image of the impact of connections on wages. In particular, better connected executives have had consistently higher salaries and higher total earnings throughout the observed 16-​year period. There is unfortunately little data on people entering into elite networks in the sample (senior executives tend to be older and if they are entrenched, they tend to be entrenched for long periods of time, and certainly prior to my first point of examination, the year 2000). A potential extension would be to get data that would allow tracing this process back in time until the 1950s, 1960s, or at least 1970s, and see if the documented increase in income inequality since the end of the 1970s was indeed preceded by the rise in power of elite networks. In other words, did US and UK senior executives started forming closer ties to the government just before the observed rise in income inequality, or is this relationship merely a persistent peculiarity of these two countries (epitomized by WASPs in the United States or Old Etonians in the United Kingdom) that is not necessarily tied to rising inequality?

28 Introduction Even in the case of obtaining a longer time-​series of political connections, it would still be necessary to observe the same relationship cross-​country to determine if it holds. The problem here, even if we assume that we can approximate for elite networks in other countries, is in the comparability of the data. In the US and UK samples, elite networks are proxied via a very specific indicator: whether executives used to work in politics or are connected to politics via personal relationships. If a similar proxy could be found in other countries it would have to measure the exact same thing in order to make the cross-​country comparison possible. Overall, despite the limitations in satisfying external validity of the argument, the empirical chapters of the book do provide a useful blueprint for further research on the topic. They provide empirical backing of the argument that political factors could be an important determinant of inequality, even if this result is limited to a specific context and time period. Any further research on this topic, examining the political causes of inequality, should approach it from the same perspective: try to get an individual-​level and firm-​level dataset across different societies to measure the impact of connections on income dispersion. Further research can be aimed at a more precise definition of elite networks, but simple enough for it to be easily measured in any society. This is a difficult task as elite network interactions are always hidden, however some interactions and particularly some outcomes based on informal networks can become visible and hence measurable. Second, a very useful extension would be to capture the rise of power of elite networks over time. Are they indeed robust to political and economic systems, as is assumed in the book? How volatile is the accumulated power of elite networks across different societies? Do they vary in influence from one decade to another, or is their influence persistent? The assumption is that elite network membership is dynamic—​all members are limited by the duration of their tenure, and ultimately by biology. What is the effect on society when these changes occur and when important members leave the group? Or is there no effect at all, given that the position of power of the person is what makes them a valuable member, not necessarily the person himself. To continue in this direction, one would have to combine the sociological and anthropological studies on elite groups with the political economy implications given in this book. Third, the findings of the book could help provide empirical backing to Olson’s argument that post–​World War II stability increased the power of special interest groups which resulted in slower economic growth. Olson’s lucid theory of interest group capture from the Rise and Decline of Nations has never really been empirically justified. This book provides a good starting point of the type of analysis that would be necessary—​to observe individual-​level connections between

Introduction  29 politicians in power and the managers and financial backers of powerful interest groups (NGOs, corporate lobby groups, advocacy groups, religious groups, etc.). Finally, a useful extension would be to examine whether elite networks carry a similar impact on social mobility and the equality of opportunity. The mechanism is slightly different than the impact on inequality presented in the book. It materializes through hiring decisions on the labor market, where if an individual is not a member of the network, he or she is prevented from climbing the social ranks. In other words, an individual can only reach the top of the corporate or political hierarchy if on that road he or she accepts the mutually dependent favor exchanges with influential individuals. This would be almost impossible to measure directly, unless we could measure the negative effects on connected individuals when a person drops out of an elite network because of, for example, imprisonment. Before any of these efforts are conducted, and before the theory can prove it is externally valid, the conclusions remain: politicians and firms in specific contexts do collude together and derive benefits from their interactions. These benefits to both groups hurt society as they make the distributions of incomes more unequal by pushing up the incomes of politically connected top earners. Politics is therefore an important factor that should not be overlooked when examining the causes of inequality.

I.6.  Structure of the Book The book is divided into three main parts, preceded by Chapter 1, which first defines the concept of elite networks. It provides a formal definition, discusses the logic behind their creation, and provides several visualizations of the most obvious examples of elite networks, with an emphasis on network superhubs in finance and politics. It then draws a map of political interactions and uses Olson’s logic of interest group behavior based on their relative power and salience to place elite networks on that spectrum. Finally, it describes how to quantify and establish elite network outcomes. Part I then sinks its teeth immediately into the main relationship of interest: does the existence of elite networks, a historical axiom, affect the distribution of income in a society and in which way? Chapter 2 delivers the cornerstone historical argument depicting the mechanism between elite networks and inequality. It follows several theoretical and empirical contributions in the inequality literature to argue that inequality exhibits incredible persistence over time. One of the reasons for this persistence of unequal income and wealth distribution in addition to low social mobility has to do with the accumulation of power at the top. Before the Industrial Revolution our societies were molded by autocratic rule in which wealth was

30 Introduction fully expropriated by the ruling elites. After the Industrial Revolution, societies experienced the rise of rentiers who earned wealth based on ownership of land and capital, but whose position was also cemented thanks to political power. In modern times a new elite network is being formed: it is no longer composed of nobility (as before the 18th century) and various rentiers (as in the 18th and 19th centuries). It is composed of corporate managers with links to political decision-​ makers. The rise of the “supermanager” in terms of salary and total compensation is the biggest explanatory factor of rising inequality in the West. The argument of the book is that this is due, in part, to political connections that supermanagers caress, and through which they signal their importance by having access to a wide network of influence. Their degree centrality enables them considerable power in justifying their own abnormal earnings. Chapter 3 focuses on contemporary trends in the rise of democracies, the rise of government spending, and the rise of inequality in the second half of the 20th century. It suggests that even with democracies evolving to make capitalism beneficial to all, and even when most people today enjoy far greater living standards than ever before (governments providing public goods and freeing entrepreneurial spirits), this mechanism has failed to sustain the rapid growth in wages of the top income earners (1% and 0.1%). It provides a theoretical discussion as to why this might be the case. This chapter also surveys the alternative explanations and evidence on rising income inequality. Chapter 4 summarizes the empirical evidence that confirms the theoretical intuition laid out in the previous two chapters. It uses a large database of corporate executives of publicly listed companies in the United States and United Kingdom in order to directly measure how they might be connected to politicians. A political connection implies any channel through which a top ranked corporate executive could form an informal interaction with an office-​holding politician, be it through their previous careers, or through membership in the same organizations such as country clubs, foundations, charities, church or religious groups, professional organizations, etc. Once such an informal relationship is established, favors can be exchanged; the executive gets beneficial regulation or government contracts, and the politician gets campaign donations or direct bribes. After the executive successfully extracts rents for his firm, he demands a higher compensation as a reward for his efforts. The empirical evidence supports this hypothesis and finds that politically connected executives, by assumption elite network members, tend to have higher salaries and total earnings than nonconnected executives. Part II is designed to show both the theory and empirical findings of why members of an elite network—​politicians on one hand and firms and their executives on the other—​have an incentive to be part of it. The first chapter of Part II, Chapter 5, builds on Chapter 1 and exposes the logic of elite network

Introduction  31 formation. It delivers a network theory and an economic theory explanation behind elite network formation with an emphasis on its high levels of betweenness centrality and homophily. Elite networks are shown to be highly topocratic, meaning that they condense wealth-​seeking opportunities and privileged information within narrow powerful groups. It then adds a cost-​benefit analysis to understand under which conditions individuals enter such groups, and finally, it dives deeper in the three historical principles of how people mitigate risk in order to survive. Given that societies were entrapped for centuries under the violence power principle it is difficult to simply break up elite networks arising from it. Chapter 6 presents theoretical and empirical justifications and explanations of why politicians would enter an elite network type of relationship. They benefit in two ways: preserving its position of power through building small winning coalitions, and by directly or indirectly extracting rents. This is where elite network theory meets the selectorate theory and paints a complete picture of the motivation of political survival. Rents can come in the form of bribes and kickbacks to direct usurpation of office for private gain. The chapter concludes by presenting empirical findings linking corruption, in the form of rent-​extraction, directly to re-​election possibilities of politicians. The data is drawn from the case of Croatia, where corruption has been directly approximated using fraudulent procurement contracts. Chapter 7 looks at the firm and how its rent-​seeking activities, in the form of exclusive government procurement contracts or favorable legislation and regulation, get realized through an elite network. It delivers the key distinction between two main types of firms: rent-​seekers and customer-​seekers. As long as elite networks render effects that favor rent-​seekers, adverse socioeconomic outcomes are inevitable. The chapter concludes by examining the empirical evidence showing how politically connected banks in the United States benefited the most from the 2008–​2009 bailout allocation. The final part deals with the remedies. Based on the main finding of the book—​ that inequality is rooted within political power—​it attempts to show how and why lowering political power is the only way to curtail the forces of wealth accumulation. Chapter 8 delivers another necessary historical context, explaining the rise of capitalist democracies, from the early unjust exploitative capitalism, led by the pure violence power principle, to its gradual convergence into a much fairer democratic capitalism. Despite its early unjust outcomes, capitalism managed to deliver an impetus for persistent social change, eventually resulting in political changes, as the old aristocratic elites were being replaced by democratic governments. This permanent incentive for change gradually started delivering basic human rights, as well as equal representation, rule of law, equality before the law, equality of opportunities, and eventually equality of gender and race.

32 Introduction All of this was achieved under capitalist democracies, through a unique trial-​ and-​error mechanism that helps societies learn from their errors, never again to repeat them. This was the reason why social progress became unstoppable after a period of terrible events of the first part of the 20th century. Democracies reformed from within, pledging not to repeat the errors they’d made. Socialist countries also achieved rapid progress in the aftermath of the Second World War. However, when faced with their first major internal crisis, they crumbled under pressure as they had no trial-​and-​error mechanism that could help them overcome the crisis. The unstoppable progress that lifted living standards and dramatically reduced poverty in the second half of the 20th century failed, however, to solve the issue of inequality. Inequality is driven by different forces, rooted deeply within political power. This does not mean it cannot be dealt with, but in order to curb its incentives successfully we must set the target toward its causes. This is why, as Chapter 9 will uncover, progressive taxation and greater redistribution will not be successful in solving the problem permanently—​they are focused on the consequences (unequal incomes), not the causes. Granting too much political power to individuals to solve the issue of inequality leads to personalization of governing institutions, adverse selection into politics and eventually declining trust in public institutions. This is the opposite effect of what we need to contain inequality. The context delivered in those two chapters opens up room for reform proposals in Chapter 10 on how to reduce and limit political power, reduce incentives for elite network formation, and consequentially lower inequality. These are presented in the form of three powerful levers whose goal is to exploit the democratic trail-​and-​error mechanism in order to rebuild interpersonal trust and trust in public institutions, and through second-​and third-​order effects increase civic engagement, improve selection into politics, depersonalize governing institutions, turn politicians into true servants of their people, and empower both citizens and their communities. The First Lever is a set of reforms aimed at lowering centralized political power. It includes full budgetary transparency, the necessity of free media and investigative journalism, the introduction of term limits and strict punishment of political transgressors, calls for lesser discretion and more rule-​based politics (introducing key performance indicators for politics), and the reduction of the scope of centralized government activities, shifting more public good decisions directly to citizens and communities. The Second Lever is a set of reforms aimed at empowering citizens. It aims to give citizens more power in allocating their tax contributions to the public goods they wish to build, in choosing exactly which public goods they want their local governments to deliver, and in offering them a credible threat in the form of punishment referendums of governing officials.

Introduction  33 Finally, the Third Lever is a set of reforms aimed at empowering the community, with the idea of giving greater redistributive power not to local governments, but to the community directly in all choices of local public good allocation. Simultaneously applied, all three levers will gradually, and with significant errors expected along the way, reduce centralized political power, improve selection into politics, improve trust and democratic capital, reduce incentives for elite network formation, and finally, reduce inequality. It is by no means a one-​ size-​fits-​all solution, nor is it a set of policies that can deliver immediate change. The very usage of the term “lever” is to signal that these reforms gradually change people’s incentives and behavior, which will eventually lead to improved social outcomes. It is a very long and difficult path, but it is the only path that enables democracies to deliver its greatest benefit—​a rule of the people, by the people, and for the people.

1

Why Study Elite Networks? Among the Davos attendees are many titans of finance who pull the levers of the global financial system. This system is not simply interlinked by institutions and transactions, but it is fundamentally a human system, because on the most basic level it is the result of human interaction. Understanding the interconnections of the key players is vital if we want to understand the system as a whole. —​Sandra Navidi (2017) Superhubs, p. 5

1.1.  Defining Elite Networks: Informal Mutually Beneficial Interactions between Political and Corporate Elites Figure 1.1 sums up the basic logic of elite network theory: informal social networks are created between politicians in power and top executives of politically connected firms so as to benefit all agents within the network. Politicians satisfy their interest by staying in power and extracting political rents in the form of donations, bribes, and other favors, while executives manage to get exclusive government contracts and favorable regulation for their firms. Based on acquiring these direct rents, firms reward their top executives with higher salaries which consequentially increases the divergence between top income earners and everyone else. This is an avenue that has been mostly neglected by the inequality literature, but its mechanism is potentially the missing link in understanding the true underlying causes of inequality. Rents for firms correspond to the standard definition of rent-​seeking according to the political economy literature.1 Firms use their connections to government to gain a significant advantage over their competitors, be it through direct government subsidies and procurement contracts, favorable legislation or regulations, or by gaining other forms of privileged access or control over a good or service. In other words, rents are an outcome of the political process. Corporate and political institutional capture ensure lower risk for network members and enable them to capture a non-​market-​generated source of income. This additional source of income (on top of the innovation-​driven or opportunity-​driven market share) is not being achieved via regular market transactions, but is rather a consequence of capturing political benefits. Rents are Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0002

Why Study Elite Networks?  35

Elite networks Officeholding politicians

Corporate executives

Firms receive rents

Higher executive salaries

Firm-level rents: Political rents: • Staying in power • Campaign donations • Bribes and favors

• Government contracts and subsidies • Favorable regulation and legislation

Higher inequality

Figure 1.1  The interaction between political and corporate interests within elite networks provides another, thus far unrecognized, avenue for explaining higher levels of inequality.

thus deliberately defined broadly to include everything from barriers to entry, favorable tax policies, subsidies, market power, protectionism, etc., as these politically generated benefits reflect themselves in the differences in nonmarket earnings realized by firms that achieve them. Obviously, they differ from industry to industry, they differ in scope and context, but the focus here is more on how they benefit those who capture them, and how they represent a direct outcome of the elite networking mechanism, as described in Figure 1.1. The focus is also on individuals who benefit from being connected to politics. The purpose of being connected is to generate politically driven nonmarket returns for firms. These politically captured returns enable the connected individuals who achieved them to claim a more important role within their firms. The individual-​level empirical findings presented in Chapter 4 suggest that political connections are a powerful mechanism of translating the misuse of power into abnormal earnings. Obviously, not every member of the top 1% or 0.1% of income earners is connected with politics. There are many, in fact, who generate their wealth purely from innovation, expertise, knowledge, or value added provided. But those who are connected tend to derive a significantly higher monetary compensation (compared to their nonconnected peers) because of the connection. It is these types of individuals, coalesced within an elite network, which are the focus of the book, and it is through these groups that we get to observe how the misuse of the political process gets translated into inequality. Rents and inequality are outcomes of the same power capture process. And just a rents can be lowered and disrupted by changing economic policies (an outcome that

36  Elite Networks has happened throughout history), so too can inequality be lowered by changing the right policies focused at curbing incentives for a power grab. The literature captures this indirect relationship between rent-​seeking and inequality. For example, Dabla-​Norris and Wade formalize a model where only wealthy agents engage in rent-​seeking to protect their wealth.2 Demand for protection is achieved through the influence wealthy agents exert over politics, implying that both rents and inequality arise as outcomes of that process, much as the elite network theory suggests. Stiglitz argues that the top 1% originates in the United States mostly originate from rent-​seeking activities instead of market innovations,3 while Furman and Orszag add some empirical backing to this story by finding that the most profitable US firms who pay the highest salaries and thus increase between-​firm inequality draw their higher returns not from innovation but from rent-​seeking.4 This finding is further strengthened by Song et al. and Autor et al.,5 who find that between-​firm differences in incomes are the key drivers of inequality given that high-​paid workers tend to cluster in bigger and more successful, superstar, firms. A lot of such firms reached the stage of their life cycle where they hire corporate executives with political connections who help them mitigate certain regulatory requirements (the Big Tech industry being a good example—​achieving market status via innovation, and protecting it via rent-​seeking). Finally, Lindsey and Teles use case studies of regulatory capture to show how firms that stifle competition divert resources and wealth toward the elites.6 Each of these findings strengthens the theoretical arguments of the book. Not a single rent-​seeking activity is possible without the necessary collusion between power-​holders in the corporate and political world. The misuse of power is not necessarily limited to interactions with politicians. It also happens within the corporate world. Piketty, Saez, and Stantcheva find that, when top tax rates are lowered, corporate CEOs resort to aggressive bargaining with their boards to push their incomes up, through a channel they also refer to as rent-​seeking. The alignment with the argument of this book is that CEO compensation—​during periods of low top tax rates—​tends to go up as a result of a power struggle rather than a productive effort. In other words, the growth of CEO compensations, which puts upward pressure on top 0.1% inequality, is not due to innovation or greater efficiency of their businesses, but due to their own clout over their boards. Elite network theory adds an additional dimension here; it’s not just the basic power play between managers and owners, it’s a much more subtle way of turning political favors into corporate bargaining power. The biggest contribution of the book is to observe the collusion between agents of power, first in its historical context to generate a theory of elite network behavior, and then empirically on an individual level, to capture the connections between individual politicians and individual corporate executives. The goal

Why Study Elite Networks?  37 is to show how agents forming informal elite networks through the misuse of power drive higher incomes for many top income earners. The misuse of political power is the crucial root source of inequality. The historical roots of high inequality have been documented by Scheidel. His argument is that high levels of inequality are a persistent socioeconomic phenomenon that can only be temporarily reduced via violent events like wars, revolutions, or epidemics.7 What we noticed in the post–​World War II period was therefore only a temporary decline of what is otherwise permanently high inequality. Things did change, however. Piketty notes that the wealthy of the second half of the 20th and the 21st centuries are no longer rentiers living off land and inherited wealth, but corporate executives, or as he calls them “supermanagers,” living off their abnormally high salaries.8 The abnormal part is the key driver of present-​day inequality, and as this book shows, it is driven by the misuse of power. Levels of inequality, however, differ among countries. Therefore, one prediction arising from this theoretical model is that we should observe higher levels of inequality in every society where corporate interests are heavily interlinked with governments and political power. Societies where its biggest corporations grow due to politically generated revenues, which they acquire through personal relationships between their executives and political decision-​makers, should have higher levels of inequality (measured as the income share of the top 1% income earners) than societies where this practice is limited. This prediction will not be tested empirically between countries in the book, as there is unfortunately not enough country-​level data on individual political connections to make that inference. Rather, the book will show and explain how modern elite networks function within countries, how they benefit their interconnected participants, and measure the extent of their influence on the increasing disparities in earnings. The focus is specifically on the United States and United Kingdom, two developed, high-​income countries with good institutions, but with higher inequality than in other developed, high-​income countries.

1.1.1.  Does It Pay Off to Be Part of the Network? An elite network is a group connecting political elites to a number of powerful individuals within a specific political jurisdiction, be it the entire country, region, city, or municipality. Individuals within such groups are all bound by their common interest: preservation of their position of power. Elite network members, in addition to politicians, can either be wealthy individuals, typically owners of big companies with a specific agenda that are usually main campaign donors;

38  Elite Networks corporate executives from public or private sector firms—​from banks and the finance industry, tech giants, oil and energy companies, or other firms with direct dependency on politics; private sector consultants or advisors who held elected office or other positions in government prior to their corporate career; and interest groups in the classical sense, represented by lobbyists and their respective boards. All of these members have an incentive to be a part of a politician’s close network from which they are free to extract rents as nonmarket revenue. Firms get them in terms of exclusive procurement contracts, corporations and interest groups get them in terms of favorable industry-​wide or group-​wide legislation and regulation, and campaign donors get them in terms of personal favors they are free to demand from politicians. This is therefore a network model of cartelization9 where even though the agents compete against each other for government favors, they all fully realize the benefits of cooperating with each other to achieve individualized-​level support from the government. In a codependent system where support and loyalty are traded for concessions, those linked with the politician, who have a lot to lose from his or her demise, will make sure he or she stays in power for as long as possible. Politicians have strong incentives to form and maintain such networks. They too can enjoy multiple benefits. They get important campaign contributions, crucial to their political survival, they can take bribes for favorable legislation or exclusive procurement contracts, they can get in-​kind benefits (dinners, golf games, sports or theater tickets, various gifts), and can even benefit after their political careers, earning money by giving speeches at industry events, or take on high-​level executive or board positions in companies they used to regulate. However, there are costs associated with elite network membership. For every agent the most important are reputational costs. If the activities of the network are uncovered, each agent suffers a cost to his or her reputation even if the activity is not illegal per se. If the activities are illegal, there is a potential cost of prison time, which increases the reputational costs before even engaging in the activity. Each agent also has the cost of effort (which includes the costs of entry). Being part of an elite network means going to a lot of social gatherings and activities, thereby spending time to build the relationship. This may last for several years, even decades, and it may yield physical costs along with the costs of time and effort. Politicians, in addition to all of the above, also have the costs of potentially losing an election if any of their shady activities are uncovered. The potential benefits from joining an elite network have to outweigh all the potential costs for each individual agent in order for them to join. This means that not every person will join an elite network, just the ones who, according to Becker, value the potential benefits more than all the potential costs.10 This is further discussed in Chapter 5 presenting a cost-​benefit analysis of elite network membership.

Why Study Elite Networks?  39

1.1.2.  The Role of Institutions Every political system is based on upholding power. Therefore, the logic of elite networks should be similar (albeit not identical) across different political systems. However, the variation in outcomes in curbing this power differs with respect to how a county’s institutions are defined, which is conditioned by historical patterns of development and a whole number of other endogenous factors. This is why we notice different consequences of elite network behavior in different institutional surroundings. But not everything is down to institutions. Consider the puzzle of why voters fail to punish corruption.11 The institutional argument fails to explain why corruption tends to exhibit high levels of variation among the highly developed institutionally stable and inclusive societies. The United States, Japan, France, or Ireland, for example, exhibit higher rates of corruption than the equally institutionally well-​endowed Scandinavian countries, New Zealand, or Switzerland. The average gap between the former and the latter group of four countries is 16 points on the Transparency International CPI index for 2020 (Table 1.1). This is almost the same average gap as between the former group and, for example, Italy, Greece, Czechia, and Georgia.12 All of the aforementioned countries fare much better than the low-​income countries in terms

Table 1.1  All CPI values are taken from Transparency International (2021) and are for the year 2020 Selected countries (top ranked; score 80–​90)

Corruption Selected Corruption Perception countries Perception Index (CPI) (close to Index (CPI) top; score 66–​77)

Selected countries (medium ranked; score 50–​60)

Corruption Perception Index (CPI)

New Zealand

88

UK

77

Israel

60

Denmark

88

Austria

76

Cyprus

57

Finland

85

Belgium

76

Latvia

57

Switzerland

85

Iceland

75

Poland

56

Sweden

85

Japan

74

Georgia

56

Norway

84

Ireland

72

Czechia

54

Netherlands

82

France

69

Italy

53

Germany

80

USA

67

Greece

50

Group average

84.6

73.2

55.3

40  Elite Networks of corruption; however, the variation among the first two groups suggests that institutions are hardly the only explanation. Political competition and the threat of electoral punishment are obviously not enough to curb incentives for corruption in all countries. In democracy’s defense, the average levels of corruption in democratic countries are much lower than in nondemocratic countries. Electoral competition therefore certainly does help, but it does not eliminate corruption. In fact, many democracies have corruption levels equal to or even worse than autocracies (Table 1.2). For example, the average corruption levels in democratic Slovakia, Croatia, Hungary, Romania, and Bulgaria (average of 45 points on the CPI index) have been similar to those in autocratic Jordan, Cuba, Belarus, Bahrain, and China (same average). Some democracies, like Brazil, Serbia, or Colombia fare even worse than many autocratic countries. Similar examples are prevalent, particularly when looking at corruption in democratic countries of Latin America, all of which have only marginally lower levels of corruption than most African anocracies or dictatorships. There is therefore a factor within democracies, even highly developed ones, that does make corruption possible, permissible, and to some extent desirable. In order to account for this variation in political economic outcomes even across rich democracies I revisit Olson’s idea of organized special interests with an important addition: interest groups are not necessarily limited to organized groups in the classical sense (such as labor unions, industry lobbies, the NRA, AARP, etc.) but also as groups of powerful individuals that collude together to preserve their status of power. The stronger the links between them, the more Table 1.2  CPI values taken from Transparency International (2021), * represents autocracies according to Polity IV (2018) Selected countries (low ranked; score 40–​50)

Corruption Perception Index (CPI)

Selected countries (low ranked; score 40–​50)

Corruption Perception Index (CPI)

Selected countries (low ranked; score 38–​40)

Corruption Perception Index (CPI)

Slovakia

49

Jordan*

49

India

40

Croatia

47

Cuba*

47

Turkey

40

Hungary

44

Belarus*

47

Colombia

39

Romania

44

Bahrain*

42

Serbia

38

Bulgaria

44

China*

42

Brazil

38

Average

45.6

Average

45.4

Average

39

Why Study Elite Networks?  41 powerful they become. This accumulation of power among elite networks makes them more successful than any other group in achieving their goals, which could explain the existence (and persistence) of corruption and favorable legislation given to certain industries and/​or individuals. Well-​defined institutional rules and even a strong legal system might not be robust to such practices and their consequential outcomes in every country. In other words, institutions matter, but institutions are not always fully robust to elite networks.

1.1.3.  Elite Theory in Sociology The definition of elite networks used throughout this book is somewhat similar to what classical sociology refers to as elite theory. According to elite theory a very small minority of economic and political elites holds the majority of power and influence within a society. The origin of power is derived either from one’s position in the corporate hierarchy or from one’s position of control over wealth-​ creating activities. It is through these positions that elites determine who will hold political power, or at least try to influence policymaking. Accordingly, a democratic system is often criticized by elite theorists as incapable to curtail the power of the elites. Similar to Olson’s logic of group behavior, elites are successful because they represent a well-​organized group with a narrow interest, whereas the nonelites are disorganized and hence unable to challenge the elites. The origin of classical elite theory can be found in the works of Max Weber, Vilfredo Pareto, Gaetano Mosca, and Robert Michels. From Weber elite theory draws the definitions of power and the three forms of domination—​charismatic, traditional, and rational-​ legal—​ through which power relations are formed within a society.13 From Pareto it determines elite selection: members of an elite originate from proximity to economic or political power, which is enabled by political connections of families and inherited wealth. In other words, if social mobility were unrestricted within a society, elite members would be drawn from a pool of its most competent members. However, because of unfair advantages driven by family wealth and connections, the selection of new elite members is skewed in favor of the current elite.14 Mosca’s contribution is similar to Olson’s interest group theory, in that the elite is a well-​organized tiny minority, as opposed to an unorganized majority (“the masses”), and therefore has an advantage in achieving its own interest. He also discusses the intellectual and moral, in addition to economic, superiority of elites over the masses.15 Finally, Michels coined the phrase the “iron law of oligarchy” to define the existence of ruling elites within any democratic organization, be it the state or the firm. He discusses the role such elites play within organizations, where they control both information and resources, therefore enabling them to concentrate power.16 Elites have

42  Elite Networks been a focus point of many contemporary research efforts ever since in sociology, economics, and political science.17 What most of them have in common was to understand who the elite is (its composition), how does one become a member of an elite (its selection), and more importantly, whether the very existence of elites hampers the idea of pluralism and democracy itself (its implications). My approach differs from the aforementioned sociological studies in that I do not draw implications of elite behavior on the functioning of democracy,18 nor am I focused on explaining within-​elite selection. Rather, I am concerned with showing how elite networks benefit the actors within them—​how they help politicians stay in power, and how they help firms gain a competitive advantage. Furthermore, I find politicians at least equally important as the corporate elites. Politicians have the power of creating a legal framework that benefits certain industries or interest groups, and the power of redistributing a huge portion of a country’s GDP—​its government budget. The position of political power therefore carries significant impact over the corporate sector, even when politicians originate from or are appointed by the corporate sector, as elite theory would assume. Nowhere is this more obvious than in autocracies, but even in democracies proximity to political power is a crucial part of what makes membership of an elite network worthwhile. Finally, I am interested in the consequences of elite network behavior on the distribution of top incomes through a very specific mechanism—​the effect of political connections, which is a proxy for elite network membership, on executive salaries.

1.1.4.  The Spontaneous Nature of Elite Networks The pursuit of self-​interest is rational, legitimate, and expected within a political order. When politicians and special interests realize they can both benefit by colluding with each other, their rational and legitimate response is to do so. Every citizen within a (democratic) political order has the right and is encouraged to pursue his or her own interests. Every individual indeed does so, in almost all cases as a part of a wider group: family, friends, school, club, society, religious organization, business organization, political party, etc. The interests can be entirely altruistic where a person gets satisfaction purely from helping others (e.g., religious organizations or charities), or they can be entirely selfish where a person gets satisfaction from furthering his or her own goals, or the goals of those close to them (e.g., winning office, getting a better job/​salary, etc.). In such an environment it is natural to expect both cooperation and conflict to occur, with varying rates of success between groups. A society, therefore, is a vibrant web of interactions between a multitude of agents organized through various groups all promoting their own interests,

Why Study Elite Networks?  43 whether selfish or altruistic. Some are more some are less successful at doing so, which means that some groups will over time acquire more power and influence than others. People with wealth and power in democracies arise as a consequence of such a dynamic exchange, and once they reach wealth and power, they have every incentive to preserve it. This is an evolutionary behavioral trait of humans: being well off compared to those around creates an incentive to maintain this status. Being worse off creates an incentive for change. If everyone in a society is equally well off or worse off, people will lack behavioral incentives to drive themselves into action. This is what explains the failures of communism in countries like Mao’s China, Pol Pot’s Cambodia, or North Korea, where equal misery among the population reduces incentives to aspire to anything better, thus retarding economic development. When everyone in a society is equally miserable, being miserable is taken as a natural state of affairs. When one knows he or she is miserable but there exist a wealthy few (such as monarchs, nobles, or contemporary wealthy elites) then there is a strong incentive to change such a social order. This is what triggers revolutions—​ inequality eventually reaches a boiling point above which people react to a sense of injustice.19 The contemporary divide between the rich and the middle classes is a natural consequence of a society in which individuals and groups fight for their own interests. This “fight” is a euphemism. The fight itself is what Adam Smith described as a persistent interaction between self-​interested individuals who will all, as if being led by an invisible hand, act to promote the collective interest.20 One thing is missing from Smith’s quintessential contribution: the fact that success and hence power varies between groups. Successful groups will be able to skew the distribution of resources toward themselves. Successful groups, with a greater stake in maintaining their position of power, will have a greater incentive to shape systemic rules and incentives. A good example of this is the elite-​ competition theory of democratization by Ansell and Samuels. They argue that democratization did not commence as a consequence of fear from uprising by the unorganized and dispersed poor who wanted political rights (as suggested by comparative political theory explaining the origins of democracy), but rather, building on Olson’s logic, from the well-​organized and concentrated wealthy who lacked political representation and feared expropriation of their newly acquired wealth by the state. Successful groups emerged to form a new elite.21 Have the elites in some way conspired to produce such outcomes? I reject the premise of a conspiracy theory.22 Elite behavior is not due to some grand intelligent design. Each self-​interested agent or group of agents formulates networks with others and acts within them to further their interest. Be it businesses, civil society groups, or the government itself. They all, albeit different, behave exactly the same way. They formulate clusters and interact to determine what is best for

44  Elite Networks each of them. Each group is prepared to vigorously defend its actions under the justification of protecting the interests of its members and/​or their ideological worldview. No group will ever consider that the furthering of its interest will in any way harm the rest of society. Quite the opposite actually. This is why there can be no conspiracy and no grand design, just random sets of clusters that are highly connected and interdependent and have a strong incentive to preserve their interest. There is no “enemy” in the classical sense of that word. People with power are, after all, only temporary. There is a persistent flux and dynamism between those who currently hold power, those who held it 20 years ago, and those who will hold it in 20 years’ time. However, in each case, no matter who holds power and who the elite is, they will always express the same behavioral pattern—​self-​preservation. The formation of elite networks within a country is therefore entirely spontaneous, serving to promote the narrow interests of its members, and limited in most cases geographically, around the centers of money and power. Figure 1.2 shows the geographical clustering of US and global Fortune 500 companies, where 85% of the top 500 global companies are located in the biggest urban centers in only nine countries. In the United States the distribution is geographically more even, however the urban centers on the East Coast, Silicon Valley, Chicago, and Texas still feature a heavy concentration of industry giants. An elite network is essentially a social network characterized by high levels of positive assortativity—​ the tendency of highly connected and powerful individuals to be closely connected with other highly connected and powerful individuals.23 Highly connected individuals tend to cluster with each other so that they maximize their influence across the broadest segment of the population. For example, Fortune 500 CEOs, sports and entertainment agents, or venture capitalists all have a strong interest (and it is part of their job description) in being connected to a large number of other highly connected individuals. For them the mere volume of individual connections is irrelevant as long as the network as a whole is important and serves a purpose. The same logic applies to politicians who, by the very definition of their jobs, become connected to a wide network of powerful and well-​connected stakeholders, whether locally or nationally. This makes every elite network in a country highly topocratic, meaning that the average compensation of an individual depends on how connected they are, and how central they are within the network. A topocratic network is the opposite of a meritocratic network—​it matters more who you know than what you know. According to Borondo, Borondo, Rodriguez-​Sickert, and Hidalgo, if a country is dominated by topocratic networks it is more likely to have higher levels of economic inequality.24 The logic is simple: the majority of people in a country are poorly connected to one another (they do not know many people outside their everyday environment) while a very small number of individuals in key

Why Study Elite Networks?  45

Figure 1.2  Geographical clustering of Fortune 500 companies in the United States (upper panel) and Global Fortune 500 (lower panel). The clustering within the United States is tied to urban centers (the East Coast corridor from Boston to Virginia has about 25% of all Fortune 500 companies, followed by the Silicon Valley, Chicago, and Dallas and Houston in Texas, each with about 6% to 7%). Globally three areas dominate in the clustering of the biggest 500 global companies: East Asia (36%; China, Japan, Korea), United States (27%), and Western Europe (22%; France, Germany, United Kingdom, Switzerland, Netherlands). Source of data: Visualize the Fortune 500, available at: https://​fort​une.com/​franch​ise-​list-​page/​visual​ize-​the-​fort​une-​500-​2022/​.

positions of power are highly connected, in most cases to other highly connected individuals in powerful occupations. This is the very essence of an elite network. One notable advantage of these connections is access to privileged information on various opportunities (for example, job listings of top management positions,

46  Elite Networks business opportunities such as private tenders, or valid stock tips) which are usually limited to a very narrow social group of trusted individuals. One obvious consequence is an increase in informational asymmetry, which skews the distribution of wealth-​gaining activities toward a narrow group of highly connected individuals.

1.2.  Visualizing Elite Networks In order to understand the topocratic nature of elite networks, it is perhaps best to visualize a few of the most notable examples. The obvious place to start is the revolving door between politics and finance, focusing specifically on individuals who achieved high levels of centrality within their networks by building tight relationships in both the corporate and the political world. These are typically individuals with a high number of connections (high network degree), and high levels of betweenness centrality, meaning they stand as a central node between two or more distant groups providing a connection between them. Their position is valuable precisely because of their level of connections and network centrality. Network theory recognizes such individuals as superhubs.25 In order to draw networks of a few selected superhubs, I use a unique database of over one million individual-​level corporate, political, and public office connections in the United States.26 This database was the cornerstone for the empirical findings presented in Chapter 4 of this book. It includes connections between corporate executives and people sitting on corporate boards of all publicly listed companies in the United States. In addition to corporate executives, the database also lists all politicians or academics sitting on various corporate boards, as well as public officials in higher office positions. The database allows the user to search across previous careers, education, or membership in various organizations (business clubs, nonprofits, charities, church groups, etc.) as a way to link different individuals and establish the strength of one’s connection to other powerful individuals. When drawing such networks, there are always a few people that stand out—​ those with a high number of connections (high degree), and those in the center of the network, connecting individuals through various links (high betweenness centrality). The graphs are drawn so as to include only first-​degree connections27 for the person of interest, and we see all the interconnections between those connected to that person, the central node. Each node in the figure represents one person, and the lines are connections/​links between them. Individuals with high levels of betweenness centrality will be at the center for a lot of groups/​ clusters that surround them. They will typically not be part of a cluster, but will stand in the middle between them and connect different clusters together.

Why Study Elite Networks?  47 The first example is an embodiment of a superhub drawing connections between Wall Street and public office, and one of the most powerful and controversial figures in finance in the 1990s and 2000s, former US Treasury Secretary Robert Rubin. Prior to becoming President Clinton’s treasury secretary in 1995, he worked at Goldman Sachs for 26 years, eventually becoming cochairman, before moving on to public service. As treasury secretary he carried out the infamous repeal of the Glass-​Steagall Act, which used to separate investment and commercial banking, and oversaw a number of loosened financial regulations concerning derivatives.28 Many have blamed these policies in particular for the calamity that ensued in 2008. After his tenure at the Treasury, he moved on to Citigroup in various positions, but controversially resigned in January 2009, at the height of the crisis, while taking $17 million in cash and $33 million in stock options, receiving in total $126 million during his time at Citigroup. He left when the bank was virtually insolvent, having to be bailed out by US taxpayers.29 He also came under investigation in 2001, when he used his former connections at the Treasury asking if they could convince credit rating agencies not to downgrade Enron’s corporate debt, of which Citigroup was a large holder. He was cleared of any conflict of interest in that case.30 Rubin, therefore, is a man who held strong ties to Wall Street elites and obviously carried significant political clout as the man who spearheaded President Clinton’s very successful economic agenda. He was a superhub of an elite network, acting as an important connector between the political establishment of the ruling Democratic party and the initially skeptical business community. Rubin had no trouble using the power granted to him by his network centrality to promote his interests, both while in office and especially after he came back to Wall Street. His network, drawn in Figure 1.3, clearly shows this central position connecting a total of six different groups, coalesced into clusters. Each cluster corresponds to a different part of his life and career. The majority of people in this network are either from politics or banking, followed by other corporate executives and wealthy investors. For example, the red cluster in the far upper corner are mostly people he knew from banking and business (connected to him throughout his career either at Goldman or Citigroup). The light blue group furthest down are people connected via Blair Effron, a cofounder of Centerview Partners, an investment banking firm. Then there in the pink cluster (up, right) are people Rubin met through politics like the former Clinton advisor and White House communications director under both Reagan and Ford, David Gergen, or Peter Orszag, another individual who made a transition from public office (during Clinton and Obama administrations) first into Citigroup (just like Rubin), and then becoming the CEO of Lazard. Also in the same cluster is the

48  Elite Networks

Figure 1.3  Network of Robert Rubin (big black middle node), former US Treasury Secretary and former co-​chairman of Goldman Sachs.

former President Bill Clinton himself, along with several other high-​ranking politicians. In the green cluster (down, right) are people from the academia that Rubin also encountered through politics and possibly the Council on Foreign Relations, like Glen Hutchins (Silver Lake, also serving in the Clinton administration), Columbia professor Lynn Thoman (who is the big connector between the green and the dark blue clusters (down, left)—​featuring lesser-​known businessmen and investors), Sheila Burke (former Dean of Harvard Kennedy School), and so on. Then there are people like Rubin’s successor at the Treasury, Larry Summers,

Why Study Elite Networks?  49 also in the middle of connecting a few clusters between the corporate world, academia, and politics. Finally, in the orange cluster (up, left) are other notable corporate CEOs and Wall Street executives that complete the deep strength of his network. Rubin finished his career as the chairman of the Council on Foreign Relations for 10 years before retiring in 2017. The Council is a nonpartisan think tank that, among other things, publishes the journal Foreign Affairs, and within its membership gathers many senior politicians, former secretaries of state, diplomats, corporate executives, bankers, and media figures. More than a suitable end-​of-​ career position for a man with such network centrality. In fact, the Council itself represents a well-​established location for an elite network hangout. It is a place where network ties between powerful individuals are established and deepened. The Council is obviously not the only such place, nor is it the one that necessarily attracts the most powerful superhubs (like the Davos World Economic Forum), but it is a place where networks are broadened and interests get aligned. Rubin’s superhub status stands at the center of such an elite network. The 2008 crisis forced Rubin out of banking, albeit with a hefty compensation, but it set the stage for three other individuals whose elite network superhub status was essential in curbing the September–​October post-​Lehman panic. Henry Paulson, the treasury secretary under President Bush, and prior to that CEO of Goldman Sachs; Ben Bernanke, the Fed Chairman; and Tim Geithner, the president of the New York Fed, and the first treasury secretary under President Obama. The three of them conceived and carried out government bailouts of major US banks during the most turbulent period for the US financial industry since the Great Depression. The crux of this episode, arguably quite successful in restoring stability and mitigating further panic, were regular daily formal and informal meetings and phone calls held between Paulson, Geithner, and the chief executives of the nine biggest banks in the country.31 The goal was to figure out the best response strategy to the ongoing post-​Lehman panic. The response came in the form of the Troubled Asset Relief Program (TARP), a bailout package that diverted over $180bn taxpayer funds to the nine biggest banks that were involved, at least informally, in drafting the package. The people in charge of the big banks had deep connections to the two key regulators who were drafting the TARP bill. Paulson used to be CEO of Goldman Sachs prior to becoming treasury secretary, while Geithner’s position as the president of the New York Fed is chosen by a board containing the chief executives of all the biggest banks in the city. This is another typical example of an elite network, one that ties together key public officials with the most powerful corporate executives in the process of drafting regulation that significantly affects those very same corporate executives and their compensations (primarily through bonus schemes). It is an

50  Elite Networks elite network outcome that directly affects top incomes of politically connected executives, thus adding to income inequality. Most importantly, the elite network relationship in this case is not illegal nor illegitimate. It is a spontaneous self-​preservation instinct that guides behavior of elite network members to pursue their goals. During the biggest financial panic of their lives, it was hardly surprising to see chief executives grabbing any opportunity to persuade policymakers, members of their own elite network, to help them. Figure 1.4 shows Paulson’s network. His network, just like those of Geithner and Bernanke, was drawn by excluding the other two in order to avoid congestion in the graph. Paulson’s network is less dense and less diverse than Rubin’s from Figure 1.3, but it does paint a very accurate image of Paulson’s importance and centrality. The closest to him, for example, is another big connector (the big green node, immediately above him), Muneer Satter, his former colleague from Goldman, and founder of Satter Investment Management.

Figure 1.4  Network of Henry Paulson (big black middle node), former US Treasury Secretary, and former CEO of Goldman Sachs.

Why Study Elite Networks?  51 There are also academics (red clusters, sparsely connected, to the right) like Anne Krueger (famous for her 1974 paper on rent-​seeking, who later worked at the World Bank and the IMF), or Chicago economics professor Michael Greenstone. In the more dispersed clusters (lower, right) are people from the corporate world like former American Express CEO Ken Chenault, or Disney CEO Bob Iger, but also many academics, journalists, and politicians Paulson encountered through his Treasury tenure. They were all connected to him through his tenure in politics, but are in separate clusters because people from the corporate world are typically not well connected to journalists or academics. Paulson was the connector. The green cluster (upper, right) contains individuals from the banking world, ranging from his former firm Goldman Sachs to executives from JP Morgan, Morgan Stanley, Citigroup, Bank of America, etc. In short, all the executives32 he had regular contact with during the banking crisis. Note the high clustering coefficient of this group, suggesting a typical elite network pattern of interconnectedness of people working in big banks. Finally, the blue and orange clusters (left), both closely interconnected into one single cluster,33 contain individuals from the academia and the NGO sector who work on environmental issues concerning climate change. This is because Paulson, after his public office career, set up the Paulson Institute with a goal of advocating cleaner environment initiatives and fighting climate change. He is also engaged in a number of similar environmental activities and NGOs, from which he draws the rest of the cluster. Figure 1.5 shows Geithner’s network. Geithner was the subject of an interesting academic research paper34 that presented the social connections during a crisis hypothesis. The idea is that every person of power has a narrow social network on which they rely during high-​stress episodes when crucial decisions need to be made. Geithner’s connections to the industry were so important that the stock price of every firm connected to him went up and delivered abnormal returns immediately after he was nominated by President-​elect Obama. Geithner’s position of treasury secretary during the Obama administration was very similar to Rubin’s appointment during Clinton—​they both sent a comforting signal to the industry that their elite network member has the highest level of decision-​making power. It was reassuring for the industry that the same person they held deep ties to during the previous administration, the person whom they’ve aided in drafting the TARP proposal, is now at an even greater position of power to potentially benefit them. This is obvious once looking at Geithner’s network. It resembles that of Bob Rubin, in that it has seven different clusters all connected through him as the central node. Interestingly, the second-​biggest node in Geithner’s network (the big red node close to the dark blue cluster, in the middle, slightly to the right) is

52  Elite Networks

Figure 1.5  Network of Tim Geithner (big black middle node), former US Treasury Secretary and former president of the New York Fed.

Jack Lew, former lawyer and congressional staffer, who became Geithner’s successor as treasury secretary, after which he left to become the COO of Citigroup. Geithner was, throughout his career, much more involved with public service and politics than both Paulson and Rubin, which his network reveals. It is almost double the size of Rubin’s and almost three times the size of Paulson’s. He joined Wall Street after his public office tenure, just like Lew, so it makes sense that their networks will be slightly different from Rubin’s and Paulson’s, despite all three being good examples of elite network clusters. A lot of the same people from banking that were connected to Paulson are also connected to Geithner, mainly through the green (upper) and red clusters (right, and middle). Incidentally, the upper green cluster also features all individuals working at Warburg Pincus, the private equity firm where Geithner was appointed as CEO after his public office

Why Study Elite Networks?  53 career. Geithner was the key connector between Warburg Pincus and JP Morgan, for example, leading to a multibillion-​dollar investment from JP Morgan into Warburg in 2016.35 All of these banking connections are present in the network. The middle red clusters, scattered all around, contain people Geithner met through his time in office; on one hand people from the Obama administration, like the aforementioned Lew, former chiefs of staff, and other cabinet members, to people from the World Bank, European Bank for Reconstruction and Development, or the IMF, like David Lipton, to foreign ministers of finance, like Vitor Gaspar from Portugal, Simeon Djankov from Bulgaria, Donald Kaberuka from Rwanda, etc. Each group is a cluster within itself; the IMF and World Bank are one, the European finance ministers are another, African and Asian finance ministers yet another. Geithner also draws connections through his membership in the Council on Foreign Relations (CFR), the aforementioned elite network hangout. These belong to the blue cluster (far left) and they include some notable names like three former secretaries of state, Condoleezza Rice, Madeleine Albright, and even Henry Kissinger, but also other notable diplomats, former bankers from JP Morgan, Goldman Sachs, Citigroup, Wells Fargo, or senior public officials such as former World Bank president James Wolfensohn. Note also that the far-​left cluster is, via Geithner and two other nodes, connected to the pink cluster (lower left) containing mainly former politicians, most notably Barack Obama and Hillary Clinton, and many public officials that served in the Obama administration (chiefs of staff, press secretaries, some cabinet members, and so on). The key connectors here, between the green (upper), purple (lower, left), and red clusters (middle), besides obviously the central node Geithner, are the billionaire David Rubenstein (chairman of the CFR and member of many boards, a true networker), Obama’s deputy chief of staff Mona Sutphen (another public servant who switched public office for a private equity firm and later ended up at UBS), Sharon Percy Rockefeller, the former first lady of Virginia, and active member of the Bilderberg Group, and the aforementioned former IMF director David Lipton. The yellow cluster (lower, right) are all current or former employees of the US Treasury, some of whom moved on to other careers in finance, the corporate sector, or diplomacy. Note that the big red node, Jack Lew, is connected to the very same people after succeeding Geithner. Geithner, therefore, successfully connects several elite networks: the CFR network, the private equity and banker network, the network of foreign ministers of finance, and the network of former or current public officials and high-​ ranking politicians. He is the embodiment of an elite network member with high betweenness centrality, and hence extreme importance. No wonder his presence was crucial during the height of the 2008 panic.

54  Elite Networks

Figure 1.6  Network of Ben Bernanke (big black middle node), former Fed Chairman.

Ben Bernanke, the last member of the trio, has an even larger network than all the three former Treasury Secretaries, depicted in Figure 1.6. However, the majority of his network, the big blue cluster in the upper right corner, consists of academics, given that he came to the Fed from the position of a tenured Princeton professor. After the Fed, he deepened his academic connections through his position as a fellow at Brookings Institution. The smaller pink nodes attached to the blue clusters are also academics, but academics Bernanke met outside Princeton or Brookings. The scattered pink connections (far right, and middle, around the center) contain people from the IMF, the World Bank, and many other central bankers across the world. Some of these nodes include people like Stanley Fisher, another academic economist who served as a Federal Reserve vice chair and a governor of the Bank of Israel, and Anne Krueger, an academic economist taking positions at the IMF and the

Why Study Elite Networks?  55 World Bank. An interesting connection is the big blue node in front of the red nodes (first big node next to the center)—​this is Bernanke’s successor and Biden’s treasury secretary, Janet Yellen. She too has numerous links between the academic blue clusters and the red cluster (down, right) which are mostly former politicians and public officials. Bernanke’s other connections are linked to his post-​Fed advisory positions at Pimco and Citadel. The Pimco cluster is green (down, far left), featuring two interesting connectors. One is former European Central Bank chairman Jean-​ Claude Trichet, who took a similar position to Bernanke’s at Pimco following the end of his term (Pimco, apparently, enjoys appointing former central bank governors), and the other is the Nobel laurate behavioral economist Richard Thaler. He links Pimco to the blue academic cluster, while Trichet links it mostly to the political red cluster. Bernanke links them to both. Finally, the yellow cluster (up, left) are people from Citadel, the market maker where Bernanke serves as an advisor. The biggest node in that group, connecting the yellow cluster with the red and pink ones in the middle, is Citadel’s founder Ken Griffin. He too sits at a very interesting intersection between politics, business, and public office. Next, we divert our attention to two very powerful figures in the financial world, both considered to be important networking nodes with high levels of betweenness centrality, and whose networks are characterized with high levels of assortativity. The first is JP Morgan CEO, Jamie Dimon, often dubbed the public’s favorite Wall Street banker, while the second is one of the most famous investors of all time, George Soros. These two are interesting as they, unlike the previous four, still yielded a powerful influence in finance and investing at the time their networks were drawn (latest data from 2019). Their networks, however, differ almost dramatically (Figure 1.7). Dimon’s network is huge, he is one of the most connected individuals in the entire database. Soros on the other hand, while still a highly connected and important node, has a much less dense network and is connected to only one-​fifth of the corporate bosses and politicians that Dimon is connected to. This is an interesting observation given that these two represent alpha personalities that attract even the most powerful elite network members at gatherings like the CFR or the Davos forum (as eloquently presented in Navidi’s Superhubs). Finally, unlike the aforementioned, who are no longer in the same position of power (although they still carry a lot of influence), Dimon and Soros still possess real power; the first through his long-​lasting position as the CEO of America’s biggest bank, and the second as an intriguing world-​famous billionaire investor, notorious for being called “the man who broke the Bank of England,” but also an often-​used scapegoat for many conservative politicians who like to emphasize his links to the Clinton family,

56  Elite Networks

Figure 1.7  Networks of Jamie Dimon (upper), and George Soros (lower).

Why Study Elite Networks?  57 or his promotion of a liberal agenda (which are the basis of frequent antisemitic attacks on him personally, especially in his native Hungary). Despite their similarity, Soros seems much less connected to the US corporate and political world than Dimon. In fact, most of Soros’s connections are people working for his foundations: academics (like Lord Robert Skidelsky or Hans-​ Werner Sinn), notable economists (like Financial Times’s John Kay or former Financial Services Authority chairman Lord Adair Turner), and one former central banker, Paul Volcker. There are very few politicians (the Clintons, former New York mayor and billionaire Mike Bloomberg, and a few ambassadors like Susan Rice) and only a handful of investors and bankers (most working for some of his funds). His network is however more international, spanning from the United Kingdom and Europe to the Middle East (Emirates and Qatar) and Southeast Asia. Dimon, on the other hand, is connected to a huge number of former or active corporate executives (the red clusters, to the left), and an even larger number of people in banking and finance (the blue clusters, to the right). Politicians and public officials are scattered in the green and pink clusters (lower left corner). His network contains some of the same aforementioned names, including former treasury secretaries Geithner, Lew, or Summers, central bank governor Bernanke, and every big bank CEO who held regular meetings with Paulson and Geithner during the September 2008 panic (Lloyd Blankefin of Goldman Sachs, Ken Lewis of Bank of America, Vikram Pandit of Citigroup, the controversial John Thain of Merrill Lynch, and so on). Dimon’s network reaches much broader than finance. He is connected to many notable corporate executives, like the long-​lasting General Electric CEO Jack Welch, Microsoft CEO Steve Ballmer, Exxon CEO Rex Tillerson (who later became Trump’s first secretary of state), in addition to some of the wealthiest men in the world like Amazon’s Jeff Bezos, LVMH’s Bernard Arnault, or Dell’s Michael Dell. The network does not stop there, it transcends into politics to include former president Barack Obama and former secretary of state Condoleezza Rice in addition to many senators, congressmen, ambassadors, and a host of other powerful people like lawyer Robert Shapiro, or James Murdoch, former CEO of 21st Century Fox and the son of media mogul Rupert Murdoch. With a powerful network like this one, Dimon is obviously proving his worth to JP Morgan. A central elite network node, a corporate executive with links to many people at key decision-​making positions, who is able to use his networking power to deliver successful outcomes to his firm. For this he is consistently rewarded, to a point where he is among the few corporate executives that have become billionaires.36 It is easy to see how this case in particular exacerbates the income inequality issue at the very top of the income distribution. Elite network mechanism at work. Unintended, surely, but highly effective.

58  Elite Networks The only person that can be compared to Dimon in terms of networking prowess and who still yields considerable business power is Larry Fink, CEO and founder of the biggest global investment fund, BlackRock. Fink runs a company with $9 trillion assets under management, which holds equity in over 5,000 of the largest US and global publicly listed companies.37 He possesses considerable power in terms of influencing management decisions of companies in which BlackRock holds equity. As Figure 1.8 shows, Fink’s network is very similar to Dimon’s. High degree, huge betweenness centrality of the central node, and high level of assortativity. In fact, Fink is connected to almost all of the same people Dimon is connected to, including Dimon himself (the big green node, close to the center, with high betweenness centrality): Paulson, Geithner, Welch, Ballmer, Tillerson, Bezos, the Murdoch family, Moynihan, Blankfein, and a number of other decision-​makers in banking, the corporate world, and politics. In fact, Fink’s exposure to politics is even greater than Dimon’s due to his board membership at the Council on Foreign Relations and the World Economic Forum. He is also on the board of the

Figure 1.8  Network of Larry Fink (big black middle node), founder and CEO of BlackRock.

Why Study Elite Networks?  59 Robin Hood Foundation, the most famous New York charity organization, a networking hub for a myriad of investors, bankers, and politicians. It is not surprising that Dimon’s and Fink’s networks are so similar. They both possess highly assortative networks, each clustering around the same powerful people, suggesting a clear elite networking pattern that makes such networks highly topocratic. Recall from the previous section that a system is topocratic if the compensation and power available to an individual is determined primarily by their position in a network. This was true for all the aforementioned public officials as well: the 2008 crisis trio and Rubin. From these examples it is easy to see how topocratic networks encourage the dissemination of privileged information, exchange of favors, and soliciting advice in times of great uncertainty and adversity (like during the 2008 crisis). However, despite such implications, the existence of elite networks is a natural occurrence in any society. People in positions of power, whether corporate or political, will always be drawn to other similar people in power. This may often yield positive consequences for societal development. Elites can use their power to deliver beneficial outcomes and even provide for certain public goods (through the gift exchange principle—​a good example is the Robin Hood Foundation). The problem is not with the motivation for entry nor with the existence or assortativity of elite networks. The problem is the extent to which elite network membership allows its members to capture power and bend rules in their favor. This is the outcome that implicates wealth concentration and high inequality, and this is the outcome that needs to be prevented. The logic of elite network formation will be further examined in Part II of the book, particularly through Chapter 7, which examines corporate cultural capture and superhub links between finance and politics. It explains how and why personal relationships and connections between the central nodes in politics, public office, and the corporate world are all but inevitable. It also shows empirical evidence of how connections to politicians and public officials made an important difference in the allocation of bank bailouts during the 2008 financial crisis.

1.3.  Putting Elite Networks on the Map of Political Interactions In order to fully understand elite network behaviors and their consequential outcomes, we should also place them in the correct context of a broader pattern of interest group behavior. According to Olson, the more narrowly defined groups are the ones with the best ability to solve the collective action problem, and given their narrow scope of interest, they are more likely to exhibit stronger

60  Elite Networks preferences and high salience for the issue at hand. This concentration of preferences coupled with their ability to eliminate the free rider problem will, over time, give them enough resources to gain considerable power in pursuing their interests. We can derive two factors that determine the relative strength, and hence success, of an organized group: (1) salience (s) of the group’s preferences toward a given set of socioeconomic and political issues—​this determines strength of members’ preferences and their ability to solve the collective action problem; and (2) the amount of resources, that is, money and power (p) the group has at their disposal—​this is required in order to advocate and fight for the collective interest of the members. Salience is defined over a set of socioeconomic and political issues instead of just for a single issue. Every group has high salience for at least one issue. They would not form a group in the first place if there was no issue over which its members’ preferences are aligned. The set of socioeconomic and political issues represents the cause every group is fighting for. The causes range from altruistic, where the group satisfies their interest by helping others, to purely selfish, where the group cares only about promoting its members’ personal goals. A member of an elite network cares very strongly about preserving its status and will use any argument (like building a better society) to justify their cause. Salience can therefore be best expressed as a measure of strength of the members’ preferences. Money and power are used interchangeably as a single category. Money refers simply to the amount of funds a group has at its disposal. Power supplements the money category. For example, unions in the 1960s in Western Europe and the United States did have a lot of resources, but the primary reason for their success was not money, but the bargaining power they held. Churches in some countries have a lot of resources, but in others their access to money is limited. In each case, however, they carry real power over a large group of churchgoers. An elected political in office has real power over the distribution of the budget. In his case, power leads to money. A dictator has even more than that. In addition to distributing resources, the dictator also makes choices over life or death; hence his power is maximized and uncorrelated to money. Based on the realization of these two factors we can draw a two-​dimensional spectrum of the category of groups we find in a society. Each category is broadly defined to include a number of individual interest groups that best fit the categorical definition. For example, categories like NGOs, church organizations, lobby groups, or labor unions include all possible individual interest groups which will obviously differ between each other with respect to their level of resources and their relative power. This is why I only focus on categories as opposed to spatially defining case examples of individual groups. The two-​dimensional spectrum is drawn on Figure 1.9.

Why Study Elite Networks?  61 1 0.8

Jetsetters and yuppies

Money/Power (p)

Dictator and his cronies Elite networks (in democracies)

0.6 Military in democracies (non US)

0.4

Lobby groups Unions (1960s)

0.2 –1

–0.8

–0.6

–0.4

–0.2

0 –0.2

0

–0.4 SMEs Dispersed groups (consumers, unemployed)

Scientific organizations (academia)

0.2

0.4

–1 Salience (s)

0.8

1

Religious organizations Unions (2000s)

–0.6 –0.8

0.6

NGOs type 1 (Greenpeace) NGOs type 2 (charities)

Sects and cults

Figure 1.9  Two-​dimensional spectrum of interest group categories by salience (s) and money/​power (p). Elite networks are part of the first quadrant, but being positioned in the upper right quadrant does not imply elite network status.

Note that each point on Figure 1.9 is the hypothesized average value of s and p for each group, meaning that for each point there is both a horizontal and a vertical distribution within which each individual group is located. Think of each point as the intersection of mean values of the distributions for salience and money/​power for each interest group category. For example, some sects are certainly more resourceful (or even much less salient) than others, but on average their clustering in terms of money/​power is on the lower right end of the spectrum. Also, some lobby groups and unions are more powerful than others, but on average they are never as powerful as an elite network (provided that they operate outside of it). There are four possible spatial positions, each corresponding to one quadrant on the figure: If, p > 0, s > 0 the group is very powerful and is able to obtain considerable resources, while its members possess strong lexicographic preferences over their preferred set of outcomes (such as maintaining their relative level of power). In this first quadrant (upper right) we would find typical examples of elite networks as defined in the book: a collusion of political and corporate elites in democracies. In dictatorships this elite network is the group that surrounds the dictator, usually referred to as his cronies. The dictator’s elite network is identical to the ones present in premodern times, represented by kings and nobility.

62  Elite Networks This first quadrant also includes political parties and various lobbyist organizations which are not always direct members of political elite networks but are close enough to render considerable influence. The majority of political party members in democracies cannot be considered as part of elite networks (regular members, volunteers, and activists), but the few that hold elected office are elite network members and therefore belong in a different category. Similarly, if a particular lobbyist is a member of an elite network, then he or she no longer belongs in the category of lobbyists. An example of a lobby group that is not necessarily part of an elite network would be the AARP. The NRA is an example of the latter, as it exhibits significant influence over politicians via campaign donations. A similar level of influence was held by the labor unions in the West in the 1960s and 1970s. They had a large and well-​organized membership with strong preferences in addition to considerable resources, which made them a powerful player on the political spectrum. Their modern counterparts however belong in the lower quadrant—​they maintain the same relative level of salience (they still fight strongly for their issues) but they do not possess nearly as much bargaining power as they did in the 1960s. If p < 0, s > 0, the group is less resourceful but still contains a considerable amount of salience toward a set of issues they find important. These groups belong in the fourth quadrant. Typical examples include religious organizations, various NGOs and modern-​day unions, and an extreme version—​various sects and cults which on average do not carry any de facto power (except over its members) but do have considerable salience over their beliefs on what a society should look like. A sect is a good example of a group whose members possess very strong preferences over a defined set of issues, but outside this very narrow group they have virtually no influence at all. If a sect increases its resources enabling it to turn into a proper religious organization (like scientology), their spatial position changes and they move up within the category of religious organizations. NGOs are the only category for which I define two types: the more aggressive ones like Greenpeace or antiabortion activists, and the more passive ones like various charity organizations or advocacy groups. If p > 0, s < 0, the group is wealthy (or powerful) but its members’ preferences on various political and socioeconomic issues are not as strong as with other groups. They usually have a very small number of things they care strongly about and, if necessary, they might use their money/​power to influence it. These include the idle rich like jetsetters (typically with inherited wealth) and to some extent yuppies (in the 1980s) and social media influencers (nowadays) whose main concern is their lifestyle, rather than societal issues. The military, on the other hand, is a very hierarchical and strongly aligned group and should, by every rationale, always belong in the first quadrant. In dictatorships they do, in fact, belong there—​as members of the dictator’s power-​holding coalition. But in

Why Study Elite Networks?  63 democracies, without exposure to war, the military usually does not get involved in current affairs. The strength of their preferences, their salience, is still very strong, but their issues are limited, which is why under normal conditions they belong in the second quadrant. The military is the best example of a group that can shift very quickly between quadrants (from the second to the first), primarily because of the relative resources that they possess. If a war occurs, they are immediately placed within the first quadrant. Finally, if p < 0, s < 0, the group is typically too dispersed and unorganized to solve the free rider problem of its members which is why very often the group is never able to successfully organize to promote its interest. The best examples are Olson’s definitions of dispersed groups such as consumers or the unemployed. They are too diverse and too geographically dispersed to be able to organize and thus act in their common interest. This makes both the consumers and the unemployed classical price-​takers. The strength of their preferences may be strong, but they are unwilling or unable to solve the collective action problem. When a consumer group, for example, becomes organized in order to demand better quality service or better regulation of a given industry, they then switch to the fourth quadrant, and enter the distribution space of type-​2 NGOs. If they feel really strongly about an issue and are able to generate more resources, they join the distribution space of a type-​1 NGO or even a lobby organization or political party. Even though Figure 1.9 represents static relationships between categories, the behavior of individual groups is always dynamic. The relative position of a group on the spectrum may change. For example, politicians are members of elite networks when and if they are in power. When they are not in power, or are not even expected to be in power, their position on the spectrum will alter, and will move most likely to the fourth quadrant (when in opposition but with a high probability to win office at one point), or very low in the fourth and converging to the third quadrant (when in opposition but with a very low probability to win office at any point). The given two-​dimensional space does not represent individual fixed points where we would locate each and every example of an interest group we can find. It represents the mean values that individual groups converge to or diverge from. The main point of Figure 1.9 is to showcase the spatial position of an interest group necessary to successfully advocate one’s interest. Individual groups within each category (NGOs, unions, religious organizations, lobbyists, political parties) will position themselves somewhere within the spatial intersection of the two distributions for each category, and may switch categories depending on their relative exposure to resources or the discipline of their membership. In order to be fully successful in promoting their members’ interests, a group should aim to be located in the first quadrant. The level of success in the fourth

64  Elite Networks quadrant depends on how close they are to the first, or in other words, on the number of resources they have. Members of the second quadrant, because they are resourceful, can easily switch to the first if they stumble on an issue of high enough salience that will trigger their membership (e.g., a war or budget cuts for the military, or an issue a group of influencers feels strongly enough to promote and change). However, members of the third quadrant will have the hardest time shifting closer to the first as they face two main obstacles—​(1) lack of resources and (2) lack of incentive to organize.

1.4.  Quantifying Elite Network Outcomes Empirical efforts at recognizing who is and who is not a member of an elite group have usually been limited to occupational studies where all leaders of business, political, or military organizations have been classified as the elite.38 This is helpful in understanding the network of connections between powerful individuals, but is not enough to tie their connections to a particular set of outcomes. This book is primarily focused on outcomes that are supposed to result from an elite network–​type of relationship, rather than listing and analyzing its individual members and their connections. The relationship that leads to an outcome is hidden (deals done behind closed doors), but the outcome itself is visible, as are the connections between powerful individuals. For example, a politician whose district is riven with fraudulent procurement contracts keeps winning elections and stays in power for a long period of time. The same politician has a carefully selected group of local businessmen and power brokers on his side, which provide him with the necessary funds and votes in exchange for favorable contracts with the government. A symmetric example is a firm whose top management has a clear connection to politics and receives a beneficial treatment from the government, where the decision is made by the very same politicians to whom they are connected. To bind these outcomes together I construct an indicator of elite membership by looking at how certain top executives use their political connections and their relative position within a social network in order to drive up their compensation packages. Essentially there is no precise way of measuring an exact elite network relationship, mostly because their interactions are hidden from the general public, but it is possible to apply a number of approximations by making inferences from direct outcomes that have the highest probability to signal an elite network–​type of relationship. As defined earlier, the reason why someone would enter into an elite network is to preserve their position of power. All elite networks throughout recorded

Why Study Elite Networks?  65 human history depend on preserving power and wealth. Modern elite networks, at least in democracies, are composed of politicians whose desire is to preserve power and rent-​extracting firms and their executives whose desire is to uphold wealth. This book shows two things: first, it establishes that both parts of an elite network benefit from proximity to each other, and second, it establishes how their relationship impacts the distribution of top incomes. In other words, it uncovers how both types of powerful individuals benefit from being included in this network of influence, and what unwanted consequences their relationship entails. What are the typical outcomes that arise from an elite network relationship and are somewhat observable in the data? Corruption is one. In particular, a specific form of corruption from procurement contracts that are allocated on a suspicious and hence potentially fraudulent procedure. Or the type of corruption where politicians change urban planning laws and other legal and regulatory procedures in order to benefit a specific set of actors. The focus is on office-​ holding politicians, and the potential benefit they can incur if they are part of informal groups to which they distribute rents and are free to take rents from. There is no way, apart from criminal investigations, to directly prove that the owner or CEO of the firm received a contract because of their connection to a politician. However, in many cases based on the firms’ relative performance and status it is possible to deduce a potentially suspicious relationship that implicates the inner dealings of an elite network. Corruption is therefore approximated through potentially fraudulent procurement contracts received by rent-​extracting firms whose primary source of income is conditional on political connections to get government contracts. Examples include cases in which firms with no employees and as a single bidder on the tender received procurement contracts worth millions, in which firms signed contracts vastly exceeding their capabilities, or when firms with substantial losses received indirect subsidies through lucrative deals. The intuition behind the formation of each of these corruption proxies is based on qualitative evidence using interviews with key stakeholders in the procurement process. Observing fraudulent public procurements is the easiest way to establish a link between corrupt politicians and rent-​seeking firms, especially if the allocation of procurements depends entirely on discretionary power of the politicians. Chapter 6 uncovers the empirical relationship between corruption driven by elite network membership and political survival. Another outcome that testifies to an elite network type of behavior is when firms that are directly and undoubtedly connected to politicians (via lobbying, campaign donations, or their high-​ranking personnel) get favorable deals from the government compared to similar firms that are not connected to politics. Although the outcome need not be a direct result of elite network

66  Elite Networks membership, any causal effect of political connections on a government subsidy again implicates the existence of an informal relationship through which the decision-​ making process was molded. Empirical evidence of this effect is presented in Chapter 7 by looking at the impact of firm-​level political connections on policy favors using the case of government bailout (TARP) funds allocated to the US finance industry during the 2008–​2009 financial crisis. Financial institutions that received TARP funds used lobbying, campaign spending during the 2008 cycle, and direct links their corporate management had with the decision-​makers (Congress, Treasury, Fed) to get a better bailout deal for themselves. Politically connected firms were clearly favored over unconnected ones during the bailout allocation process. Arguably the most efficient way of quantifying an elite network is to examine individual-​level data of top corporate executives, specifically their memberships in various organizations in addition to their career paths to see if at any point they could have formed a relationship with the political decision-​makers. For example, a corporate CEO might have been working at a top position in government prior to their position in the firm, which would grant them preferential access to former colleagues if a firm needs a government favor. Also, corporate bosses and politicians can be members of the same country clubs or various societies, charities, religious groups, can be tied together through think tanks and NGOs, and can often meet in informal gatherings where they have direct access to privileged information. Once such an informal relationship is established and caressed for a while, favors can be exchanged. Once the executive successfully extracts rents for his firm, he demands a higher compensation as a reward for his efforts. If this theory holds, we would notice a clear difference in earnings between executives who carry some type of political connection and those who do not. An elite network member should always be compensated more than a nonmember. The empirical findings of Chapter 4 confirm this theoretical intuition. Politically connected executives, by assumption elite network members, tend to have higher salaries and total earnings than nonconnected executives. They are also better connected within the corporate world; however, the network effect does not offset the political connection effect—​if anything, it reinforces it. The implication this carries for income inequality is striking. It suggests that one of the main drivers of top wages of today’s supermanagers are their political connections. Being a member of an elite network directly implicates earnings of such individuals and hence drives up income inequality. *** Now that we have a proper contextual definition of an elite network, before we can measure its direct impact on inequality, in Chapter 4, we first need to offer a deeper historical portrayal of the evolution and persistence of elite networks.

Why Study Elite Networks?  67 Elites are an omnipresent feature throughout human history. Whenever there was wealth and power there were elites that commanded it. And whenever elites commanded wealth and power, unequal distribution of wealth and income was a natural consequence. Chapter 2 delivers this cornerstone argument by re-​ examining the history of mankind through the lens of elite rule and the logic of power, while Chapter 3 focuses on democratization trends arising in the late 20th century that made no significant long-​term impact on structural inequality primarily because they could not disrupt the logic of power.

PART I

T HE IMPAC T OF E LI T E N ET WOR KS ON IN E QUA LI T Y The first part of the book delivers the cornerstone historical and empirical arguments that explain the direct impact of elite networks on inequality. The existence of elite networks is contingent on having or being in proximity to political power. As Chapter 2 shows, they are a persistent phenomenon of human history and have existed ever since mankind started accumulating wealth through surpluses from agricultural production. These initial instances of wealth drove demand for the first proto-​states and, by extension, the first demand for condensing political power within a narrow set of individuals. These were the primordial elites, granted a divine right to protect the population and their food surpluses from outside threats. As civilizations progressed, the societal role of the ruling elites has expanded to unprecedented levels of authority and power. An unequal distribution of wealth, condensed within the ruling elites, was a natural consequence. Even as societies progressed further and got rid of the shackles of authoritarianism and Malthusian poverty, elite networks remained. In the 20th century, when democratization finally emerged, there was no significant, long-​lasting impact on structural inequality, as Chapter 3 shows. Democracies, contrary to expectations, had no definitive impact on lowering inequality. They introduced, for the first time ever, the concept of the welfare state, and significantly expanded the size and scope of governments to provide the basic public goods like health care and education. However, the forces of wealth concentration among the privileged, well-​connected groups were never subdued. None of the great reforms and positive social trends of the 20th century solved the real problem behind inequality—​the proximity of elites to political power. Today, elite networks no longer represent a relationship between the kings and the nobles, or between the rulers and the owners of land and capital. Today, they are part of an interdependent relationship between a legitimately elected politician and executives and/​or owners of the most powerful corporations in a country (except for dictatorships, where they are manifested more or less the same way as during premodern times). Chapter 4 shows how modern-​day

70  Elite Networks “supermanagers” affect the distribution of incomes. They exploit their proximity to political power in order to secure rents for their firms, which reward such connected executives with higher salaries and bonuses. This impact on the top 1% and 0.1% of income earners is the main driver of higher inequality in modern-​day societies. The same forces of wealth concentration that have driven elite networking incentives for centuries are present in modern-​day democracies. Their impact on inequality cannot be much different.

2

Evolution of Elite Networks and Inequality Inequality in the Very Long Run

Inequality was regarded as a normal condition and injustice as a personal misfortune or even as an individual’s just deserts rather than as a social evil [. . .] the general pervasiveness of inequality ensured that its legitimacy went unquestioned. —​Bruce G. Trigger (2003) Understanding Early Civilizations, p. 142

Elite networks are driven by the logic of power. Their perseverance throughout human history, centered on the assumption that a powerful well-​organized minority controls the majority of a society’s resources and hence wealth-​seeking opportunities, implies that income and wealth inequality were persistent features of recorded human history, at least ever since the origin of the Agricultural Revolution some 10,000 years ago. Whenever there was wealth, there were elites that wielded control over the extraction and distribution of wealth. And every elite was focused on achieving authority, or to put it simply, on achieving power over others: the power that granted them both de jure and de facto rights over the lives and livelihoods of all others beneath them. But no ruler could ever rule alone. Every ruler rested on a close group of powerful subordinates who were rewarded for their loyalty or, alternatively, who mounted coups against a ruler who no longer held their support. Proximity to the ruler created incentives for the first elite networking. It is here where alliances and enemies were forged. It is from this relationship that the distribution of wealth was decided on. According to Gerhard Lenski’s 1966 seminal book Power and Privilege, inequality varied significantly throughout history and was determined by two major forces: concentration of power and the size of surplus in a society (in terms of available resources and technology). The rate of concentration of power determined how much wealth (surplus) could have been extracted by the elites for themselves. The more powerful the ruler and his clique, the greater the level of inequality.1 This logic is confirmed by a more recent research paper from Milanović, Lindert, and Williamson, who find that all preindustrial societies were characterized by very high potential inequality, or a very high “extraction ratio.”2 An extraction ratio measures the potential level of inequality when wealth Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0003

72  Elite Networks is concentrated in the hands of the elites. High extraction ratios imply a greater probability that wealth is to be extracted by the top income groups, and hence is the main characteristic of an unequal society. The more repressive the ruling elite, the more unequal the society. The many historical portrayals of inequality make it an inescapable fact of life of all premodern societies, as exemplified by the opening quote to this chapter.3 The theory of elite networks offers a new structural explanation behind long-​ term cyclical inequality, one that is deeply rooted in the history of mankind and molded by the insatiable thirst for wealth and power. We must therefore go back in time to when wealth was first being created to understand why elite networks were formed and why they yielded an almost evolutionary incentive for a small ruling elite to confiscate societal wealth. From then on, we can explain the cyclical nature of inequality and how it evolved through millennia, subject to the constraints of autocratic and extractive stationary bandit political orders, Malthusian economic orders, and subject to the interchanging forces of wealth concentration and occasional destruction.

2.1.  The Agricultural Revolution: From Hunter-​Gatherers to Settlers Our journey begins just before the Agricultural Revolution, the first major source of inequality,4 commencing some 12,000 years ago (10,000 years B.C.). By then our human ancestors lived mostly in hunter-​gatherer societies, foraging food and only moving from one place to another if the food supply of their territory was exhausted or if a change in climate forced them to move on. They were already living in kinship-​based bands and tribes and had established social ties between them. The early Homo sapiens, even as hunter-​gatherers and foragers, were very territorial. According to recent anthropological findings, the effect of changing climate conditions, a global cooling phase that lasted from 195,000 to 125,000 years ago, pushed the Homo sapiens species from inside Africa to its coasts.5 There they discovered ample amounts of coastal food, which to them became a dense and predictable source of food supply. This triggered high levels of territoriality among early humans, which led to intergroup conflict. The conflict between tribes enabled conditions for the formation of deeper social behaviors and norms within groups. They worked together to protect their food sources in order to preserve exclusive access to this resource. Early humans were therefore driven both by cooperation (within tribes) and competition (between tribes). Cooperative behavior, in turn, fostered innovation in new technologies, primarily in weaponry. The first projectile weapons—​ spears, bows, and arrows—​developed for the first time around 71,000 years ago.6

Evolution of Elite Networks and Inequality  73 Within-​tribe cooperation was crucial for the success in between-​tribal competition for resources. These technological advancements in weaponry coupled with within-​tribal cooperative behavior allowed the Homo sapiens species to easily adapt to new environments, and encouraged their spread to other continents. They first reached the Middle East around 55,000 years ago, which led to the extinction of the Homo erectus. This was followed by further expansion into Southeast Asia, which led to the extinction of the Denisovan humans, and then to Europe some 45,000 years ago, leading to the extinction of the Neanderthal. Their further spread to all other continents (Australia about 45,000 years ago, North America 14,000 years ago, and South America 13,500 years ago) was in all cases followed by extinction of anyone who lived there—​whether it was another Homo species or the megafauna (large mammals and plants) that existed in the territory. The Homo sapiens, due to their knowledge in the use of advanced tools and weapons, as well as cooperation in hunting, were a constant threat to all big animals and other species of humans they found wherever they settled. The violent and destructive tendencies of our early ancestors were inscribed in their DNA eons ago.7 Opposed to a somewhat romantic portrayal of the early hunter-​gatherers as peaceful idle humans, they were nothing of the sort. Violence and conflict were not unknowns in the world of foragers. According to archaeological excavations it seems that the early humans exhibited constant conflicts between and possibly within groups. All this happened even before the Homo sapiens started to spread and eliminate all other species of humans on Earth. This obviously varied from one group to another, but evidence of violent deaths of early humans certainly does suggest that violence and tribal connections were powerful motivational factors behind early human settlement. Then came the Agricultural Revolution. Contrary to popular belief enlisted in the name itself (revolution commonly suggests a swift change of an old system with a new one), the Agricultural Revolution was a long and gradual process where domestication of plants and animals and the consequential sedentary lifestyle took centuries, even millennia to occur. The process itself can be dated to sometime around 8500 B.C. in the Fertile Crescent and about 7000 B.C. in China, reaching other continents only later. Most importantly, it happened randomly and without the early humans ever realizing it. Sedentary lifestyle was not a consequence of human ingenuity or deliberate action—​it was a consequence of chance. Jared Diamond describes this accidental process of discovery: the plants the foragers consumed (like wheat or various fruits) while spitting and defecating their seeds started to grow around their temporary settlements, particularly if they were located close to water and fertile lands (rivers like Euphrates and Tigris in Mesopotamia or Yangtze in China).8 Once they started to realize this, they turned gradually to cultivating

74  Elite Networks those plants, but still remained nomadic. Not until they started planting crops deliberately were they forced to settle more permanently, a paradigm shift that lasted at least a few thousand years. As more effort was needed to cultivate plants, there was less time available to gather other food. This slow gradual change affected the settlers’ procreation (populations began to grow) as well as their diets, making it difficult to say when exactly the change occurred. What we can be relatively certain of is that during its gradual adoption, the Homo sapiens have hardly noticed it. Each new generation simply lived like the old one, making only minor improvements. Getting used to a luxury often means not being able to imagine life without it. Farming was the luxury, in terms of a stable and predictable source of food supply, and it set the stage for the rest of our development as a civilization. Not all hunter-​gatherers became farmers. Many gave up on it after even a few centuries and continued to develop as nomadic tribes. Yuval Harari develops the argument of how evolution played an important role in switching the Homo sapiens from foragers to farmers: “the currency of evolution is neither hunger nor pain, but rather copies of DNA.”9 In other words, the higher the number of humans, the more successful the species. The basic reproductive instincts of early humans changed once they settled down. Farmers living a sedentary lifestyle could give birth to more babies and in a shorter time span than nomadic hunter-​gatherers. This implied growth of population and considerably larger tribes of settled farmers than nomadic foragers. Furthermore, sedentary lifestyle brought innovation in terms of tools, and with the origination of food surpluses, stronger incentives for protection. Tribes with strong within-​group protection (in the form of first organized units, guards, or even proto-​armies) with better weapons could have easily defeated the roving, nomadic hunter-​gatherers, who had a choice to adapt and become farmers themselves, face starvation (as their food sources were being exhausted), be killed by the farmers, or leave to an isolated territory and continue living detached and in peace (as some did until the modern age, remaining relatively egalitarian). Therefore, one can say that evolutionary motives led us to gradually accept a sedentary lifestyle as this implied a larger number of units of our species. Evolution molded the Homo sapiens into a sedentary species so as to ensure rising reproduction: Man, like every other animal, has no doubt advanced to his present high condition through a struggle for existence consequent on his rapid multiplication; and if he is to advance still higher, it is to be feared that he must remain subject to a severe struggle.

—​Charles Darwin (1871) The Descent of Man, p. 618

Human existence is consequent on rapid multiplication. Because their new agricultural lifestyle made humans multiply as a species, their biological urges

Evolution of Elite Networks and Inequality  75 made sure they kept on multiplying. The sedentary destiny of our ancestors was sealed.

2.2.  How Exactly Did Farming Translate to Inequality? The sedentary lifestyle that turned foragers into farmers gradually changed their behavioral patterns. According to archaeological excavations from the early stages of the Agricultural Revolution, it was obvious that early settlers already experienced inequality even before proto-​states started to ascend. There is ample evidence of burial sites where some bodies were buried with vast amount of gold ornaments and other fine objects, suggesting the importance of this person relative to everyone else in the community. Especially noteworthy are burial sites of children who were buried with lavish decorative objects, most likely suggesting that they were children of someone important and/​or where themselves the object of praise and awe. This implies that in the early stages of the Agricultural Revolution there was a tribal hierarchy in place where tribal leaders and their families enjoyed greater benefits and hence a skewed distribution of wealth. Inequality therefore did not commence with the rise of the state—​it existed well before it. The rise of the state only exacerbated early inequality due to an even steeper hierarchical structure. This too was a gradual process, by no means linear and simple, and driven by a host of factors: population growth, demand for protection of scarce resources, the necessity for organization and leadership, technological progress, and so on. However, patterns differed between territories, regardless of hierarchy. In some there was undoubtedly an increase in inequality compared to the hunter-​ gatherer times whereas in others things remained relatively egalitarian. New evidence suggests the domestication of large animals is Eurasia played a crucial role in generating an unequal distribution of wealth in early sedentary societies, and that this can explain the differences between settlements across the continents.10 The impact of geography and climate was crucial. The Fertile Crescent had a number of advantages for being the first territory with early settler societies already in 8500 B.C. The most important factor was not the fertility of the land, but its biodiversity. This was the largest area in size which contained the largest amount of both plant and animal species suitable for domestication. It had 32 suitable large-​seed plant species compared to only 6 in East Asia, 5 in Mesoamerica, 4 in North America and Sub-​Saharan Africa, 2 in South America, and 1 in South Africa. Similar for large animals. The Eurasian continent had in total 14 large mammal species suitable for domestication (from the major ones like cows, sheep, goats, pigs, and horses, to the minor ones like camels, llamas, donkeys, reindeers, water buffalos, etc.), compared to only 1 in the Americas and

76  Elite Networks 0 in Africa and Australia, as all the big animals in these other territories were either extinct (in the Americas and Australia) or unsuitable for domestication (zebras, hypos, rhinos, elephants, and giraffes are hard to tame, while the carnivorous animals are simply too costly to feed). Mesopotamia therefore had the luxury of having the perfect combination of mammal species suitable not just for pulling plows but also in providing ample supply of meat, clothing material (wool, skin) and easier transport. It also had the greatest variation in climate, which favored the biodiversity, and a wide range of altitudes (from mountains to lowland rivers) suitable for harvesting the different varieties of plants and raising different varieties of cattle. This was important as it did not confine the hunter-​ gatherers to stay in one territory but enabled them to spread out and continue their unconscious gradual settlement.11 The advantage of suitable animal and plant species was not only crucial in societal development of early settlers but also in the development of patterns of inequality. According to new archaeological research the availability of livestock to pull plows and transport both goods and people made land a valuable resource.12 Places that did not have the luxury of large mammal species relied on human labor, meaning that their development was necessarily slower. You could cover much more ground much more efficiently using animal force. Livestock virtually became the first primitive form of capital. Many excavated graves revealed that the most important people within a community regularly got buried not just with their material possessions but also often with their cattle—​in other words, their property. Livestock and land were being transferred to future generation as inheritance, and were also rented to families that could not afford to have them. Access to livestock and land were therefore the first sources of wealth and the first form of distinction based on wealth for early human settler societies. The same research concludes that societies in which farming was labor-​ intensive (based on human labor rather than animals) were much more egalitarian than societies in which farming was land-​intensive (where land ownership was concentrated and livestock was the essential tool necessary for harvesting and transportation). The average difference in estimated Gini coefficients between the two farming systems (on a sample of 90 Euroasian archaeological sites) was 0.3 in favor of the land-​intensive systems (an average of 0.5 in land-​intensive versus an average of 0.2 in labor-​intensive).13 These findings are confirmed in previous research efforts looking at the size of dwellings to establish within-​society equality.14 They confirm that the rise of early settler inequality was primarily due to restricted land usage and that a land-​ limited societal order experienced much higher inequality than a labor-​limited order. This too is attributable primarily to the existence of large mammals suitable for domestication in areas with higher levels of wealth inequality. According to this study the average Gini coefficients were 0.17 for the hunter-​gatherers

Evolution of Elite Networks and Inequality  77 compared to 0.35 for agricultural settlements. The ability of the settlers to adapt mammals for transportation combined with innovations in bronze metallurgy enabled the development of advanced warrior groups that were able not only to defend resources but also expand the settler’s territorial presence.15 The stage was set for the next logical step in human development: the rise of the state.

2.3.  Proto-​States and the Formation of Hierarchical Orders The sedentary lifestyle contributed to humans becoming even more territorial and protective of their fragile food supplies. This change brought to one very important difference in human behavior—​planning and thinking about the future. Foragers made no plans for the future. They preserved the food they could carry and had very limited expectations about the food supply they were to encounter tomorrow. If the food source was exhausted, they simply moved on—​a life very similar to that of our animal antecedents. Sedentary farmers, on the other hand, had no such opportunity. They had to learn how to plan ahead, how to manage to grow enough food to keep everyone in the family unit fed through the winter, and most importantly how to adapt to nature’s uncertainty regarding agriculture (the seasonal cycles—​will it rain, will the river rise enough to irrigate the crops, will there be a flood, how long will the drought last, what if the season was poor, will there be enough food, etc.). In time, people innovated to become less dependent on weather and climate conditions. They adapted. They irrigated fields, they harvested new crops to make extra supply in case the original crop fails, and invented new tools that helped them adapt to weather conditions. How did this affect their lives? Given that they invested a lot of time and resources in their crops they were adamant in protecting them against roving bandits (potential predators and freeloaders) and other tribes. If the roving bandits would rob them of their entire food supply, this literally meant condemning the farmers to starvation. A response had to arise. And it came in the form of first proto-​governments with protective armies, which would be given a part of the farmers’ food supply (surplus) and offer their protection as compensation. In Mancur Olson’s words, the roving bandits were replaced by the stationary bandits16—​the proto-​state, and its ruling elites: the kings, the priests, and the military. The stationary bandit has what Olson calls “encompassing interest” within the territory that he controls. He has the authority to extract and distribute resources and has incentives to curb or promote violence. In other words, a stationary bandit monopolizes violence and crime in one territory. This reduces the risk of outside coercion and defines societies within the limits of expected coercion. It is from these incentives that the first autocratic states were born, from ancient King Sargon of the Akkadian Empire to all future autocrats

78  Elite Networks and ruling elites. Autocracy, operating under the same assumptions as it did in ancient times, was the dominant social order for the vast majority of recorded human history.17 Geography again played an important role. States, as a necessity of protection against outside violence, rose in densely populated river valleys (like Mesopotamia or the Nile valley), where it was relatively easy for predators and other tribes to attack, and which were bounded by a desert or an ocean so that it was difficult for people to flee and escape violence.18 In low-​density areas people would run away, while in areas difficult to reach attacks occurred with a lower frequency. Thus, the demand for protection, and hence hierarchical organization that gave rise to autocratic elites, was strongest in larger, denser, and geographically bounded areas where the population had no choice but to stay put and invoke a protective agent. In the first proto-​states with the first forms of ruling elites, incentives were created for the first primitive form of labor specialization. The majority worked in agriculture, but some, arguably stronger individuals, had to be used to protect the surpluses and thus became soldiers. The strongest among the soldiers became army leaders and, by extension, leaders of the community (this was nothing new, tribal communities always chose leaders based on strength and appeal), while the smartest were used to write, to keep track of the surpluses and later on debts. Early societies started forming their own classes and the first hierarchical dominance structures. Rulers, bureaucrats, and soldiers all started to have an active role: protecting and organizing production so as to feed the community. The newly created occupations relied on the farmers to feed them, while the farmers relied on them for protection. This social contract obviously worked better in some places than in others, but the very basis of the relationship served as a crucial enabler of greater food production and by extension greater population growth. Communities that grew larger in size were more equipped to protect themselves from outsiders. This made them grow even stronger over time. Communities were gradually turning into larger and denser societies. As this change was occurring other things were necessary to bind the rising population, to force them to cooperate and organize them into a disciplined unit that can fulfill the societal function of food production and further procreation. Two conditions were crucial to enable mass cooperation between people who were no longer tied to a small community based on kin. The first was a hierarchical dominance structure based on religion but also trust and confidence in the upper echelons of the hierarchy to determine the rules of a society. The second was the invention and usage of writing. Harari refers to the first condition as an imagined order, a type of subjective, made-​up reality that enabled mass cooperation of early humans.19 People needed a hierarchy to believe in, so they invented one. They invented gods and

Evolution of Elite Networks and Inequality  79 they invented the rulers. The process was autocatalytic, a positive feedback loop; the ruling elites defined what was acceptable and what was not, and drew their influence from divine power. The concept of religion was not invented by the elites (as there is evidence that people believed in animal gods during hunter-​ gatherer times), but it was certainly exploited in order to preserve their position of power. Egyptian pharaohs declared themselves direct descendants of the gods. As did many other rulers across ancient history. This was the easiest way to convince people that the ruler is to be trusted and obeyed without question. There were strong incentives [. . .] for rulers to invest in the right spiritual infrastructure, such as building of lavish places of worship. This offered a lever over internal control, allowing leaders to form a mutually strengthening relationship with the priesthood who, across all the principal religions, wielded substantial moral authority and political power. This did not mean that the rulers were passive, responding to doctrines laid out by an independent class (or in some cases caste). On the contrary, determined rulers could reinforce their authority and dominance by introducing new religious practices.

—​Peter Frankopan (2015) The Silk Roads, p. 28

Religious and political leaders in the first proto-​states therefore created the first ruling elites, all based on an imagined hierarchical order where divine power is bestowed on a ruler who defines how a society and his subjects should conduct themselves. It is exactly through this type of order that cooperation is enabled among complete strangers. If a means of payment is defined by the ruler by having his face printed on the coin, this implies a certain value to that piece of metal that forces two parties who engage in trade to respect its institutional character. They believe in the ruler and thus they believe in his coin. Imagined hierarchical orders, defined this way as beliefs in common myths and common gods, are not necessarily deliberate manipulative mechanisms; instead “they are the only way to make a large number of humans cooperate effectively.”20 They constitute a powerful propaganda machine, as one does not dare to question the motives of the ruling elite just as no one dares to question the motives of the gods. Religious propaganda mechanisms enabled the ruling elites to justify their power throughout human history; from ancient times until today. The invention of writing was the second most important binding condition in fostering cooperation and speeding up the expansion process of ancient empires. Instructing social orders required a way to easily transmit them, in ways other than gossip or word of mouth. Having a larger society implied rapid expansion of information, all of which had to somehow be aggregated. This initial demand for aggregation of information and the spread of rules led to the invention of writing. Writing was first invented and used in ancient Sumer between 3500 and 3000

80  Elite Networks B.C. by bureaucrats to note surpluses in food and who owed what to whom. It was used for accounting purposes, which later advanced into purposes of ruling a kingdom or an empire. Writing was a crucial prerequisite of a well-​functioning state, a state that can successfully maintain political hierarchy and enforce order. The two conditions, writing and a hierarchical order, also formed a positive feedback loop. Writing enabled stronger political organization developed in terms of a well-​known hierarchy. The hierarchy created back then, with the onset of states and rulers, was the one that was prevalent throughout the rest of human history. A hierarchy between classes—​the elites consisted of rulers, priests, and soldiers versus the workers and the farmers—​evolved into a hierarchy between the rich and poor, a hierarchy between owners and slaves, between whites and blacks, and even between men and women. Harari suggests all of these hierarchies were a product of our imagination—​ this is how it is supposed to be. There is no natural distinction between any of these imagined hierarchical orders, only what we have been imagining throughout history. It is immensely difficult to rid ourselves of these burdens as they are by now deeply embedded in our behavior. It is how our parents were raised, how they raise us, and how we will raise our children. When American forefathers signed the Declaration of Independence emphasizing freedom, they were all slave-​owners. Interestingly, none of them saw this as a logical inconsistency, given that slaves were a fact of life, in a well-​established hierarchical order. The justification provided by white slavers in the United States was that black slaves were less intelligent than white owners, which is precisely why they were slaves. No matter how ridiculous this sounds from today’s perspective, merely a few centuries ago this was the dominant political hierarchical order and pseudoscientific studies were being done to “prove” it. Slaves were merely subpar humans. And yet there was never any natural, any biological reason for humans to behave this way. Gender is a similar story. From the onset of civilization women were treated as the property of men. In many legal systems rape was not considered a crime against a woman, but a legal breach of property of the man who “owned” her. The initial divergence between men and women, where men used to hunt and go to wars against other tribes to protect their territories while women gave birth and cared about infants, set the stage for this very long imagined hierarchical order which is still dominant in many countries in the world today. The reason for its persistence was the obvious initial biological difference between men and women.21 The hierarchies that enabled order and dominance of one small group over the entire population meant that at the top of the distribution there was always an elite consistent of rulers, their vassals, and their priests who all had complete control over the extraction and distribution of resources. Belief in a dominance hierarchy enabled the elites to amass fortune gradually over centuries—​long enough

Evolution of Elite Networks and Inequality  81 for anyone to stop questioning what or who gives them the right to seize control over the distribution of resources. Just as no one, not even the ancient philosophical moral authorities like Plato or Aristotle, asked why one should have the right to own another human being, whether a slave or a woman. It was simply the belief system that existed in society that everyone accepted and took for granted. A belief system that developed for millennia. An imagined dominance hierarchy that gave the ruling elites justification over the usage of force and violence in amassing and redistributing wealth, enabled them to define ownership rights over land and all forms of property—​including other human beings like slaves and women—​and enabled them to use religion as a way of upholding their divine right to hold power and justify all their acts of violence and extraction.

2.4.  Concentration of Wealth in Ancient Civilizations: Rise of the Malthusian Economic Model The state was thus born out of demand for protection against freeloading predators. People preferred the certainty of a stationary bandit—​one ruler to whom they had to give a predefined part (albeit large) of their food surplus—​to the uncertainty of having a number of roving bandits randomly killing, raping, and pillaging their settlements. The demand for protection had the effect of creating the first dominant hierarchical power structures where the leader of the settlement was given unlimited power and the incentive to use violence as a method of both protection and expropriation. This was a bad deal for the farmers who still lived barely above subsistence levels given that the newly formed elites usually usurped a huge amount of the food the farmers produced. But at least they had protection and were able to alleviate, to some extent, uncertainty about the future. All of this evolved gradually, over several millennia. The first villages (like Jericho) sprung around 8500 B.C. in the Fertile Crescent shortly after the domestication of plants. By 5000 and 4000 B.C. there were cities across the Mesopotamia (early Sumer and Babylon settlements), and by 3000 B.C. the valley of the Nile was united into the first Egyptian kingdom, counting several hundred thousand people—​already an impressive early state. By 2250 B.C. the first empire was formed in Mesopotamia, the Akkadian Empire, having over a million people and a strong standing army of over 5,000 soldiers. Between 1000 and 500 B.C. the world featured several impressive empires and kingdoms; in addition to the Egyptian Kingdom and the Late Assyrian Empire, there was also the Babylonian Empire and the Persian Empire. At around the same period, in 508 B.C., Ancient Greece was developing a democratic order in its city-​states. After Alexander the Great conquered the Persian empire in 330 B.C., the Qin dynasty united China

82  Elite Networks in 221 B.C., while in Europe, at around the same time, Rome was starting its Republic and was beginning its expansion throughout the Mediterranean. The Roman Empire (from 27 B.C.) was the most impressive of the ancient empires, counting over 100 million subjects, and carrying an army of between 250 and 500 thousand soldiers, not to mention developing a road network still in use today in addition to a multitude of other innovations and cultural contributions. The Agricultural Revolution took about 8,000 years to take us from the hunter-​ gatherers to the pinnacle of early civilization—​the civilizations from which we still, until this very day, draw inspiration and influence. Precise data on income and wealth distributions during ancient times is very difficult to find. We can only rely on archaeological findings and historiographic material to try and understand the patterns of economic and political development. From existing evidence, it is safe to conclude that the rise of early empires merely exacerbated the already existing patterns in the distribution of wealth. Inequality existed before the state, motivated by ownership over land and livestock. Adding a clear hierarchical power structure enabled it to rise even higher. In ancient times the majority of the population lived barely above subsistence income, while a small elite enjoyed significantly greater living standards. This is exemplified by archaeological excavations of dwellings in ancient societies (like for example in Pompei or ancient Egypt), which showed that over 90% of the population lived in small, cramped houses, whereas a select few lived in lavish mansions and large estates. Further evidence from burial sites confirms that those living much better than the rest were members of ancient elite networks—​people in power and close to power.22 Incentives for organizing into elite networks to wield control over the distribution of wealth are as old as society itself. No matter which empire we observe they all express the same pattern: proximity to political power generated wealth. Premodern states generated unprecedented opportunities for the accumulation and contraction of material resources in the hands of the few, both by providing a measure of protection for commercial activity and by opening up new sources of personal gain for those most closely associated with the exercise of political power.

—​Walter Scheidel (2017) The Great Leveller, p. 43.

In other words, the concentration of wealth in ancient states was in the hands of those holding power and their loyal elite networks. Given that this elite—​the “original 1%” comprising the military, the clerics, and the rulers—​controlled the allocation of resources in ancient societies they were effectively turning political inequality into economic inequality.23 Political power was the sole determinant of wealth in ancient times, a pattern that continued well until the 20th century.

Evolution of Elite Networks and Inequality  83 To understand why political power rendered such a strong incentive for wealth concentration, we must contextualize the economic model of preindustrial times. The economies in ancient times (and in general during preindustrial times) were all bounded by the Malthusian trap. In preindustrial times both inequality and wages were driven by exogenous events like wars, epidemics, or new discoveries that would primarily affect population growth. Mean incomes stagnated relatively over time so only changes in population growth would affect inequality. Lower inequality and temporarily higher wages led to higher population growth among the poor, which exerted negative pressures on their wages, increased inequality, and subsequently increased mortality among the poor, driving population growth down. Societies did grow slowly over time, and innovations in technology certainly did occur over the centuries, as did changes in social norms and cultures, but mean incomes were stagnant and living standards were tied to mere subsistence. Any innovation in technology that would increase incomes and make life easier would trigger population growth and the negative cycle would be reinforced: greater population among the poor would increase inequality, decrease wages, increase mortality, and again lower population growth.24 Within such a constrained economic model, coupled with a stationary bandit political model, neither of which allowed the voluntary market exchange principle to fully evolve, the best way of enrichment and achieving life beyond mere subsistence for an individual was to become part of the elite network: to exploit the violence power principle. However, becoming part of an elite network was not something made by choice. It was a privilege attained at birth, by being tied to the right family with connections to power. Proximity to power implied several potential methods of enrichment: controlling surplus production, transportation links, trade flows (by being granted permission from the state) or participation in conquest. Authoritarian empires had a particular impetus for conquest. Conquest meant further expansion and control over new territories and new riches; from land to capital to slave labor. In the case of each conquest the ruler and those in proximity to the ruler benefited the most: attaining new lands, amassing new wealth, and seizing control over the newly acquired territories as a reward for participating in battle. It was common for the conquerors to award control over new territories to their closest generals and other elite members. Alexander the Great gave each new territory he conquered to one of his generals. The expansion of the Roman Empire was also contingent on rewarding war veterans with land once they finished their service, and they were also allowed to loot and pillage all newly conquered territories. The higher the ranking within the army the greater the potential reward. Being rewarded for engaging in conquest remained a powerful motive for military conscription throughout ancient history and beyond.

84  Elite Networks The Arabs and the spread of Islam since the 8th century, the Viking conquests in the 9th and 10th centuries, the Crusades in the 12th century, the Mongol expansion in the 13th century, the conquistadores in the New World in the 16th century, the Ottoman expansion since the 15th century, etc.—​in each case the motivation was the promise of enrichment and a better life (even if it were in the afterlife, as it was promised to or believed by the Arabs, the Crusaders, and the Vikings). Enlisting in battle was a good example of how engaging in something that carried a high risk of getting killed was still a better choice than living a life barely above subsistence in ancient and preindustrial societies. The incentives for engaging in violence drawn from power offered a much better chance of attaining a better life than by engaging in voluntary exchange (this argument is further developed in Chapter 5). The decision to engage in conflict, just like the decision to accumulate wealth, was in each case confined on the rulers and their close elite networks. In order to benefit from conflict or from controlling production surpluses it was really necessary to be part of an elite network. Ancient China, for example, the first state with a multilevel administrative bureaucracy had a highly hierarchical social order with the concentration of power entirely in the hands of the emperor and loyal highly ranked imperial officers. Between them they exchanged favors and were particularly prone to corruption and violence, deciding who gets control over land, labor, and means of production (capital). The high-​ranking elite network members enjoyed protection from the emperor in that they were safe from any prosecutions, could freely exploit anyone around them, and had virtually no constraints on their rent-​seeking activities. As a consequence, nobles and high-​ranking officials established powerful coalitions to support each other and often intermarried to limit wealth dispersion and ensure that civil service jobs in the Empire were almost exclusively hereditary.25 Powerful families ruled China for centuries. The extreme lack of social mobility meant that there was very little hope for anyone born outside the upper classes to ever achieve an above-​ subsistence living standard. It was a typical stationary bandit political model that became embedded in the very core of Chinese politics, a trait that is visible even today. Clark’s research on social mobility based on rare elite surnames shows a remarkable level of persistence of elites in the upper income and wealth brackets that tends to last for several centuries. His research in China shows that despite the Communist revolution in the 20th century that had the explicit goal to rid the country of its century-​ long class enemies (killing or impoverishing tens of thousands of landlords, businessmen, and powerful families), Chinese elites are still overrepresented in the top income groups in China today. Even within the highest ranks of the Communist Party the prerevolution elites are counted among its most powerful and richest members.26 To borrow an expression from economic theory, elite persistence tends to be very “sticky” over a long period of time.

Evolution of Elite Networks and Inequality  85 Rome experienced an elite network hierarchy very similar to that of ancient China. As Rome expanded its power, first as a Republic and then as an Empire, its rise was supported by an aristocracy in the Senate who made all the decisions on military conquests, on the allocation of senior officials in public office, including appointing the prefects of the Roman provinces, and widely encouraged rent-​extraction among the elites. Personal wealth accumulation among the elite networks in Rome was entirely based on proximity to the Senate and later to the emperor. It was a heavily stratified society where the top 1% controlled the majority of all wealth above subsistence.27 Rome, however, was politically volatile. Periods of great distress in the final decades of the Republic and periods of great uncertainty under some of its most ludicrous emperors implied vast instability for the elite networks. Internal conflicts often changed the fortunes of aristocratic families, as did the violent redistributions during many civil wars and fights for power. This most likely had a negative effect on the forces of wealth concentration and disabled them from pushing wealth inequality to astronomically high levels in Roman times. Stability is the best impetus for wealth concentration in the hands of the elites. If these elites keep changing during periods of violent upswings it is unlikely that inequality reached permanently high plateaus during Roman times, despite the extractive elements of its political stationary bandit model. It was most likely cyclical, as it was during the entire preindustrial age. The downfall of the Empire in 476 obviously further destabilized the accumulation of wealth among the existing elite networks (except for those in the Eastern Empire, Byzantium), and drove inequality down in yet another cyclical swing. The forces of occasional destruction were quite frequent in Rome, which explains why despite obvious political forces that allowed heavy wealth concentration, the Gini estimates based on social tables from Roman times were not as high as expected (around 0.4 in income inequality28). In ancient times, the evolution of incentives for coercion and wealth accumulation within the constraints of the Malthusian economic model suggests that the forces of wealth concentration were already very strong and heavily contingent on the position of power. The stationary bandit political model coupled with a Malthusian economic model ensured the persistence of wealth concentration among the elites. As societies moved on to feudalism, mercantilism and colonialism, the forces of wealth concentration only grew stronger.

2.5.  The Long Malthusian Trap: Inequality and Elite Concentration during the Middle Ages until the Industrial Revolution The underlying mechanism behind development patterns of societies after ancient times during the Middle Ages and all the way up until the Industrial

86  Elite Networks Revolution was not much different from what had happened since the emergence of first empires. After the fall of the Roman Empire medieval kingdoms and the allegiance they had to the Catholic Church in Europe expanded the extractive stationary bandit model even further. The nobility in Europe formed the feudalistic model of extraction, which was little different from the patterns of development of the first proto-​states: a feudal lord owned the land on which peasants were allowed to live barely on subsistence income, having a social status not too different from slaves. This system, however, was more decentralized than in times of ancient empires, but instead of one stationary bandit ruling over a vast territory there were multitudes of local stationary bandits all of which formed nobility within a single state; a nobility which had to pledge loyalty to the king of the land, to the Church, or both. The Catholic Church of the Middle Ages was perhaps the best example of a stationary bandit model. It featured a strong dominance hierarchy with a well-​oiled quasi-​bureaucracy supported by a diverse network of priests across the Christian world, who during the Middle Ages were often the only literate people in a community. Coupled with their monopoly over interpreting religious doctrines, this gave them significant localized power, similar to the power held by feudal lords. Furthermore, the Church had the best possible religious justification of the pope’s power, and it could rely on a well-​defined elite network in terms of kings and nobles loyal to the Church. The pope had the exclusive power of giving legitimacy to each and every king in the Christian world. It was very easy for the Church to form alliances this way and become the crucial source of power during feudalism. Most importantly, the Church carried a strong desire for conquest and territorial expansion. Their goal was simple—​spread Christianity as far as possible and maintain a strong monopoly over its interpretation and its direct link to God. The Crusades were a typical example of the power of the Catholic Church during those times. A massive military conquest which was justified by religious ideology of returning the city of Jerusalem into the hands of the Christians and away from Muslim “infidels,” and which offered rewards in terms of eternal redemption for anyone who would engage in conquest. Behind it all was the desire of the Church to control the trade links between the East and West and establish territorial and ideological dominance in the world.29 The Church relied on a powerful elite network comprising the kings of England, France, the Holy Roman Empire, Italian city-​states, and all territories under its dominion. Once the Church realized its immense power through the ability to raise a massive army on several occasions during the 12th and 13th centuries (regardless of the fact that many Crusades were unsuccessful and often complete fiascos), it established this practice throughout subsequent centuries and the papacy became one of the most powerful political positions in the world.30

Evolution of Elite Networks and Inequality  87 The Church and its elite network of kings and nobles owned the majority of wealth during the Middle Ages and beyond in Europe. A feudalistic economic extraction model ensured that wealth concentration among elite networks was extremely high until the 19th century, reaching up to 0.9 in certain cities and periods. However, as Figure 2.1 suggests, wealth inequality in Europe during those times, over a period of 500 years, was gradually rising but with cyclical swings, within a range between 0.5 and 0.9. The reason behind the cyclical nature of inequality throughout the Middle Ages and even after the discoveries of the New World and the Enlightenment movement, was in the realization of random events that occasionally destroyed wealth and disrupted existing elite networks. The stationary bandit political economic model of Malthusian times suggests that the forces of wealth concentration should have kept the concentration of wealth in the hands of the few a persistent phenomenon over time. The trend in Figure 2.1 does hint that inequality had the tendency to gradually rise over time. The forces of wealth concentration work best when uninterrupted by periods of instability. But Europe during the Middle Ages and beyond was anything but stable, with persistent episodes of conflict, state failures, and pandemics that caused certain dents to the gradual increase of inequality. Such was the strength of the forces of wealth concentration that these random events had to have been of very significant magnitude to cause a dent in the trend. The impact of the 14th-​century Black Death is a good example, visible through its immediate and Average wealth inequality in Europe, 1300–1800 0.9 0.85 0.8 Gini coefficient

0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 1300

1350

1400

1450

1500

1550

1600

1650

1700

1750

Figure 2.1  Average wealth inequality in Europe from 1300 to 1800. Included are estimates of wealth Gini coefficients for Italian city-​states, and selected cities in England, France, Germany, Netherlands, and Denmark. Sources: Alfani and Ammannati (2017), Sussman (2006), and Scheidel (2017).

1800

88  Elite Networks sudden impact on wealth inequality, shown in Figure 2.1 for Europe on average and in Figure 2.2 for selected Italian city-​states. After this one-​off random event wealth inequality rebounded and continued to increase, absent any other major structural shocks. A similar impact on inequality in later periods was attributed to war, state failures, or any other type of pandemic. In the case of Augsburg in Figure 2.2, for example, the decline of wealth inequality in the 17th century was due to the impact of the Thirty Years’ War, which resulted in killing 20% of the German population. Another very powerful leveling force. The forces of occasional destruction led to massive leveling of inequality primarily because they adversely affected the rulers and their elite networks. The greater the destruction of wealth or human lives, the greater the negative impact on inequality. Once a society recovers from the forces of occasional destruction, its forces of wealth concentration continue to provide the same incentives for extractive

Wealth inequality in selected European cities, 1300–1800 1 0.9

Gini coefficient

0.8 0.7 0.6 0.5 0.4 0.3 1300

1350

1400

1450

1500

1550

1600

Piedmont

Poggibonsi

Augsburg

Northen Netherlands

Lucca

1650

1700

1750

1800

Tuscany Southern Netherlands

Figure 2.2  Wealth inequality in selected European cities: Italian city states (Piedmont, Poggibonsi, Lucca, Tuscany), Augsburg in Germany, and the northern and southern parts of The Netherlands. Source: Scheidel (2017), c­ hapters 10 and 11.

Evolution of Elite Networks and Inequality  89 elites. It happened throughout the preindustrial world, regardless of cornerstone events such as the New Discoveries or the Enlightenment. This is once again best exemplified by Clark’s research on social mobility based on rare elite surnames. The descendants of the nobility that was created throughout the Middle Ages are still in the upper tail of income and wealth distributions in many societies today. Clark defines a law of social mobility that predicts a family’s regression to the mean gradually over time in terms of its fortune and relative status. The logic here is simple: new generations are not necessarily as competent and successful as the previous ones, suggesting that their share of good luck runs out, and the family gradually loses its wealth and its importance. However, for most elite families this regression tends to be very slow and can last up to five hundred years. For some families the regression never happens, meaning that certain medieval elites have successfully preserved their relative hierarchical position within a society over several centuries. This remarkable finding is persistent across a wide range of societies that Clark and his associates look at, societies with very different historical patterns of development. Their argument is first confirmed in Europe in Sweden and England, extended surprisingly to the United States, where family dynasties were much younger than in the Old World, and further verified in Asia on the cases of China, Japan, Korea, and India, but also in South America in the case of Chile. Most interestingly their findings were confirmed among specific minority religious or ethnic groups like the Jews, Copts, Roma, Protestants, or Muslims in selected societies who all preserved their status through intermarriage (similarly to the castes in India), meaning that they either sealed their status at the top of the income distribution or at its bottom.31 The origins of contemporary inequality and lack of social mobility in many nations can therefore directly be traced to Malthusian times of stationary bandit extraction-​based politics. This is the legacy of elite networks on the distribution of income and wealth, a legacy that spawns across several centuries. The same patterns existed beyond Europe. In Asia, religion and culture were different but the logic of elite rule was almost identical. In the Arab world and Persia, the same type of authoritarian stationary bandit model ensured that wealth was heavily concentrated in the hands of the caliph/​sultan, their dynasties, and the nobility that supported their rule. The Chinese imperial rule and its corresponding bureaucratic elites were hardly any different. The Mongol conquests of the 13th century that were characterized by brutal violence and resulted in the second-​largest empire in history (topped only by the British Empire in the 19th century) were a perfect example of the impetus for conquest resting on the violence power principle. Their rule was also heavily contingent on an elite network formed by the heirs of Genghis Khan to later rule the empire. It was no different from ancient China or ancient Rome. The cultures of the Mongol empire, the

90  Elite Networks Arab world, or China were very different from European medieval structures, and were in many instances more advanced and carried greater ability to innovate and design new technologies. The Arab world and China were the centers of innovation and places where lavish wealth could have been found for centuries.32 However, their economic and political models were still molded by identical incentives used by the medieval elites in Europe: the incentive to coerce using power allowing the forces of wealth concentration to take full effect. It would be wrong to conclude that societies did not change during Malthusian times. They almost certainly did: they evolved, innovated, adapted new technologies, built lavish infrastructural objects, and grew in size, but progress was slow, daunting, and entrapped. Preindustrial societies before and after the Great Discoveries were all bounded by the Malthusian economic model, enabled by the extractive stationary bandit political model, allowing the forces of wealth concentration to develop virtually unhinged, halted only by random forces of occasional destruction, and guided by the violence power principle. Did these incentives change at all with the Industrial Revolution? The world, as we know it, changed completely. Wealth creation for the first time in history became inclusive, which gradually led to forces of democratization and an unprecedented rise in living standards and technological innovation. The Malthusian trap was broken. But inequality persisted. Elite networks persisted.

2.6.  Wealth Concentration and Elite Power in the 19th and 20th Centuries: From a “Society of Rentiers” to a “Society of Managers” The early years of the Industrial Revolution followed a familiar pattern of elite enrichment at the expense of the rest of society. The transition from agriculture to industry and the breakthrough from the Malthusian trap should have enabled inclusive wealth-​seeking opportunities and increased social mobility for everyone. This did happen, but it developed gradually over a very long period of time and was not entirely endogenous. It took two of the most vicious wars in human history, followed by a battle of ideas between socialism and capitalism, to start bringing down inequality and increasing opportunity through a more inclusive democratic capitalist system. A bourgeois middle class started emerging and exerting pressures for reforms and gradual democratization (exemplified by the Reform Acts in England). But all this took time. The initial benefactor from the early rise of capitalism in the mid-​19th century was neither the poor disenfranchised majority nor the emerging middle classes; the initial benefactors were owners of land and capital, the preexisting elite networks.

Evolution of Elite Networks and Inequality  91 In his book Capital in the 21st Century Thomas Piketty delivers an accurate portrayal of inequality trends during and after the Industrial Revolution. Before the 20th century inequality in Europe was driven by income from capital and land, extracted through rents. The vast majority of capital and land were almost exclusively in the hands of the elites. The kings and the highest ranks of the nobility were the top 1% of earners, and they owned between 50% and 60% of all wealth in a country. The other 9% of top earners, consisting of landed gentry (lesser ranks of the nobility) and clergy, owned between 30% and 40% of all wealth, meaning that the top decile of earners in a typical early industrial society owned over 90% of all wealth (see Figure 2.3). Such a distribution of wealth was a common characteristic of preindustrial and early industrial societies.33 The logic of elite networks did not change with the onset of the Industrial Revolution. Only its structure changed. In the 18th and early 19th century the best way to attain above-​subsistence wealth was still through ownership of land or capital, a privilege held only by members of elite networks. Wealth was preserved through inheritance and between-​nobility intermarriage. The vast majority of the population owned almost no wealth at all (90% of earners owned less than 10% of all wealth in the 19th century), so their opportunities in life were not too different from ancient times. The existence of basically two classes in preindustrial times translated into the struggle between the owners

Share of total wealth among top 10% of wealth holders, 1810–2010

100%

% of total wealth

90% 80% 70% 60% 50% 40%

1810

1830

1850

1870

England

1890 France

1910

1930 Sweden

1950

1970

1990

2010

United States

Figure 2.3  Share of total wealth among top 10% of wealth holders from 1810 to 2010 for England, France, Sweden, and the United States. Source: Piketty (2014).

92  Elite Networks of capital and the working classes as societies made the transition from agriculture to manufacturing. The rise of the merchant middle class had not yet made its impact. What were feudal peasants in the earlier era now became the low-​ paid expropriated workers in the 19th century. The early Industrial Revolution era became the origin for the battle of ideas between capitalism and socialism, exemplified not only through the works of Karl Marx and the rise of communist ideology but also in culture through novels of Jane Austin, Charles Dickens, Honoré de Balzac, Émile Zola, etc. Capitalism—​attaining wealth based on capital ownership—​became the dominant economic model of postindustrial times, while socialism—​deriving attractiveness to the masses from its emphasis on worker oppression—​became its main alternative, to be fully achieved only in the aftermath of the 20th-​century wars. Internal pressures for democratization (in the form of rising middle-​class power34 and interelite competition35) made gradual and irreversible changes over time, but they never usurped the logic of elite networks; proximity to power was as important as ever. Ferguson testifies to this high concentration of global wealth and power within the select few European empires,36 all of them under the control of even fewer powerful families (the two most prominent being the Rothschilds on the corporate/​financial end and the Saxe-​Coburg-​Gotha noble family on the imperial/​dynastic end):37 National economies fell increasingly under the sway of large industrial corporations, whose owners and managers, along with the bankers who financed them, began to constitute a new social and political elite, albeit one intimately connected to the old regime. The map of the world by 1900 was an imperial jigsaw, with eleven Western empires controlling disproportionate shares (58 per cent, in all) of earth’s territory, not to mention its population (57 per cent) and economic output (74 per cent).

—​Niall Ferguson (2017) The Square and the Tower, p. 151.

The new capitalist economic model eventually did create new opportunities for disruptive innovation and capital-​intensive production in a number of industries. This was by far the most important long-​run impact of the Industrial Revolution and is exemplified by the rise of the middle classes. Gradually, new elites emerged as well by exploiting the new vast opportunities, but attaining significant wealth was still highly contingent on political power. A person who lacked political backing could not build a business empire and become part of an elite network. They could pursue opportunities on their own and achieve decent living standards, but they could not become rich without being close to power. Gilded Age robber barons hardly could have achieved their dominance without the backing of the government, nor could any large corporation in England,

Evolution of Elite Networks and Inequality  93 France, Spain, Austria-​Hungary, or Prussia. For the first time in history, it was becoming possible for a nonnoble person to achieve above-​subsistence wealth through the realization of the free exchange principle, but success was still contingent on the violence power principle. The consequence of the Industrial Revolution was to enable inclusive wealth creation to all, however it also set the stage for the rise of another powerful elite network: between government and business. The internal composition of elite networks changed. Kings and nobles were slowly being replaced by politicians and owners of big corporations. The Tammany Hall political machine mentioned in the Introduction was a typical example of 19th-​century collusion between politics and business. The levers of power had shifted. The kings and queens were being overthrown by revolutions (violent or peaceful) and replaced by modern governments, while the nobles who owned land and capital were being replaced by powerful industrialists carrying monopoly licenses granted by the state, or better yet, by friendly politicians, thus constituting a new elite network. The structure has changed, but the logic remained.38 It is thus no wonder that capital continued to dominate wealth concentration of pre-​20th-​century elites. Even after capitalism developed into a much more inclusive economic model in the 20th century—​a consequence of continued internal pressures, the rise of the middle classes, and greater demand for public goods and redistribution—​the logic behind elite wealth concentration remained unchanged. One thing did change: income from capital was no longer the main source of wealth. In the 20th century income from labor took over. What caused this dramatic switch from capital to labor? Figures 2.3 and 2.4 paint a very clear image of the development of elite wealth in the 19th and 20th centuries. The top 10% of earners owned 90% of wealth in the Old World (selected European countries: France, England, and Sweden), and 60% in the New World (United States), although the share of top decile wealth in the United States rose to over 80% by the beginning of the 20th century. The numbers for the top 1% are even more staggering: the very pinnacle of elite networks, the wealthiest nobility and capital owners, by the end of the 20th century owned almost 70% of all wealth in England, up to 60% in France and Sweden, and around 40% in the United States. Then came Scheidel’s “four horsemen”:39 the two World Wars, a post–​World War I influenza pandemic (which killed even more people than the War itself), a major economic crisis and hyperinflation, dissolutions of former empires, and communist revolutions—​all four major forces of occasional destruction acted together within a relatively very short time span of only 30 years. The composition of wealth owned by those at the top changed dramatically, as did the very structure of elite networks. The outcome was a more equal society by the end of the 20th century than it was after the onset of the Industrial Revolution. The top 1% earners in Europe

94  Elite Networks Share of total wealth among top 1% of wealth holders, 1810–2010 80% 70%

% of total wealth

60% 50% 40% 30% 20% 10% 0%

1810

1830

1850

1870

England

1890 France

1910

1930

Sweden

1950

1970

1990

2010

United States

Figure 2.4  Share of total wealth among top 1% of wealth holders from 1810 to 2010 for England, France, Sweden, and the United States. Source: Piketty (2014).

owned about 20% of all wealth (in America it went up to 30%), while the top decile owned around 60% of all wealth in France and Sweden and almost 70% in the United States and United Kingdom.40 Despite having lower numbers than in the 19th century, for many people this is still too high, particularly when considering the rising trend over the last few decades of the 20th and the first decade of the 21st century. A period of relative stability once again allowed the forces of wealth concentration to generate higher levels of inequality. The relationship between capital and labor changed, and this reflected the patterns of selection in elite networks. No longer were elite networks composed of owners of capital and land, living off rents. In the 20th century they started to include those whose incomes depended on wages rather than rents. Societies made a switch from what Piketty calls a “society of rentiers” to a “society of managers.”41 To exemplify this shift Piketty uses Balzac’s famous 19th-​century novel Père Goriot, where one of its shady characters, Vautrin, offers his life lesson to the main protagonist. Vautrin’s lesson is that an average person’s best chance of achieving above-​subsistence living standards in the 19th century was to inherit wealth (or marry someone who is an inheritor or rentier). Income from labor was simply not rewarding enough. Figure 2.5 shows this logic: if a person wanted to be rich,

Evolution of Elite Networks and Inequality  95 inheritance (or marrying a rich widow) was much more likely to generate wealth than income earned from labor. An inheritor in the top 1% of income earners (living off rents) had about 2.5 times more income than a top 1% labor earner. Ownership of land or capital, attained usually through the violence power principle (as either a benefactor of political power or an inheritor of wealth that was once created through coercion) was more lucrative than any job a person might have held at that time.42 As Figure 2.5 further suggests, this relationship started to change already by the end of the 19th century and was completely reversed in the first part of the 20th century. For the first time in history labor-​based income became the main source of wealth. This happened primarily because the role of capital-​based income experienced a massive decline. The reason for this was most likely the aforementioned gradual rise of the middle classes followed by the impacts of the world wars. Families from lower income brackets slowly started to accumulate some wealth of their own, thus reducing the ratio of inheritance of the top 1% to the bottom 50%. Another interesting observation from Figures 2.3 to 2.5 is that the French Revolution of the late 18th century did not really affect the patterns of inequality or wealth concentration in France. The king was overthrown, as were many of his loyalists, followed by several years of brutal internal power struggles. However,

Inheritance vs. Labor, 1790–2010 Ratio between top 1% and bottom 50%

30 25 20 15 10 5 0

1790

1810

1830

1850

1870

1890

Top 1% inheritance earners

1910

1930

1950

1970

1990

2010

Top 1% labor earners

Figure 2.5  The relationship between income from rents (in this case in the form of inherited capital) and income from wages in France. The graph compares ratios between top 1% and bottom 50% of income earners from both inheritance (rents) and labor (wages) from the late 18th to the 21st century. Source: Piketty (2014).

96  Elite Networks elite networks obviously persisted, as their wealth was unharmed. Even the age of Napoleon did not have any adverse effect on wealth inequality in France—​on the contrary, the top deciles expanded their wealth concentration during and after those times. The largest impact on inequality in Europe (and France by extension) was still made by the 20th century’s four brutal forces of destruction: two major wars, a pandemic, revolutions, and dissolutions of empires. The most important conclusion from Figures 2.3 to 2.5 is that societies in the 20th century made the switch from rentiers to managers, from capital-​based income to labor-​based income. Today’s elite networks are different from the pre-​ 20th-​century elites in that their source of wealth is no longer exclusively tied to capital or land ownership but is instead driven by their relatively higher wages. Today’s top 1%, or more precisely the top 0.1% of earners—​the pool from which new elite networks are created—​comprises senior corporate executives (CEOs, board members and presidents) of publicly traded firms in the real sector, finance executives and professionals (from banking to hedge funds), top corporate lawyers and consultants, disruptive entrepreneurs (like Musk, Bezos, Gates, Zuckerberg, Jobs, etc.), various superstars (from the performing arts, media, or sports), and the occasional real estate mogul.43 Most individuals in this category draw their incomes primarily from labor, even though they obviously also benefit from exposure to financial markets through shares in their expanding empires and other investments. Structurally the elites are different from before; they are no longer made up of rentiers (inheritors and rentiers are still present but are far from being the majority group), but their outcomes are still in many cases contingent on their proximity to political power (as will be shown empirically in Chapter 4). The postindustrial era thus saw a complete change in the composition of elite networks. First, the nobles and the aristocracy were replaced by industrialists running huge corporate empires. They were the new aristocracy, still drawing their wealth from capital ownership (and to a much lesser extent from land ownership). A crucial transformation here is that for the first time in history elite membership was no longer exclusively hereditary. The new owners of capital, 19th-​century industrialists, became wealthy because they took advantage of existing opportunities. It was the emergence of the voluntary exchange principle, although with an important caveat—​building a monopoly in steel, oil, or railroads was heavily contingent on having access to political power. The second big transformation was from the industrialists to the managers, triggered by the shift from capital-​based income to labor-​based income as exemplified in Figures 2.3 to 2.5. Competence, talent, and hard work became more important than ever, hereditary status became relatively marginal, but proximity to power was still immensely important. Both of these transformations dramatically affected the internal structure of elite networks—​kings and nobles were replaced by elected

Evolution of Elite Networks and Inequality  97 politicians and corporate executives—​ but their internal dynamics did not change: the emergence of the free exchange principle did not annul the violence power principle as the key driving mechanism. We see this in the consequences of rising income inequality in the late 20th century, when the forces of wealth concentration during periods of prolonged stability once again exacerbated the logic of elite networks. The rise of income inequality since the 1970s, mostly in Anglo-​Saxon countries but also across the Western world, was primarily a consequence of the rise in wages and nonwage incomes of the top 1% and top 0.1% of income earners. Salaries, bonuses, and capital gains from investments in the financial markets were responsible for a large part of this increase in top incomes.44 Many economists describe this rise of salaries as abnormal because their growth was larger than the productivity gains of the companies. The increase of top managerial salaries was a consequence of many factors, but mainly it was due to an increasing demand for top talent (the so-​called superstar effect) which significantly affected their position in the social hierarchy. Their talents led them to become part of the new elite network. The pre-​20th-​century elites built their wealth on coercion and ownership of land, capital, and often people. The 20th-​and arguably 21st-​century elites built their wealth based on talent, but also, as we shall see in many cases, proximity to political power. Nevertheless, talent and competence are not the most important determinants of the new elite’s incomes. Something else is playing a crucial role, as it did throughout human history: the network effect. Top corporate executives in publicly listed firms, despite having quite dictatorial tendencies, do face some constraints on their power. Many CEOs successfully avert this by playing office politics and appointing a Board of loyal members who will never question their authority. Occasionally, due to pressure from shareholders, a CEO does get punished if he or she does not deliver; if revenues are stagnating and/​or share prices are going down. The best way of proving their worth to the company and satisfying their shareholders is by maintaining connections to the key decision-​ makers in power. Corporate executives tend to exhibit a high level of degree centrality within their networks. Degree centrality implies that they position themselves in a key space within a network where they can maximize their connections to the widest number of important people (executives of other companies and politicians). Their connections directly enable them to secure exclusive contracts and big jobs that benefit their companies. Hence, they signal their value to the shareholders and are thus able to justify their high salaries and bonuses. In other words, they draw their higher earnings by being well connected, among others, to politics. It is the same logic of elite networks that existed throughout human history. The only exception is that today the connections within elite networks are not

98  Elite Networks rewarded directly through spoils from plunder or control over land, but indirectly through exclusive contracts from public tenders that deliver abnormal compensations to the key nodes in the network. The cycle of history continues. Elite networks once again draw their influence from political power. As it was in ancient history and throughout preindustrial times, the dominant position in the social hierarchy is still somewhat dependent on proximity to political power. It is hardly the only determinant of wealth. Talent, competence, and a host of other factors are still important—​which is the crucial difference between today and preindustrial times—​but proximity to power still remains a very important determinant of wealth for someone in the very top of the income distribution, someone who is a member of an elite network. The role of politics has also changed in terms of the tools the rulers have at their disposal. No longer are leaders in democracies allowed to use coercion and violence (at least when it comes to domestic affairs, where constraints do exist), but they still hold enough power in distributing huge government budgets. In fact, the size of modern-​day budgets gives the politicians in office immense power in the distribution of wealth. Anyone who depends on the allocation of government budgets, on procurement contracts with the government, or on regulatory or legislative decisions granting a privileged position on the market, or anyone seeking to promote a certain policy will have an interest in being connected to and influencing the decision-​makers in power—​through lobbying, campaign spending, funding various pressure groups, or directly affecting allocation decisions. All such actions necessarily lead to a rise in top incomes, thus widening income inequality. The foundation of an elite network is to gradually build influence through informal gatherings and events where decision-​making politicians meet with industry executives. Political influence is rarely bought openly and directly (at least in institutionally well defined democracies). It is more often gained indirectly, where an executive enters into a long-​term mutually beneficial relationship with a politician. Important government allocation decisions are therefore rarely done through a predefined institutional mechanism. It takes time to forge relationships and gain trust among the power structures, but once trust is achieved, favors tend to be exchanged at a very fast pace. Most importantly—​ everyone benefits. *** Inequality is a persistent phenomenon of human history. It originated with the onset of agricultural production and has been sustained through the use of political power. It spreads across all economic and political systems. High levels of inequality characterized all agrarian societies in the past, where autocratic elites

Evolution of Elite Networks and Inequality  99 condensed all the wealth. The Malthusian economic model coupled with the stationary bandit extractive political model did not give many opportunities for wealth accumulation to anyone outside of the dominant elite network. Political inequality, or the inequality in the distribution of power, was the primary source of income and wealth inequality throughout preindustrial history. This artifact of societal development was, and still is, a key characteristic of many states and nations—​all, in fact, that have had or still have some type of authoritarian rule. High levels of inequality were found both in socialist countries of the late 20th century, where the elite is realized through partocratic or autocratic rule, and in modern democratic capitalist countries, where the elite exists as a combination of political and corporate power. The common denominator across different socioeconomic systems and different time periods is a powerful elite and its network—​be it a network of dictators and their cronies, a network of politicians and corporate bosses, or a network of kings and nobles. In the 20th century the levels of inequality fell to historic lows, particularly in the aftermath of the two world wars. The rapid rise of living standards and huge increases of productivity lifted hundreds of millions from relative poverty and introduced them to a living standard greater than that of the kings and queens of the earlier ages, or any elite network that existed thus far. Rapid economic growth in the 20th century, provided by the emergence of the voluntary exchange principle, sent shockwaves throughout civilization making each new generation much better off than the one before. There is no doubt that this encompassing economic progress increased social mobility and lowered the levels of income and wealth inequality in the 20th century, compared to the centuries before. The rise of capitalism and the subsequent enrichment of the middle and upper classes who demanded more democracy and franchise extension during the 19th and the 20th centuries had a gradual, century-​long impact on broadening the participation rates in wealth creation. As more people were allowed to engage in creating wealth and making a decent living from their labor input, an exploitational model of 18th-​century capitalism was gradually being replaced by a more inclusive model of capitalism (at least in the countries of the Western core—​as long as their colonies still provided enough resources that made the colonizers wealthy). However, the rise of capitalism that was followed by a gradual rise of democracy failed to curtail the structural causes of income and wealth inequality—​the forces of wealth concentration. A much bigger impact on leveling inequality in the 20th century was made by the two world wars, state failures, and revolutions, given that a huge amount of wealth had been destroyed and violently redistributed throughout Europe. It did not take long for these trends to be reversed in the final two decades of the 20th century, meaning that we are once again facing an issue of relatively high income inequality, or in other words a level

100  Elite Networks of inequality that the majority of society no longer sees as fair and justifiable. The next chapter looks at the three parallel trends that had happened in the 20th century: an emphatic switch to democracies in the final decades of the 20th century, a consequential expansion of the welfare state from the mid-​20th century, and a simultaneous increase of income inequality. It seeks to provide an answer as to why the relationship between democracy and inequality is neither linear nor negative, as it is often theoretically assumed. Then we will move on to the empirical evidence of elite concentration among top corporate executives in Chapter 4.

3

Democracy and Inequality in the Short Run We must make our choice. We may have democracy, or we may have wealth concentrated in the hands of a few, but we can’t have both. —​Louis D. Brandeis, United States Supreme Court Justice, 1916–​1939

3.1.  The Switch toward Democracy Up until the 20th century and throughout history, the majority of countries in the world were autocratic. As outlined in the previous chapter, the logic of elite rule was heavily contingent on the context of an extractive autocratic system. Today most countries in the world are democracies, albeit with vast differences between them. Some democracies presume a constitutional liberal order, while others descend into various forms of illiberal populism. Some enact strong institutions and rule of law, others allow their institutions to be captured by partial interests. Nevertheless, the trend even before the fall of the Soviet Union has been clear. Since the 1970s until today the ratio of democracies to nondemocracies has almost completely shifted. In 1970 over 70% of countries in the world were nondemocratic (a number that has been consistent since the mid-​ 19th century), whereas in 2020 almost 60% of all countries were democracies (see Figures 3.1 and 3.2). Furthermore, almost all rich countries in the world today are democracies (with the exception of a few oil exporters), while the vast majority of poor countries have some form of autocratic government. The legacy of the Cold War era is a resounding victory not only for capitalism over communism, but also for democracy over autocracy. The introduction of both capitalism and democracy has been the dominant political-​economic transition many of the former autocratic socialist countries have embraced. The transition from socialist autocracies to capitalist democracies has brought ills as well as gains. While the introduction of the market system has undoubtedly risen living standards, life expectancy, wages, and incomes; improved the quality of public goods; increased government efficiency and accountability; and improved human rights standards throughout (obviously to a varying extent depending on Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0004

102  Elite Networks 100 90

Number of countries

80 70 60 50 40 30 20 10 0

1950

1960

1970 Autocracies

1980

1990

Anocracies

2000

2010

2020

Democracies

Figure 3.1  Democracies, autocracies, and anocracies in the world, 1950–​2020. Source: Polity IV Project, 2018. Democracies and dictatorships aren’t binary categories. There are many shades of gray between the definitions of a fully consolidated democracy (such as the United States and western Europe) and a hereditary dictatorship (such as North Korea). The Polity IV database ranks countries from an index score of −​10 (hereditary autocracy) to +​10 (consolidated democracy). In the middle of its ranking scale they define an anocracy, a political system characterized by economic and political instability and inefficiency. In anocracies human rights are still endangered for various social groups (religious, ethnic, sexual minorities, etc.), there is a lack of freedom of the press, there is considerable violence and crime present (on average even more than in autocracies), but elections are allowed, as is the multiparty system, and everyone in the population freely participates in elections. Although the opposition exists it is often subdued by being forbidden from public speaking, it lacks media attention, and there are often signs of elections being rigged. The best examples of anocracies would be Russia, Turkey, Egypt, Venezuela, and Thailand in the recent decade, or Mexico before the 1990s.

the country at hand),1 critics have pointed to rising social injustice, inequality, poverty, and both economic and personal insecurity as obvious consequences of the transition process.2 Many have struggled with the introduction of the market system, the benefit of which has been captured by the well-​entrenched few (politicians and “the mafia”), rather than shared by the disenfranchised many.3 Transitional countries have seen eroding trust as a consequence of this process, where a significant amount of the population in each country has been dubbed the “losers of transition.”4

Democracy and Inequality in the Short Run  103

Figure 3.2  Global regime change from 1950 to 2022. In 1950 only 24 countries in the world were democracies, and three times as many were autocracies with a few anocracies and colonies. Africa, Latin America, Eastern Europe, and a huge part of Asia (the Soviet Union included) were all under some form of an autocratic regime. In 2022 the situation is radically different: the vast majority of countries are democracies, with only a dozen autocracies and two dozen closed or open anocracies left (mostly in Asia and parts of Africa). Source: Herre (2021).

104  Elite Networks On the other hand, a similar argument can be made in some Western countries in the past 40 years. Rising inequality, lack of social mobility, increasing sense of insecurity, depleting levels of trust, and an entire blue-​collar generation being proclaimed the “losers of globalization” are contemporary issues in Western democracies that almost mimic the concerns facing many transitional societies to date. Why have the ills usually associated with transitional economies been so easily transferred to the West, within countries which have strong institutional and constitutional rules that should have prevented precisely these types of outcomes? Why has the West failed to become fully robust to cronyism and interest group state capture? This chapter focuses on the two parallel trends that occurred in the West in the second part of the 20th century (hence dubbed “the short run”): the rise of government spending initiated by the expansion of the welfare state, and the simultaneous rise of income and wealth inequality. It examines why inequality failed to subside after one of the strongest periods of living standard expansion and productivity growth (from the 1920s to the 1970s in the United States and from the 1950s to 1970s–​80s in Europe and Japan) coupled with the rise of the welfare state.

3.2.  The Relationship between Democracy and Inequality The long-​run switch to democracy unfortunately had no lasting impact on reversing inequality trends. The rise of the welfare state, one of the most significant sociopolitical outcomes of the 20th century, coinciding with the rise of democracy and greater political representation, did not put a dent in the forces of inequality, at least not as a long-​term outcome. What we observe today are much higher living standards than ever before. Economic progress over the 20th century brought about incredible riches and possibilities for newer generations to engage in occupations that their ancestors could not even envision. This unprecedented economic progress was driven, for the most part, by historically high productivity growth from the 1920s to the 1970s (in the United States, and from the 1950s to 1980s in Western Europe and Japan), an era during which wages for the low and middle classes steadily increased. Economic progress was undoubtedly followed by a rise in democratic standards (more on this in Chapter 8). As shown in Figure 3.1, countries were rapidly switching to democracies, but their democracies evolved internally as well. Take the United States, for example. In the 1950s and 1960s, despite strong productivity growth and rising living standards, large parts of the population, like women and African Americans, were partially or fully excluded from this progress. The 1960s were a particularly turbulent decade in the United States with

Democracy and Inequality in the Short Run  105 mounting support for social change, which suggests that despite seeing lower inequality in the numbers5 a lot of people felt they were being excluded from progress. The civil rights movement, originating a full 100 years after President Lincoln abolished slavery, perfectly encapsulates the state of disenfranchisement of the African American community, not only in the South, where they were treated like third-​grade citizens and were used as cheap labor, but also across the country, where they struggled to move up the social ladder even with access to higher education. It was not until the mid-​1970s and throughout the 1980s and 1990s that the consequences of the civil rights movement and the women’s liberation movement bore fruit and allowed the previously disenfranchised groups to climb up the social ladder and yield high(er)-​paying positions in the corporate, academic, and public sectors.6 The emancipation movements were not without problems. The 1980s and ’90s experienced issues of sexual harassment at work, and had to feature affirmative action campaigns to increase workplace equality (the rap and hip-​hop culture of the 1990s implies that the black community was still experiencing difficulties related to poverty, violence, and crime). Even 40 years later, the highest ranks of the corporate and even academic milieu were still dominated by white men. Nevertheless, all of these changes were undoubtedly a significant improvement over the supposedly low inequality/​high social mobility era of the 1960s. American democracy therefore gradually improved over the decades of its largest economic expansion. Even though disenfranchised and minority groups gained more rights and were climbing up the social hierarchy, the expansion of rights did not reduce the inequality trends of the 1980s and 1990s. We must therefore ask ourselves, should there be any implicit relationship between democracy and inequality? In one of the most famous papers on political economy, Meltzer and Richard proposed a theoretical model in which more democracy implies more redistribution and hence lower inequality.7 In a typical societal income distribution curve, the median voter is usually to the left of the mean income voter, which is always slightly tilted to the right—​meaning there are more poor people than rich people in a society (see Figure 3.3). The larger the gap between the mean income voter and the median population voter (i.e., the poorer the median voter), under the assumptions of the Downsian model of electoral competition,8 more redistribution will be demanded. This unique theoretical proposition implies that more unequal countries should have or strive for larger governments, as a result of the voters voting for greater redistribution. This might have been true in the beginning of the 20th century, when most Western countries started introducing democracy by expanding the franchise, which saw a corresponding increase of government spending and revenues (see section 3.3). However, in the past 50 years both the

106  Elite Networks f(y)

Poor

m

µ

Rich

Income (y)

Figure 3.3  The distribution of income in society. The x-​axis represents income, while the y-​axis represents how many people there are for each distribution of income. Point m represents the median income person, the person that splits the population into two equal halves, while μ represents mean income. The distribution of income in society is never a normal distribution but is always right skewed given that there are more poor and middle-​income people than rich people in a society. Notice how in any society the mean income is always greater than median income. According to Meltzer and Richard (1981) the greater the distance between mean and median income, i.e., the greater the level of inequality in a country, the more likely that a country will demand more government spending.

size of government and the levels of income and wealth inequality significantly increased. Why didn’t the increase in government size, largely attributed to the expansion of the welfare state, lower inequality as the Downsian and Meltzer and Richard models suggest? Did the behavior of well-​organized special interests deplete the resources available for redistribution to the poorer ends of society? Did democracy fail in its attempt to lower inequality and provide the basic living standards for the poor? Or is favoritism of the few at the expense of the many an inevitable outcome of a democratic system? Looking at the distribution of inequality across countries, there is no consistent relationship between a democratic system of governance and the Gini coefficient (Figure 3.4). The poorest and the most unequal societies in the world today are predominately not democracies (or at least not complete democracies), meaning that their ruling elites do not really care about the position of the median voter. On the other hand, some of the countries with the lowest level of inequality have a strong democratic tradition. There are also so many exceptions suggesting the opposite: relatively unequal democracies (like the United States

Democracy and Inequality in the Short Run  107 Average Gini coefficient, 1980–2020

Average Polity democracy index, 1980–2020

0.2 10

0.25

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

8 6 4 2 0 –2 –4 –6 –8 –10

Figure 3.4  Democracy versus inequality. The Polity IV democracy ranking is from -​10 (autocracy) to 10 (democracy). Higher Gini implies higher inequality. The graph deploys averages of the democracy index and the Gini since 1980 to 2020. Source of data: Gini: World Bank, https://​data.worldb​ank.org/​indica​tor/​SI.POV. GINI. Democracy: Polity IV.

or the United Kingdom), and relatively equal autocracies or anocracies (like Singapore). Furthermore, the data does not support the theoretical predictions that inequality in democracies should be lower as the voters demand more redistribution. The models fail to explain the long-​term and short-​term dynamics of inequality. While it remains relatively easy to explain the persistence of inequality in institutionally weak societies, the question remains: Why is the median voter in the West still disenfranchised? Shouldn’t the persistence and consolidation of democracy imply a gradual decline of inequality over the past generation? The empirical literature on the relationship between democracy and inequality is vast but inconclusive.9 While earlier efforts were mostly cross-​sectional regressions linking inequality to democracy, some of them emphasized that in order for inequality to decrease in a democracy enough time has to pass for the democracy to consolidate.10 More contemporary studies used a panel dataset to test the same relationship, but have found varying impacts of democracy on inequality. While some authors confirm a positive effect of democracy on lowering inequality,11 others find no conclusive evidence that democracy lowers inequality.12 Most of the early literature tends to suffer from endogeneity problems: the omitted variable bias (a variable affecting both democracy and

108  Elite Networks inequality that could be driving the results) and the reverse causality problem (inequality leads to democracy, not vice versa). Furthermore, the literature is mostly focused around the initial impact of democratization on inequality, and how much time it took for inequality to decline in a well-​consolidated democratic environment, a direction of research inspired by Kuznets’s theory of inequality.13 Democratization arguably carried an additional effect on inequality, an effect driven by the forces of wealth concentration and realized through the rapid rise of interest group power. As interest groups in democracies become better organized, they become more successful at increasing the size of government, but bias that increase toward themselves. This leaves relatively less money for redistribution programs aimed at the poorer ends of society, particularly in terms of education and health care. Olson’s theory of collective action explains this phenomenon where small (privileged) groups possess enough information, have low enough organization costs, and are far more homogeneous in distributing the potential benefits to successfully solve the public good allocation problem.14 As the number of interest groups fighting for state redistribution increases, this becomes a wider burden for society, whose productive resources are being suboptimally allocated. The phenomenon is called interest group state capture, and is unfortunately characterizing more and more democratic societies today—​both rich and poor. The concentration of power among interest groups in democracies epitomizes the internal logic of elite networks: small, privileged, well-​organized, with access to key information, and most importantly, with proximity to power.

3.3.  The Rise of Government Spending in Democracies With the rising allure of democracy commencing in the 1970s, two other trends originated at approximately the same time in the already-​established democracies of the developed world. The first was a significant rise in size and scope of government (starting already in the mid-​1960s for most countries), while the second was an almost parallel rise in income inequality (starting in mid-​1970s and picking up pace after the 1980s). There is no implied causal relationship, of course; however, it is puzzling that higher government spending, driven primarily by an expanding welfare state, did not, within a democratic setting, gradually lead to lower inequality. In other words, the theoretical implication that democratization should over time accumulate lower income inequality has not been backed up by empirical evidence. Let us examine the empirics. Table 3.1 shows the size of government spending as share of GDP for 16 selected developed countries, within and outside of

Democracy and Inequality in the Short Run  109 Table 3.1  Size of government expenditure as share of GDP for selected countries, 1910–​2010. Source: IMF (2013) Dataset “Public Finances in Modern History” based on Mauro et al. (2013) Country

1910 1925 1935 1950 1960 1970 1980 1995 2005 2010

Austria

17.3

13.7

17.2

26.2

36.3

39.5

50.1

56.3

49.9

52.5

6.4

10.9

11.3

10.9

16.4

42.1

52.6

59.2

52.7

58.4

France

10.5

11.2

20.2

24.1

22.2

20.5

45.9

54.4

53.5

56.6

Germany

19.4

14.2

12.6

27.1

29.2

39.4

48.2

54.8

46.9

47.8

18.3

21.7

25.8

29.1

44.1

53.7

41.3

33.7

46.8

13.8

11.4

24.5

17.3

18.2

32.3

40.7

52.4

47.9

50.2

Netherlands 10.2

7.6

11.7

23.9

20.4

43.1

55.2

56.4

44.7

51.2

Denmark

Ireland Italy Spain

9.9

11.8

13.9

11.7

11.1

13.9

21.4

44.4

38.4

45.6

Sweden

8.5

9.5

12.1

18.2

24.4

29.5

41.1

64.9

53.8

52.8

Switzerland

4.1

4.2

6.2

8.5

7.1

25.7

29.9

35

35.2

34.1

United Kingdom

8.2

18.1

19.4

32.6

33.1

42.1

47.6

43.9

44.1

50.5

Average Europe

10.8

11.9

15.5

20.6

22.5

33.8

44.2

51.2

45.5

49.7

United States

2.2

3.2

10.1

14.7

28.2

32.3

34.2

37.1

36.3

42.4

Canada

6.1

7.1

12.4

14.9

30.6

36.1

41.5

48.5

39.1

43.8

Australia

2.6

4.3

5.4

12.2

21.8

25.3

33.2

36.8

33.9

37.1

New Zealand

13.5

17.7

17.9

21.8

25.3

24.8

38.1

33.4

29.7

35.1

Japan

3.1

2.1

3.5

15.8

18.3

20.2

33.4

34

34.2

39.8

Average Non-E ​ urope

5.5

6.9

9.9

15.9

24.8

27.7

36.1

38

34.6

39.6

Europe, with averages for each group.15 Government spending data includes only spending of the central government (excluding local government and the rest of the public sector), and in this case it also excludes interest payments on public debt. Figure 3.5 presents the full trend of government spending over the past century for the same group of countries, while Figure 3.6 presents the same long-​ term trend for central government revenues (also excluding local government and the rest of the public sector).

110  Elite Networks Government spending to GDP (%) 1900–2010 80 70 60 50 40 30 20

2005

2000

1995

1990

1985

1980

1975

1970

1965

1960

1955

1950

1945

1940

1935

1930

1925

1920

1915

1910

1905

0

1900

10

Australia Austria Canada Denmark France Germany Ireland Italy Japan Netherlands New Zealand Spain Sweden Switzerland United Kingdom United States

Figure 3.5  Central government spending to GDP for selected developed countries, 1900–​2010. Source: IMF (2013) Dataset “Public Finances in Modern History” based on Mauro et al. (2013). Government revenue to GDP (%) 1900–2010 70 60 50 40 30 20

2005

2000

1995

1990

1985

1980

1975

1970

1965

1960

1955

1950

1945

1940

1935

1930

1925

1920

1915

1910

1905

0

1900

10

Australia Austria Canada Denmark France Germany Ireland Italy Japan Netherlands New Zealand Spain Sweden Switzerland United Kingdom United States

Figure 3.6  Central government revenues to GDP for selected developed countries, 1900–​2010. Source: IMF (2013) Dataset “Public Finances in Modern History” based on Mauro et al. (2013).

The overall trends are clear. There has been a significant increase in the size of government in the past century among the world’s richest democracies, in most cases increasing rapidly after World War II, particularly in the 1960s and 1970s. This growth in the size of government has slowed down during

Democracy and Inequality in the Short Run  111 the 1980s, peaked in the mid-​1990s, and slightly regressed in the early 2000s, only to increase once more as a consequence of the 2008 financial crisis (fiscal stimuli and bailouts as an immediate response to the crisis). The growth in both government spending and revenues (Figure 3.6) exhibits the same pattern, except for the post-​2008 crisis years where the rise in government spending overtook that of revenues, leading to increasing budget deficits in many developed democracies, all of which acted as a trigger for austerity policies starting in 2010. Interestingly, European countries have, throughout the observed 110-​year period, had on average higher levels of government size than non-​ European countries. All of these trends are expected. The beginning of the 20th century was characterized in most part by a laissez-​faire approach to the role government in the economy, limiting its size and scope significantly, and limiting tax revenues to tariffs and duties. The interwar period introduced first social security systems, while the onset of the Great Depression gave further rise to the scope of government intervention in the economy, almost doubling it in size by 1937. In countries that descended into fascism (Germany, Italy) the national-​socialist model rested on an encompassing role of the state with an aim to allocate resources into the war machine. The size of government doubled in those countries in the interwar period. The post–​World War II period was characterized by the dominance of the Keynesian economic consensus. The welfare state started expanding rapidly, as did all other forms of government intervention in the economy. Tanzi and Schuknecht attribute this rise of government spending to the battle of ideas over the role of government in the economy.16 In economic theory the dominant paradigm was that of welfare economics,17 which revolved around a social welfare function and having the state as the perfect social planner that maximizes social welfare by efficiently allocating, redistributing, and stabilizing the economy.18 The role of the social planner was to prevent the re-​emergence of structural factors that contributed to the Great Depression and its subsequent outcomes, unemployment in particular. Hence the desire of Western policymakers to adopt traditionally socialist policies, and to promote the idea of a mixed economy transcending beyond pure socialism and pure laissez faire. Social security programs expanded the most during the postwar period as they gradually started to turn into modern welfare states.19 This was happening all across the Western developed democracies, transcending party lines. In the United States, Social Security was established as part of President Roosevelt’s New Deal in 1935, but it was the Republican government of Dwight Eisenhower that expanded Social Security and continued all the major New Deal programs from the Roosevelt administrations. Medicare and Medicaid were both introduced

112  Elite Networks later during Lyndon Johnson’s Great Society program. In the United Kingdom, the postwar Atlee government created the UK welfare state; they introduced national insurance; founded the National Health Service (NHS); introduced public housing, family allowances, and planning laws; improved worker’s rights; and even nationalized industries. Other European countries followed in the exact same way: Germany established the social market economy, very similar to the Nordic model of social democracy combining a market economy with a strong welfare state; France enacted its social protection and welfare state system immediately after the War, as did the Netherlands and many other countries throughout Europe (some sooner, some later), but also beyond Europe (such as Canada, Australia, New Zealand, and Japan). It was therefore no surprise to witness the significant expansion in the size of governments in the postwar period. The role of the state in the economy changed and modern welfare states were created. Based on skepticism toward government intervention in the economy, the 1980s delivered political change characterized by a reduction in the scope of government intervention and a slowdown in the growth of government size. The dominant Keynesian economic consensus was disrupted through three major developments, one of which was external: in the 1970s the West was faced with a major oil shock. This was a direct consequence of political instability in the Middle East driven by the Yom Kippur War, after which Arab oil producers decided to cut supply and hence increase oil prices as a political lever against the West. The second development was the shift in the intellectual battle of ideas. The monetarist tradition of Milton Friedman, the rise of public choice theory, and the recognition of Hayek’s work with a Nobel prize in 1974 convinced many that governments can fail too, and that they are not necessarily the benevolent social planner they were assumed to be.20 Hayek, Friedman, and the Chicago School economists provided intellectual ground for the policies enacted by the new Thatcher and Reagan governments in the United Kingdom and United States. The battle of ideas shifted against big government and the Keynesian paradigm. The third factor that made the switch was the 1970s stagflation, a phenomenon of simultaneously increasing inflation and unemployment, which was inexplicable by mainstream economic theory at the time. Given the success of the Thatcher and Reagan administrations in reigniting economic growth in their countries, the new economic mainstream consequentially inspired policy changes throughout Europe and beyond, with varying interpretations and consequentially with varying rates of success, particularly when it came to transitional postcommunist economies and the countries of Latin America. Market reforms and privatization projects did commence in the 1980s, but their effects were not fully apparent until the mid-​1990s, which is why it is not

Democracy and Inequality in the Short Run  113 surprising to see government size peaking in the mid-​1990s. Thatcher’s and Reagan’s governments’ legacy was given further momentum by the fall of communism, as even their political opponents (Democrats in the United States and New Labour in the United Kingdom) continued with promarket reforms reducing both the scope and size of government. This decrease in size was, however, only marginal overall (from an average of 50% down to 45%), bringing the size of government back to its 1980 levels by 2005. After the 2008 crisis, the size of government increased once again, mainly as a result of countercyclical fiscal policies implemented during the onset of the crisis. The United States and many European countries engaged in massive bank bailouts to prevent further panic on the markets and a potential new depression (or at least, that was the prevailing narrative), and have subsequently continued with large fiscal stimuli designed to initiate a stronger recovery. The politics of austerity did not start until 2010 in most countries, primarily as a response to a rapid increase of public debts and deficits, in most cases attributed to the initial responses to the crisis. However, even with austerity policies the size of government did not go down by much at all. Then in 2020 we received a stimulus package on steroids as a response to COVID-​induced lockdowns, bringing government size (and levels of public debt) to historic highs. The data on government size can only tell us so much. A more interesting story is to uncover the scope of government, or in other words, the composition of government spending. According to Tanzi and Schuknecht, during the post–​World War II period all OECD governments experienced a rapid increase in government spending on subsidies and transfers, mostly due to an expansion of the welfare state. On average total spending on subsidies and transfers (entitlements) went from a mere 4.5% before the war to 23.2% in 1995, an increase by a factor of five. Social expenditures on health, education, and pensions explain the majority of this increase (Figure 3.7). Education and health spending went from 2.1% and 0.4% before the war to 6.1% and 6.4% respectively in the mid-​1990s. Unemployment insurance and other income transfer programs remained mostly marginal parts of the budgets throughout. The category that expanded the most in this period were pension expenditures. They went up from 1.9% prewar to 9.6% in the mid-​1990s. The rapid expansion of both pensions and public health expenditures were a natural outcome of an aging population, but it was a policy decision of solidarity that made these programs possible and indeed achievable.21 Other spending categories mostly remained relatively stable throughout. Defense spending peaked during the Cold War, but never went past 4% and gradually declined to less than 2% over the past 20 years. Spending on public investment has mostly circulated between 3% and 4% during the entire period. However, spending on interest on public debt has quadrupled since the 1970s

114  Elite Networks Social spending categories as share in total spending (%), 1937–1995 12 9.6

10

8.4

8 5.8

6

2 0

2.1

1.9

6.1

6.4

4.5

3.5

4

5.8

2.4

0.4 1937

1960 Education

1980 Health

1995

Pensions

Figure 3.7  Education, health, and pension spending from 1937 to 1995. Entitlement spending over the past 70 years increased rapidly primarily in these three categories. Source: Tanzi and Schuknecht (2000).

(corresponding to the trend of rising public debts in the past four decades22). It went from a mere 1.4% to 4.5% in 1995, and over 6% in 2010.23 Such dramatic increases have understandably caused fiscal instabilities in many countries, particularly after the 2008 crisis. Finally, governments expanded not only in their programs but also in their services, which needed to be properly administered. The expansion of the bureaucracy was thus far from surprising. The size of public sector employment in total employment went from 5.2% before the war to 18.4% in the mid-​1990s (Figure 3.8). Public sector wages thus became the second-​biggest expenditure category, after social expenditures, in almost every developing economy.24 All the reported numbers are averages across the developed countries, and while there exists considerable variation between countries on each particular category of spending, the trends are the same for all, just as they have been for the growth of government spending. To conclude, government has steadily increased in size throughout the past 60 years, while also changing considerably in structure and hence scope. This trend has been slowed down from time to time, but it is a trend that undoubtedly shows how in contemporary societies the role of government is as important as it has ever been. The post-​COVID fiscal and monetary response exemplifies this better than anything.

Democracy and Inequality in the Short Run  115 Share of total social spending and relative public sector employment (%), 1937–1995 25

23.2

21.4 20 15

12.3 9.7

10 5 0

18.4

17.5

5.2

4.5

1937

1960 Subsidies and transfers

1980

1995

Public sector employment

Figure 3.8  Total spending on subsidies and transfers (social spending), and the share of public sector employment in total employment. Source: Tanzi and Schuknecht (2000).

3.4.  The Rise of Inequality in Democracies One of the first questions that come to mind while going through the spending numbers is, Have the governments been successful? Has an increase in government size and scope made the difference it was supposed to make? Did it make people healthier and more educated? Did it decrease unemployment and smooth out economic shocks? Did it provide the necessary infrastructure and public goods? And finally, did it decrease poverty and the rates of economic inequality? In other words, how good was government performance of its three main Musgravian functions: allocation, redistribution, and stabilization? While unemployment, inflation, and deficits are cyclical categories for which we cannot make any plausible inference, data on health, education, and inequality do tend to show sensible trends over time. Infant mortality dropped significantly, life expectancy increased throughout, average years of schooling doubled, illiteracy rates are down to almost zero, pensions dramatically reduced poverty among the elderly, and poverty overall has declined as well.25 It is obviously difficult to claim a causal relationship between the role of government and any of these trends, nor do I intend to engage in such a debate. However, one particular trend does stand out, stubbornly defying expectations in light of evidence of an expanding welfare state—​the rise of income inequality.

116  Elite Networks Earnings at top decile as % median, 1945–2013 250 230 210 190 170 150

Australia Canada* France Germany* Italy Japan Netherlands* New Zealand* Spain Sweden Switzerland United Kingdom United States

130 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 3.9  Gross earnings at the top decile as % of the median for selected developed countries, 1945–​2013. Source: Atkinson et al (2017) Chartbook of Economic Inequality. Countries with an asterisk (*) have two different series merged into one.

Figure 3.9 testifies to an increasing trend in gross earnings of the top decile of the population (top 10% income earners) as compared to the median income earners for selected developed countries, starting by the end of the 1970s, with the exception of France, which has not experienced a relative increase of top earnings at all, and Japan, where inequality did not start increasing until the 1990s. Atkinson suggests that the distribution (or dispersion) of earnings is the most precise measure of income inequality as it captures only the differences in earned incomes (as opposed to capital, rents, transfers, etc.) and it focuses on the relative performances of top income earners compared to the rest of society (this way it accounts for the fact that some people do indeed deserve higher wages due to their greater job responsibility or risk). If the growth of income among the top 10% of earners outpaces the growth of the median income in the population, inequality goes up.26 Before drawing out even remote implications, one should note that earnings data is not perfectly comparable between countries. This is mainly because of different ways of defining and reporting income and earnings across countries, particularly over time.27 However, in this particular case the point of interest is the trend within each country, and whether such trends of a rising dispersion of top incomes happened in other rich democracies at the same time. The point is therefore not to show that earnings at the top grew higher in the United States compared to other countries, or to point out the differences in levels of earnings across countries, but only to imply a general trend where top earnings relative to the median increased in the past 30–​40 years for the entire observed, yet deliberately limited, sample. The short conclusion is: they have. This has been well established in the literature for each of the countries used in the sample. Atkinson offers the most

Democracy and Inequality in the Short Run  117 detailed summary of inequality trends across 20 OECD countries. He finds that over a period of 25 years, from the 1980s to 2014, only France and Ireland have seen a significant (over 10%) growth in incomes of the bottom earners that has outpaced the growth of the top decile earners. All other countries have experienced high growth rates in top decile incomes coupled with either modest or declining growth rates of the bottom deciles. In the United States for example, top decile incomes increased by 15%, while bottom decile incomes fell by 5%. Similar trends, although with a lower magnitude in both directions, occurred in the United Kingdom, Netherlands, Sweden, New Zealand, and Australia, while other countries all experienced much higher growth rates in earnings of the top decile compared to growth in earnings of the bottom decile. Atkinson also points out that inequality in the dispersion of earnings mostly originated because of changes in top incomes, a trend that coincided in almost all observed countries.28 The Gini coefficients, a standardized measure of overall inequality shown in Figure 3.10, indicate a similar trend as do the earnings numbers from Figure 3.9. Across the observed sample inequality started going up since the 1980s onward, with the average increase across the sample being around five percentage points (excluding the United States and the United Kingdom the average increase is still high—​around four percentage points). The Gini index is designed to be more comparable across countries, meaning that we can make inferences of how the United States is a particularly interesting outlier in the developed world with significantly higher inequality than the rest Gini household gross income, 1945–2013 50 45

Australia Canada

40

France Germany

35

Italy Japan

30

Netherlands New Zealand

25

Spain* Sweden

20

Switzerland United Kingdom

15 10

United States 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Figure 3.10  Gini coefficients for equivalized household gross income, for selected developed countries, 1945–​2013. Source: Atkinson et al (2017) Chartbook of Economic Inequality. Countries with an asterisk (*) have two different series merged into one.

118  Elite Networks of the sample. On the other hand, the United States has throughout history always had higher inequality than European countries, persistently trading off inequality for efficiency. Back in 1975, in the midst of a decade of historically low levels of inequality in the United States, when average top income taxes were 75%, Arthur Okun still reports “alarmingly” high inequality in the country: “The richest 1 percent of American families have about one-​third of the wealth, while they receive about 6 percent of after-​tax income. The bottom half of all families hold only 5 percent of total wealth.”29 Today this figure is even higher for the richest 1%, holding around 18% of total income, and about the same in total wealth. The “cut-​throat” model of US capitalism based on an innovation-​driven economy versus the “cuddly” European capitalism30 will always imply greater relative inequality in the United States. The issue therefore, once again, is not in the relative differences between countries, but in the trends pointing to the same direction. Income inequality has been steadily increasing throughout the rich, developed democracies, with only a few exceptions. It should be noted that the data for individual countries reports an increase of inequality measured by any indicator used. The Chartbook of Economic Inequality database31 contains indicators of top 1% and top 0.1% of income earner dispersion but also indicators of wealth inequality and poverty.32 Figure 3.11 presents the trends of an increase in inequality at the very top of the income distribution: Piketty’s supermanagers. The share of top 1% earners’ Share of top 1% in gross income, 1900–2013 30 25 20 15 10 5

Australia Canada France Germany* Italy Japan Netherlands New Zealand Spain Sweden Switzerland United Kingdom United States

19

00 19 05 19 10 19 15 19 20 19 25 19 30 19 35 19 40 19 45 19 50 19 55 19 60 19 65 19 70 19 75 19 80 19 85 19 90 19 95 20 00 20 05 20 10

0

Figure 3.11  The rise of the supermanager, exemplified by the increase in the share of top 1% earners’ income in total gross income from 1900 to 2013. A share of 18%, for example, in the United States implies that 1% of top income earners captured 18% of all total gross income. Countries with an asterisk (*) have two different series merged into one. Source: Atkinson et al. (2017) Chartbook of Economic Inequality.

Democracy and Inequality in the Short Run  119 income in total gross income has seen two major trends in the 20th century—​a gradual decline from the beginning of the century culminating in World War II (confirming Scheidel’s argument that wars tend to be great equalizers), followed by a 30-​year relative33 stagnation and a major resurgence since the mid-​1980s. It should be noted that the resurgence trend is mostly driven by Anglo-​Saxon countries; United States, United Kingdom, Canada, New Zealand, and Australia. However, all other countries have also experienced an increase in the share of top 1% earners’ income, albeit much smaller than in the aforementioned countries. The average increase in inequality, measured through the share of top 1% income growth, was 55% from 1980 to 2008 for the entire sample (shown in Figure 3.12). But when dividing it between Anglo-​Saxon and non-​Anglo-​Saxon countries the difference is clear: the average increase in inequality in the former group of countries was over 90%, while in the latter group it was 33%. Two countries that particularly stand out are the United States and the United Kingdom with an average increase in top 1% inequality of 130% over the past 30 years. The trend across all countries supports Piketty’s supermanager argument: the rise of income inequality was primarily due to the rise of corporate executive earnings (the majority of those in the top 1% of earners). While all income indicators suggest the same conclusion of increasing income inequality across the developed world, poverty indicators paint a somewhat Average increase in top 1% inequality

140.0%

129.9%

120.0% 100.0%

90.6%

80.0% 60.0%

55.5% 33.6%

40.0% 20.0% 0.0%

Entire sample

Anglo-Saxon countries

Others (excluding Anglo-Saxon)

US and UK only

Figure 3.12  Average increase in inequality from 1980 to 2008 when measured as the share of top 1% earners’ income in total gross income across the sample of 13 chosen countries. Source: Atkinson et al. (2017) Chartbook of Economic Inequality.

120  Elite Networks different picture. On average across all countries, poverty levels have, since the War, dropped down from above 30% to slightly below 13% by the early 1990s, but have since then started to gradually increase. For some countries the rise in poverty exacerbated after the financial crisis of 2008, bringing the average poverty levels across the sample back to an average of around 17%. However, this increase has been much lower than the corresponding rise in earnings inequality, reinforcing the notion that the main culprit for the rise of income inequality has not been a decline of bottom incomes, but rather a rapid rise of top incomes. Finally, wealth inequality also experienced a steady decrease since its high prewar levels across the entire sample (wealth was always much more concentrated than income, so wealth inequality levels were historically always higher than income inequality), and has for most countries evened out and stagnated in the past two decades. In the United States and the United Kingdom however, according to some measures, wealth inequality has in the same period again started to slightly increase. To sum up, two very interesting trends occurred simultaneously over the past 60 years. There was a continuing expansion in government size—​its expenditures, revenues, social spending, public debt—​a trend that started in the 1960s, peaked in the mid-​1990s, declined slightly afterward, and has again started rising after the 2008 crisis. During the same period, originating after the 1980s, there was a significant increase of income inequality, mostly driven by large increases in incomes of top earners (1% and 0.1%). There is hardly any causal inference to be made at this point, but the trend of rising inequality does pose a question on the predictions of classical political economy models on the relationship between democracy and inequality. What could have been the reasons behind this? Did the behavior of well-​ organized special interests deplete the resources available for redistribution to the poorer ends of society by channeling excess government spending toward themselves? In other words, has the expansion in government size been tilted toward narrow groups with more power, thus creating a rent-​seeking society? Is favoritism of the few at the expense of the many an inevitable outcome of our political system? In order to shed a new light on these questions, we need to implicate politics and understand the role of political decision-​making in shaping economic outcomes. But before that, it will be useful to survey the numerous empirical studies done by economists to uncover and understand all the economic factors that could help explain the rise of income inequality.

3.5.  Economic Explanations of the Rise of Inequality According to Atkinson, Piketty, and Saez, annual average real income growth for the bottom 99% of income earners in the United States was only 0.6% over a period of 40 years (from 1976 to 2007). In the same time, incomes for the top 1%

Democracy and Inequality in the Short Run  121 grew at an annual rate of 4.4%. Most of this income growth did not come from salaries (even though their growth was also substantial), but from things like capital gains and business income (investments, dividends, business profits, etc.) This long-​run trend led to massive accumulation of wealth at the top, where the top 10% of income earners earned 47% of all income in the United States, while the top 1% earned about 19% of all income34 (their share has slightly declined as an immediate consequence of the financial crisis, but has continued to expand since). Even though the United States is an outlier among rich democracies, similar trends of rising incomes at the top and stagnating incomes at the middle and the bottom happened all across the West. Why have real wages stagnated particularly among the lower income groups and the middle classes, while income earnings at the top grew rapidly? Economists have provided a number of explanations. The most usual ones are the impact of education, technology, and globalization. Technological progress in the past 30 years was attributed to the IT revolution (which many call the Third Industrial Revolution), and has been tilted in favor of those with greater skills and better education. Economists call this the skills-​based technological progress hypothesis,35 fully capturing the creative destruction impact of technology on jobs and wages. The supply side has adjusted to this change. More and more people started getting better education across the developed world. University enrollments as well as the share of people with a university degree are substantially higher in every rich democracy compared to pre–​World War II levels. The skills premium, according to which greater education implies better paying jobs, has made a further dent in the divergence between top and bottom incomes, which are becoming clustered with respect to attained education levels. Globalization was another important factor. The rise of China and the supply chain economies of East Asia have created an outsourcing trend where low-​ skilled manufacturing jobs in the West were being shifted to low labor cost Asian countries. Free trade created clear winners and losers from globalization. Domestically, immigration took the rest of the low-​skilled jobs, pushing the labor market equilibrium toward lower wages. The low-​skilled, poorly educated blue-​collar workers who could not adjust faced either long-​term unemployment or a significant cut in their incomes by accepting jobs in the emerging services industries, which paid less than their former manufacturing jobs.36 Economic theories do recognize the importance of certain political factors that depressed low and medium-​income earnings. Minimum wage laws, always a controversial political issue, have not followed inflation trends, thus further depressing real wages of the poor. Changes in tax codes became less progressive over time, and opened up new opportunities for wealthier citizens to shift their wage incomes into things like capital gains. This consequentially increased the share of capital and decreased the share of wages in total national income. On

122  Elite Networks the other hand, transfers to the poor became more streamlined toward groups with more political power (such as the elderly). Atkinson also cites various cuts in welfare state programs throughout the 1980s and 1990s, leading to a scaling back of the welfare state, rising levels of long-​term unemployment, rising interest rates on student loans, and of course the declining bargaining power of labor unions (membership in unions, particularly in the private sector, fell drastically since the mid-​1970s).37 Deaton adds to this the gender wage gap, which particularly affected single-​mother households, and the fact that poorer, nonvoting immigrants are not politically represented well enough to successfully fight for more redistribution for themselves.38 On the other hand of the spectrum, positively affected by most of these changes, were the top income earners. As emphasized earlier, the rise in earnings dispersion came primarily as a result of higher growth of top incomes, mainly through higher gains from capital investments, dividends, and business profits (for innovative entrepreneurs). The high-​skilled, innovative part of the economy received their rewards. Emerging billionaires in the tech sector created new jobs and entirely new industries. The problem was that such industries required a high-​skilled workforce and could not replace the jobs lost due to outsourcing, at least not for the same people. While the tech sector captured most of the innovation-​led growth, other industries turned toward trade and managed to achieve large economies of scale operating multinationally. This could explain the exponential growth of CEO incomes (both their salaries and their capital gains), the competition for which was no longer local or national, but global. Gabaix and Landier have directly linked this expansion in firm size to CEO pay. They found that, since the 1980s in the United States, a sixfold increase in CEO pay can largely be explained by the sixfold increase in market capitalization of the biggest companies.39 An increase in industry concentration40 surely contributed to this effect, as large firms became even larger, increasing the bargaining power of their CEOs in demanding greater compensation. As some CEOs secured higher compensations, the contagion effect made sure that other firms also raised salaries for their CEOs in order to attract similar high-​quality candidates or risk losing their best people to the competition. The multinational orientation of US firms can explain even the within-​firm differences where workers get paid the local rate of their city labor market while CEOs get paid the global rate in the market for top talent. A similar logic is at hand for earnings of other superstars, either in sports or show-​business. High demand for top quality talent, either by spectators or shareholders, combined with extremely scarce skills will make sure that earnings of superstars keep on rising. However, in some industries the superstar effect seems to be clouded by luck. In finance the best example are bonuses given to high-​performance employees as a consequence of high stock market returns. Kahneman showed that there was

Democracy and Inequality in the Short Run  123 no consistency in year-​to-​year rankings of top performers in the stock market within financial firms.41 In other words, finance companies were rewarding luck instead of skills. A similar phenomenon occurs with CEO rewards in the oil industry. An oil shock that pushes oil prices up and has nothing to do with individual skills of a CEO (usually triggered by political or exogenous market factors) will still deliver a large bonus to that CEO.42 Another factor could have been policy changes. Extensive countrywide research conducted by Piketty, Saez, and Stantcheva suggests that tax policies since the 1980s have been a major contributor to the rise of inequality in OECD countries. They find a strong negative correlation between top tax rates and top 1% income shares, meaning that top rate tax cuts since the 1980s in selected OECD economies impacted the rise of top income shares. Furthermore, when top marginal tax rates fell, they’ve increased the bargaining power of top executives vis-​à-​vis their boards, thus increasing their total compensation. The bargaining effects were the key factor in explaining inequality at the very top.43 Frequent changes in the tax code, instead of solving the issue typically only add to its complexity. In almost every country the tax code is deliberately cumbersome, with so many exceptions and potential loopholes that are typically only exploited by those with privileged access to information and resources necessary to seize the loopholes. Complexity exacerbates inequality because the poor and middle classes lack the resources to exploit tax loopholes and end up paying a disproportionate burden.44 This is part of a wider trend of increasing policy complexity in general, particularly since the second half of the 20th century, epitomized by a significant increase of regulators and lawyers and the buoyant bureaucracy.45 As technological progress makes societies exponentially more complex, there is an increasing, yet futile, necessity to regulate everything. The typical regulatory responses to economic and social complexity, other than always being introduced ex post, can also produce adverse effects by increasing informational asymmetry to benefit those close to the decision-​makers. Fukuyama sees this as one of the key issues of US political decay, as domestic political institutions are becoming more susceptible to usurpation.46 Informational asymmetry arising from policy complexity further restricts access to opportunities, adding to the inequality problem. Inequality is typically portrayed as a within-​ firm phenomenon, where widening disparities in earnings between the executives and the workers act as the main driver of income disparities. However, in a recent paper, Song and his colleagues use microlevel data for the United States and find that the inequality effect, particularly among the bottom 99% as well as for the majority of the top 1% (except those in the top 0.02% of earners), is more likely to arise between firms, rather than within firms. The reason for this is clustering of high-​paid workers in bigger and more successful firms, and low-​paid workers in smaller

124  Elite Networks and less successful ones.47 This is confirmed by Autor and his colleagues, who find that the biggest firms, the so-​called superstars (like Big Tech) act as the main driver of compensations at the very top.48 The concentration in firm power and by extension executive compensation is nowhere more obvious than in finance. The individual elite networks drawn in Chapter 1 all originate from the finance industry, the expansion of which has been unprecedented since the 1980s, as have the salaries of their top executives. Research has shown that finance employees get 50% higher earnings in the United States compared to their nonfinancial counterparts, while the wage premium for top finance executives is 250% compared to nonfinance top executives.49 This trend of divergent wage premiums for finance versus everyone else started in the 1990s and has continued to increase steadily over the next 20 years. It is hardly surprising therefore that 14% of the increase in the US Gini index since the 1990s can be attributed to executive incomes in finance.50 In the United Kingdom, research has shown that finance industry employees, which constitute around 5% of the entire workforce, captured 60% of all growth in top incomes from 1998 to 2008.51 Even in France the same trend is noticeable since the mid-​1990s, despite finance not having such a prominent role as in the United States or the United Kingdom. About half of the increase of the top 0.1% of incomes was captured by people working in finance.52 These effects are hardly surprising when taking into account the significant expansion of the finance industry and its contribution to GDP in many countries worldwide. In the United States, its share of GDP has increased from 4.9% in 1980 to 8.3% in 2006, growing at a much faster rate than in the previous 30 years.53 This growth in earnings has mostly been captured in asset management and household credit provision (specifically mortgages). In other words, housing loans pooled into various types of bonds and bond-​backed derivatives, repackaged and resold to various asset managers and big banks (the culprits of the 2008 financial crisis). As finance captured a larger part of economic growth, it is no surprise to see salaries going up across the industry, especially for its top executives. The rising power of finance was not only an issue in developed countries, but carries an impact on developing countries too. A study has found that the most unequal developing countries face further constraints from special interest capture and high barriers for access to finance. Specifically, the political and corporate elites (i.e., elite networks) capture financial liberalization reforms so as to divert access to finance for themselves, while entirely socializing their risks. Unequal access to politics leads to unequal access to finance, which skews the distribution of opportunities, and by extension incomes and wealth. As a consequence, this creates further political backlash against any kind of reform, pushing such societies further down the negative feedback loop.54 Think of the

Democracy and Inequality in the Short Run  125 privatization scandals in transitional Eastern Europe or Latin America and the huge negative perception their domestic populations have against reforms like privatization. Privatization schemes across transitional economies were typically organized to favor entrenched domestic elites close to politics, very often originating from the old regime elites.55 In developing countries therefore, the capture of financial institutions by elite networks is the main mechanism through which the elites exacerbate inequality. This is becoming an ever more prominent trend in developed countries as well. It is a natural consequence of expanding elite network power. To sum up, the economic culprits for higher inequality can be found in the interaction of several factors. Rapid technological progress in the past 30 years resulted in a typical creative destruction process where new jobs and careers made certain types of old jobs obsolete (automated work). Some of these obsolete jobs were outsourced to Asia, others either did not re-​emerge or were limited to low-​skilled immigrants. The low-​skilled workers failed to adapt to the changes and were left stranded either at lower paid jobs or became long-​term unemployed. Poor education played an important role, while stagnating wages in the “dying” sectors only widened the gap. On the other hand, the innovative part of the equation was working quite well taking advantage of the new technological wave, thus further raising the income of the top 10% and the top 1% (hence the great disparity between college and non-​college-​degree workers). It is not hard to imagine how these two forces, one pulling the low-​skilled downward, the other pulling the high-​skilled upward, widened the income inequality gap in the West. However, the standard skills-​ based and globalization and technology hypotheses explain only one part of the rise in unequal incomes. They are insufficient to explain the drastic changes at the very top of the income distribution. First and foremost, all developed countries experienced the same impact of better education that improved skills, and a similar impact of globalization and new technologies, but only some saw a huge rise in top 1% incomes. These were societies which obviously had a higher share of the largest global corporations, the superstar firms. Furthermore, sectors in which superstar firms are becoming more and more concentrated—​like the finance industry—​are seeking to preserve their status mainly through the political process. An obvious way is through lobbying and campaign spending, but a much more subtle and effective way is through elite networks, where corporate executives befriend politicians, who become more reluctant to pass legislation that hurts their friends’ cause. It is important to note here, as shown in Chapter 7, that industry concentration is another artifact of rent-​seeking firms preserving their status through political connections. While the increase in firm size can be a good explanatory factor of the rise in wages, and while an increase of firm concentration and the rise of superstar firms adds to that effect where their top corporate executives are

126  Elite Networks being paid higher salaries, the missing link here is the political connection these firms needed to have to allow such concentration and subsequent rise and preservation of corporate power.

3.6.  Political Factors That Could Help Explain the Rise of Inequality In addition to the innovative and superstar part of the top 1% (and a few lucky ones), there were others who instead of working for their wealth either inherited it or have benefited from the political system of capturing rents. Stiglitz refers to them as rent-​seekers, those who receive “income not as a reward for creating wealth but by grabbing a larger share of the wealth that would have otherwise been produced without their effort.”56 The problem is that even though the economic pie is growing, an even bigger amount of that pie is being captured by rent-​seekers instead of wealth generators, which could be a potential culprit for inequality. Furman and Orszag cite an increasing concentration of high returns among the most profitable firms which all consequentially pay the highest salaries and drive the between-​firm inequality in earnings. Most of these concentrated returns are not based on innovation, leading them to conclude that the returns could be a result of rent-​seeking.57 Lindsey and Teles suggest that high levels of industry concentration, an increasing importance of intangible assets in corporate balance sheets, and growing barriers to entry in many industries all point toward increasing rent-​seeking that is driving up inequality.58 While all these authors are referring to the United States, Svejnar and Bagchi examine the hypothesis worldwide. They find that in countries where billionaires made their fortune thanks to political connections allowing them to control and build monopolies, the effect of their accumulated wealth and the consequential income inequality on economic growth is extremely negative. In cases where the billionaires weren’t politically connected there is no effect between inequality and growth. Also, more unequal countries, where individual fortunes are conditioned on political connections, are all characterized by high levels of corruption.59 Corruption and political connections tend to go hand in hand. And they seem to be pulling inequality with them.60 Hence the importance of politics in explaining the phenomenon of rising inequality. If the political system is captured by crony rent-​seeking elites who make sure that the gains from economic growth are limited to the selected few, if the majority of the population believe that the system is skewed against them (“rigged”) and that even in their greatest effort they still cannot move up the social ladder or even escape poverty, this will fuel anger and create political discontent. One direct and noticeable consequence of this across the West has been

Democracy and Inequality in the Short Run  127 the ascent of extremist populist political parties, feeding on the justified anger of disenfranchised voters. Consider again the example of the finance industry. It is no surprise to learn that banking in particular has been very effective in using campaign expenditures, lobbying, and their accumulated political influence to guide regulatory decisions, block unfavorable and get favorable legislation passed in Congress, and as a consequence achieve abnormal market returns.61 In fact, banks with greater lobbying expenditures were shown to be taking on more risk, particularly prior to the recent mortgage crisis,62 while greater political influence clearly helped some banks secure much better bailout deals.63 The main issue with such activities of the finance industry is that lobbying done by banks to save themselves imposed a huge cost on taxpayers. This happened in Europe as much as it did in the United States. The consequence was a sovereign debt crisis as European countries massively accumulated debt to save systemically important banks from bankruptcies, thus imposing a double cost on taxpayers when it came time for harsh austerity measures to address the debt and deficit issues. These activities have increased both inequality and government spending. The problem therefore is that lobbying activities of one powerful group (even if done on an individual firm-​by-​firm basis) have imposed direct costs on the rest of society. In the fight for government redistribution the well-​organized and the powerful groups will prevail. This goes beyond individual industry examples. A number of papers confirm this effect, where greater lobbying expenditures enable interest groups to get a large part of the budget in several countries.64 To illustrate how this works, consider the example set out by Deaton on poverty rates in the United States.65 Poverty trends have genuinely been favorable in the United States ever since the 1960s. They have experienced steady declines for all social groups, the old, the young, the African Americans, and all minorities. In the 2000s, even prior to the crisis, poverty rates for all groups have started to slowly increase, except for one group—​the elderly (to be exact, all those over 65 years of age). The reason the elderly were the only group that experienced a consistent decline in poverty rates, down to 10% according to the official Census data,66 is entirely attributed to the success of the Social Security (Medicare) program for senior citizens. And why has the Social Security program been so successful for citizens over 65? Because their interests are being protected by one of the most powerful lobby groups in Washington, the AARP (formerly the American Association of Retired Persons). The AARP has consistently been ranked among the top three most powerful interest groups in America by the Fortune’s “Power 25” interest groups ranking,67 and it has spent over $260 million on lobbying to Congress since 1998.68 Its lobbying efforts have certainly paid off as its members have enjoyed a better standard of living and lower poverty

128  Elite Networks rates compared to all other groups. Lowering poverty and improving living standards of senior citizens is certainly a commendable effort; however, the problem is that when one group successfully diverts budgetary resources toward their own interests, it manages to do so at the expense of all the other groups dependent on public funding. In other words, successful political lobbying from the AARP for its members lowers the funds available for all other groups that are becoming more exposed to the risk of poverty. Political scientists have uncovered a myriad of ways special interests tend to capture the political system. Some of them link this directly to inequality, others simply point out an indivertible logic of a system gone wrong. Findings have however mostly been limited to the United States, so there is a necessity to provide a compelling, overreaching theory that will be able to explain why the political system tends to get captured by special interests regardless of a specific country’s institutional strength.69 An insightful paper by Bonica, McCarty, Poole, and Rosenthal tackles the issue of inequality and democracy, asking why democratic forces haven’t lowered inequality. They stress five possible reasons as to why the US political system failed to address the issue of rising income inequality. The first is an ideological battle of ideas that lowered support for transfers, taxation, and financial regulation. The second was immigration and low turnout among the poor, in addition to administrative measures that made it more difficult for the poor to vote. The third was that rising wealth of the population made a larger fraction of them less supportive of social transfers (e.g., difference between Social Security, Medicare, and Obamacare). Fourth is that the wealthy have used their resources to influence legislative and regulatory decisions via lobbying and campaign contributions, and the fifth is that the political process has further been undermined and distorted by things like gerrymandering, filibusters, and political polarization resulting in gridlocks, all designed to maintain the status quo of American politics, which clearly benefits the few at the (indirect) expense of the many.70 Bermeo however elaborates that democratic mechanisms may even prevent the formation of broad-​based support for more redistribution. Democracy is not about achieving economic but political equality, a desire for the state to offer protection against violence and of property rights. Therefore, many voters do not see inequality as an issue of high enough salience for it to be directly tackled by their governments. In other words, democracies can survive with high inequality.71 Hacker and Pierson directly attribute the rise of inequality in the United States to political causes, rather than globalization or the skills-​based technological change hypothesis. According to them the most important change since the 1970s was the rise in organizational capacity of various special interest groups representing the wealthy elites (big business organizing to reduce regulation and

Democracy and Inequality in the Short Run  129 taxes).72 The decisions on resource redistribution were not being made by the voters as suggested by the Meltzer and Richard median voter hypothesis, but rather by a myriad of special interests, as suggested by Olson. The elites have blocked reforms that could hurt them and have promoted beneficial tax and regulatory policies, tilting the gains from globalization in their favor (and hence causing the huge increase in top incomes which drove income inequality up). The status quo favored the elites, so they lobbied to preserve it, as McChesney has shown describing this relationship as “money for nothing.”73 While Hacker and Pierson assign the blame for rising inequality to both political parties in the United States, Bartels finds that political partisanship is the main culprit. He too claims that inequality is, for the most part, a political phenomenon but links its rise to the policy choices made during the terms of Republican presidents. He finds that real incomes of middle classes have doubled during a Democratic president compared to a Republican one, while incomes of the poor and working classes have increased tenfold under Democratic presidents compared to Republican presidents, controlling for all other factors. Bartels even goes so far to suggest that income inequality would have stagnated during the past 60 years had there always been a Democratic president in power, and that it would have been twice as high as it currently is had the Republicans been in power for all this time. These claims hardly constitute a causal link between Republican presidents and inequality, as the statistical analysis is overly simplified, suffers from endogeneity issues, and is based on a very small sample to be able to adequately convince the reader in the validity of the argument. However, there are clear policy differences between the two main US political parties that arguably have made different impacts on inequality in the past several decades (for example, in tax cuts, minimum wage legislation, labor union power, etc.) His explanation is therefore that inequality was driven by partisan policy decisions.74 He also finds evidence of politicians in power being more likely to represent the preferences of the wealthy rather than the general population, a conclusion that has further been examined and confirmed by Gilens. These findings testify to a significant increase in political inequality in the United States, suggesting the lack of representation of the poor and middle classes whose preferences are almost always subverted to those of the rich. This too could have contributed to the rise of income inequality as the wealthy (roughly approximated and underrepresented by the top 10% income earners) consistently get the policies they prefer (economic, social, religious, or foreign), whereas policy preferences of the median income voters and those in the lowest 90th income percentile are consistently ignored. The same finding is confirmed for interest group policy preferences. The poor and the middle classes don’t always lose—​they get the policies they prefer only if their preferences align with those of the wealthy or with a given

130  Elite Networks interest group (however interest groups tend to represent very narrow groups of citizens in most cases, so the likelihood of their alignment with the general population is low). Accordingly, US democracy became the service of the wealthy and organized few instead of the disorganized many.75 Moss, through a series of case studies and interviews with Washington insiders, delivers a similar finding, suggesting that policymaking decisions in the United States are not driven by what the voters want, but are a consequence of diligent interactions between politicians and special interests who both achieve their goals at the expense of the rest of society.76 Overall, such findings are striking. Not only do they imply that when it comes to affecting policy change the elites and special interests fare most successfully, they also indirectly imply that such groups could have an important say in setting the agenda that, in most cases, preserves the status quo. This means that the institution of majority rule in America is to a large extent disturbed. Citizens do not get to influence policy. Their preferences only get represented if they happen to align with those of the elites and special interests. Such a democratic order, skewed in favor of elite power, explains the status quo bias of American politics and why is it so hard to implement policy change. *** The problem with the contemporary political system is that it has morphed to such an extent that it can allow those with power to bend the very institutions that serve to hold it accountable. The political elites and special interests tied with them managed to bend the rules to their liking, further skewing budgetary resources, legislation, regulation, and hence the distribution of income toward themselves. It is surprising to see this happening in a rich and institutionally well-​endowed country like the United States, particularly since such outcomes were usually limited to either transitional economies of Eastern Europe or Latin America, and to an even more grotesque extent to developing countries in Africa and the Middle East. The capture of governments by powerful elites is a pattern all too common throughout the developing world, so one must ask why a similar pattern is apparently occurring in the developed world. Why haven’t the institutional rules and decades of accumulating democratic capital77 been able to prevent rent-​seeking and state capture? Why haven’t the forces of democracy embedded by a strong constitutional and institutional order prevented the rise of inequality? The same logic of power operates throughout developed democracies and throughout political-​economic systems. This is not an effect caused by democracies, but rather by the quest for power. Even the strongly embedded institutional safeguards in democracies cannot always stay robust to the logic of elite network power, which is why we must look within such networks to get a better understanding of how they operate.

Democracy and Inequality in the Short Run  131 These arguments get us closer in being able to attribute rising inequality to elite networks, which distort the political process to serve their private interests, but there is still one step missing—​a coherent individual-​level empirical proof on how the mechanism works. More precisely, we need to establish a mechanism on the level of individual politicians and individual corporate executives through which their distortion of the political process widens the gap between the top income earners and the rest. This is the topic of the following chapter.

4

Political Networks and Wages of Top Corporate Income Earners Too often, executive compensation in the US is ridiculously out of line with performance. That won’t change, moreover, because the deck is stacked against investors when it comes to CEO pay. —​Warren Buffet

Inequality, as emphasized in the previous chapters, is a product of many different interrelated factors. This chapter focuses its attention on a factor that has thus far mostly been overlooked in discussions on inequality—​the impact of political connections of corporate executives. I introduce a new way of measuring elite networks while also directly estimating the wage premium that executives receive if they are politically connected. Following the definition from Chapter 1, an elite network is formed between politicians in power (holding executive elected positions, cabinet positions, or key positions in parliamentary committees) and senior executives of the biggest publicly listed companies in a country. The network is formed based on proximity to each other through a similar career path or membership in the same organizations (clubs, societies, foundations, religious groups, etc.). Within an elite network, favors are exchanged between actors. These usually imply beneficial regulations or government contracts (rents) for the firm in exchange for political campaign donations or other form of support for the politician. The executive, who had successfully extracted rents for the firm, can demand a higher compensation for his efforts. An empirically observable consequence of this relationship is that executives who are members of elite networks have higher average annual earnings than executives who are not members of elite networks—​they receive a wage premium, which is a direct consequence of their political connection. In order to approximate the effect of elite networks I use a unique individual-​level and firm-​ level database from BoardEx across a 16-​year time span for the United States and the United Kingdom, and identify which individuals had connections to politics either via their previous careers in the public sector (if they held any high-​level positions in government) or if they were members of the same organizations as their connected politicians (this includes country clubs, foundations, charities, Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0005

Political Networks and Wages  133 trusts, church or religious groups, professional organizations, etc.). The theoretical implication of why and how these networks form between individuals is the topic of Chapter 5. This chapter delivers the cornerstone empirical argument of the book. It first describes the data and looks at the trends of inequality over the stipulated time period for both countries, after which it presents the results of the empirical analysis. The empirical conclusions suggest that politically connected executives, i.e., members of elite networks, tend to be more valued by their companies and are indeed rewarded with higher compensations. Furthermore, elite network members tend to have a larger total number of people they are connected to, meaning that any member of an elite network is highly likely to be on average more (and hence better) connected than anyone outside an elite network. Both of these factors matter as an explanation of higher executive earnings, however direct political connections matter more. Being connected clearly matters in the business world, and it tends to get substantially rewarded.

4.1.  Data and Variables The main source of data are individual-​level connections and earnings of top corporate executives in the United States and the United Kingdom from 2000 to 2015, collected by the private company BoardEx, accessed via the Wharton Research Data Service.1 The BoardEx database is assembled by looking at all publicly listed companies of a given country in addition to selected large private companies (not listed on the stock market but reaching considerable size in terms of both employees and revenues). The quality of the dataset is one of its main advantages. They use official sources to gather information on individual profiles for each senior executive in their sample of firms. When doing so they cross-​reference whether the data has been assigned correctly to each profile, and they make sure that each profile is completed, subject to data availability in the official registries. In the United States they use Annual reports from the Securities and Exchange Commission (SEC), US stock exchange data (NYSE and NASDAQ), press releases and official data from corporate websites, and various other registries to track individual careers over time. For the United Kingdom they use official reports from Companies House, the London Stock Exchange, the Regulatory News Service, press releases and official reports from corporate websites, and various other registries of personal data. They have data on 11,800 firms in the United States and 3,500 firms in the United Kingdom (some of which got de-​listed over time), while the total sample contains two-​thirds publicly listed companies and one-​third large private companies. In total they have data on over 750,000 individuals worldwide

134  Elite Networks collected over a number of different periods. In each case they collect full information for each director and senior manager in the firm (director, CEO, CFO, COO, president, chairman, VP), including their wages and total earnings (containing bonuses and stock options), employment history, gender, age, education, total experience, performance, the size of their network, and membership in other activities outside their workplace. The size of the network represents the total number of people an individual executive is connected to through various business and personal interactions. For example, two executives that served on the same board in another company before, graduated on the same year from the same university degree, or are members of same organizations (business clubs, professional organizations, nonprofit associations, church groups, charities, school or university boards, etc.), are considered to be members of each other’s networks. They might not be friends or interact frequently, but there is a connection that can enable one of them to easily get in touch with the other. From this extensive dataset I focused on the period from 2000 to 2015 (the year 2000 is when the dataset starts). I extracted individual-​level data on over 26,000 individuals working at 1,300 firms in the United States, in addition to over 21,000 individuals working at 2,500 firms in the United Kingdom.2 From the full dataset I created four panels, two individual-​level and two firm-​level. The initial datasets had to be further reduced by focusing only at senior top-​level executives and the company’s highest earners (president, chairman, director, CEO, CFO, COO) thereby excluding all lower-​level earners in the corporate hierarchy, many of which were included in the original dataset. The final individual-​level datasets contain 46,299 observations for the United States, and 31,422 observations for the United Kingdom. The firm-​level datasets contain 12,829 observations for the United States (around 800 firms), and 19,908 observations for the United Kingdom (around 1,200 firms).

4.1.1.  Measuring Networks and Connections of Individuals and Firms 4.1.1.1. Network Size Measuring the size and scope of individual connections is the most valuable part of the BoardEx database. It allows a detailed insight into how important it is to be a part of a particular network. An individual executive’s network is measured as the number of connections (links) he or she has with every other executive in the dataset. The larger the number of an individual’s connections (the degree of a network), the more influential he or she tends to be. These connections are achieved either through previous jobs, memberships on boards in other companies, education (same generation and/​or university major), or membership in various

Political Networks and Wages  135 other organizations (which include clubs, charities, church groups, university or school boards, nonprofits, etc.). An individual’s network, or total number of connections, therefore includes everyone he or she was or currently is connected to throughout their career, education, or other activities. Despite the effort the BoardEx team applies in making sure the data is entered correctly, there is still a possibility of measurement error. In particular, some individuals tend to have very large networks (over ten thousand people), while others seem to be connected to only a few people. An executive at a high position with such few connections is surely an underestimation, however given that BoardEx links executives only within its dataset it is entirely plausible for some to have such low levels of connections (less than 10). On the other hand, individuals with over 10,000 connections are not necessarily superconnected. These are mostly lawyers belonging to bar associations of their state, so the dataset lists them as connected. Every person that wants to practice law in a state has to pass a bar exam and therefore become a member of the state bar association, which means that these individuals are highly unlikely to be connected to so many people. The point is to look at individuals with several thousand connections—​these are more likely to be the central hubs of their networks, and discard the lawyers with an unrealistically high number of connected individuals. Consider a few examples. Jamie Dimon, the popular CEO of JP Morgan whose network was drawn in Chapter 1 (Figure 1.7.), has 5,776 connections in the database through various organizations he belongs to (most of which through his job, but also through charity organizations, business groups, and the New York Fed). Brian Moynihan, the CEO of Bank of America has 6,184 connections and belongs to a lot of similar organizations as Dimon. Lloyd Blankfein (4,499 connections) and David Solomon (2,656 connections), former and current CEOs of Goldman Sachs, both, for example, belong to the Robin Hood Foundation, a charity that helps fight poverty in New York, and are obviously connected to a lot of the same individuals through their jobs. Outside of finance, the CEOs of Microsoft and Alphabet, Satya Nadella and Sundar Pichai, have 5,699 and 4,167 connections respectively, more than most other big business CEOs (Tim Cook of Apple has a network of 2,559 connections). Interestingly, the wealthiest individuals in the database, such as Warren Buffet (1,465), Jeff Bezos (2,222), Elon Musk (1,054), or Mark Zuckerberg (1,473) all have less connections than the average bank or corporate CEO, which is expected—​the CEOs need to network much more than the most powerful business leaders or investors (George Soros, Bill Ackman, and Ray Dalio are also averaging around 1,000 connections, despite being investor superhubs). Those in between politics, academia, and business also tend to have many connections. Larry Summers has 4,888, Ben Bernanke 3,128 (pictured in Figure 1.6. in Chapter 1), while Peter Orszag and Jack Lew top them all with 8,052

136  Elite Networks and 9,149 connections respectively (although Lew’s are slightly overestimated due to him also being a member of the DC Bar association). Network size is therefore the first explanatory variable, attempting to estimate how the mere size of an executive’s total personal and business connections influences his or her earnings. The hypothesis is that the larger the size of an individual executive’s network, the greater his or her earnings, resting on the assumption that a well-​connected executive has a greater chance of securing a better deal for the firm.

4.1.1.2. Political Connections Another unique advantage of the BoardEx database is the ability to examine career trajectories of executives and board members prior to their top positions in the firm, as well as their memberships in various organizations. These two facts enable me to match individuals to their previous positions in government and to each other in order to define the main variable of interest: political connections, defined as an indicator variable POLCON it = {0,1}. The main POLCON variable is composed of two specific measures of connections: previous government experience, denoted POLCON Gov, and membership in the same organization as the relevant politician,3 denoted POLCON Org . I first explain how I coded each of these to form my unique measure of political connections, and then I describe the dataset matching process through which I ensure that each individual is uniquely and correctly matched. The first subvariable, POLCON Gov, defines an individual to be politically connected if that individual held a decision-​making position in government at any time prior to the start of his current executive tenure in the private sector. Establishing this connection required additional cleaning of the database, which implied cross-​checking which type of position a senior executive held while working in the public sector. If this included medium or low-​level bureaucratic positions (e.g., analyst, officer, civil servant, assistant, researcher, clerk, etc.) these individuals were coded as not being connected, POLCON Gov = 0. However, if they were members of various state and national committees, councils, or assemblies, not to mention if they held direct executive positions in government (manager, director, chief of staff, chairman, minister, congressman, senator), this does qualify as a political connection, POLCON Gov = 1. This variable encapsulates the idea that after leaving their position in government, a senior executive continues their membership in the social network of existing political decision-​makers. The second subvariable, POLCON Org , looks at whether an executive was connected to a politician via membership in the same organization. What classifies as an organization? The BoardEx Data Directory defines these through “other activities” and provides a detailed list of all organizations every

Political Networks and Wages  137 executive is or was an active member of. The list is large, and it includes a wide variety of organizations: professional organizations (excluding bar associations in this case, but including any type of industry lobby group), clubs (country clubs, Rotary clubs, various societies), university of school boards (and their alumni organizations), NGOs, institutes and think tanks, trusts, foundations, charities, church groups and various other religious organizations, veteran or military organizations, various political advocacy groups, and even political parties.4 Coding these two subvariables required a lot of attention to detail. Establishing government experience was relatively straightforward: it required matching two separate datasets, one that tracked previous executive careers (manually excluding all lower-​level bureaucratic positions first), and the main dataset that contained data on executive earnings and firm-​level details. However, coding POLCON Org was more difficult. The first step was to recognize which entry in the sample is still a career politician (this included congressmen, senators, members of parliament, assembly legislators, mayors, governors, cabinet ministers). Then I divided the samples into (active) politicians and nonpoliticians (just executives) and matched the two samples by the organization ID. This enabled me to see which executives and politicians are members of the same group, and to how many politicians in total an individual executive is connected. I’ve excluded all cases where the timing between the links does not fit to make sure that I correctly coded the connections. This means that any national or local-​level politician (from cabinet ministers to mayors, including the legislators like members of Congress or Parliament) had to be a member of the same organization as the corporate executive in order for the executive to be coded as politically connected. This relationship is taken as prima facie evidence of a connection within an elite network. Given that BoardEx also has data on political figures and their affiliations it was straightforward to pick out which individuals belonged to the same organizations. The final POLCON it variable was then created by simply adding the two subvariables together, with the indicator taking the value of 1 if an individual worked at a decision-​making position in the government at any time prior to his or her executive job, and/​or if he is a member of the same organization as the local or national legislator (assemblymen, congressmen, and senators) or executive politician (mayor, governor, secretary of state, or cabinet minister). In total for the United States there were 4,561 cases (9.8%) of previous government experience (across 16 years), and 30,921 cases (66.7%) of same organization membership, however there were about 3,000 cases where an individual had both previous government experience and was member of the same organization as the politician. These were also coded as 1. In total for the United States, across 16 years, there were 32,412 entries coded as politically connected (70% of the

138  Elite Networks sample). In the United Kingdom, there were 2,490 entries (7.9%) with previous government experience, and 12,063 cases (38%) of same organization membership. In total there were 12,850 entries coded as politically connected (40.1% of the sample) across 16 years. This may seem like a lot, particularly in the United States, but apparently executives from both countries realize the importance of being connected to politics. These connections need not however imply direct friendships. They just suggest that it is easier to reach a politician if he or she is a member of your organization.

4.1.1.3. Firm-​Level Political Connections When aggregated on a firm level, I use two versions of the POLCON classification, each constructed by summing up individual executives by unique firm ID for each firm in the dataset. The first indicator of connections is constructed if a firm had at least one senior executive that was politically connected, denoted as POLCON mt (POLCON Firm in the results section). However, for this first case I had to be more restrictive and focus only on previous government experience. In other words, I use only the POLCON Gov variable to construct POLCON Firm. If I take into account same organization membership there would be virtually no variation in the US case, as almost every firm had at least one person that was politically connected via same organization membership. In this case, when looking only at executives with previous government experience, about 60% of US firms can be coded as politically connected. This allows enough variation to get comparable estimates as with the individual-​level data. In the United Kingdom I get about 40% of firms that had a politically connected executive with previous government experience. For this reason, I also construct the second POLCON variable for firms, where I look at the total number of all politically connected senior executives in a firm, denoted as POLCON Total . The idea is to distinguish between firms with low and high numbers of connected executives in their boards. In reality all it takes is a single well-​connected executive to land an exclusive contract with the government. However, firms where the entire board is filled with connected executives are clearly better examples of corporate rent-​seekers than those with only a single connected executive. Theoretically they should both be considered as politically connected, but this way I get a better distribution of connected individuals within them, and hence a more precise estimate of the marginal effect of political connections—​the value of an additional connected board member. There is a clear difference between US and UK firms. In the United States the distribution is wider (there are firms with up to 15 connected executives) and the majority of firms have between 1 and 5 connected executives. In the United Kingdom the distribution is narrower, and politically connected firms have in most cases between 1 and 2 connected executives.

Political Networks and Wages  139 Table 4.1  Descriptive statistics and t-​tests for individual network size with respect to political connections Mean

St.Dev.

Min

Max

N

Politically connected =​1

1539.7

1574.8

13

16338

31803

Politically connected =​0

673.7

814.1

10

6185

13621

Politically connected =​1

1473.2

1539.8

7

9792

12692

Politically connected =​0

399.6

701.1

4

7967

18300

United States network size

United Kingdom network size

T-​test connected vs. unconnected

T-​test connected vs. unconnected

(US network size)

(UK network size)

866.1***

1073.6***

(60.8)

(82.7)

4.1.2.  Correlation between Political Connections and Network Size How do political connections correlate with network size? The first interesting descriptive finding is that a person who is politically connected has on average a much larger network. Table 4.1 shows t-​tests for network size according to the indicator of political connections for the two countries.5 In both cases the network of politically connected executives is two to three times the size of politically unconnected executives.6 The firm-​level data for both countries indicates exactly the same conclusion—​executives in politically connected firms have significantly larger networks than executives in nonconnected firms. This adequately sums up the main explanatory variables—​being part of elite networks also implies being better connected, and most likely more influential. This could imply that elite networks have a tendency to be topocratic,7 where it matters who you know in the decision-​making hierarchy.

4.2.  Inequality in Top Executive Incomes Another useful part of the BoardEx database is extensive data on annual individual earnings, including data on salaries, bonuses, stock options, pension contributions, and total direct compensation for each executive. Two outcome

140  Elite Networks variables are created from this data: the logarithm of salary (log salary) for each individual, and the logarithm of total direct compensation (log earnings) as a measure of total annual pretax earnings for each executive, which includes salary, bonuses received (if any), capital gains (if any), and pension contributions. The focus on top executive earnings is contingent on the fact that top incomes (of 0.1% of income earners) are the key driver of income inequalities over the past 40 years, while over 60% of those in the top 0.1% are typically corporate executives. The presumption is therefore that the majority of the increase in top incomes, and hence income inequality, can be explained by the increase of top executive incomes (as was made clear in the previous two chapters). The first useful thing would be to plot this data to see if executive earnings as reported in the database followed the same trends of reported increases in top incomes over the past years. Figures 4.1 and 4.2 show the distribution of individual executive earnings for both countries, and they suggest a similar finding to the majority of income inequality literature: even among top earners, the top 1% of executives have much higher total earnings than the rest. In these examples, the top 1% of highest-​paid executives in the United States had higher total annual earnings than 99% of all other executives combined (keep in mind that the minimum total compensation in the sample was $150,000) (Figure 4.1, lower panel). For salaries the distribution is flatter, but it still suggests that the top 5% of highest-​paid executives in the United States had higher total annual earnings than 95% of all other executives combined (Figure 4.1, upper panel). For the United Kingdom the distributions are almost identical, with the difference being that in the United Kingdom the top 1% of executives had slightly higher relative salaries than in the United States, but slightly lower relative total earnings. Once again, the top 1% of executives have higher total annual earnings than the other 99% of executives combined (Figure 4.2, lower panel), while 5% of executives have higher annual salaries than the other 95% combined (Figure 4.2, upper panel). Next, we look at the trends for log earnings over the observed 16-​year period. Overall, they suggest that executive incomes have continued their rise over the observed decade and a half, and that even after a correction in the trend during the crisis and precrisis years the general trajectory hasn’t changed. In the United States there is a noticeable difference in trends for salaries (Figure 4.3, upper panel), which carry an almost uninterrupted upward trajectory during the entire period, and total earnings (Figure 4.3, lower panel) which underwent a large decline that started two years before the crisis, only to recover after 2009 but growing somewhat slower than before the crisis. The trends for the United Kingdom are slightly different; both salaries and total earnings peaked just before the crisis in 2007, experienced a correction which lasted for two years, and then, ever since 2010, continued to grow at almost the same pace as before the crisis (Figure 4.4). By 2015 both salaries and total earnings in the United Kingdom were higher than

Political Networks and Wages  141 (a)

10

Salary share (density)

8

6

4

2

0 0

20

40

60

80

100

80

100

Percentile of individual earners (b)

20

Total earnings share (density)

16

12

8

4

0 0

20

40

60

Percentile of individual earners

Figure 4.1  Percentile histogram for individual salaries (upper panel) and individual total earnings (lower panel), United States.

142  Elite Networks (a) 10

Salary share (density)

8

6

4

2

0 0

20

40

60

80

100

80

100

Percentile of individual earners

Total earnings share (density)

(b)

15

10

5

0 0

20

40

60

Percentile of individual earners

Figure 4.2  Percentile histogram for individual salaries (upper panel) and individual total earnings (lower panel), United Kingdom.

Political Networks and Wages  143 (a)

6.6

Log salary

6.5

6.4

6.3

6.2

6.1 2000

2005

Year Log salary

(b)

2010

2015

95% C.I.

7.1

Log total earnings

7

6.9

6.8

6.7 2000

2005

2010

2015

Year Log earnings

95% C.I.

Figure 4.3  Annual changes in individual executive salaries (upper panel) and total earnings (lower panel), United States, 2000–​2015. Logarithmic values of salaries and earnings are used.

144  Elite Networks (a) 6.3

Log salary

6.2

6.1

6

5.9 2000

2005

2010

2015

Year Log salary

95% C.I.

(b)

Log total earnings

6.8

6.7

6.6

6.5

6.4 2000

2005

2010

2015

Year Log total earnings

95% C.I.

Figure 4.4  Annual changes in individual executive salaries (upper panel) and total earnings (lower panel), United Kingdom, 2000–​2015. Logarithmic values of salaries and earnings are used.

Political Networks and Wages  145 their precrisis peak. For the United States this was only the case for executive salaries, as total earnings still haven’t bounced back to their precrisis levels. Looking at the trends of top incomes on a firm level, similar conclusions can be inferred for the United States: a steady rise of executive salaries (Figure 4.5, upper panel), and a similar correction of executive total earnings (Figure 4.5, lower panel) followed however by a stronger rebound after the crisis, reaching its precrisis peak in 2015. In the United Kingdom the firm-​level data paints an entirely different picture (Figure 4.6). It suggests that both salaries and total earnings of executives have, at least on a firm level, been declining since 2000. However, since 2009 the trend has picked up and is moving upward again. The firm-​level data in the United Kingdom suggests that average firm-​level inequality has been gradually declining until after the crisis, in spite the fact that individual executive top incomes have been steadily rising in the same period. This fact is in alignment with the UK Gini coefficient and the UK top income share data,8 which suggest stagnating income inequality during the observed 16-​year period. Given that the focus of this book is to examine the impact of political connections on inequality, it would be useful to observe the same trends of increasing executive incomes with respect to individual and firm political connections, as defined in the previous section. Figures 4.7 and 4.8 decompose individual-​level executive earnings by political connections (the main variable POLCON) for both countries, while Figures 4.9 and 4.10 decompose firm-​ level executive earnings by political connections (again POLCON) for both countries. In all four cases the conclusion is identical: politically connected individuals and firms (red lines, denoted as 1) have had higher salaries and total earnings than unconnected individuals and firms (blue lines, denoted 0) during the entire observed period. In some cases, even the growth of earnings for connected executives was higher. For example, the trend of total executive earnings in the United States (Figure 4.7, lower panel) clearly shows that the growth of earnings for connected executives was faster than for nonconnected executives pre-​2005 but also since 2013. However according to only this evidence it is inconclusive whether political connections actually caused the increase in top incomes. It seems instead that the difference in top earnings by levels of political connections is a time-​invariant characteristic of US and UK labor markets. Other variations of this relationship with the POLCON variable unpacked by origin of connections (previous political experience, POLCON Gov or membership in the same organization POLCON Org ) all confirm the same story: no matter how we define political connections, being connected implies a higher level of executive salaries and total earnings.9 These results are presented in greater detail in Tables 4.3 and 4.7 and they suggest that in the United States membership

146  Elite Networks (a)

8.4

Log Salary

8.2

8

7.8

7.6

7.4 2000

2005

2010

2015

Year Log salary

Log Total earnings

(b)

95% C.I.

8.6

8.4

8.2

8 2000

2005

Year

Log earnings

2010

2015

95% C.I.

Figure 4.5  Annual changes in firm-​level executive salaries (upper panel) and total earnings (lower panel), United States, 2000–​2015. Logarithmic values of salaries and earnings are used.

Political Networks and Wages  147 (a)

7

Log Salary

6.8

6.6

6.4

6.2 2000

2005

2010

2015

Year Log salary

95% C.I.

(b) 7.4

Log Total earnings

7.2

7

6.8

6.6 2000

2005

2010

2015

Year Log earnings

95% C.I.

Figure 4.6  Annual changes in firm-​level executive salaries (upper panel) and total earnings (lower panel), United Kingdom, 2000–​2015. Logarithmic values of salaries and earnings are used.

148  Elite Networks (a)

6.8

Log salary

6.6

6.4

6.2

6 2000

2005

2010

2015

Year 0

1

95% C.I.

(b)

Log total earnings

7.2

7

6.8

6.6 2000

2005

2010

2015

Year 0

1

95% C.I.

Figure 4.7  Annual changes in individual executive salaries (upper panel) and total earnings (lower panel), decomposed by political connections, United States, 2000–​ 2015. Logarithmic values of salaries and earnings are used.

Political Networks and Wages  149 (a) 6.4

Log salary

6.3 6.2 6.1 6 5.9 2000

2005

2010

2015

Year 0

Log total earnings

(b)

1

95% C.I.

7

6.8

6.6

6.4 2000

2005

2010

2015

Year 0

1

95% C.I.

Figure 4.8  Annual changes in individual executive salaries (upper panel) and total earnings (lower panel), decomposed by political connections, United Kingdom, 2000–​2015. Logarithmic values of salaries and earnings are used.

150  Elite Networks (a)

8.5

Log Salary

8

7.5

7 2000

2005

2010

2015

Year 0

Log Total earnings

(b)

1

95% C.I.

9

8.5

8

7.5 2000

2005

2010

2015

Year 0

1

95% C.I.

Figure 4.9  Annual changes in firm-​level executive salaries (upper panel) and total earnings (lower panel), decomposed by political connections, United States, 2000–​ 2015. Logarithmic values of salaries and earnings are used.

Political Networks and Wages  151 (a) 7.5

Log Salary

7

6.5

6 2000

2005

2010

2015

Year 0 (b)

1

95% C.I.

8

Log Total earnings

7.5

7

6.5

6 2000

2005

2010

2015

Year 0

1

95% C.I.

Figure 4.10  Annual changes in firm-​level executive salaries (upper panel) and total earnings (lower panel), decomposed by political connections, United Kingdom, 2000–​2015. Logarithmic values of salaries and earnings are used.

152  Elite Networks in the same organization is a more important driver of the earnings difference between connected and unconnected executives, while in the United Kingdom previous government experience drives a bigger wedge between connected and unconnected executive earnings.

4.3.  Empirical Strategy and Results It is difficult to impose a causal relationship between political connections, network size, and executive earnings. First of all, a decision of a firm or individual to become connected and join an elite network is highly endogenous, meaning that connected individuals and firms self-​ select into the treatment group. Second, some firms are simply large in size and employ a lot of people in their local communities, which makes them more likely to capture the attention of politicians, regardless of whether they really benefit from any connections. And third, there could be a number of factors that affect both individual executive earnings and someone’s incentive to join an elite network. For example, individual ability or competence (or even likability) makes someone more likely to earn a higher salary and at the same time more likely to engage with a larger network of other individuals. For firms, the decision on earnings and connections can be related to their relative stage in the life cycle. Firms in the later stages of the life cycle are less responsive to disruptive technologies, which forces them to turn to political support and capturing rents in order to survive. In this case the position in the phase of the life cycle will determine the assignment of firms into connected and nonconnected and will simultaneously also affect the earnings they pay to their executives. Using fixed effects on a panel dataset for a period of 16 years helps resolve some of these difficulties. Fixed effects are useful in controlling for any time-​ invariant factors that cannot be observed (such as executive ability) or are difficult to measure precisely (firms that capture political attention; stage of the firm life cycle) but can be a source of endogeneity and between-​unit differences. When observing the effect of changes in political connections or network size on the changes in earnings over time all such time-​invariant differences get canceled out allowing us to estimate a clear within effect. In this particular case I use firm-​level fixed effects to estimate the effect of being politically connected vs. being politically unconnected within firms over time. I therefore answer the question of whether it pays off for an executive within a firm to be politically connected. The logic behind including time fixed effects is similar—​time fixed effects tend to capture all time-​varying variables and trends that affect outcomes for all units in the same way. For example, in the figures above it was obvious that earnings

Political Networks and Wages  153 growth was interrupted during the financial crisis. Including time fixed effects controls for this as the crisis was an event that had an impact on all units in the same way. The following regression is estimated for individual-​level data for both countries.

logWit = α m + γ t + τ1 POLCON it + τ 2 d ( g )it + Xϑ + ηit

(4.1)

The outcome variable logWit indicates either log salary or log total earnings for individual i in year t. The two main explanatory variables denote the indicator of political connections for an individual, POLCON it , and the size of his or her network, or the degree of the network, d ( g )it , where g represents the total network. These two estimates represent the wage premium as suggested by equation 4.1. α m is the firm fixed effect and γ t is the year fixed effect. Finally, Xϑ represents a set of covariates that include education level, age, gender, total experience (time in current firm plus time in other firms), total number of boards the person sits on, bonus ratio (size of bonus relative to total earnings), and equity ratio (size of equity-​based compensation relative to total earnings). The same equation is estimated for firm-​level data for both countries, using the same indicators and covariates but aggregated on a firm (m) rather than an individual level. The only difference is that two versions of the POLCON variable are used, an indicator version POLCON mt , and a continuous version POLCON Total.

log Wmt = α m + γ t + π1 POLCON total ,mt + π 2 d ( g )mt + Xϑ + ηmt

(4.2)

The validity of the fixed effects approach rests on the assumption that any unobservable factors related to the treatment are indeed fixed over time. The estimation tends to control for a number of factors that could affect executive earnings on both an individual and firm level, plus it tends to cancel out unobservable factors that remains fixed over time, but it can still be vulnerable to any unobservable factor that does vary across time and is not controlled out by the time fixed effect.10 In order to at least partially address these concerns on the firm level I use industry-​level averages of political connections as firm-​level instruments, a strategy suggested by Angrist and Krueger to resolve measurement error issues,11 and applied by others to control for firm-​level unobservables.12 This strategy suggests taking average levels of connections (when defined as a continuous variable, POLCON Total ) across each industry and use it as an instrumental variable (IV) for firm-​level connections within that industry. The assumption is that any variation across industries is not likely to be driven by unobserved firm-​specific factors, but rather by industry-​level characteristics that are likely to be exogenous to firm-​level choices over its wage structure. In

154  Elite Networks other words, all firm-​level unobservables that are likely to be correlated with its executives’ earnings (and that vary over time) are not affecting the outcome if we use industry-​level connections. Given that the IV is an industry-​ level variable the variation only remains on an industry-​level, rather than firm-​level. Simultaneously, this is the biggest disadvantage of the proposed IV approach. The first stage is satisfied given that firm-​level connections are correlated to industry-​level connections, as any firm-​level influence will depend on its industry-​specifics (such as the regulatory status of that industry, its strategic importance to politics, sensitivity to foreign shocks, etc.). The 2SLS estimation procedure is used, where equation 4.3 estimates the first stage relationship while equation 4.4 estimates the second stage relationship:

POLCON POLCON Total , mt = α + ς INDTotal , mt + Xϑ + ε mt (4.3)



 Total ,mt + Xϑ + ε (4.4) log Wmt = α + ρPOLCON mt

POLCON Where INDTotal , mt is the industry-​level average of political connections for each firm m in year t. All the other variables are the same as in equation 4.2.

4.3.1.  Results: United States Table 4.2 shows the results for the effect of political connections and network size on individual-​level executive earnings in the United States. The first two columns show results for log values of the outcome variables: log total earnings and log salary, while the final two columns show results for absolute levels of executive total earnings and salaries so as to get a better idea of the monetary value of the effect. The results show a strong positive effect of both political connections and network size on individual executive total earnings and salaries. The interpretation of the main effect is focused on the difference between individuals within firms. It helps us answer the question: does it pay off for an executive within a firm to be politically connected? Being a politically connected executive results in 12.6% total higher individual earnings,13 and about 12.8% higher salaries than nonconnected executives. For an average executive compensation of around $1.18 million per year, this corresponds to an annual pay rise of about $150,000. The absolute value of the effect (estimated in columns 3 and 4) is similar. A switch from a nonconnected to a connected individual within a firm increases total earnings by almost $154,000, while it increases salaries by

Political Networks and Wages  155 Table 4.2  Individual level executive earnings and political connections, United States. Covariates include the following: education, experience, membership on other boards, bonus ratio, equity ratio, gender, age. T-​value shown in parentheses. *** denotes significance at 0.1% (1)

(2)

(3)

(4)

Log total earnings

Log salary

Total earnings

Salary

0.119***

0.121***

153.9***

73.16***

(23.8)

(31.02)

(9.62)

(20.9)

0.017***

0.016***

30.51***

11.36***

(16.9)

(16.0)

(6.07)

(10.34)

Covariates

YES

YES

YES

YES

Firm fixed effects

YES

YES

YES

YES

Time fixed effects

YES

YES

YES

YES

Observations

44071

44071

44071

44071

R squared

0.653

0.577

0.455

0.506

POLCON Network size

$73,160. The substantial size of the estimated effect in each case suggests that it certainly does pays off to be politically connected within publicly listed firms in the United States. The network size effect also has to be taken into account, even though it is more modest. The variable network size was divided by 1,000 for easier interpretation of the coefficients. The results from Table 4.2 suggest that for every increase in an individual’s network size by 1,000 people (which is two-​thirds of the standard deviation increase of 1,435), salaries and total compensation go up by around 1.7%. In absolute terms this corresponds to $30,500 higher total compensations and $11,360 higher salaries. Given that it isn’t likely for executives to expand their networks by 1,000 people each year, the network size effect is even more modest. It is however suggestive that better connected individuals do carry a greater weight in firms’ reward structures. It just happens that political connections carry an even greater one. The main driver of the rent-​seeking premium are therefore direct political connections. The size of this premium for individual executives is about $70,000 in salaries, and about $150,000 in total earnings. The results reported in Table 4.2 only look at the total value of the POLCON variable. Table 4.3 unpacks the POLCON by the origin of an executive’s connection—​is it due to his or her previous experience in government, or is it

156  Elite Networks Table 4.3  Individual level executive earnings and political connections by origin of connection, United States. Covariates same as in Table 4.2. T-​value shown in parentheses. *** denotes significance at 0.1%

POLCON

(1)

(2)

(3)

(4)

(5)

(6)

Log total earnings

Log total earnings

Log total earnings

Log salary

Log salary

Log salary

0.119***

0.121***

(23.8)

(30.25)

POLCON_​Gov

0.059***

0.054***

(9.83)

(10.8)

POLCON_​Org Network size

0.096***

0.101***

(23.9)

(33.67)

0.017***

0.023***

0.019***

0.016*** 0.022***

0.018***

(16.9)

(22.9)

(19.1)

(16.1)

(21.8)

(17.6)

YES

YES

YES

YES

YES

YES

Firm fixed effects YES

YES

YES

YES

YES

YES

Time fixed effects

YES

YES

YES

YES

YES

YES

Observations

44071

44071

44071

44071

44071

44071

R squared

0.653

0.648

0.651

0.577

0.569

0.575

Covariates

due to membership in the same organization as the relevant politician? All three cases are reported, where columns (1) and (4) merely repeat the same findings as columns (1) and (2) in Table 4.2. In the US case it is clear that the same organization membership effect is stronger than the previous experience in government effect. In particular, a political connection of an executive resulting from previous government experience increases total compensation by 6.2% and salaries by 5.6%. On the other hand, a political connection of an executive resulting from membership in the same organization as a politician increases total compensation by 10.1% and total salaries by 10.6%, which is closer to the overall POLCON effect. Clearly both of these forms of connections increase executive earnings, however, at least in the US case, being a member of the same social group as a politician is more important (and more rewarding) than being connected to them through previous jobs.

Political Networks and Wages  157 The same disentangled effect is estimated in absolute terms (not shown here), and even though it implies the same conclusion for salaries—​$60,000 higher salary for the organization effect and $40,000 for the experience effect—​it suggests the opposite for total compensation. Total earnings increase by $165,000 for an executive with previous government experience, and by only half as much for an executive who is a member of the same organization as a politician. The network effect in each case is similar to the one estimated in Table 4.2. Table 4.4 presents the estimates of the same relationship but on a firm level, where all the main variables and covariates are aggregated from the individual-​ level dataset. The table does not show results for absolute level earnings any more, but instead shows versions with and without time fixed effects. Many versions were estimated using different combinations of the covariates and interaction terms, and different versions of the outcome variables, but the results remain robust in each case. They yield an almost identical conclusion to the individual-​ level estimates: political connections on a firm level carry a strong, significant, and positive effect on top executive earnings. The only difference is that the effect is more modest on a firm level, which is expected. Top earnings across all executives in the board should even out and produce a smaller effect than when looking only at individuals. A politically Table 4.4  Firm-​level executive earnings and political connections, United States. Same covariates as in Table 4.2 but on a firm level. T-​value shown in parentheses. *** denotes significance at 0.1% (1)

(2)

(3)

(4)

Log total earnings

Log total earnings

Log salary

Log salary

0.027***

0.025***

0.019***

0.018**

(3.87)

(3.68)

(3.55)

(3.42)

0.027***

0.023***

0.022***

0.018***

(15.97)

(13.74)

(16.82)

(14.47)

Covariates

YES

YES

YES

YES

Firm fixed effects

YES

YES

YES

YES

Time fixed effects

NO

YES

NO

YES

Observations

12810

12810

12809

12809

R squared

0.462

0.509

0.351

0.421

POLCON Firm Firm network size

158  Elite Networks connected firm compared to a politically unconnected firm pays its executives between 2.5% and 2.7% more in total earnings, and between 1.8% and 2% more in salaries. Although this seems modest in absolute terms it translates to about $30,000 higher annual earnings paid to executives in connected firms. The network effect is similar as before, between 2.3% and 2.7% higher total earnings, and around 2% higher salaries for every increase in network size by 1,000 people. Finally, the IV estimates on a firm level support the earlier conclusions. They also indicate that firm-​level political connections and network size exhibit a large and positive effect on top executive earnings in the United States. In Table 4.5 the first two columns report the effect of using political connections as a continuous variable, counting the total number of politically connected board members for each firm. The final two columns use the indicator value of firm connections, as before. The IV estimates, correcting for potential measurement errors by using industry-​level connections as an IV, produce much higher effects than reported in Tables 4.2 and 4.4.14 According to the first two columns, the more politically connected directors a firm had, the greater their earnings. In particular by adding only a single director that is politically connected to a firm’s board Table 4.5  IV estimates of firm-​level executive earnings and political connections, United States. Same covariates as in Table 4.2 but on a firm level. T-​value shown in parentheses. *** denotes significance at 0.1%

POLCON Total

(1)

(2)

(3)

(4)

Log total earnings

Log salary

Log total earnings

Log salary

0.221***

0.165***

(14.02)

(12.54) 0.842***

0.628***

(13.67)

(12.18)

POLCON Firm Firm network size

0.018***

0.016***

0.025***

0.021***

(13.77)

(14.59)

(25.46)

(25.63)

2.655***

2.654***

0.696***

0.696***

(30.72)

(30.7)

(25.7)

(25.67)

Covariates

YES

YES

YES

YES

Observations

12810

12809

12810

12809

R squared

0.415

0.35

0.362

0.296

First stage

Political Networks and Wages  159 (which is a bit smaller than a one standard deviation increase of 1.5) yields a 16.5% higher salary and 22% higher total executive earnings for that firm. The indicator value estimates of political connections of a firm are even larger. They suggest that by switching from an unconnected to a connected firm (by hiring a director with political connections), total executive earnings in that firm will go up by 132%, while salaries will go up by 87%. In both cases this is an enormous effect, much larger than the very modest ones reported in Table 4.4. The network size effect remains roughly the same however, between 1.6% and 2.5% for salaries and earnings. As with the case of individuals, direct political connections seem to be the biggest contributor to the wage premium.

4.3.2.  Results: United Kingdom For the United Kingdom we once again start with individual-​level results shown in Table 4.6. As before, the first two columns show results for log values of total earnings and salaries, while the final two columns show results for absolute values of the outcome variables. Similar to the US individual-​ level results, the UK estimates also suggest a strong positive effect of political connections on individual executive Table 4.6  Individual-​level executive earnings and political connections, United Kingdom. Covariates include the following: education, experience, membership on other boards, bonus ratio, equity ratio, gender, age. T-​value shown in parentheses. *** denotes significance at 0.1% (1)

(2)

(3)

(4)

Log total earnings

Log salary

Total earnings

Salary

0.046***

0.054***

89.39**

69.48*

(7.66)

(10.8)

(2.59)

(2.34)

−0.003

−0.002

2.827

−​10.45

(−​1.3)

(−0.9)

(0.212)

(−0.91)

Covariates

YES

YES

YES

YES

Firm fixed effects

YES

YES

YES

YES

Time fixed effects

YES

YES

YES

YES

Observations

27236

27236

27236

27236

R squared

0.646

0.55

0.155

0.124

POLCON Network size

160  Elite Networks earnings. The effects, however, are smaller than in the United States. A politically connected executive in the United Kingdom gets a 4.7% higher total compensation and a 5.5% higher salary than a nonconnected executive within the same firm. In the UK sample the average executive total compensation was £965,000, meaning that being politically connected can result in a raise of almost £50,000 annually. The absolute numbers suggest an even larger effect, around £90,000 more in total earnings, and almost £70,000 in salaries. What is surprising in the United Kingdom is that the network effect for individuals vanishes once political connections are involved, and even points to an opposite direction (implying that having a larger network hurts an executive’s earnings), however in neither case is it statistically significant. This surprising result needs to be examined more closely by again unpacking the origin of an executive’s political connections. Is it due to one’s strong social network built in clubs, societies, and NGOs, or is it due to one’s previous government experience, where all the important political connections were initially established? Table 4.7 seems to suggest that the latter is the case in the United Kingdom, exactly the opposite to the United States. Previous government experience in the United Kingdom yields a stronger effect than the overall POLCON estimate—​ it increases total compensation by 7.5% and salaries by 8.25% (about £75,000 on average). The organization membership effect is smaller than the overall POLCON estimate—​it increases total compensation by 3.4% and salaries by 4.2% (about £38,000 on average). When looking at the disentangled effect in absolute terms it makes an even stronger case. The previous government experience effect increases total executive earnings by £230,000 annually, and salaries by £135,000 annually. The organization membership effect in absolute terms is not even statistically significant. The wage premium in the United Kingdom is therefore fully explained by political connections derived from previous government experience. These findings could help explain the insignificance of the network effect in the United Kingdom. Clearly the predominant way political connections are formed in the United Kingdom is through direct government experience. The clubs, societies, and various organizations may attract a large membership but these are most likely clustered around the same type of people. In other words, it is less likely in the United Kingdom that a club of corporate executives invites an executive politician or legislator to be a member, than it is in the United States. Jointly these results confirm the intuition from the US case that political connections are important determinants of executive earnings, and that political connections matter far more than the pure size of one’s network. In the United States, political connections formed through memberships in same organizations are the key drivers of higher earnings for executives, while in the

Political Networks and Wages  161 Table 4.7  Individual-​level executive earnings and political connections by origin of connection, United Kingdom. Covariates same as in Table 4.6. T-​value shown in parentheses. *** denotes significance at 0.1%

POLCON

(1)

(2)

(3)

(4)

(5)

(6)

Log total earnings

Log total earnings

Log total earnings

Log salary

Log salary

Log salary

0.046***

0.054***

(7.67)

(10.8)

POLCON_​ Gov

0.072***

0.079***

(7.98)

(9.87)

POLCON_​ Org Network size

0.034***

0.041***

(5.67)

(8.2)

−​0.003

0.002

−0.001

−​0.002

0.004*

-​0.0003

(−​1.3)

(0.98)

(−​0.5)

(−​1.0)

(1.98)

(−​0.15)

Covariates

YES

YES

YES

YES

YES

YES

Firm fixed effects

YES

YES

YES

YES

YES

YES

Time fixed effects

YES

YES

YES

YES

YES

YES

Observations

27236

27236

27236

27236

27236

27236

R squared

0.646

0.646

0.646

0.55

0.55

0.549

United Kingdom previous government experience is more likely to increase an executive’s earnings. Table 4.8 presents the same relationship on a firm-​level. As in the US case it does not report absolute earnings any more, but only different estimates with and without time fixed effects. All possible calculations and combinations produced similar effects yielding the same conclusion as in the US case and in line with the UK individual-​level findings: political connections within a firm yield a positive and significant effect on top executive earnings. What is interesting is that in all four cases reported here (with and without time fixed effects for salaries and total earnings) the effect is the same: around 10%. This is significantly higher than the modest firm-​level estimates in the United States, but in line with the results for individual level estimates in the United Kingdom. Switching from a politically unconnected to a politically connected firm increases both total earnings and

162  Elite Networks Table 4.8  Firm-​level executive earnings and political connections, United Kingdom. Same covariates as in Table 4.6 but on a firm level. T-​value shown in parentheses. *** denotes significance at 0.1% (1)

(2)

(3)

(4)

Log total earnings

Log total earnings

Log salary

Log salary

0.0982***

0.0994***

0.0981***

0.0988***

(8.46)

(8.61)

(8.7)

(8.76)

0.056***

0.055***

0.052***

0.051***

(17.04)

(16.32)

(16.31)

(15.76)

Covariates

YES

YES

YES

YES

Firm fixed effects

YES

YES

YES

YES

Time fixed effects

NO

YES

NO

YES

Observations

19835

19835

19835

19835

R squared

0.234

0.236

0.116

0.116

POLCON Firm Firm network size

salaries of firm executives by an average 10%. In absolute terms this translates to almost £100,000 higher executive annual earnings. The network effect is now statistically significant and points to the right direction, however its effect is still modest compared to the political connection effect. Across all four estimates it stands around 5%, meaning that a 1,000-​ person increase in network size increases executives’ salaries and earnings around 5% annually on average. This may seem like a fairly decent effect size, however we must keep in mind that not many executives can expand their network by a 1,000 people each year. This implies that the overall effect of adding individuals to one’s network is almost negligible. Connections matter more on a firm level than an individual level in the United Kingdom, although the total size of the network effect is still not as impactful as the effect of political connections. The IV estimates for UK firms are more modest than the corresponding estimates in the US case, although they also confirm all the earlier made conclusions. Results are presented in Table 4.9, where the first two columns report the effect when using political connections as a continuous variable, counting the total number of politically connected board members for each firm, while the final two columns use the indicator value of firm connections. Even though they are smaller than the corresponding US IV estimates, they do support the UK firm-​level findings from Table 4.8. According to the first two

Political Networks and Wages  163 Table 4.9  IV estimates of firm-​level executive earnings and political connections, United Kingdom. Same covariates as in Table 4.6 but on a firm level. T-​value shown in parentheses. *** denotes significance at 0.1%

POLCON Total

(1)

(2)

(3)

(4)

Log total earnings

Log salary

Log total earnings

Log salary

0.125**

0.164***

(3.26)

(4.52) 0.250***

0.329***

(3.3)

(4.58)

POLCON Firm Firm network size

0.097***

0.084***

0.11***

0.089***

(27.08)

(24.99)

(40.11)

(38.13)

1.328***

1.328***

0.6638***

0.6638***

(29.2)

(29.2)

(24.35)

(24.35)

Covariates

YES

YES

YES

YES

Observations

19835

19835

19835

19835

R squared

0.523

0.418

0.523

0.417

First stage

columns, adding one politically connected director increases salaries by 16% and total earnings by 12.5%. The estimates using the indicator value of political connections are higher, however much more in line with all the earlier presented results. Switching from a nonconnected to a connected firm increases total executive earnings in a firm by 28%, and executive salaries by an average 39%. This is a large effect indeed, but nowhere near the US levels of 87% and 132%. What explains such results overall? Executives who join firms coming from high-​level government positions and who are members of elite networks tend to be more valued and hence better compensated by their firms. One of the obvious reasons for this, as shown already in Table 4.1, is that politically connected individuals have larger personal networks. In the business world where networking is an extremely important feature of securing new jobs and helping the company succeed, this is a valuable asset that tends to be rewarded. However, the network effect is much more modest than the political connections effect, and for the United Kingdom it is not even statistically significant. It seems that in the United Kingdom social networks matter less than direct links to the government. They still do matter, particularly on a firm level, but their impact in the United States is much stronger and hence more important.

164  Elite Networks

4.4.  The Impact of Political Connections on Executive Salaries in Other Countries The findings for the United States and the United Kingdom are striking. Unfortunately, there is not enough data to fully validate the hypothesis in other countries. Ideally, for each country of interest we would need to have the same database of top executives and politicians which includes their educational background, organization membership, and previous employment to draw links between them. This way we could make inferences on whether political connections are a stronger determinant of top corporate salaries in high-​inequality countries compared to lower-​inequality countries. Although this type of validation is, at the moment, unfeasible, a similar pattern of political connections being an important determinant of top executive earnings can be found in several other countries, at least according to the, albeit limited, literature on this subject. This section surveys the few empirical papers linking political connections to executive compensation. Many research efforts empirically examined the link between political connections and firm performance on a number of countries and different time periods.15 They’ve found that political connections enable firms to mitigate regulatory burdens, prevent market entry for new competitors, increase their value, have a higher probability of being bailed out during a crisis, get better loans, pay less taxes, take on higher leverage, and enjoy greater market power and abnormal market returns. However, these papers focused only on the impact of firm performance when their corporate boards had politically connected executives within them. They did not look into the microlevel data of corporate executives and how their personal connections impact their compensations. In the United Kingdom, this was shown for the second part of the equation: Members of Parliament benefited from their corporate connections after serving their time in office. They were awarded lucrative positions on the boards of companies they used to regulate.16 Although this affected their total wealth and boosted their abnormal earnings, thus adding to overall earnings inequality, the effect is not as large as the impact of corporate executives’ political connections. Most studies that did examine the direct link between political connections and corporate compensation were done in China, another economic powerhouse with a different political system, but similar impact on inequality. All such studies confirm the positive impact of connections on executive compensation, either directly or indirectly. One study finds the effect of political connections on higher earnings to hold for private firms, but not for state-​owned firms. Executives are valuable when they can provide the necessary link to politics and benefit their companies, especially when the owners of private firms are not politically connected themselves. In state-​owned firms this mechanism

Political Networks and Wages  165 is not even necessary given that direct political oversight already exists.17 A later study confirms these findings and also shows how CEO political connections increase both firm performance and their own salaries. This effect, interestingly, is found to be stronger in less-​developed regions, except if a CEO has local political connections, in which case the regional development does not affect the positive impact of connections on salaries.18 A more recent study on a sample of over 3,700 Chinese firms over a period of 12 years finds that politically connected CEOs of private firms—​where a connection is specifically labeled as being part of the Chinese “elite”—​have up to 20% higher salaries than nonconnected ones, and are also much less likely to experience turnover.19 Another recent study confirms this effect in light of anticorruption investigations. An anticorruption campaign reduces the within-​firm pay gap. This is attributed to reducing the agent problem in firms that are more corrupt as a consequence of their political protection.20 Going back to Europe, one study confirms the effect in Austria, and another in the United Kingdom. In Austria, the authors establish a positive causal relationship between political connections of board members of public sector organizations and their executive compensation,21 while in the United Kingdom, on a sample of over 700 nonfinancial companies from 2000 to 2012, the finding is that politically connected CEOs are clearly paid higher total compensations compared to their nonconnected peers.22 A finding with almost identical implications to the one presented in this chapter. Corporate executives are thus well positioned to benefit from political connections. What about owners of capital? This group is typically examined through the impact political connections have on their firms (which is the subject of Chapter 7). However, one useful approximation would be to look at a limited sample of global billionaires and a subjective but useful measure of their political connections from Svejnar and Bagchi, mentioned in the previous section. Their finding confirms that countries with higher levels of inequality have a greater number of politically connected billionaires. For example, in Canada, Netherlands, and Belgium, or in Scandinavian countries, the authors found no billionaires that gained their wealth through political connections over a time span of two decades. In Japan, Germany, France, and Australia they found only one or two, and in the United States and Singapore a handful. On the other hand, in less developed countries where politics plays an essential role in determining large-​scale corporate success, it was hard finding billionaires who weren’t politically connected. Russia is an obvious example with 100% of billionaires (the oligarchs) coded as politically connected. Similar stories are found in Malaysia, Indonesia, Thailand, India, Mexico, Argentina, Colombia, and other countries with higher levels of income inequality than average.

166  Elite Networks What this might imply, however, is reverse causality—​in very unequal countries, wealth-​seeking motives are achieved through political protection and the violence power principle. Political connections therefore do not cause inequality, but are a product of inequality. However, given the historical overview presented in Chapter 2, a more convincing argument is that usurpation of power came first, and inequality followed as a natural consequence. Countries that lack the proper institutional incentives to curb misuse of power are forced to witness both of these phenomena amplify each other over decades. *** This chapter untangled the direct impact of political connections on income inequality. It showed that being a member of an elite network, representing the collusion of interests between in-​office politicians and senior executives of publicly listed companies, affects the earnings of executives that are members of such networks. An elite network is measured directly, by examining the career trajectories of senior corporate executives and where and how they intertwine with politicians. Two main definitions of connections were used: (1) if an executive had direct previous experience in senior government positions, and (2) if an executive is a member of the same social group as the politician (which includes various clubs, societies, professional organizations, church groups, trusts and foundations, charities, NGOs, political parties, etc.). The dataset was made using 16 years of individual-​level and firm-​level data for two countries with the highest share of top 1% earnings in total incomes—​ the United States and the United Kingdom—​to confirm the hypothesis that senior executives within firms do get paid a wage premium if they are politically connected, compared to their nonconnected colleagues. This wage premium is large and quite consistent over time. It is estimated to be around $150,000 of total compensation in the United States and around £90,000 of total compensation in the United Kingdom. Even though the network effect is an important determinant of the wage premium, its main contributors are direct connections between executives and politicians. There is a difference between the United States and United Kingdom—​direct connections to politics via previous government experience matter much more in the United Kingdom than in the United States. In the United States connections via same organization membership matter more in driving the wage premium. This suggests that in order to extract rents in the United Kingdom in the forms of exclusive government contracts one should have some experience in government beforehand. In the United States one is better off in joining a social group with the politician they intend to lobby. The findings of Chapter 4 should add yet another important but thus far overlooked factor to the income inequality debate—​the role of social networks formed on the highest level between executives and politicians. One of the

Political Networks and Wages  167 biggest causes of increasing income inequality over the past 30 years have been rising earnings of the top 1% of income earners. Most of these are executives of publicly listed companies, and their earnings, at least ever since 2000, have been characterized by a wage differential resulting from direct political connections. Elite network membership certainly pays off.

PART II

INSIDE T HE LO G IC OF A N E L IT E N ET WOR K The first part of the book defined the idea of elite networks and presented the main empirical and historical arguments of how concentration of power and consequentially wealth within a narrow elite group widens the distribution of income and wealth in societies. The second part will focus on specific motivations behind each piece of an elite network: office-​holding politicians on one end of the spectrum, and corporations represented by their executives on the other. It will thus directly build on the findings of Chapter 4 and focus only on modern elite networks, between elected politicians and corporate executives. Only briefly, in Chapter 5, while presenting the main theoretical argument, will I provide examples of historical elite networks and how their interactions generated benefits to those within them. Chapter 5 will also present network theory and economic theory justifications of entering into elite networks. It will show why these networks exhibit particularly high levels of betweenness centrality and assortativity (homophily), and why this makes them highly topocratic as opposed to meritocratic, where wealth-​seeking opportunities and access to privileged information are condensed within a narrow group. It will also show the economic cost-​benefit reasoning of when and under which conditions individuals decide to enter such networks. Finally, in order to fully understand why elite networks emerged and persisted, it will revisit the three main principles of how people used to mitigate risk in order to survive—​voluntary exchange, gift exchange, and violence based on power—​helping us better understand the motivation for entry into elite networks. Centuries of entrapment under the violence power principle made incentives for elite networking difficult to subdue, which is why we still have forces of wealth accumulation present today, driving up inequality. Chapter 6 focuses on modern-​day elected politicians. Their benefit from entering into elite network relationships is to preserve their position of power and extract rents. Rents can come in various forms: small favors, nepotistic arrangements, places in corporate boards after leaving office, bribes and kickbacks, and direct usurpation of office for private gains. Firm motivation for

170  Elite Networks entry, presented in Chapter 7, is also centered around extracting rents for the firm, and strengthening the position of the executive within the firm. Their rents are realized in the form of exclusive government procurement contracts or favorable legislation and regulation, for which their executives are rewarded with greater compensation, and can even include conspiracy acts with politicians to take profits from fraudulent deals. This chapter also delivers the crucial theoretical difference between rent-​seeking and customer-​seeking firms. As long as the system is preserved via elite networks to favor the rent-​seekers, we will have adverse socioeconomic outcomes. The remedies for such outcomes are the subject of Part III.

5

The Internal Logic of an Elite Network The richest people in the world look for and build networks, everyone else looks for work. —​Robert T. Kiyosaki

The first chapter introduced the concept of elite networks and defined them as mutually beneficial interactions between political and corporate elites. Top corporate executives form very close friendships or connections with the key decision-​making political leaders. Both parties extract some form of rent for themselves. Firms get rewarded through exclusive government procurement contracts (like Odebrecht’s major infrastructure projects across South America), subsidies, favorable regulation, and legislation (like finance industry lobbying in the early 2000s), or can even take bloated fees for brokering deals (like Goldman Sachs in the 1MDB scandal). The corporate executives who made the connection are rewarded with higher salaries and bonuses, as shown in Chapter 4, and solidify their stature within the company hierarchy. Politicians also benefit from such relationships. In developed democracies they draw greater campaign donations, or are rewarded with lucrative postcareer positions, typically in corporate boards of the firms they used to regulate. A good example is former US Senator Evan Bayh, who earned $3.8m for sitting on four corporate boards, one of which was a bank that received considerable bailout funding for which he voted while a senator.1 In less developed democracies, where institutions are personalized, corruption is high, and accountability low, politicians take direct bribes and other favors, and very often bind their corporate friends in a strong interdependent network making sure that everyone within the network loses if the central hub—​the office-​holding politician—​loses their seat. A network of powerful people has a strong incentive to preserve their central connecting hub. Joining powerful networks is legitimate. People have every right to enter into relationships that can serve or promote their interests. Problems arise when promoting narrow interests produces negative externalities, or when it implies breaking or bending the law. In those cases, the outcomes arising from such relationships retard economic growth, lower economic opportunity, and increase inequality. Very often we see elite networks becoming too powerful, bending the Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0006

172  Elite Networks laws and institutions to their will, and completely neglecting society’s institutional or judicial constraints. When the judiciary is merged to be a part of the political and corporate elite network, the capture is complete. With full political capture of institutions elite networks start living under different rules. The laws of society cease to apply to them. They get so comfortable in misusing their position of power that they manage to justify any behavior, from petty crime to massive corporate scandals, from sexual predation to involuntary murder (such as in traffic accidents). When being part of an elite network, unless your transgression somehow hurts the network, you will avoid punishment. Enabling this behavior constrains the majority of wealth-​seeking opportunities to a narrow group of people, thus exerting strong pressures on inequality. To fully understand this concept, we must present a valid theory of elite networks that includes an explanation of why they emerged and what motivates their formation, from a network analysis perspective, an economic cost-​benefit perspective, and a historical perspective of explaining the principles of risk mitigation when being part of the network.

5.1.  Network Theory of Elites How are elite networks formed? Mostly they arise spontaneously and geographically dispersed around the centers of money and power, leading to multiple different elite networks existing within a country. There is no centralized elite network system, although they do differ with respect to status and level of power, where some elite networks are local, some national, and some global (like elite gatherings at the World Economic Forum in Davos). The point however is not in their organization, but in their systemic spontaneity of formation. Elite networks are social networks, forming between persons of power. Like all social networks, they are characterized by high levels of positive assortativity2 of well-​connected agents. Elite networks are therefore generated by highly connected individuals (high-​degree nodes) who tend to associate and cluster with like-​minded highly connected individuals with similar interests. As emphasized in Chapter 1, this makes any elite network highly topocratic. In a topocratic system average compensation depends on how connected an individual is, and what their position is within a network.3 In other words, it is the complete opposite of a meritocratic system—​it matters more who you know than what you know. An economic system of a country is more likely to be topocratic and hence have higher economic inequality if there exists a large number of individuals who are poorly connected and a very small number of individuals who are highly connected. This is, in fact, what most societies look like.

The Internal Logic of an Elite Network  173 Elite networks represent the most central, highly connected nodes within any topocratic system. Having a powerful politician (where power is determined by the politician’s network centrality) linked to the majority of other important nodes in a network is a necessary condition for it to succeed, whereas having corporate executives linked to each other is a sufficient condition. Consider a group of corporate executives attempting to organize a cartel to achieve dominant market positioning and set high prices. They have an incentive to cooperate and form a network between them, however the success of their rent-​seeking venture will depend on political and regulatory approval, or more precisely, if no regulation prevents them from colluding. The more powerful their political connection, the more likely their venture will be successful. Special interests use lobbying and campaign donations to reach politicians, and this is always conditional on having a direct connection to the right politician. Figure 5.1, showing only a portion of the huge network of over one million individuals and politicians based on the panel analysis in Chapter 4, showcases exactly how this looks. The big gray nodes represent the central hubs, the most highly connected individuals (nodes with the highest degree centrality). These are either politicians or top corporate executives (CEOs). The bigger the size of a node, the more connected they are. The colors of other linked individuals represent different companies and various corporate executives within them (note that not all corporate executives are CEOs, many are board members, presidents, or vice presidents, as explained in the previous chapter). The most interesting nodes are the ones with central positions within a network, connecting several firms and executives at once. The more central a node, the more powerful the person, as he or she may draw favors from many firms and act as an important connector. Notice also the importance of betweenness

Figure 5.1  Clustering of superhubs within elite networks. Each dot represents one person, where the size of the dot is proportional to the number of connections and individual has, while its color represents different firms. Source: BoardEx data.

174  Elite Networks centrality—​the number of times each node can connect other nodes in the network. Individuals can therefore draw influence not only from their size of connections but also from their central position within the network (such as the big gray nodes linked to each cluster in Figure 5.1). The most powerful ones, however, are those in between different clusters, carrying both superior network size and the ability to connect various individuals from distant groups (a good example is the central gray node between the orange, purple, blue, red, and green firms). The existence of such highly connected nodes with high levels of betweenness centrality, particularly in positions of political power, means they possess superior (or have privileged access to) information on potential opportunities4 and will only process this information among their own narrow social group. Betweenness centrality is usually considered the most important factor in the flow of information between nodes, even between the high degree ones. An executive with privileged access to a high-​degree politician (and vice versa) will be the first to benefit from that proximity. This increases informational asymmetry and undermines the market mechanism in the distribution of opportunities, making it more likely that the distribution of wealth and top incomes will also be captured within elite networks.

5.1.1.  The Link between Clustering and Size of Networks To understand this mechanism better we need to examine the link between network size and clustering. Imagine a purely random network with a large number of nodes (N =​10 million). Imagine that the average degree (the number of links each node has) in such a network is 150. This would correspond to each person on average being connected to 150 other people.5 This being the average number suggests that some people are connected to only a few people, while others are connected to thousands (recall the descriptive statistics of Chapter 4). In this very realistic scenario of average social connections in a lower-​medium-​sized country of 10 million people, the probability of any random link forming in this country is 150/​9,999,999 (average degree divided by N-​1), which is 1.5*10-​5, or about 0.000015. However, clustering within social networks tends to be much higher than the probability of any random link formation. Clustering measures the extent to which a network is closely connected, and usually real-​life social networks are very dense and have a high degree of clustered nodes. For example, the average clustering coefficient for citations of papers in economic journals is 0.16 on a sample of over 80,000 nodes.6 This means that there is a probability of 0.16 for two economists to be connected in this network of academic publishing. If publishing

The Internal Logic of an Elite Network  175 were a random network with the same degree as the economic citations network (of 1.7), the probability of a random link forming would be a mere 2.1*10-​5 for the network of same size and connectivity, or about 7,600 times smaller than in the actual network. The clustering coefficient for web pages measuring how linked they are, is estimated at 0.1078, while for actor networks it is estimated to be 0.79.7 Going back to the initial example, the real-​life clustering within a network of 10 million people is likely to be several thousand times greater than the probability of forming links if this were a purely random network. This striking finding is a normal feature of social networks. Social networks, as shown in Figure 5.1, are characterized by high levels of clustering. However, the rate of clustering differs with respect to the degree of a node. Low-​degree nodes (those with low levels of connections) aren’t likely to have high levels of clustering, while better-​connected nodes are more likely to be clustered among those with whom they share opinions or social values, for example. However, the highest-​degree nodes exhibit lower levels of clustering than high-​degree nodes, meaning that the relationship between degree size and level of clustering is concave: low-​degree nodes have low levels of clustering (close to 0; these are most people in a society), some high-​degree nodes exhibit the highest level of clustering (such as any high-​salience interest group located in the two right quadrants of Figure 1.3 in Chapter 1), while the highest-​degree nodes exhibit a lower level of clustering simply because their connections span across multiple fields, groups, or subcultures (agents in sports or in entertainment, corporate CEOs, high-​end venture capitalists, etc.). In addition, the highest-​degree nodes are also the ones with the largest levels of betweenness centrality (as is obvious from Figure 5.1). Taking all this into consideration it is easy to see how social networks, because of their unique patterns of clustering and positive assortativity among well-​ connected nodes, generate inequality and low social mobility: they condense opportunities within highly topocratic networks. An elite network is an extreme example of both of these phenomena of social networks. They are generated by high-​degree nodes who tend to associate with like-​minded high-​degree nodes with similar interests, but will have lower average clustering and much higher betweenness centrality than high-​salience interest groups for example, and hence greater opportunity for the spread of their vast network of influence. Figure 5.2 shows a relationship between clustering levels and degrees of a node for group-​level connections. Each point on Figure 5.2 represents a phase transition point. At point D the degree is low, as is the level of clustering. At point A the degree is slightly higher, as is the level of clustering. The overwhelming majority of people belong somewhere between these two points: widely dispersed and with only a few connections, consisting of family, a few friends, or neighbors.8 Between these two points people start making connections and generating

176  Elite Networks Clustering, C(g) B

C A

D

Degree, d(g)

Figure 5.2  The different phase transitions depicting the relationship between clustering and degrees within group-​level connections.

clusters—​they self-​select into social groups or subcultures, generate connections in schools and in the workplace, join societies, become fans of sports clubs. Some people will be more connected, and have a greater network of friends, some will be less, but in general none of these two points imply significant connectivity. After point A there are two potential paths—​each increases the level of clustering as well as the size of the network, however they result in two different types of networks. After point B is when people start to reach maximum levels of clustering. High-​salience groups tend to form at or approaching point B. Examples include any powerful interest group, positioned in the right-​ hand side quadrants of Figure 1.3 in Chapter 1. These are groups formed between individuals with very similar interests, behaviors, or ideologies who are connected primarily based on network homophily. Examples vary from political parties to labor unions, from country clubs to book clubs, from the NRA to Greenpeace. An extreme version, with a much lower degree would be a sect or a cult—​highly devoted and connected members, but in relatively low numbers (the curve would end up to the left of point A). At point C the level of clustering is higher than before, but the size of the network is larger than on the B-​path. This is characteristic of a small group of highly connected people; basically any person who within his or her career is considered a superstar—​corporate CEOs, bestselling authors, famous academics, or those in a directed path (one-​sided connections, where you are connected to them, but they are not connected to you)—​social media influencers, film stars,

The Internal Logic of an Elite Network  177 rock-​stars, athletes. These people have too many connections to exhibit high-​ level clustering, yet they tend to be connected to many individuals, and hence have very high betweenness centrality. Elite networks are more likely to develop on the C-​path, rather than the B-​ path. B-​path type behavior typically characterizes interest groups who, because of their high levels of clustering and assortativity, tend to fall into the echo chamber trap. This type of social behavior is encouraged in such groups and they have a higher likelihood of forming negative socioeconomic outcomes. This is the key difference between interest groups in the classic sense and elite networks: the level of clustering is lower, degree size is higher, and betweenness centrality much higher for members of elite networks, meaning that their potential influence is greater. Their members are always the most important nodes in any network.

5.1.2.  The Centrality of Elite Network Superhubs In her book Superhubs, Sandra Navidi explains the immense power of financial elite networks. She uses the term “superhub” to describe the central nodes of a network that carry the greatest networking power, and by extension the greatest economic power. She limits her analysis mostly to financial institutions containing some of the most well-​connected and powerful individuals in the world. Superhubs are extremely valuable to their companies as they almost guarantee their dominant market status.9 The CEO superhub of a financial institution is a typical example of an executive from Chapter 4 in terms of their effect on the dispersion of earnings. Their value to the firm is due to their immense network and their political connections, for which they receive lucrative compensation packages. The more powerful the network, the better the information a person is exposed to, which works to solidify their position/​status and their wealth. Every superhub in finance is highly interconnected with every other superhub. Through these connections they share privileged information and access to opportunities, thus solidifying the topocratic nature of their elite networks. An elite network knows no checks and balances. It is not a mechanism that can be regulated because it is not a typical organization. It is a relationship between friends, many of whom are extremely powerful and can be depended on for crucial political or business decisions. Such powerful networks are self-​sustaining—​ being attached to one is a desire of those who wish to emulate the same level of success and power. This persistence makes elite networks even more powerful and influential. As networks grow in size (and by extension power and wealth), each new node wants to be connected to the most powerful (well-​connected) node. This

178  Elite Networks tendency benefits the best connected and most senior nodes, increasing their power even further, making them indispensable to the network (like a pyramid scheme, where the ones on top get the most money). This happens in both business and politics, whenever there is considerable power given to a single individual. A highly connected person, the highest degree node, becomes too powerful and too important. Removing this person would render a cascading effect down the entire network, which is why each node within the network, particularly those closely connected to the central node, has a strong incentive not to allow the person to lose their position (unless, of course, they themselves can replace the central node). This is especially true in corrupt political networks, where every corrupt node is linked to the central decision-​maker—​the elected official—​and has a powerful incentive to keep them in power. They will donate to campaigns, organize fundraisers, or act as direct brokers to engage more voters. Their existence is jeopardized (Maslow’s basic needs of security are endangered), so they fight hard to protect the central node. The logic of clustering and homophily is also true for corporate networks, particularly in finance. This is why companies like McKinsey or Goldman Sachs seek to push their partners or senior executives into high-​ranking corporate and government positions. From those positions the well-​connected individuals can benefit their former employers. All of this is perfectly legal, bribes are unnecessary. Firms obviously do benefit from casting a wide net of their former employees to various other corporate positions. These employees have no obligation to benefit their former companies, nor is this ever explicitly asked from them. The reason they do so is their friendship links, the homophily effect. In business you like to work with people you can trust. If these are your close personal connections, all the better. This is not a negative consequence of some deep systemic fault. It is natural human behavior. We trust our closest friends and connections, and choose to work with them whenever we have the opportunity. During the height of the 2008 panic on financial markets, the three men in charge, Treasury Secretary Henry Paulson, Fed Chairman Ben Bernanke, and New York Fed Chairman Timothy Geithner trusted each other and relied on personal connections from their former jobs—​mostly big bank CEOs (see more details in Chapter 7). It is hard to accuse them of collusion or justify the moral outrage simply because they were asking their friends for advice. This is a natural reaction, particularly during times of panic when quick and important decisions need to be made. You talk to people involved, most of whom are your personal friends, and try to find a solution. The problem, therefore, is not network homophily or clustering, because these are natural occurrences. The problem lies in granting too much decision-​making power to certain positions. Individuals at the helm of these positions are prone

The Internal Logic of an Elite Network  179 to typical human errors and cognitive biases, or are prone to emulating typical human behavior—​in this case, the idea that you ask people you trust what they think should be done. It is normal for executives and politicians to network; this is, after all, the key requirement of their careers. In this networking they get closer to each other and rely on each other. The network bonds only grow stronger over time, and this always benefits the most highly connected individuals, meaning that the distribution of wealth at the top is necessarily getting more and more concentrated. Solving this problem does not necessitate eliminating the motivation for networking, as this is impossible. It implies lowering motivation for elite network membership by lowering the scope of political power.

5.1.3.  Historical Examples of Elite Networks The main focus of the book are modern-​day elite networks, today more closely connected than ever before due to the impact of technology on information flows and the levels of globalization. However, the mechanism of an elite network effect that causes betweenness centrality and assortativity among its members is hardly different today from what it was throughout history. Examples are plentiful; from 17th-​ century French establishment elites building powerful alliances based on friendships and proximity, to 15th-​century Italian financial firms seeking to be connected to politics and the pope in a complex array of competing city-​states, to 18th-​century British nobility and capital owners being closely connected to the Crown. None is more emblematic however than the level of influence reached by the 19th-​century Rothschild family, establishing themselves as the most powerful family in finance and banking of all time, purely due to their impressive and wide-​reaching network of connections to rulers and all other powerful people in imperial Europe. In fact, according to Niall Ferguson’s detailed biographic overview of the Rothschilds, they were not just the biggest global bank that dominated the international bond market, they were also a factor of stability in international finance, without whose support it was impossible to wage wars or make major political decisions (the analogy for today is imagining a merger of the top four biggest US banks, the Federal Reserve, and the IMF into one body).10 It was an elite network where the financiers were arguably more powerful than the rulers. Ferguson doubles down on this claim in his book The Square and the Tower, explaining how the industrial and credit networks built by the Rothschilds were much more effective than the hierarchical imperial orders of 19th-​century Europe. Even when compared to the Saxe-​Coburg-​Gotha noble dynasty, which gave rulers to almost every European Empire at the time, the Rothschilds carried greater political clout.11

180  Elite Networks Unfortunately, not a lot of data exist that can be used to visualize these networks and make inferences about their impact. There are a few exceptions that cannot provide much depth over the inferences but can be helpful in visualizing how elite networks used to look in history, in times of kings and noble families, before they started implicating corporate bosses and elected politicians. I present an example of each, the royal network centered around the 16th-​century Spanish Crown, and a network of connections of a noble family in 18th-​century Scotland. I use the datasets assembled by the Visualizing Historical Networks project of the Harvard Center for History and Economics (2020).12 The first is a visual representation of the privileges given by the Spanish Crown for the colonization of the New World. It is centered around King Charles V and the individuals he granted the rights to settle, populate, and evangelize new territories. The project uncovering this relationship found over 500 privileges given to 250 individuals for colonizing the new territory,13 and drew them on a map of interactions, shown in Figure 5.3. The size of a node in the network depicts the total number of privileges an individual received, while its color depicts the territory over which the privilege was granted (as shown in the legend). The central node on the right part of the graph is Charles V, while on the left is his wife, Isabella of Portugal, who was the regent of Spain due to Charles’s frequent travels throughout Europe, meaning that she was in charge of the kingdom while he was away. Both of them represent important central nodes, granting privileges to individuals close to them. What is interesting is that individuals negotiating between both sovereigns got greater privileges over new territories than those connected only to a single one. Given that King Charles was often absent, leaving his wife as the regent, waiting for him to come back would result in gaining less privileges then if simultaneously courting his wife, Queen Isabella. As emphasized earlier, the betweenness centrality of these nodes was the key feature that made them more powerful and gained them higher privilege. The entire network is an elite network, but its most powerful individuals, apart from the king and queen, were those in between them, the proper elite network superhubs. A good example were the members of the Welser family, a German banking family from Augsburg, who were financiers of Charles V,14 but even in his absence managed to accumulate significant wealth from the colonization of the New World. The main reason for this was their high level of betweenness centrality. The second example shows network connections of the Johnstone family, a noble family in 18th-​century Scotland, drawn based on a book The Inner Life of Empires by the British historian Emma Rothschild.15 A large family of seven brothers and four sisters, with a network spanning across their close friends, like famous philosophers, judges, or other noble families, rose to high stature in

Figure 5.3  A network of privileges granted by the Spanish Crown to colonize new territories. The two monarchs granting the privilege were King of Spain and Holy Roman Emperor Charles V, and his wife Queen of Spain and Holy Roman Empress Isabella of Portugal. The size of the circle depicts the number of privileges over new territories gained by each individual. Source: Cachero, and Rodriguez-​Modrono (2022) and the Center for History and Economics, https://​histe​con.fas.harv​ard.edu/​visu​aliz​ing/​pri​vile​ges/​index.html.

182  Elite Networks British society during the height of the British Empire and the expansion of the Atlantic slave trade. Figure 5.4 shows the intricate web of connections of the family including their friends, their business associates, and even their slaves and servants. The members of the Johnstone family are green, the rest are in red. The links between the nodes are also colored differently, based on different types of relationships. Personal relationships (friendships) are red, professional are green, family are purple, acquaintances are blue, while the relations between slaves and their owners are yellow. The size of the node is proportional to their level of interconnectedness, and hence importance in the network. A few interesting relationships come to mind. First, notice close personal relationships between some of the brothers and the most famous Scottish philosophers of the time, like Adam Smith, David Hume, or Adam Ferguson. William Johnstone, for example, studied with Adam Smith in Edinburgh, and later went on to become one of the richest men in England, marring an English heiress Frances Pulteney (taking her family name, tying him to her estate), and being a Member of Parliament for 36 years (in the 18th century, only the nobility could be Members of Parliament). He owned property all over the world and was engaged in the slave trade. It was, much like the Tammany Hall machine of 19th-​century New York, a merger of politics and business into one. His brother George, a naval officer, also served as a Member of Parliament, was a governor of the British colony of West Florida, and was heavily involved with the British East India Company. John Johnstone was the second-​richest member of the family, earning the most of his fortune also within the East India Company, living for 15 years in India, in charge of the Company’s Persian correspondence. At least four members of the family were all connected or worked at the East India Company, solidifying the elite network relationship between the Crown and its nobility. Both of these networks are typical examples of premodern, nobility-​based elite networks. Highly clustered and centered around the ruler, or with high levels of homophily connected to other powerful and respected men of their time. Political and business functions condensed into a few powerful high-​ degree nodes with particularly large levels of betweenness centrality. Even though political systems changed, the essence of an elite network remained the same: powerful and wealthy individuals connected to other powerful and wealthy individuals; in the premodern age centered around the rulers, today centered around the most important office-​holders. And in each case, the heavily clustered elites condensed the majority of wealth gaining opportunities within their own networks.

Figure 5.4  A network of connections of the Johnstone family, 18th century, Scotland. Source: Rothschild, E. (2012)

184  Elite Networks

5.1.4.  Incentives for Firms and the Impact on Inequality Are all these network theory implications rooted in empirical conclusions of the literature? Certainly. Many empirical findings suggest that politically connected individuals, and by extension their firms, incur excess benefits from their favorable position and proximity to power.16 On the other hand, the literature also suggests that such benefits come at a cost for the connected firms—​ they shift their activities and resources from satisfying customers to lobbying for political favors, they are discouraged from investing in new technologies, and they have no incentive to acquire top talent to keep their business growing. One such research uses a sample of 8,000 firms in 40 developing countries to measure direct costs of political influence.17 It finds that politically favored firms are less likely to invest and innovate (engage in R&D, restructure operations, open new plants or production lines), which as a consequence lowers their productivity and sales growth. Regardless of their privileged position that renders them substantial benefits (such as protection from expropriation, regulatory and tax benefits, or better access to finance), politically connected firms have disincentives to innovate and are therefore worse performers that nonconnected firms.18 As one potential extension to this mechanism proposed in the literature, I classify the politically connected firms as rent-​seekers (or rent-​extractors), and nonconnected firms as customer-​seekers. The distinctions between the two are presented in greater detail in Chapter 7, with the emphasis on how firms benefit from elite network connections, but for now just a few main characteristics. The difference between the two types of firms is their focus of interest: do they reward executives for their political connections or do they reward executives for market innovation? The customer-​ seeking firms have to constantly improve their product space in order to remain present on the market, whereas rent-​seekers have lesser incentives to innovate and compete for customers, as they can count on political rents. Rent-​extraction implies either classic rent-​seeking: protectionism, barriers to new entrants, and privileged status,19 or receiving fixed procurement contracts and subsidies.20 In the case of privileged status and barriers to new entrants this means that rent-​extracting firms get to keep their large market share and therefore still realize revenues through gaining customers, but would not be able to do so without the help of political power. It should be noted that the definitions of innovation and investment here are not confined only to the implementation of new technologies. Innovation and investment imply regular competitive-​based activities of firms: introducing better sales techniques, better positioning on the market, constant improvement of the quality of the product or the service based on customer feedback, and most importantly attracting and hiring top talent to ensure the company’s

The Internal Logic of an Elite Network  185 long-​term success. Lack of such incentives to innovate and invest will shift the hiring decisions in rent-​seeking firms toward executives who are members of elite networks that could benefit the firm. This mechanism can also work in reverse—​the executive who has already self-​selected into an elite network pushes the firm to become a rent-​seeker in order to showcase his or her strength. The connected executive is generating benefits to the firm by securing favorable regulations or exclusive government contracts, and consequentially demanding higher compensation. This extension is important in trying to address the reasons behind top executive income growth. The most basic interpretation of the inequality literature focused on firm-​level analysis is that the rise in top incomes is mostly a within-​firm phenomenon. In other words, the widening disparities in earnings between the executives and the workers were the main driver of inequality. However according to a recent paper21 using extensive microlevel data for the United States, the main driver of inequality seem to be differences between firms, rather than within firms.22 The entire increase of inequality for the bottom 99% has been driven by between-​firm inequality, and almost all of the increase of inequality of the top 1% has also been driven by between-​firm inequality (except for those in the top 0.02% of the income distribution). The authors suggest that a possible explanation could be the clustering of high-​ paid workers in more successful firms and low-​paid in less successful ones.23 The question that remains unanswered is the composition of these clustered high-​paid employees, specifically their connections to politics. Could between-​firm inequality also be driven by rent-​seeking motives of corporate executives? My hypothesis is that the decision to hire workers and thus enable the observed clustering effects is contingent on how a firm survives on the market (i.e., how it realizes its revenues). If it survives by extracting rents on the political market it is indifferent in hiring workers to boost its productivity, but it will hire well-​connected executives to extract rents. Therefore, an obvious distinction between a rent-​seeker and a customer-​seeker is how well-​connected their top executives are to politicians, and whether they are members of close networks with the politicians in power. This is how I define the indicator variable of political connections as used in Chapter 4. A politically connected firm will have top executives with deep links to politics, i.e., CEOs and board members who used to work at top-​level government jobs and who are direct members of the same social networks as the politicians in power. The clustering of highly connected individuals in the biggest firms explains the abovementioned between-​firm effect on the rise in the dispersion of earnings. In general, both the within-​firm effect found in Chapter 4 and the between-​firm effect found in the literature are parts of the larger elite network effect.

186  Elite Networks

5.2.  Economic Theory: Costs and Benefits of Elite Network Membership Thus far I have mostly discussed the advantages of elite network membership, taking it for granted that it represents a stable equilibrium for each agent to join the network and never deviate from it. However, costs of membership certainly exist and may not be low. Costs include reputation, effort, the threat of prison time if the activity is illegal, and an additional cost of losing elections for the politician. In order for an individual agent to join an elite network the benefits of membership have to outweigh the costs. This logic is shown in Figure 5.5, representing the cost-​benefit structure of elite network membership with respect to the total amount of units consumed (think of them as money, goods, and services gained through rent-​seeking) when part of an elite network. It also shows the marginal costs and marginal benefits functions in the lower panel, while the absolute costs and benefit functions are shown in the upper panel. The cost function for each individual agent f (c ) (dashed line) is concave up to the point M , after which it becomes convex. It reflects the idea that costs initially grow at a diminishing rate. Initial effort and reputation are sacrificed to enter the network. These entry costs are high, but once they are realized, further consumption of elite network benefits does not imply great additional effort or threats to reputation, at least until point M , when by assumption most benefits enjoyed by agents are not (yet) illegal. Hence marginal costs (also dashed line) decline at a steady rate until point M , where they reach their lowest value. After further greater consumption of rent-​seeking units, and after agents start getting involved in illegal activities, costs of elite network membership exponentially increase the threat of reputation, particularly under the possibility of a prison sentence. After point H, the costs have gone up too high, and this small margin is the level at which an individual is punished (e.g., by going to prison24). At point H is where the marginal costs significantly exceed marginal benefits. The benefits function f (r ) is convex at first, reflecting the idea that the benefits of elite network membership have to increase exponentially to attract individuals inside (e.g., earning more money to afford a better lifestyle). Once the consumption of units reaches point M , more exposure to the network still increases benefits, albeit at a diminishing rate. At this point the marginal benefits from elite network membership are highest. Once the initial thirst for a different lifestyle has been satisfied, additional units do not increase marginal utility as fast as before. The equilibrium conditions are given in the lower panel of Figure 5.5. An agent decides to join an elite network only when his or her marginal benefits

The Internal Logic of an Elite Network  187 Benefits/costs

H f (r) r>c M f (c) r MC ).

exceed his or her marginal costs of membership (MB > MC ). Because of the specific shapes of the total costs and benefit curves we have two equilibria for which marginal costs equal marginal benefits, E1 and E2 . A decision to join an elite network happens only between those two points. It is interesting to notice that the decision to join an elite network happens even when r < c (costs exceed benefits), at least after point E1. Before point E1 an individual knows that a substantial effort is necessary to become part of a network as there is a limit to how many units he may consume. In other words, an

188  Elite Networks individual is not high enough in the hierarchy to even have an opportunity to join the network, so any costs of effort in trying to get in significantly outweigh the benefits. This may change in a dynamic setting, as agents may expect to climb up the hierarchy once a part of the network. Therefore after point E1, after a certain effort has been invested (going to social gatherings and meeting the “right” people, earning a large sum of money, winning an election, or gaining more experience that enables a person to climb the corporate hierarchy and therefore meet the people with power), marginal benefits start exceeding marginal costs, making it attractive for a person to join the network, even if the total costs temporarily exceed total benefits (e.g., paying a high initial price to cater to a politician, the benefits of which will only be visible later). Soon enough the agent will start gaining marginal utility, for which total benefits will exceed total costs, r > c happening after point M.25 This situation will last only until point E2 , after which, despite still receiving benefits from network membership, the marginal costs start to exceed marginal benefits. In colloquial terms, the agent is “in too deep,” as there is an increasing likelihood that the activities of the group might be uncovered and that punishment will occur. Finally, after point H, costs again outweigh the benefits as the agent is punished for illegal activities. At this point membership in an elite network stops. An extension to this setting would be to model the within-​ network relationships of agents. Not all agents are the same, nor are they valued the same way in the network. Some carry more influence than others (although this obviously changes over time), so to adjust for this I would have to draw different cost-​ benefit curves for different agents depending on their network centrality (how important they are in the system). Even though I assume, for simplicity, that all agents that engage in the network are equally important as each of them is driven by an incentive to pursue their interest, it is safe to assume that the theory holds for both high-​degree and low-​degree nodes. In other words, each node will only join if his or her expected benefits are high enough. For example, when high degree nodes reach their central position in the network, their costs are certainly lower and benefits much higher, but in order to build a position of the central node it takes time and a lot of effort, hence the costs of getting there are still significant. Before they reach their level of centrality and can extract the maximum benefits from point M onward, they have to go through the r < c period, when their costs are high and benefits not immediately obvious. Every person with even the slightest experience in networking knows this. You make connections without immediately being aware how they might benefit you in the long run. Many of them never do, but as you amass a greater network, get connected to more people and extort a significant effort in doing so, over time, if you persist in your efforts, you find yourself as a central node

The Internal Logic of an Elite Network  189 with greater abilities to connect people than before, and greater opportunities to benefit.

5.3.  Joining Elite Networks to Mitigate Risk and Achieve Status: The Three Principles The previous sections explained the network theory and economic theory justifications behind elite network formation, and why this would affect inequality. To close the circle behind explaining incentives for membership we need a historical context of how humans formed mechanisms of mitigating risk and achieving status. As mentioned in the Introduction and throughout Chapter 2, we can boil it down to three principles of how people secure their existence and status, both today and throughout history. Securing one’s existence is at the very bottom of Maslow’s pyramid of needs. It is just above the basic physiological needs (food, water, sleep, reproduction), making it a very important motivational factor for each person. Securing existence means mitigating risk. In any point in history, the risk of being subject to some form of violence was high. Opportunities for making money and achieving any form of above-​subsistence living standards were scarcely available to anyone outside the top 10% of the wealthiest population, until basically the second part of the 20th century. During the Middle Ages for example, one way to escape Malthusian subsistence was to go to war, like the Crusades, and rob and pillage your way to some wealth. Still, most of what was stolen would end up with the nobles under whose flag you fought. To become part of an elite network was incredibly difficult for anyone not already being born to a family of means. But even these elite networks had to have had an origin. Just like the rulers of the first proto-​states, as described in Chapter 2, the rise of the nobility, the second part the network, had to have been rooted in some motivational factor, besides just being in proximity to the ruler. To understand what it was, let us build on the three principles of satisfying Maslow’s needs, as described in the Introduction: the voluntary exchange principle, the gift exchange principle, and the violence power principle. The first principle, voluntary exchange, implies that people satisfy their existential needs by exchanging goods and services on a voluntary basis, without coercion, and based on consensual institutions and rules agreed on by all actors (like exchange of money or other means, legal contracts, social contracts, etc.). In the modern age such institutions are purposely designed and provided as public goods (protection of human rights, legal system, monetary system, market regulations), but throughout history the institutions necessary to promote voluntary exchange arose spontaneously. The motivation to exchange surpluses with

190  Elite Networks other members of the community originated with the Agricultural Revolution. As described in Chapter 2, the first proto-​states emerged as a system to protect surpluses from roving bandits, replacing them with stationary bandits and the first hierarchical power structures. However, the hierarchical structures gave little institutional support to promote free exchange, which is why economic growth was severely constrained, but the impulse to trade and exchange goods voluntarily was nevertheless present in all societies through all ages. Human existence was always contingent on various forms of organized societies, which determine rules of how collective decisions are being made. From the primitive tribal systems to modern states there was a search for an efficient and functional form of political order as a medium of making the right collective decisions: from autocracies to various forms of democratic orders whose institutions become more and more inclusive. History teaches us that the rise of liberal and democratic institutions, ascending in parallel with the adoption of the voluntary exchange principle, enabled the greatest progress and growth of wealth in human history, the largest amount of the world’s population escaping poverty, but also the largest demographic expansion in history of mankind. We can arguably claim that societal development depended on how the voluntary exchange principle was adopted. When it was protected and encouraged by spontaneous social norms, societies flourished. When it was subverted by autocratic rulers or periods of war and uncertainty, societies regressed. During the long period of gradual adaptation of democratic norms and institutions in the post–​Industrial Revolution era, the free exchange principle was finally getting supported by proper institutional incentives and protection. It took a while for those institutions to yield their full effect (as is described in greater detail in Chapter 8), but once they did, it allowed societies to finally reap maximum benefits from the principle of free exchange. Governments were switching from despotic autocracies to inclusive democracies, they were limiting the motivation to engage in coercion and violence, they made sure that property rights were respected, that human rights were protected, and that people could feel safe. The superiority of the free exchange principle is also visible through much more efficient usage of human resources. Having the freedom of expression encouraged social mobility and inspired talented individuals to fulfill their potential, rendering massive positive impacts on innovation in business, science, and art, much more so than ever before in history. The voluntary exchange principle was therefore an omnipresent motivation for securing one’s existence, but it was not until the 20th century, enabled by the historical factors responsible for the rise of inclusive democratic capitalism, and protected institutionally via all the necessary public goods, that the principle yielded maximum benefits for societies. People could finally resort to voluntary exchange to secure their existence without being afraid of coercion, or without

The Internal Logic of an Elite Network  191 having to be drawn to elite networks. Elite networks persisted as well, preserved by the forces of wealth accumulation, and still promoting unequal distributions of wealth and income, but they no longer subvert voluntary exchange for all individuals. The second principle, gift exchange, has also been a persistent factor throughout human history. The gift exchange culture was even found to be a predecessor of voluntary exchange, where people in close knit communities exchanged goods as gifts. During ancient times, throughout the Middle Ages, and even during the Enlightenment period, the gift exchange principle was adopted through the patron–​client relationship, where wealthy individuals would fund artists or explorers. Today, in the modern age, this principle is maintained in the form of philanthropy and through various charities and foundations that are often indispensable in financing universities, think thanks, and the arts; fighting diseases; or promoting other cultural and social issues. Such organizations are not usually present on the market, nor do they necessarily get funded by governments. They depend on gift exchange, where the donor receives benefits in such an exchange through improved reputation, prestige, or simply the feel-​ good factor. Another form of the gift exchange principle is inheritance. Sometimes inheritance fortunes can be used to promote the aforementioned philanthropic activities, but much more often inheritance has been a transaction limited to a family unit, particularly since inheritance fortunes were not large for the vast majority of the population. Inheritance is a good example of an altruistic exchange, not driven by market mechanisms, and not imposed by governments (it is regulated by governments, but not imposed). It serves the purpose of solidarity, where parents and children take care of each other in various stages of life, and parents have an internal motivation to leave something for their children after their passing. It is a natural tendency that requires no outside enforcement.26 The problem is when inheritance itself becomes the source of inequality, in instances where a large fortune is given to the benefactor. The main issue is with regard to the origin of the inherited wealth. If it was created based on the free exchange principle (e.g., fortune created by an entrepreneur whose wealth was attained without coercion and based on voluntary market exchange), where the benefactor must preserve the wealth via the same principle, it can be considered just.27 If, on the other hand, it was created based on proximity to political power or through violence, then the inheritance itself is unjust and deepens inequality. This brings us to the third principle of securing one’s existence: the violence power principle, the key motivation for elite network membership and usurpation of power. This was a principle always applied by a minority of individuals who had the legitimacy of using coercion over others. The usage of violence, robbery and pillaging of property, or coercion of people into slaves was a

192  Elite Networks dominant pattern of behavior throughout human history until the modern age. It was a successful risk-​mitigating strategy. Being in power or in proximity to power was in every society in history a much better strategy than being subjected to the whims of those with power. Using violence and coercion as methods of gaining wealth and securing one’s existence was a much more successful strategy than resorting to voluntary or gift exchange. For one, the costs and risks of using violence were lower than the anticipated benefits. There was no universal and morally acceptable punishment of violent transgressors, provided that these transgressors held any kind of power. There were no human rights that protected the population from being coerced at will by their rulers. The risk of theft or pillaging was also low given that the population being coerced had no way to successfully defend itself against armies or guards without weapons. Societies where violence legitimized by power is seldom punished yielded a minority group that used violence to ensure its existence. This was the origin of an elite network: everyone with a legitimate claim to power (kinds, nobles, the clergy) used coercion to secure their existence and ultimately status. It is no wonder that societal elites attempted to preserve their position and make it hereditary and permanent. It was the most optimal way of securing one’s existence and did not include the uncertainty of hard labor, or having the products of that labor violently destroyed. There was little way for anyone to disrupt an elite network from the bottom-​up. The only way for someone to lose privilege was to have a falling out with the ruler (losing the benefits given by proximity to power), be deposed by a mutiny of other nobles or pretenders to the throne, or to be deposed by an outside threat in the form of an enemy army (like the Huns for the Roman Empire, Mongols for China and Asia, Vikings throughout Europe, etc.). In short, the usual way of ending someone’s position in an elite network was death. There certainly was a risk of elite network membership, but that risk was tied to the limits of sovereign power (both internal and external). This molded political and economic institutions in societies with a clear goal of strengthening extractive political power. Religion was used as an apt justification of the divine right to hold power, which is why the clergy class also enjoyed elite network privileges. This is how societies were organized for several millennia, ever since the dawn of ancient civilizations, up until the modern age, with very seldom and secluded exceptions. Slavery societies, feudalism, or modern-​day despotism all had the same internal structure and motivation. Slavery was a normal condition, widely accepted, as was the idea that any ruler had the perfect legitimacy to murder, pillage, or otherwise coerce anyone else in society. An obvious consequence to this uncertainty and elite power capture, encapsulated by the violence power principle, were Hobbesian living standards,

The Internal Logic of an Elite Network  193 a long-​term Malthusian trap, and the perpetuation of the forces of wealth accumulation yielding constant upward pressure on inequality. Societies did manage to gradually change. It took several critical junctures in human history for things to start moving in a different direction. One of these was the Black Death plague in the 14th century, followed by the transitionary period of the Renaissance in the 15th and 16th centuries and the Ages of Discovery. Another was the invention of the printing press in the 15th century, which enabled a faster spread of new ideas via books and for the first time allowed people in Europe to express themselves differently and pursue their happiness outside the confounds of the traditional Malthusian trap. The stage was set for the Enlightenment era. This embraced the Church as well, with the Reformation movement having the goal to bring the Christian doctrine closer to everyday life, thus endangering the monopoly of the Catholic Church, itself operating under the violent power principle throughout the Middle Ages (Crusades, inquisition, conquering territories by the papacy, etc.) This further encouraged the competition of religious ideas and gave rise to the free exchange principle as an alternative to the violence power principle. The emphasis on individual freedom started replacing the dominant coercive collectivism, which signaled a prelude to the Industrial Revolution.28 Despite these initial major disruptions, from the Enlightenment, New World discoveries, all the way to the Industrial Revolution, every state was still guided by the violence power principle well until the 20th century. But the change in rhetoric and the perception of individual freedom that motivated 18th-​and 19th-​century revolutions spurred century-​long philosophical and political economic debates on the limits of state power that carry on until today. The voluntary exchange principle finally won over in the second half of the 20th century, becoming fully adopted in at least one part of the world, the one embracing democratic capitalism with a reinvented role of the state in providing public goods, and no longer relying on violence and coercion (the other part of the world started adapting the same some 40 years later). The violence power principle was subdued but was never completely eliminated from human behavior. It is difficult to completely disrupt a principle that has guided and motivated human behavior for centuries, even millennia, always latently present within societies, lurking for an opportunity to manifest itself. Whenever a society starts to disintegrate during a crisis, the latent violence power principle starts to resurface. Calls for autocratic (typically populist, one-​size-​fits-​all) solutions are particularly vocal in times of crisis, when many people feel under pressure and experience hopelessness, thus turning to simple solutions. The most recent example was the aftermath of the European sovereign debt crisis, which saw a dangerous move toward authoritarianism in many European countries, typically as a result of internal demand of the population.

194  Elite Networks Full re-​emergence of the violence power principle rooted in predatory behavior, lawlessness and destruction thankfully did not materialize, but its motivation is omnipresent. It will take a lot of time living under peaceful and prosperous conditions, driven by rapid technological progress to fully subdue this latent driver of human behavior. The current rate of societal progress, empowering and enriching more people than ever before, and at a faster pace than ever before, is the main reason for optimism that societies could finally break away from the violence power principle, forever limiting it to the fringes of society. In order to help them achieve this, a proper set of reforms aimed at lowering political power and empowering the citizens, described in detail in the final chapter, should be implemented. Only after eliminating the violence power incentives can we hope to eliminate the motivation of the elites to capture and misuse power and solve the issue of unfair distribution of incomes. Elite networks have been persistent throughout history, and the motivation for joining them, rooted in the violence power principle has always been present. People could secure their existence if they were engaged in voluntary exchange or if they were granted a gift, either from a patron or in the form of inheritance. However, securing one’s existence in either of these two ways was, throughout history, riskier than the violence power principle. The reason is because it lacked any institutional protection. There were no rules guiding it and people were subject to unpredictable forces (often of nature), or at the mercy of roving or stationary bandits. One could lose their entire property, wealth, and family without any protection whatsoever. This is why the market exchange principle needed a strong set of institutional rules to make it work and deliver the greatest benefit to society. It was only then, when governments started to provide a system that helps foster market exchange, that the greatest levels of economic progress were achieved. The progress of our societies finally happened when we learned how to subdue the violence power principle. But it was never entirely destroyed. What we have today is still a threat of elite networks usurping power, pushing up inequality and constraining opportunity. The motivation of joining them is still strong, even though being part of an elite network today does not grant you the same legitimacy to use violence or coercion as it did in the past. But it does give you access to unique information and opportunities, and enables the network effect to work to your advantage. Before presenting a detailed overview of helpful reforms in the third section, culminating in Chapter 10, we must first see how the violence power principle still exists in the modern world, no longer being used to murder or destroy, but still strong enough to motivate both politicians and firms to extract rents. This will be shown in the next two chapters: Chapter 6, which looks at the political motivations and benefits of elite network membership, and Chapter 7, which looks at the same for the firm.

6

Motivation for Politicians Extracting Rents and Staying in Power

Ninety percent of the politicians give the other ten percent a bad reputation. —​Henry Kissinger

Now that we understand why people enter into elite networks, we can examine the benefits of membership for each group. We start with politicians. What benefits can politicians achieve by entering into such relationships? In high-​ corruption countries two benefits stand out in particular: (1) avoiding punishment if they are able to generate a minimum winning coalition of interests, which is very often consolidated by the elite network they belong to; and (2) the ability to extract rents for themselves, in the form of various monetary or nonmonetary concessions. In low-​corruption countries, the risk of taking bribes or otherwise embezzling public money is too high, so politicians receive other types of rewards. They get appointed to corporate boards after leaving office, enjoy small favors while in office, or, in cases of low accountability and oversight, can engage in extortion and similar practices, and may still build powerful local coalitions to remain in power. The political agency literature, in addition to the selectorate theory, provides the best explanations of how this mechanism works, showing that office longevity and rent-​extraction need not be mutually exclusive. When this is the case, a politician can stay in office for long periods of time, supported by an elite network they’ve helped set up, contingent on all of them receiving their share of the rent. A more direct approach to estimating rent-​extraction in the form of corruption is also possible. A number of empirical research efforts have been made, across several countries, that aim to approximate corruption using fraudulent procurement contracts. When measured this way, corruption is shown to have a direct positive implication on political re-​election. This is worrisome, but as this chapter will show, it is not entirely surprising.

Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0007

196  Elite Networks

6.1.  Corruption, Clientelism, and Rent-​Extraction Politics is an arena for resource redistribution. It is a tool of representative democracies necessary to satisfy the redistributive and allocative roles of governments. It is also a tool that can be misused to promote special interests. A big part of the political process involves using distributional powers of executive office to fulfill personal goals of self-​interested actors. Be it those of the politicians themselves or the interests of those around them. In their influential book Brokers, Voters, and Clientelism, a group of four political scientists, Stokes, Dunning, Nazareno, and Brusco, deliver a comprehensive definition of such wasteful distributive strategies, referring to them as models of nonprogrammatic distribution. Whenever there are no clear public criteria for distribution or when these criteria are abandoned to promote private interests, such distribution is considered to be nonprogrammatic. The “most toxic” form of nonprogrammatic distribution is clientelism; a quid pro quo relationship between voters and politicians where votes are exchanged for direct material benefits.1 In addition, noncompliance from voters is subject to punishment, implying that clientelism can take a form of coercion, thus inhibiting the proper functioning of democracy.2 Clientelism, however, is not necessarily just about influencing, buying, or coercing votes from individual voters. It can also extend to direct material benefits, exclusive contracts, and favorable regulation given to private businesses. The premise is the same—​distributional benefits are shared to interested actors within the same network of interests who in return promise political loyalty. These activities however, unlike regular vote-​ buying through pork-​barrel spending or populist policies, are often illegal. It is thus more precise to expose such activities for what they really are—​political corruption.3 Corruption can take many forms: (1) misallocation of public resources—​ exclusive contracts are given to firms based on political connections, (2) nepotism—​a variant of patronage where public sector jobs are given to insufficiently qualified relatives or friends, (3) bribery—​the classic compensation mechanism for any corrupt activity, (4) conflict of interest—​public servants use their connections and access to information to generate benefits for themselves or their clients, and finally (5) institutional corruption—​the implementation of laws and regulations that directly favor private or public enterprises.4 Therefore, according to Shleifer and Vishny, corruption is “not some moral aberration, but a general and natural consequence of the operations of the grabbing hand government.”5 There is an inherent incentive for politicians holding power to design a system that maximizes rent-​seeking behavior based on close personal networks of interests.6

Motivation for Politicians  197 The key to political survival lies in the effectiveness of this network of interests in preserving systemic corruption that enables all connected actors to keep extracting rents. In practice this means that politicians holding power will enter into numerous deals with various special interests from which they can expect to receive electoral support.7 In a codependent system based on close personal ties where concessions are traded in return for votes, bribes, or other kind of support, those linked with politicians, who have a lot to lose from their demise, will make sure they stay in office for as long as possible. In such an institutional framework, well-​organized ruling elites manipulate the economy by generating privileges based on the personalization of governing institutions. Whenever intrapersonal relationships between the powerful and the political elites determine political and economic outcomes, societies exhibit political capture of public institutions.8 What I am specifically interested to show in this chapter is empirical evidence of when corruption, formed through a clientelistic elite network type of relationship, carries a positive electoral effect for politicians. In other words, politicians are encouraged to build and enter into elite networks as this enables them not only to collect rents but also to stay in power. The most accurate way to capture corruption, as shall be explained below, is to approximate it through potentially fraudulent procurement contracts administered by the local or national government, given to selected politically connected private firms. The hypothesis is that due to personal connections and mutual dependence between rent-​seeking firms and the political establishment, politicians can win elections without decreasing the scope of their corrupt behavior. Moreover, politicians have a greater probability of staying in power if they create an informal network of interests that enables and encourages corruption.

6.1.1.  Political Agency These hypotheses are drawn from the standard theoretical and empirical findings of the political agency literature. Political agency assumes that politicians in power have strong incentives to misuse that power for private gain. Political accountability in front of voters (principals) should prevent politicians (agents) from fully expropriating the public budget. However, due to lack of transparency and an informational advantage, politicians often do get away with allocating a fraction of public funds to their private benefit. These activities, whether done at a local level or a grand-​scale national level, are widely known as corruption, as defined above, but in the terminology of political agency models it is called rent-​extraction. Rent-​extraction, unlike the corporate practice of rent-​seeking, includes excess payments (bribes) extracted through public good expenditures on various

198  Elite Networks pork-​barrel and white elephant projects obtained by an incumbent politician. For example, while building a road or a bridge a politician can conceal his rent-​ extraction by presenting one price to the public while paying a different (lower) price to the contractor, thus taking the difference for himself. Furthermore, there are also instances of fraud in public procurements, diversion of public funds (expenditures without proof of purchase), and overinvoicing (buying goods above market price), which tend to be more easily and frequently done on a local rather than a national level. Rents are therefore hidden within budgetary expenditures that provide the easiest rent-​extracting opportunities—​such as public investments on infrastructure projects, IT projects, or defense spending. In each of these cases, a politician needs close collaboration with the business owners or CEOs, thus closing the elite network circle. Political agency is mostly concerned with politicians as nonbenevolent rent-​ seekers driven purely by self-​preservation. They seek to maximize their private benefit from holding office by implementing their preferred policies. While providing the general public goods for the satisfaction of voter preferences and thus generating favorable public outcomes, they have a strong incentive to divert some of the budgetary allocation toward wasteful spending from which they aim to extract rents. This assumption is aligned with Buchanan and Tullock’s and Brennan and Buchanan’s definition of politicians as self-​interested utility maximizers9 or the more recent definition from Bueno De Mesquita et al.,10 who assume that: . . . all political leaders, regardless of their institutional setting, have a common utility function that emphasizes first holding onto (or gaining) office and second maximizing their personal income while in office.

—​Bueno De Mesquita et al. (2005), p. 21

Selection of politicians is therefore adverse and tends to produces “bad politicians.”11 The opportunity costs of working in the market sector are too high for high-​ability individuals, so the selection of candidates into politics will be skewed toward low-​ability or corrupt individuals. Even if high-​ability citizens were allowed to enter office in order capitalize on their prepolitical experience, there are still incentives for these individuals to enter the political market in order to obtain private benefits, meaning they will still engage in a trade-​off with the voters over optimal policies. Political agency models therefore describe a general setting in which a rational agent’s maximization problem is to capture political rents.12 The voters are unable to observe the budgetary allocation process directly, creating the problem of electoral accountability of politicians (the monitoring problem). Uncertainty and asymmetric information give further incentives to politicians to misrepresent themselves and pursue their own interests. Due to such behavior of agents,

Motivation for Politicians  199 there exists a trade-​off between voter utility (policies appealing to voters) and rent-​extraction (policies appealing to politicians in power). The central issue is whether or not electoral competition and the discipline effect of the voters will induce politicians to announce voter optimal policies or rent-​maximizing policies. The models are often characterized by a two-​period setting in which a politician’s term ends in the second period.13 In order to stay in office and reach the second period, an incumbent politician should limit his rent-​extraction in t = 1, since retrospective voters reward congruent behavior. The re-​election incentive should improve the discipline of politicians. However, in the second and final period (t = 2), a moral hazard problem arises since bad politicians are free to divert the entire budget toward their private means. In classical moral hazard models, the homogeneous voter observes the politician’s action but with a noise. The politician observes this noise before making his action (or level of effort), which depends on the re-​election rule chosen by the voters to limit the incumbent’s incentives for rent-​extraction. In expanding the moral hazard problem newer models introduced adverse selection14 concerning how good politicians should distinguish themselves from bad ones, where the first period behavior of bad politicians implies “mimicking” the behavior of good politicians and sacrificing first period rents in order to remain in office and expropriate the entire budget for rents in t = 2. The probability of a politician doing so depends on his time preference for money (the discount factor). The selection effect is added to the discipline effect, where candidate types determine the competency of politicians in providing public goods, or whether or not they are likely to extract more rents. The candidate’s choice of policy will determine his type and send a signal to voters on re-​election. When voter preferences are taken into account, things get more complicated, given that voters themselves are far from perfect agents that punish every transgression.

6.2.  Do Voters Punish Corruption? There is one puzzle that appears to characterize electorates regardless of their institutional background and strength of democratic institutions. The impact of corruption on chances of re-​election should by any reasonable assumption be negative. However, the majority of the literature suggests otherwise, leaving an open question of why voters do not punish corruption at the polls, at least not as much as one should expect. The empirical literature supports this worrying pattern across many developed countries, from the United States to Japan, from Greece to Italy: voters often fail to punish corrupt politicians by electing them out of office.15 Even

200  Elite Networks when the literature does confirm a negative impact of corruption on chances of re-​election, the general finding is that incumbent politicians are either not punished at all or not punished enough.16 The underlying assumption beneath this puzzle is that the vast majority of voters are honest people who dislike corruption. Accordingly, they should punish corruption when they become aware of it. But being aware of it in a world of informational asymmetries is not straightforward. Politicians have every incentive to hide their illegal activates, and particularly the strength of their powerful networks, from the public eye. This has led to the origination of two competing hypotheses explaining the conundrum of why voters fail to punish corruption, even when they know it is likely to be present.17 These can be defined as the information hypothesis and the trade-​off (material benefits) hypothesis. The information hypothesis follows the Downsian logic of rationally ignorant voters.18 Voters either rationally chose to stay misinformed or they suffer from informational asymmetries. According to this hypothesis there are two ways to overcome the information asymmetry: (1) voters receive information from sources they consider to be credible and trustworthy (the media would be one mechanism for closing the informational asymmetry, however voter trust in media is selective at best), or (2) if this information is easily accessible to all voters.19 The trade-​off, or material benefits, hypothesis on the other hand claims that voters are fully aware of the extent of political corruption in a country, but tend to overlook it due to personal benefits they might receive (like patronage or pork spending) and/​or ideological proximity to their candidate. This hypothesis envisions politicians in a quasi–​Robin Hood fashion where voters know they will steal, “but at least they are giving (some of it) back to the people.” Furthermore, due to uncertainty regarding the challenger amid adverse selection into politics, voters often fear replacing the incumbent as they are not sure that the challenger will keep delivering the same benefits as the incumbent, nor are they certain that the challenger won’t be even more corrupt than the incumbent. This is particularly relevant in electorates where voters perceive all politicians as bad. If corrupt politicians deliver successful policies (infrastructure, growth, etc.), if they are perceived to be of high competence, corruption is irrelevant to the voters, and politicians can avoid punishment for years. This is one mechanism by which corruption can become entrenched in democracies. In one of my papers with colleagues Podobnik and Stanley, examining phase transitions of when corruption prevails in democracies, we show that democracies do not cause corruption but can serve as a mechanism that preserves it.20 If voters rationally decide to be ignorant about political corruption, or if they rationally support a known corrupt candidate, then the mechanisms of political competition do not work, and democratic institutions are inefficient in combating corruption.

Motivation for Politicians  201 With this in mind, the given two hypotheses are not necessarily mutually exclusive. For example, core party supporters choose to remain rationally ignorant about corruption and dismiss any evidence against their candidates as politically motivated accusations from the opposition or the biased media. In this case, ideology of a candidate and his perceived competence keep a candidate safe from punishment. If a politician lacks competence or likability, or if he failed to deliver the expected benefits, then voters will punish him at first implication of corruption. The combination of the two hypotheses suggests that voters find both of these factors important: competence (in terms of likability, ideological proximity, and delivered benefits) and corruption, the importance of which is triggered only if competence is low. Voters thus react to corruption only if the candidate is of low quality, or ideologically distant.

6.3.  Elite Networks and the Selectorate Theory There is another factor that can amplify informational asymmetries among voters—​the existence of elite networks, which thrive on having information skewed toward the selected few, as shown in Chapter 5. Elite networks are formed to promote the partial interests of its members. When politicians and firms collude to build infrastructure projects while splitting the profits, voters receive only a partial signal—​they usually cannot tell whether the project’s inflated price was due to corruption or some other exogenous shock. The motivation for hidden rent-​extraction delivered through elite network membership is therefore immense: rents are created and distributed among its members, and the politician delivers a public good (in this context the better term is public bad) for which he may be rewarded by the voters. From the politician’s point of view, the motivation for being part of an elite network is fourfold. First, such relationships reduce risk and uncertainty for all connected members.21 Being part of an embedded network of interest encourages all parties to stay in, be loyal, and protect the patron, knowing that his demise might endanger one’s entire enterprise and its expected revenues. Politicians count on the loyalty of their network, whose motivation to keep him in power is greater the more interconnected they get. The biggest risk is to lose the patron, and all members will actively seek to reduce this risk. The second motivation, derived from risk-​aversion is the ability to extract rents.22 As emphasized earlier, political rents vary from bribes, to various embezzlement schemes, to receiving campaign funding. Third, being part of the network breeds nepotism as it relies on a number of “small favors” that vary from playing golf and getting tickets to sporting events, to directly hiring friends and

202  Elite Networks relatives of the politician in the firm as a favor to the patron.23 All actors have an incentive to perform the favor as this will further improve trust and strengthen the ties of the network. This is particularly true when substantial criminal activities are involved. The commitment device becomes even stronger in that case: none of the actors involved in the criminal act have any incentive to voluntarily break the network, as they know that the downfall of others could also be their own. This makes ties within the network even more exclusive, as the group accumulates power. Such groups do eventually break down, particularly in institutionally well-​defined democracies where punishment occurs through an impartial legal system. In more extractive institutional environments however, such groups can persist for quite a long time, and can be notoriously difficult to break. The final motivation is that elite networks embed minimum winning coalitions that help politicians stay in power very long. To understand this mechanism fully, we must turn to the selectorate theory, presented in the book The Logic of Political Survival by four political scientists, Bueno De Mesquita, Smith, Siverson, and Morrow.24 According to the selectorate theory, politicians lacking serious accountability and media oversight can stay in power for very long if they manage to create a large enough group of supporters they can reward with various concessions. The group of loyal supporters that directly benefit from their politician or party staying in power is defined as the minimum winning coalition (W). A winning coalition is a subset of a larger group called the selectorate (S), representing all voters eligible to vote in a society. The smaller the winning coalition with respect to the size of the selectorate (the so-​called W/​S ratio), the greater the chances of political survival. This is due to the fact that members of small winning coalitions can easily be replaced by members outside of the coalition if they fail to remain loyal. The costs of defection of members within the winning coalition are too large (they lose their privileges and benefits). This testifies to a mutually dependent relationship between political elites and the groups they choose to include in their winning coalitions. In dictatorships, winning coalitions are typically made of senior army officials which help preserve the regime, or are otherwise made up of elite networks that strengthen the dictator’s rule and benefit themselves from proximity to the dictator (similar to Ben Ali’s family and close associates, as portrayed in the Introduction). In democracies, politicians cannot reward their coalitions in the same fashion, so they resort to vote-​buying practices or offer jobs in the public service. They may also collaborate with corporate interests where campaign donations are exchanged for favorable legislation and other benefits, and where many politically connected firms can get directly involved in encouraging voters (this is particularly evident in less developed, high-​corruption democracies).

Motivation for Politicians  203 Democratic politicians often use the help of key members such as union, religious, or ethnic group leaders who can secure votes through their large membership. The leaders enjoy the rewards, while the members provide the votes. The logic among both democratic and autocratic leaders, however, is the same: create a mutually beneficial and interdependent relationship with a group of core loyal supporters who have much to lose if their politicians lose office and thus have a very strong incentive to keep them in power. These coalitions are very stable but they are highly dependent on a single person. Everyone stands to lose if the central node of the network—​the corrupt mayor—​loses his position. The network makes sure this rarely happens. Environments where very small but powerful winning coalitions are enough for the politician to stay in power for very long are more likely to have their institutions personalized, lack any accountability toward the broad voter base, suffer from higher corruption, emblematic clientelism, and consequentially higher inequality, lower living standards, and very low incentives for igniting economic growth and development. In democracies, even developed ones, such scenarios are much more likely to happen on a local rather than a national level, simply because of much lower accountability existing locally. For one, not all local areas have free media to report transgressions and local scandals. If a scandal does occur it will rarely result in punishment unless picked up by national media. It is thus no wonder that longevity in office in local government is highly correlated to instances of corruption.25 The longer someone holds power, the more likely they will successfully build their coalitions, and the more likely they are to redistribute resources among themselves and their loyal supporters or cronies. Local elite networks might not be as powerful as national elite networks, but within their local environments they do hold significant clout and can seriously undermine local development and increase inequality. The mechanism described by the selectorate theory is to a large extent similar to the overreaching hypothesis of this book, with the difference that the selectorate theory emphasizes that large winning coalition environments (with respect to the electorate, so most democracies) will deliver beneficial public policies, and that the creation of small powerful cliques will be limited to autocracies or at best local levels of government. I posit that such cliques arise everywhere, and at all levels of government. They are most likely to be noticed and examined at a local level, however their spontaneous emergence will characterize any system where political power can be misused to promote partial interests. Therefore, when merging elite network theory with the selectorate theory we can see why elite networks are so lucrative and attractive for politicians. They are able to turn their position of power into a machine for extracting rents and accumulating wealth. At the same time, they can rely on their elite network to help them keep their position, from which they all may benefit.

204  Elite Networks

6.3.1.  The Nonlinearity between Corruption and Political Survival The more people involved within a network of a politician’s key supporters, the more likely that corruption will be noticed by outsiders, rendering a nonlinear effect on political survival. How? Given that corruption is hidden, the mechanism through which the process works could be the following: voters are unaware of actual corruption and are only aware of rumors and media reports. These rumors and media reports increase in intensity as the mayor gets involved in more corrupt practices (e.g., allocates more fraudulent procurement contracts). The more corrupt activities there are, the more difficult it is to hide them from the public’s eye. Voters and reporters however do not see the full extent of corruption. They only see signals if other people talk more about them. A mayor who is corrupt manages to deliver a substantial amount of public goods through his cronies, and this is perceived to be beneficial by the voters (and rewarded with re-​election), as they do not notice any direct corruption. However, as more and more jobs are distributed to a narrow group of core supporters, voters (and reporters) do start to notice and the information spreads. This is why I assume a nonlinear relationship between corruption and re-​ election probabilities. With corruption being measured through fraudulent public procurements, this is not easily observed by the majority of the public, particularly at the level of local government. A mayor easily gets away with some small(er) level of corruption without antagonizing too many constituents. As more jobs are distributed within the elite network, more and more local constituents tend to notice that the distribution of procurement contracts (or jobs in the public sector) is unfairly skewed toward the same people, those with closer ties to the mayor. This triggers media reports and spreads information to a wider number of voters, who, on receiving this new information, update their beliefs over the mayor’s competence and behavior, and decide whether the level of corruption is indeed too high and should be punished. When too much corruption becomes visible, mayors have a lower probability of survival. Intuitively, there is no justification of why corruption should entail a linear effect on the probability of winning (so that it is either a strictly increasing or decreasing function of corruption). It is much more likely that the relationship depends on a specific situation the incumbent is in. For example, is the incumbent electorally safe, or does he have to fight for his or her seat? The effect of corruption on re-​election can depend on the uncertainty surrounding the electoral race. If a politician expects a close race, he is less likely to engage in corruption (to “loot”) but more likely to try and disturb the election process to win (to “cheat”).26 In other words, the behavior of a politician depends on how well he performs on previous elections and to which extent this allows him to cheat.

Motivation for Politicians  205 To sum up, an elite network type of relationship, which reinforces the creation of corruption, entails a concave effect on a politician’s re-​election chances. Voters see the creation of public goods and reward the politician without knowing whether he delivered the public goods with some corruption. This partially corresponds to political agency models, where political types (good or bad) are hidden, so voters must infer type based on action—​whether or not the politician provides enough public goods and at what cost. Public goods can be provided under a high cost by a bad politician who extracts rents (and still gets re-​elected) if the voters believe he faced a high-​cost shock. Extending this logic further, a bad politician can fool the people only a few times before such practices get uncovered. Hence the concave effect.

6.4.  Case Study: Using Corruption and Minimum Winning Coalitions to Stay in Power Why is it important to establish that there is a nonlinear, concave effect between corruption and probabilities of political re-​election? This is the main mechanism that politicians rely on in order to preserve their position. They are intuitively aware of an upper limit of suspicious allocations they are allowed to make, before being uncovered and punished by the voters. In more developed democracies, this limit is very low, and is inversely related to the level of accountability. Low accountability and low transparency environments in highly developed democracies are similar in their levels of corruption and their impact on re-​election to high-​corruption, less developed democracies. In their case, the upper limit of “allowed” corruption can be quite high. A good example is the empirical case of Croatia, an EU member state with particularly high levels of corruption, and a relatively young democracy, developing in the aftermath of a devastating war in the 1990s. Croatia is interesting since it provides a case-​in-​point of both the selectorate theory, in terms of how politicians build minimum winning coalitions to stay in power, and the elite network theory, in terms of how both politicians and their connected firms can benefit from entering into mutually dependent relationships. To paint the picture, consider the high-​profile cases of corruption uncovered in the country. The former prime minister (who was in power for six years) was found guilty on corruption charges for forcing state-​owned companies to give exclusive procurement contracts to a media company, from which the PM himself, and a few of his close cronies extracted funds directly. His political party, the conservative HDZ, was also found guilty in the same process and was ordered to return the illegally acquired funds of around €4 million, used to finance their campaign. The former PM was sentenced to nine years in prison, but the

206  Elite Networks country’s Supreme Court brought down the entire sentence two years later based on a technicality, returning the process back to the city court. On a local level, instances of corruption were even worse and have in many cases gone without major punishment. Mayors from big cities, originating from the entire spectrum of political parties (right, left, liberals) were arrested or accused for abuse of power, bribery, and striking favorable deals with selected private firms. The mayor of Zagreb stands out as an emblematic case. In power for 20 years, he was arrested, released immediately after paying the largest ever bail of €2 million, arrested again for breaking probation by intimidating witnesses, and five months later released from jail thanks to the quickest ever decision of the Constitutional court in his favor. He ran the city from jail through his many proxies, and even managed to depose his deputy, a figurehead temporarily placed in charge, while still in jail. After his release, and during a series of trials, he managed to get elected twice as a member of parliament with his new party (only a few months after being released from jail), and won a new term as mayor in the 2017 elections (two years after his prison stint). He never lost an election afterward. He died just before the 2021 election, and was hence never punished by the voters for any of his transgressions. He is not the only mayor that did the same. At least two more mayors got out of jail on corruption charges, and managed to get re-​elected with ongoing trials. There is something particularly problematic in the ability of Croatian voters to effectively punish their politicians. Many local politicians hold office for more than 20 years (with the average tenure being 12 years, three full terms). Even the case of the prime minister was not a punishment enacted by the voters. The process of his incarceration started only after an internal party struggle, where he was hoping to pull strings from the shadows. He was stopped by political power, not by the will of the voters or by the justice system. The way political institutions are defined in the country encourages such outcomes. Local office knows no term limits, mayors have virtually unchecked decision-​making power in their local constituencies (in terms of urban planning laws or administering procurement contracts), and the motivation for corruption in general is high. I have therefore used the case of Croatia to test the first part of the elite network theory: the benefits of staying in power while engaging in procurement-​based corruption, the proxy for elite network level connections between politicians and connected firms. This was published in my paper titled “Corruption and Re-​Election: How Much Can Politicians Steal before Getting Punished?”27 The paper made two important contributions that help us understand the mechanism of the elite network theory. First, it reinforced the idea of measuring grand-​ scale corruption directly through fraudulent public procurement contracts. This is a good proxy of seeking a connection between firms

Motivation for Politicians  207 and office-​ holders. The procurement contracts database consisted of over 400 thousand contracts administered on a local level of government, for 556 municipalities over an eight-​ year period. This database was connected to firm-​ level financial statements, thus enabling direct recognition of suspicious allocations. “Suspicion” is defined precisely, and verified based on official interviews with people involved directly in the procurement process (from special police investigators, to national and local government officials, former mayors, civil society watchdogs, and firms participating in procurements). The definitions included cases in which firms with no employees and as a single bidder on the tender received procurement contracts worth millions, or in which firms signed contracts that vastly exceeded their capabilities (e.g., miniscule average revenues in the years before the tender, followed by a multimillion contract from the tender), or when firms with substantial losses received indirect subsidies through lucrative deals. Quantifying corrupt behavior this way yields more precise insights into the actual state of affairs in the political system of a country than can be done by using survey perception indices. Not to mention that it provides a direct implication of elite-​network-​induced corruption. The second contribution is the confirmation of a causal impact of corruption on re-​election, proven to be a nonlinear relationship. Corrupt practices entail a concave effect on re-​election chances, meaning that corruption can increase the probability of re-​election, however only until a certain level, after which corruption is too high and a politician gets punished. By showing this I successfully bridge the gap between two competing hypotheses of why voters do not punish corruption—​they do, but not until it becomes too large and hence too noticeable. I also find an optimal level of corruption to keep a politician in power. Probability of re-​election is maximized at about 20% of all funds corruptly allocated from public procurement. If more than 50% of all procurement is allocated illegally, a politician very likely loses power (see Figure 6.1). As emphasized earlier, the level at which Croatian voters tolerate corruption is considerably high. There is strong reason to believe that similar countries of eastern Europe or southern Europe, with equal levels of democratic capital or levels of corruption, would have similarly high levels of tolerance for corruption.28 In order to make a causal implication, ideally, we would need to randomly assign corruption in similar cities and municipalities and compare political outcomes with respect to the level of corruption. But achieving a random allocation is impossible since corruption for each city or municipality is determined endogenously by the politicians and voters. A way around it is to use quasi-​experimental designs where we can assume randomness over the observed outcome. In this particular case I exploit the fact that local government law defines a fixed number of city or municipality council members based on prespecified population thresholds. As the population size of

208  Elite Networks (a)

Re-election (predicted)

1

.5

0 0

.2

.4 Corruption - Total index

.6

.8

Figure 6.1  The predicted concave relationship between corruption (measured as a composite index of the different measures of suspicious procurements) and the chances of re-​election. Probability of re-​election is maximized at around 20% of total procurement corruption (20% of all contracts allocated suspiciously). Source: Vuković (2020).

a city or municipality increases, the number of council members increases automatically by two to four. The smallest municipality with up to 500 inhabitants has 7 council members, the next threshold is 1,000 inhabitants with 9 members, the next threshold is 2,500 with 11 members, and so on. This is an exogenously imposed rule (local mayors cannot affect it, nor do they have a particular incentive to increase population size to bring on more council members) which makes it as good as random over the observed outcome. We can then compare cities and municipalities with similar size around the threshold where the only difference is the number of local council members. For example, a municipality with 990 inhabitants will have 7 council members, while a municipality with 1010 inhabitants will have 9. These two municipalities are, by assumption, very similar in their fiscal capacity (tax revenues, expenditures) and several other observable and nonobservable characteristics. The point is to specifically compare similar cities and municipalities, based on size, revenues and expenditures, tax rates, population distribution, location, exposure to war, and so on, where the only differences are the exogenously imposed number of council members and the levels of corruption. This brings us close to approximating the experimental method based on which we can have some say over cause and effect.

Motivation for Politicians  209 Why would the number of council members matter? I find that council size in each city and municipality is highly correlated to corruption. The logic is simple: higher city or municipal size implies higher budgets, which increases the volume and opportunities for suspicious procurements. Very small municipalities have very low budgets and low opportunities to embezzle large funds. Larger cities have a much greater scope for corruption. Furthermore, uncovered cases of local corruption in Croatia are very often linked to mayors buying support from council members to get a majority in the local council, typically through nepotism or by handing them procurement contracts with the city. The pure size of council members is therefore a good instrumental variable (IV) to use in order to measure how procurement corruption affects electoral outcomes. When comparing two cities with the same population size and roughly the same budget there should no big difference in procurement corruption, unless it is indirectly induced through buying votes of local council members. To uncover this, I only compare cities and municipalities close to each population threshold (see Figure 6.2), where the relative difference in council size is much bigger than

0.16

Corruption (total index)

0.14 0.12 0.1 0.08 0.06 0.04 0.02 0

0

2500

5000

7500

10000 12500 Population

15000

17500

20000

22500

Figure 6.2  The concave implication is also dependent on city/​municipal size. In very similar smaller and midsized cities and municipalities corruption is higher for units just above the population threshold compared to those just below the population threshold. For big cities the opposite is true. The dotted lines denote the six population thresholds defined by law. Each dot represents average corruption for cities or municipalities just above or just below the population threshold (within ±5% population size). Source: Vuković (2020).

210  Elite Networks the difference in population size. Notice that the concave relationship is also confirmed with respect to city and municipal size. In smaller and midsized cities and municipalities corruption is higher for units just above the population threshold compared to those just below the threshold. Therefore, a local mayor in a city just above a population threshold that allocates him two more members in the council, in order to stay in power, has to broaden his coalition by marginally increasing his level of corruption. As city size increases, you do not need to corrupt as many additional council members. The reason is that in smaller municipalities two additional members mean a lot in relative terms. In big cities with 30 or 50 members, two additional members are less important for keeping a governing majority. In each case, we can see that mayors certainly do achieve significant benefits when allocating suspicious procurement contracts to their loyal winning coalition members, or in other words, to their elite network.

6.4.1.  The Impact of Minimum Winning Coalitions The final implication is related to the relationship between corruption, longevity in office, and the impact of minimum winning coalitions. An extension to the aforementioned findings is a paper looking at local winning coalitions in Croatia.29 The W/​S ratio, measuring relative winning coalition size, is approximated by looking at the absolute number of votes the winner gained in each race divided by the total number of eligible voters. The average W/​S ratio in this case was 27%, meaning that for an average city or municipality, a politician wins elections if he gets only 27% of the total eligible electorate to vote for him. This is not surprising given that turnout for local elections is typically very low, between 40% and 50%. It also means that local politicians in Croatia need to get only a small number of people in their winning coalition to stay in power. For example, for an average small city of 3,000 inhabitants, where 2,000 people have the right to vote, and turnout is only 40%, a politician needs 400 votes to win. This is not a difficult task to achieve. It is easy to see how local level minimum winning coalitions are able to keep politicians in power for long periods of time. Overall, a higher W/​S ratio—​more people needed to win an election—​is shown to decrease total tenure in office. Across three terms in office, a politician in a low W/​S ratio city or municipality is likely to have one full term in office more than a politician in a high W/​S ratio city or municipality (12 years instead of 8 years). Interestingly, this is highly correlated with corruption, as the same areas with a low W/​S ratio, where politicians are in office for a longer period of time, also have higher levels of local corruption, approximated again through suspicious procurements.

Motivation for Politicians  211 Finally, small coalition environments are also shown to have higher local tax rates, and the magnitude of the effect is quite large (Figure 6.3). Moving from a high W/​S ratio unit to a low W/​S ratio unit, local tax rates are 30% higher (around 1.5 percentage points, which is a large effect given that the average local tax rate is 4%). Interestingly, tax rates are also highly correlated to levels of corruption, (a)

.2

Corruption (predicted)

.15

.1

.05

0

0

.2

.4

.6

.8

.6

.8

W/S ratio (b)

.1

Tax rate (predicted)

.08

.06

.04

.02

0 0

.2

.4 W/S ratio

Figure 6.3  Negative correlations between the W/​S ratio and corruption (upper panel) and tax rates (lower panel).

212  Elite Networks arguably through the W/​S ratio. Moving from a low corruption to a high corruption unit increases local tax rates by 2 points, or around 50%. These findings are striking and suggest a particularly problematic environment where political longevity in office is closely tied to having a small but powerful winning coalition, where corruption in the form of procurement contracts allocated to satisfy the coalition is high, as are the local tax rates to finance excess spending. This has vast implications for fiscal profligacy of corrupt governments—​greater corruption is used as an effective mechanism of satisfying loyal elite networks in order to stay in office longer. Just as the predictions of the political agency literature imply, politicians can extract rents while maximizing their probability of staying in power. The missing mechanism explaining this effect is the impact of elite networks. *** Even though the empirical results presented in this section are mostly limited to the case of Croatia, the logic of political survival applied here is hardly different in countries or local regions where accountability is low and corruption is high. The incentives for politicians to engage with elite networks in order to build their coalitions are similar throughout, however the difference between successful and unsuccessful democracies is in the institutions that either prevent or encourage corrupt behavior. When rules are clear and enforceable, when the judicial system is not politically captured, power-​hungry individuals can be prevented from misusing political office to deliver benefits to their elite networks. Countries that lack or are unable to enforce such rules, or whose institutions are completely or partially captured by political elites, are unable to break away from the system of elite network dominance. The failure of a country to limit the power of its elite networks, both political and corporate, is the first step toward social disintegration that eventually releases the forces of wealth concentration. The failure is therefore entirely political. The remedy is to reduce political power. Before developing that argument in the final section, the next chapter looks at the second part of elite networks—​firms, and how they benefit from being in close contact with in-​office politicians. This chapter has shown their benefit being manifested only through procurement deals. But firms can benefit in many more ways.

7

The Role of the Firm The difficulties of many European countries derive from their corporatism: state projects serving cronies and vast social protection programs, both run by elites. These surged in the 1970s and 1980s. —​Edmund Phelps

In this quote the Nobel prize–​winning economist Edmund Phelps painted an accurate picture of what went wrong with most European societies in the final decades of the 20th century—​too much corporatism serving crony and elite network interests. In his book Mass Flourishing, he lauds economic dynamism that delivered the greatest progress in human history via rapid innovations occurring in the aftermath of the Industrial Revolution (the topic of the next chapter). Corporatism of the late 20th century was recognized as the biggest constraint to that dynamism and, by extension, the key cause of rising inequality over the past 30 to 40 years.1 Corporatism, in the classical sense, advocates that a society should be organized and managed through various corporate groups—​business, labor, military, academic—​protecting their interests and competing for government favors. It is based on the idea of having a strong centralized government that successfully balances between various competing interests of different corporate groups and hence carefully manages economic progress. There are many variants of corporatism but the one Phelps had in mind—​the one that acts as a negation of dynamic innovation—​is corporatism in which corporations and societies demand too much from governments, which consequentially stifles and overburdens the system with excessive regulation that benefits the well-​organized corporate groups. This is somewhat similar to Olson’s argument from The Rise and Decline of Nations, in which excessive pressures of privileged interest groups constrain economic growth. Phelps argues that the loss of dynamism in our economies increased inequality since it hit lower-​income workers disproportionately more than high-​income workers. Innovation, in the sense of developing new products, is labor-​intensive, meaning that a decline in innovation reduces labor-​ intensive jobs, which decreases opportunity for the low and middle classes and allows the continuation of wealth concentration at the top. Furthermore, due to a long period of Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0008

214  Elite Networks low interest rates, the middle classes could no longer significantly increase their wealth through savings. Investing in equities brings much greater fortune, which is again disproportionally concentrated among the already wealthy. Inequality, according to this theory, did not reduce productivity, as some may argue; it was a decline of productivity due to corporatism that exacerbated inequality. Solidarity, security, and stability are cited as typical corporatist values that require an expanded power of governments to be upheld. They drive the need for competition-​stifling regulation, for interest group state capture, protectionist industrial policy, special tax laws, and other regulations that favor narrow corporate interests. Through promoting and sustaining such corporatist values innovation and dynamism are severely constrained, so companies resort to rent-​seeking practices. Profits of traditional companies are starting to depend more and more on political protection, while executives are increasingly guided by short-​term thinking revolving around boosting share prices and capturing bonuses. Achieving political protection is further rewarded by the boards, which drives the dispersion of incomes at the very top, as shown in Chapter 4. A potential consequence was a parallel decline in total factor productivity growth, coupled with record-​high levels of corporate profits as share of GDP, as shown in Figure 7.1. Total factor productivity in the United States has been in relative stagnation ever since the mid-​1970s (with the exception of one decade, from the mid-​1990s US corporate profits vs total factor productivity growth, by decade

12.00% 10.00% 8.00% 6.00% 4.00% 2.00% 0.00%

2.69%

2.64% 1.91%

1950s

1960s

1970s

Corporate profits as % of GDP

1.22%

1980s

1.87%

2.07% 0.95%

1990s

2000s

2010s

Average annual productivity growth

Figure 7.1  US corporate profits as share of GDP compared with average annual growth rates of total factor productivity. Source of data: ILOSTAT (2020), and St. Louis Fed, FRED (2021).

The Role of the Firm  215 to the mid-​2000s). It was around 2.2% from the 1940s until the 1970s, and only 0.8% since the 1980s until 2020. Corporate profits on the other hand surged in the past two decades. They were on average 10% of GDP in the 2010s, double the amount of the 1980s. Economic theory is clear on why productivity matters. As new technology is introduced and innovations adapted, better methods of production increase output for less work input, thus increasing worker productivity and consequentially increasing their wages. Technological progress and innovations since the Industrial Revolution increased societal progress primarily through increases in productivity. Therefore, we shouldn’t worry about inequality created by technological progress, as its impact is only temporary. The productivity bandwagon soon delivers progress and better living standards for all. While this is true on aggregate and in the long run, Acemoglu and Johnson argue that this process of translating innovation into higher wages and greater living standards for all is by no means automatic. New technologies work and deliver progress only when they increase marginal productivity of workers. This was very much the trend since the late 19th century, and especially after World War II. All major technological advances back then led to increases in productivity, making sure that progress was shared by all (think of the benefits of electricity or production automation that created new tasks and many new jobs). However, since the 1980s, the authors argue, advances in technology were primarily focused on cutting costs, thus boosting corporate profits, but delivered no real benefits to worker marginal productivity. In fact, the authors directly blame automation for the decline in average total factor productivity and consequential rise of inequality of the past four decades.2 This explains what we see in Figure 7.1, the share of corporate profits went up due to cost cutting, not better innovations or greater productivity. But this is not the full story. As profits no longer depended solely on achieving disruptive innovation, this created incentives for the rise of rent-​seeking firms. Firms with deep connections to the state are by no means a historical exception or a creation of the modern age. The powerful mercantile monopolies like the British East India Company, the South Sea Company, the French Mississippi Company, or the Dutch East India Company (VOC) were extensions of their governments in controlling trade links and colonization of new territories. We saw in Chapter 5 how this was done in 16th-​century Spain, where the king directly granted privileges to selected nobles. Similar processes happened in other empires and kingdoms, but the political protection of monopoly rights continued until the modern era. Even today some companies are still justified as natural monopolies, providing a public good that the market is unable to provide fairly and efficiently. But outside the scope of natural monopolies, many modern firms still seek political

216  Elite Networks protection to grow and expand their business. Such rent-​seeking firms are the interest of this chapter. Firms that rely on elite networks to achieve their benefits from governments. Firms that are privately owned, but instead of competing on the market for customers, compete for political support. Specifically, firms whose primary source of income is conditional not on government contracts, but on political connections to get those contracts.

7.1.  Rent-​Seeking vs. Customer-​Seeking Firms Expanding on the findings of Chapters 4 and 5, we can split firms into two basic types based on their willingness to engage with politics and hence elite networks: rent-​seekers and customer-​seekers.3 There are obviously shades of gray in between them, but to keep things simple any firm whose revenues predominantly depend on political favors granted via lobbying or friendship links, and which manage to change regulations or secure a non-​market-​generated source of income is defined as a rent-​seeker. On the other hand, any firm whose revenues purely depend on satisfying some market demand, and does not at all depend on having a political connection in order to survive, is defined as a customer-​seeker. A customer-​seeking firm is not the subject of attention in this book. These are regular firms that exist to provide an unmet market demand within a certain niche or territory. The vast majority of micro, small, and medium-​sized businesses, as well as some large firms like major disruptive innovators, are customer-​seeking firms. They typically advocate lower government regulation and a lower tax burden to help them grow and expand. However, they are notoriously difficult to organize, and have very little salience with which to influence policy, which is why they are placed on the lower left quadrant of Figure 1.3 in Chapter 1. In corrupt countries, anocracies, and weak democracies, depending on the industry they operate in, they can be subject to political or criminal extortion (e.g., paying money for protection). But their underlying principle is not to engage with politics or seek political favors to increase their revenues. Rent-​seeking firms are diametric opposites of customer-​seeking firms. They seek to find a nonmarket way of capturing a bigger market position, and enter into elite network relationships to achieve this. They are encouraged by economic rents they secure. Their revenues can arise from market power, but their market power has been granted to them by the virtue of political power (good examples here are monopolies or oligopolies4). In many cases they do compete for and sell to customers, but their main battleground is the market for political favors. Political decisions can grant them exclusive privilege to sell to customers, preventing market access to competitors. There is no distinction in terms of size,

The Role of the Firm  217 however larger firms in particular industries (like defense or major construction companies) are more likely to depend on government decisions and hence more likely to ask for political favors. This does not imply that all companies in such specific industries are rent-​seekers, only those that actively seek and depend on political protection, usually through personalized relationships with the decision-​makers. Finally, rent-​seeking firms need not organize as a collective and fight for their interests this way. All it takes is for the executive to befriend a politician and they enter into an elite network relationship, where they are free to exchange favors. They employ CEOs and other executives with deep ties to the political power-​holders, and reward their CEOs for deals they get because of politics. As the findings of Chapter 4 confirm, such connected executives have higher salaries than nonconnected ones. Furthermore, firms with politically connected executives pay higher salaries than firms with nonconnected executives. This mechanism is directly responsible for increasing earnings of the top 1% and top 0.1% of income earners, thus contributing to the rise of inequality. Despite the fact that there is always a risk for a firm to get uncovered for shady business practices or for its politician to lose office, firms enter elite networks to essentially lower risk (as shown by the cost-​benefit analysis in Chapter 5). They seek protection from politics in terms of legal or regulatory acts, or they seek direct rents from government contracts. From the perspective of the firm this interaction is economically rational as it involves trading off low cost (of effort and engagement) for a greater potential return that involves reduced risk exposure. The acts of lobbying and campaign spending are completely legal and legitimate ways of influencing government decision-​making. Exchanging personal favors between connected individuals in positions of power is not necessarily always legal but it is in most cases legitimate. Using political connections implies that you trade off costs of engagement for rents. When making the trade-​off and securing rents, firms invertedly increase the dispersion of earnings at the top of the income distribution. There is obviously a between-​firm difference in both cases. Firms in certain industries, where governments are the only buyer, have no choice but to do business with governments. The aforementioned defense industry is one example, health equipment industry is another (in countries where health care is mostly public). Yet, there is a particular distinction between firms that win tenders in a fair and open process, thanks to their competitive advantage or state-​of-​the-​ art products, and firms that use political connections to affect the outcome of public tenders. Corrupt countries with weak democratic institutions are particularly sensitive to the latter. That being said, even the most developed democracies are not immune to having powerful corporations organize cartels, affect legislation, or use political connections to win procurement contracts. Some of these

218  Elite Networks companies can even be market leaders with top quality products, yet they still resort to rent-​seeking to preserve their position. This keeps their shareholders happy who, in turn, reward the politically connected CEOs for providing decent returns. One possible explanation is that the propensity toward rent-​seeking is contingent on the stage of the company’s life cycle. An innovative customer-​seeking firm, in its early stages of development, when seeking to establish a niche market position, hardly engages with politics and hopes to attract customers with a unique selling point. Some may receive seed or venture capital funding that helps them position in the market and prepare them for the growth stage. Once reaching growth stage, they are entirely focused on scaling up their products and reaching as wide as possible a customer base. During the maturity stage, with sales relatively consistent and linear, the focus is on reducing costs to prop up profits, but some firms may find it profitable to turn to executives with political connections in order to preserve their market leader position from competitors. A good example here are the US Big Tech companies. Initially all these companies were major disruptors of their industries, expanding vastly and exponentially, creating huge value added for consumers. However, during the 2010s, as they reached a certain size, they have increasingly started to hire former politicians,5 presumably to preserve their status but also to influence regulatory policy. Political appointments alone do not necessarily make them rent-​seekers, but as the demand to regulate Big Tech increases, the lobbying activities of these companies will surely increase as well.6 Having former politicians on their Boards or at executive positions is certainly expected to help them in curtailing new regulations to their benefit. Therefore, even the disruptive customer-​seeking companies may eventually turn into rent-​seekers in order to preserve their dominant position on the market.

7.2.  How Firms Benefit from Collusion with Politics Now that we have differentiated between two basic types of firms, we need to turn to the theoretical and empirical findings that explain the motivation of firms to engage in deep connections with politics. The seminal theoretical contributions explaining the link between firms and politicians come from the economic theory of regulation. George Stigler’s initial contribution, resting on Olson’s logic of collective action, was a demand-​side explanation showing the effect of cartelization: large, well-​organized firms (led by small, homogeneous groups) will lobby most effectively and get the benefit of favorable regulation, which will always come at the expense of consumers (large, dispersed groups which are unable to organize).7 Sam Peltzman expanded that

The Role of the Firm  219 model to include the role of legislators who are decision-​making agents that seek to remain in power, and will guide their decisions to fulfill that goal. The electoral concerns of legislators should explain why big businesses can sometimes be punished, or why protectionism can sometimes be relaxed.8 Gary Becker then expands Peltzman’s model by introducing competition between interest groups where size and within-​group free riding determine success in receiving government transfers.9 A more precise definition of these activities is encapsulated by Tullock’s and Krueger’s concepts of rent-​seeking:10 the process of gaining private benefits for firms by misusing the political process (via lobbying, campaign spending, personal connections, or even bribes to influence legislators). Rent-​seeking is usually linked to gaining protection for a particular firm or industry, which can vary from achieving monopoly status, to barriers to entry for newcomers that reduce competition (e.g., via licensing or patents), to imposing tariffs on foreign goods to protect domestic incumbents, or picking winners via government subsidies. In each case rent-​seeking activities hamper market outcomes by reducing its allocative efficiency and end up destroying wealth. In the context of this book, rent-​seeking is the crucial ingredient linking collusion between firms and politicians to the negative consequences of that collusion. In the previous chapter I expanded the rent-​seeking definition to a broader concept of rent-​ extraction, in alignment with the political agency literature, to include political benefits of collusion. But the classic concept of rent-​seeking has always been focused on benefits gained by firms. Following these seminal theoretical contributions, the next step is to explain what exactly the interaction between firms and politicians looks like and, more importantly, what its social welfare implications would be. An interesting concept that expanded on the economic regulation and rent-​seeking literature was McChesney’s “money for nothing’ ” model of political extortion. According to this model firms and interest groups that employ lobbyists pay money to prevent legislative outcomes that could hurt them. The “money for nothing” paradigm implies paying the price of political extortion. Politicians propose legislation and regulations that would hurt a given industry, which is then forced to pay the rent in order for a politician to withdraw this piece of legislation.11 McChesney argues that many regulatory wealth-​reducing threats are only announced for this particular purpose—​to extract rents from private sector firms. He shows this empirically for the United States and calls on empirical evidence of a similar rent-​extraction practice in Canada.12 In similar fashion, Shleifer and Vishny describe a clear-​cut set of grabbing hand government models that encapsulate the full benefits of an elite network relationship: each agent pursues their own objectives, none of which promote the public’s interest. Their analysis is particularly suitable to explain the negative

220  Elite Networks social outcomes following dubious privatization schemes in postcommunist countries. Politicians privatize profitable public firms by giving them to their cronies in return for continued bribes. They keep less profitable public firms in the government’s hands and continue to demand the pursuit of political goals from these firms in exchange for subsidies.13 Whenever loyal political support is conditioned on various material and nonmaterial benefits the political system is likely to deliver subpar outcomes, as the benefits tend to favor the interest of the few (usually the elite) over those of the many. In his seminal 1982 book The Rise and Decline of Nations, Mancur Olson warned of the negative implications of the rising power of cartels and lobbies over time. Their actions, fully legitimate outcomes of the democratic process, have resulted in the rising power of lobbyists, lawyers, and bureaucrats who exercise control over the redistributive flow of resources. As the struggles for redistribution and favorable regulation between interest groups surpass those of the productive sectors of the economy, economic growth becomes undermined and decline is imminent.14 These sets of theories present an accurate depiction of how the elite network mechanism works, and it clearly recognizes a negative effect of collusion between politicians and firms on social welfare. What are the empirical implications? And what is the size of distortion generated by such networks? Empirical research has mostly been focused on confirming Stigler and Peltzman’s contributions. For example, Bellettini, Ceroni, and Prarolo use a panel of 62 countries and find that the negative impact of political longevity on economic growth can be explained through the actions of politically connected firms that use their power to mitigate regulatory costs, prevent entry of higher quality competitors, and preserve their market dominance.15 Fisman shows, on the case of Suharto’s Indonesia, that a large part of a connected firm’s value can be explained by its political connections. He finds that 30% of Indonesia’s GDP depended on politically connected firms.16 Faccio, using a panel of 47 countries, also confirmed that greater political connections increase firm value. She finds that connected firms (firms with a large shareholder either in office or connected to the ruling party) represent almost 8% of the world’s market capitalization. Not surprisingly, connected firms are more common in countries with greater corruption and lower levels of transparency (e.g., in Russia they amounted to 87% of all market capitalization during the 1990s).17 A similar finding is confirmed in the United States, where announcements of politically connected individuals in the board of directors resulted in abnormal market returns.18 Conservative MPs in the United Kingdom benefit personally from industry connections after their time in office, being awarded by lucrative positions on company boards.19 Two other papers make the same inference but for the opposite effect of when the political connection is lost (i.e.,

The Role of the Firm  221 when a politician exits politics, either through retirement or death). When this happens, both connected firms and connected lobbyists are faced with losses.20 Furthermore, on a group of 35 countries, it has been confirmed that politically connected firms tend to receive more government bailouts during times of crisis than nonconnected firms, that government-​owned banks increase lending in election years to help party-​affiliated borrowers, and that politically connected firms get better loan deals from public banks.21 In China, examining the impact of political connections on corporate performance of China’s newly privatized firms, where a quarter of appointed CEOs were government bureaucrats, the authors find that firms with politically connected CEOs underperformed the unconnected ones in terms of post–​IPO market valuation, earnings, and sales growth. However, such firms are more likely to appoint other bureaucrats to the board, which might be the reason for derailed business performance.22 Institutions can matter as well. In has been shown in the case of Russia that weak institutional environments serve to promote partial interests of firms, and that in environments which fail to demand political accountability businessmen themselves run for office to reduce their lobbying costs and directly promote policies that favor their interests.23 A similar finding is confirmed on the case of the Ukrainian Orange Revolution. Political turnovers exercise a significant impact on politically connected firms in weak institutional environments.24 A comprehensive study of political-​business networks in the cases of Poland, Bulgaria, and Romania shows that the impact of such networks on the formation of postcommunist institutions was critical to the success of a country’s path through transition. The key implication is that broad networks in combination with greater political competition and hence greater political uncertainty generate better institutions. Narrow networks of elite interests will, on the other hand, result in weak institutions and worse economic outcomes.25 Firms can therefore clearly benefit from being in close proximity to politics, regardless of which country or institutional setting we are looking at. The next step is to again look within firms and how their powerful executives can benefit even more.

7.3.  The Revolving Door between Politics and Finance Who you know can be much more important than what you know when the system revolves around elite network superhubs. This has been empirically proven on a number of occasions. Lobbyists tend to follow their politicians when they move to different committees, while the impact of a politician leaving office has resulted in a 24% decline of revenues for a lobbyist connected to this politician.26 Another study has shown that the US Securities and Exchange

222  Elite Networks Commission (SEC), the key regulatory body behind securities fraud, is less likely to prosecute politically connected firms, and even if they do, such firms will face lower fines.27 Political donations are particularly helpful here, especially if being given to politicians in the SEC oversight committees or those in executive government. The SEC is therefore more likely to punish firms without the backing of politics. This is a woeful practice that practically legalizes corruption. You are allowed to cheat as long as you have close ties to those in power. As shown in the network graphs in Chapter 1, finance is particularly prone to being closely connected with politics and thus exploit its close ties to power. Politically connected firms engaging in securities fraud that the SEC is not punishing are, in most cases, in the finance industry. Bear in mind that financial corporations also own considerable parts of other corporations, both domestically and worldwide. A network analysis study in 2011 has shown that a handful of financial institutions dominate shareholding positions in most other firms, thus potentially exacerbating systemic risk for the global economy. In total, 737 top corporate shareholders hold 80% of ownership control over all other international corporations, while the core, a very densely clustered group of only 147 firms, three-​quarters of which are financial institutions, holds 40% of ownership control. The authors refer to them as “super-​entities in the global network of connections.”28 One of the biggest corporate owners in the world is Larry Fink’s BlackRock, giving him immense elite networking power (see his network in Figure 1.8 in Chapter 1). At the helm of such “super-​entities” are financial sector CEO superhubs, typically the most connected nodes in the corporate world, exhibiting high levels of betweenness centrality and maintaining strong friendship links with politicians in power.29 Money and favors are exchanged frequently, but are uncovered only occasionally. One practice that can verify their close ties is the revolving door between the two industries. The process starts with regulatory and cultural capture, where the regulators of the industry start adopting the same position of the industry they regulate, and are themselves often on a lookout for jobs in that industry. Nowhere is this more evident than between the finance industry and its regulators. The very process of deregulation in finance that increased excessive risk-​taking before the 2008 financial crisis, in terms of relaxed capital requirements for example, was driven in part by the process of cultural capture.30 What starts as cultural capture tends to evolve either into an employment position within the industry or, during a politician’s mandate, into an elite network relationship, depending obviously on the position of power of the captured individual. The cultural capture of people in frequent contact influences the way of thinking in both directions. More importantly, their clustering develops personal relationships that are often called on when problems need to be solved. At

The Role of the Firm  223 that point it is no longer a case of cultural capture, lobbying, or persuasion by the industry over the regulators. It is a friendship relationship, where important decisions are exchanged as favors between friends. This is obviously a normal human condition; we will always treat our friends more favorably. However, when the stakes are high, objectivity is what’s valued. Achieving such objectivity in positions of high power becomes difficult when the outcome of a decision could burden a friend or frequent associate. Even so, it is impossible to prevent politicians and corporate executives from networking, making friendships, and entering each other’s career paths. This seems to be an inevitable trend in high positions of power. Politicians from all levels, from high executive office to parliamentary committees with direct supervision over the financial sector, very often end up working in finance after their careers. A few high-​profile examples stand out. Larry Summers worked in academia and at the World Bank before entering the Clinton administration, eventually becoming treasury secretary succeeding his mentor Robert Rubin, himself a career investment banker from Goldman Sachs. Before rejoining public service in the Obama administration, he held a job at a hedge fund, D.E. Shaw, and had various advisor roles at a number of Wall Street firms. Timothy Geithner was in public service most of his career, however after being the chairman of the New York Fed, followed by treasury secretary under President Obama, he ended up as president and managing director of a private equity firm Warburg Pincus. Jack Lew went in and out between politics and the corporate world. Lawyer by training, he started working for congressmen in the House, then went to a private legal practice, and came back to serve in the Clinton administration. Before joining the Obama administration first as deputy secretary of state, and in Obama’s second term as the treasury secretary, he was the COO of Citigroup. Peter Orszag, serving in both the Clinton and the Obama administrations, got a job at Citigroup, and later at a financial advisory firm, Lazard. William Daley, also serving in both of these administrations, had a job at JP Morgan in between. The revolving door also counts outgoing appointments. Goldman Sachs pushed Robert Rubin, Henry Paulson, and Steve Mnuchin to become treasury secretaries under three different presidents: the Clinton administration, the Bush administration, and the Trump administration. All three came from top executive positions at Goldman, with Paulson being the CEO for six years before taking the position. Goldman is famous for having their employees take on central roles in government, even more important than treasury secretary. Their list of alumni in leading government positions include Australia’s Prime Minister Malcom Turnbull, former president of the ECB and later Italian prime minister Mario Draghi, and two former technocratic prime ministers, Lucas Papademos in Greece and Mario Monti in Italy. The list of other, lower-​level government positions, is even larger.

224  Elite Networks The revolving door works equally well for central bank governors. The former governor of the Bank of England, Mervyn King, much to everyone’s surprise given his frequent criticism of the banking industry, became a senior advisor at Citigroup. King’s successor Mark Carney worked in Goldman Sachs for 13 years before first becoming the governor of the Bank of Canada, and then taking over the Bank of England. The long serving US Fed Chairman Alan Greenspan took various advisory roles on Wall Street after his term ended. Ben Bernanke, his successor, took advisory positions at the market maker firm Citadel and investment company Pimco. These are just the most famous and easily traceable cases. Many more people in every country move regularly in between the industry and public service. And it is certainly not limited to finance, with Big Tech also entering the arena as their fates are becoming more dependent on political decisions. None of this is illegal or illegitimate. It isn’t even unethical. It makes sense to hire people with deep industry experience to a delicate task such as the treasury secretary or central bank governor. It also makes sense that these individuals want to cash in their vast knowledge, experience, and networking gained from being in the position of political power. After all, which firm wouldn’t want to have a superhub with all their powerful connections as their advisor or on their board? The problem with this practice, which can never be forbidden or broken up, is that former (or future) top industry executives are highly unlikely to significantly constrain the power of industrial giants, nor are they necessarily in the best position to recognize the difference between what is in the public interest and what is in the private interest. They are necessarily biased simply due to the close proximity of the people they have to regulate. Consider a few examples that showcase how proximity to either corporate or political power affects decisions. During the heyday of the financial crisis in September and October 2008, while the Troubled Asset Relief Program (TARP) was being drafted and conceived, all chief executives of the eight biggest banks in the country held regular formal and informal meetings and exchanged daily phone calls with New York Fed Chairman Timothy Geithner and Treasury Secretary Henry Paulson.31 These meetings and calls suggested deep personal connections of the bank CEOs to Geithner and Paulson, both of whom were heavily criticized at the time for being Wall Street insiders. The deal that struck a nerve, albeit not until March 2009 when it was uncovered, was the fact that AIG paid $12.9bn to Goldman Sachs (in addition to a few other big banks) with the money it got for its own bailout.32 Geithner and Paulson, along with Fed Chairman Ben Bernanke were the key decision-​makers on all policies that were implemented during the financial crisis, including the bank bailouts.33 They did not make any individual-​level allocation decisions over the distribution of the bailouts, but they created the process

The Role of the Firm  225 that would serve the interest of their network at the helm of the banks. The very idea behind the TARP was drafted alongside the big bank executives, which was necessary due to the severity of the crisis. The big banks later claimed they were forced to participate in the TARP to send a positive signal to the economy and complained about the stigma attached to them for participating. However, during the height of the panic in September and October, every single big bank was afraid it might repeat the Lehman or Bear Stearns scenario, and was in high demand for government help. I confirm this in my own research,34 based on a number of interviews conducted with legislators and regulators included in the process. Big banks were not forced to participate. The TARP was by their own design, agreed within their elite network with top government officials in charge of the rescue. The big banks did not engage in illicit or illegitimate activities during this process, they were merely offering advice and complaining to their friends in high office about their dire situation. Naturally the office-​holders reacted with a sweeping policy. Such personal relationships forged between the industry giants and its key regulators can perhaps best be explained by the social connections during a crisis hypothesis.35 According to this theory every person in a position of power has a narrow social network of friends whose opinion they rely on when making decisions. An interesting empirical test of this hypothesis concerns, coincidentally, the appointment of Timothy Geithner as Obama’s treasury secretary. When he was announced to be the president-​elect’s nominee, all firms that were connected to him, either through frequent official meetings or belonging to the same nonprofit boards or business groups, gained an abnormal return on the markets on the first few days of trading after the announcement. Then, when news broke out that Geithner might not get the position due to issues over his tax returns, the same connected firms realized negative abnormal returns on the markets. These abnormal effects are explained via the social connections in a crisis hypothesis, given that Geithner’s appointment was important for a lot of banks in the midst of the crisis. Quick decisions were necessary and having someone you can trust in the new administration is certainly reason for optimism. It is not unusual to observe such outcomes during times of crisis. Research has found that rent-​seeking activities are likely to increase during episodes of great economic distress given that governments tend to increase discretionary spending and media oversight is lower.36 A good example here is the British government’s allocation of procurement contracts to politically connected firms during the COVID pandemic, uncovered by New York Times journalists.37 They found that firms that were politically close to the governing Conservative party, but without much prior experience or even with prior controversies, received $11bn worth of procurement contracts for protective equipment, COVID tests,

226  Elite Networks and ventilators. Half of that amount went to companies that employed former Conservative ministers or gave donations to the party. The decisions needed to be made quickly, bypassing standard procedure, given that the stakes were high (battle with the pandemic). Politically connected firms benefited in this episode of distress simply due to their proximity to political power.

7.4.  Case Study: Political Allocation of TARP Funds during 2008 and 2009 Proximity to political power is clearly important for firms. It is often enough to be in the same network of friends to trigger a decision in the firm’s favor, particularly in times of economic distress. To see this on another intriguing case study, we must once again return to the 2008–​2009 US financial crisis. As emphasized earlier, efforts to find a solution to the liquidity and solvency crisis of the US banking system were initially made within a small elite network of big bank CEOs and the three office-​holders in charge of the crisis. There was no explicit collusion in this case, just a sign that the key decision was made within a group of friends and close acquaintances in positions of power. When this deal was made it created a positive externality for many smaller financial institutions down the line who then used their own political connections to get a better part of the bargain. Lower-​level elite networks were activated by individual banks to gain a short-​run advantage over the competition in the form of a one-​off government stimulus. This is where the interindustry competition for the allocation of bailouts started. The stage was set for massive campaign spending and lobbying by an unprecedented number of financial institutions to ensure that each would get their own piece of the pie. This was the motivation for my paper on the impact of political connections on bailouts given to the finance industry during the crisis.38 The financial crisis offers a particularly good setting for testing the elite networking hypothesis, for two main reasons. First, it was an unprecedented event, where major financial institutions found themselves in the midst of a huge panic in September 2008, many of which worried about their liquidity and even survival (particularly after the fall of Lehman Brothers). This made them desperate for government bailouts and refinancing, meaning that the event provided good motivation for executives to call in favors from their friends in politics, or to increase their campaign donations and lobbying efforts.39 Second, in the midst of the panic and TARP drafts there was an election campaign for the 111th Congress and the 44th president. Elections enable the utilization of close electoral races for individual congressmen within a quasi-​ experimental design. Comparing politicians in very close races (within ±1% of

The Role of the Firm  227 the 50% electoral threshold) means that outcomes of such races are as-​good-​as randomly assigned, given that such candidates are considered similar in ability and all other unobservable factors, and that the outcome might have been affected by a factor of pure chance (for example, rain on election day). As-​if random assignment in this case is the closest we can come to estimating a causal effect (since we cannot randomly assign firms into political connections), at least for bailout recipients that were connected to close electoral winners or losers. The point is to see whether a political connection delivered a better bailout deal to firms that were connected to narrow electoral winners as opposed to those connected to narrow electoral losers. This means we need a good way to measure potential political connections. I do it in three ways: by estimating abnormal lobbying spending40 during 2008 and 2009 for each firm, campaign spending during the 2008 cycle, and most importantly, direct connections between senior bank executives and Congressmen, using the same database of political and corporate connections as in Chapter 4. Direct connections assessed whether Congressmen and bank executives used to work together at least five years prior to the crisis, or whether they belonged to the same social networks (clubs, alumni organizations, etc.). The given three ways of measurement were compiled into a single indicator variable of political connections. When separating bailout recipients into politically connected and unconnected, there was a clear indication that politically connected firms got a better bailout deal, as shown in Figure 7.2. This, by itself, does not carry any causal implication that politically connected firms indeed got higher bailouts, given that we do not know anything about other firm-​specific characteristics that could have driven the allocation of bailouts.41 In order to make these comparisons more realistic and draw causal inference from them, I applied the aforementioned quasi-​experimental design of comparing firms connected to politicians in close races. I used four different methods to show the impact of political connections on the relative distribution of TARP funds. The OLS and matching estimators, both asymptotically similar, used firms as units of analysis and compared bailout packages between connected and unconnected firms. They showcase the same thing as the correlation in Figure 7.2: the effect of a TARP recipient being politically connected on the size of its received bailout funds. A more advanced regression discontinuity design (RDD) analysis, which depends on the as-​if random assignment of firms to politicians in close races, only focused on the subsample of connected firms to generate the firm-​politician unit of observation. It estimates whether a close electoral victory of a connected politician, within a 1% margin, increased the relative bailout allocation for the politician’s connected firm. It measures the effect of an additional politician

228  Elite Networks 22

Log Bailout 2008–2009

20

18

16

14

12 Politically unconnected (P = 0)

Politically connected (P = 1)

Figure 7.2  The distribution of the logarithm of received bailouts for two subsamples, politically unconnected firms on the left, and politically connected firms on the right. The red dot denotes the mean values, while the gray area represents 95% of all observations.

winning a close election, holding constant all other politicians’ performance connected to the same firm. The fourth is the instrumental variable (IV) analysis, which slightly alters the relationship and instead of looking at an individual politician’s electoral performance across each connected firm, it measures the total number of politicians connected to each firm who won or lost a close election. The effect is the impact of having one additional connected politician, who won a close election, on the size of bailout received. The results for each estimator are shown in Figure 7.3, given that each of them has the same dependent variable: relative size of bailouts. The first obvious implication is difference in effect size with the selection on observables approaches and the RDD and IV strategies (OLS and matching effects shown on left axis, IV and RDD on right axis). This should not be surprising given that OLS and matching estimate an overall firm-​level effect, while RDD and IV estimate the marginal effect of an additional connected politician. To make them remotely comparable one should multiply the individual marginal effect of a connected politician to the total number of connected politicians for each firm. For example, in the IV sample the standard deviation of connected politicians who won a close election is 6.8. Moving up by one standard deviation, i.e., donating to seven more politicians who would win a close election,

The Role of the Firm  229 300.0%

20.0% 18.0%

250.0%

16.0% 14.0%

200.0%

12.0% 150.0%

10.0% 8.0%

100.0%

6.0% 4.0%

500.0%

2.0% 0.0%

0.0% OLS

Matching

RDD

IV

Figure 7.3  Comparison of different model estimates on the relationship between political connections and the logarithm of bailouts. The OLS and Matching estimates show a large average effect of around 150% (left axis), but when methods that rely on proving causality are used, comparing in both cases the results of close electoral races (± 1% of the 50% threshold), the effects are lower but more realistic (right axis): a close victory of a connected politician increases the log bailout allocation for that politician’s connected firm by around 4% on average (about 20% higher measured as bailout to total assets); and, firms with more connected politicians winning close elections received about 7% more bailout funds for each connected politician (or about 12% more measured as bailouts to total assets).

would increase the firm’s bailout funds by 47.6%, which is a much larger effect. It is easy to see how this effect grows stronger for firms that had more connected politicians, meaning that those at the upper extremes of the distribution will yield the most benefits from their political connections. The same is true for the RDD estimation. The OLS and Matching can only give us results for the mean values of the samples and are hence much less informative. These joint results suggest that politically connected firms had an advantage in the bailout allocation when compared to politically unconnected companies. This does not imply that the decision-​makers broke the law, or explicitly gave more funds to the banks they were connected to, but it does suggest that it paid off to be part of a congressman’s social network during turbulent times of the financial crisis. Social connections played a key role in the political decision-​ making process, strongly influenced by elite networks. Such relationships represent a spontaneous survival instinct, which determines how elite network members react in preserving their privilege or position of power, an instinct that

230  Elite Networks is particularly strong during times of economic hardship or episodes of widespread panic. The main take-​away from the case study was summarized in the article’s abstract: “it generally pays off for firms to be connected to politics, but in times of crisis—​it pays off even more.”42 *** What was described in this chapter is a close variant of the corporatism Phelps was concerned about. Through their collusion with politics, achieved via personal networking and friendship links, firms are able to gain considerable rents, both for themselves and for their corporate executives. This is, after all, the foundation of the economic theory of regulation. The current literature therefore provides ample evidence on how and why connections between corporate and political interests form and what consequences they entail to society. This book looks at corporate executives and their incentives as just one part of an elite network. Without pandering to politics, they can hardly promote their narrow goals. The explanation offered by the economic theory of regulation and its extension is therefore only half of the motivation to be included in elite networks. Elite networking, in the form of revolving doors between public service and the corporate sector, or relying on powerful friends in times of crisis, or simply gaining access to privileged information or opportunity due to proximity to politics, has always existed. It is not a distinctive feature of the 20th century alone. It is a long-​lasting force of gradual wealth accumulation that persistently pushes inequality upward. As Phelps correctly emphasized, the rise of corporatism in the late 20th century reduced incentives for economic dynamism that created the vast progress after the Industrial Revolution. But this was part of a much longer trend, not a novelty of the 20th century. In Part III, I first show the unequivocal success of capitalist democracies that took time to deliver the greatest benefits to societies. Despite this massive success, it failed to solve the issue of inequality, as it failed to curtail the elite network-​driven persistent force of wealth accumulation. It is important to understand why capitalist democracies succeeded in delivering progress and how the same mechanism can be relied on to also solve the issue of wealth concentration among elite networks. This cannot be done by expanding government powers as elite networks themselves are part of the government. It has to be done differently from what we have tried before.

PART III

RE DU C ING P OLI T IC A L POWE R , T HE RO OT C AU SE OF IN E QUA LI T Y The final part starts with another historical long-​term overview, presenting in Chapter 8 the rise and development of capitalism ever since the Industrial Revolution. Capitalism was the first economic system that produced constant incentives for progress in human societies, progress that eventually led to the origination of democracy. I start by showing how it led to massive, unprecedented improvements in living standards, but I also acknowledge its failures, particularly in the early stages, when its critics did raise valid points. The problem was that the failures typically attributed to capitalism, like inequality, corporatism, or monopoly power, existed long before capitalism and are driven by forces that have existed for several millennia in human societies. There was something different with capitalism, however. Primarily, it offered people different incentives. It was within capitalist systems, even in the early unjust ones characterized by huge levels of inequality and captured states, where democracy originated. The Industrial Revolution was a paradigm-​ shifting event. For the first time ever, people experienced rapid continued progress, something that modern societies today take for granted. Innovations were changing the world faster than ever before in human history. During a single generation people witnessed the incredible power of the locomotive, the steamboat, the spinning jenny, and the rise of machines and early factories. The rate of inventions increased exponentially over time, as did their value to society. It was a signal to the people that the world was changing, that the Old World was in a decline and a New World was rising. The Industrial Revolution, through its disruptive innovations, introduced something that had not existed before—​a persistent incentive for change. It was that incentive for change which inspired major 18th-​and 19th-​century revolutions that gave rise to the concept of basic human rights. It was the incentive for change which delivered equal representation, rule of law, equality before the law, equality of opportunities, and eventually equality of gender and race.

232  Elite Networks All of these changes still failed to solve the issue of inequality, as presented in Chapter 3, because inequality is driven by different forces. It is a persistent phenomenon rooted within the misuse of political power. That, however, does not imply it is an issue that cannot be solved. On the contrary. In order to solve the problem of unacceptably high inequality, to offer a solution, we need to focus on the root causes, not the consequences. Standard proposals like taxation and redistribution, covered in Chapter 9, deal with the consequences. We need to dig deeper and think about how we can continue down the path of societal progress to change incentives for the accumulation of power. This is why an initial brief lesson in history presented in the next chapter is important to understand how we can move forward in designing reform. Chapter 10 presents a set of structural reforms. These are defined through Three Levers aimed at lowering centralized political power and curbing the forces of persistent accumulation of wealth and power among elite networks. This unique set of reforms is arguably only possible to achieve within developed democratic capitalist systems. A few clarifications are necessary here. First, the focus is on developed rather than developing countries given that developing countries typically still have a problem with achieving basic state capacity to deliver public goods. Once they enter into later stages of development and achieve a functioning state with arguably higher levels of inequality (following the Kuznets curve), only then can they move toward enacting long-​term incentive shifts that help curb elite network power and wealth accumulation. These forces are difficult to curtail while a country is undergoing rapid development, however they can be controlled through well-​designed institutional mechanisms that build interpersonal trust and social capital, crucial factors that help maintain functional communities and societies. Second, the focus is on democratic capitalist countries, given that democratic capitalism still provides strong incentives for progressive change.1 It evolved through many forms and it can still be improved. That, however, cannot be done via a one-​size-​fits-​all approach or through one magical policy change. It needs to be done by enacting several simultaneous reforms that strengthen one another and lock the system in so that it cannot regress to the previous steady-​state. Hence the usage of the term “lever” to introduce the reforms—​what I propose are small yet powerful changes that deliver indirect incentives to change the system over the long run, thus helping the democratic trial-​and-​error process achieve its full potential. The Three Levers proposed in Chapter 10 provide a fitting conclusion of the book, as they focus on second-​and third-​order effects that change incentives for selection into politics and for entry into elite networks. The reason elite networks are powerful is due to strong incentives for membership, as shown in Chapters 1 and 5. We must first think how we can reduce these incentives, both

Reducing Political Power, the Root Cause of Inequality  233 from a corporate and from a political perspective. The final chapter is therefore about redistribution. Not redistribution of incomes, but redistribution of power. Finally, decentralization and empowering communities hardly makes sense in societies where the country is already divided between tribal communities and lacks any central authority to raise revenues and provide even the basic public goods. These are typically failed or low-​capacity states whose societies are stranded in poverty. For them the goal is first to reach the next development phase and unleash the spirit of capitalist democracies before being able to use the democratic trial and error. For undemocratic countries the initial problem lies elsewhere. They lack the democratic capital necessary to sustain the benefits of the proposed reforms. They too first need to unleash the idea of democracy and follow its trial-​and-​error process before engaging in reforms aimed at curtailing political and corporate power. Changing incentives for power and wealth accumulation at the top is impossible in autocratic or anocratic countries as their very existence is contingent on a narrow elite network.

8

Capitalism and Democracy 8.1.  How Far Have We Come? The world today is in a state of constant progress. An unprecedented rise in living standards across the globe is a testament to this. Life expectancy at birth globally went up from 46 years in 1950 to 74 years in 2019 (for the developed world, life expectancy at birth is close to 80 years, up from 62 years in 1950). In every single country life expectancy is much higher today than at any time in history, with many developing countries doubling life expectancy over the past half century. Child mortality is less than 1% in the developed world and down to 3.9% globally, the lowest ever in the history of mankind, whichever country we look at. The number of people worldwide that live in extreme poverty (less than $1.90 per day) is also lower than ever, down to 10% from 40% back in 1980. While the majority of this decrease can be attributed to China and India lifting a billion people out of poverty in the past 40 years, even in Africa and across Central and East Asia we are witnessing a steady decline of people living in extreme poverty and a significant decrease of the poverty gap (the amount of money a poor household needs to reach above the poverty line). Obviously correlated with this are much higher levels of GDP per capita across the globe, especially among the developed nations. In western Europe GDP per capita increased from $6,000 in 1950 to $40,000 in 2016. In North America the increase was from $15,000 to $53,000 during the same period, led by significant increases of productivity as a result of rapid technological progress. Globally GDP per capita went from $3,800 in 1950 to $14,500 in 2016, and it is reassuring that almost every country experienced progress on that front, particularly in the past 20 years when African countries started growing more robustly (Figure 8.1).1 These global trends have also had an impact on the quality of life, especially in the developed world. People are working fewer hours (and have much more time for leisure, which is an important measure of life quality), consuming more calories with a much more diverse composition of diets, and there are lower instances of famine and undernourishment than ever before in the history of mankind (Figure 8.2). One consequence of this is that people are taller: men went from an average 5’4” (162 cm) in 1896 to 5’7” (171 cm) in 1996, while women grew from an average 4’11” to 5’3” (150 cm to 159 cm) in the same period worldwide. The rise in Europe and North America was even greater, from an average 5’5” to 5’10” Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0009

236  Elite Networks (a)

(b)

Figure 8.1  Life expectancy from 1770 to 2019 (panel A), and Life expectancy vs. GDP p/​c in 2018 (Panel B), where the size of a circle shows population size for each country. Life expectancy has been on the rise since the mid-​19th century for the developed world, while the developing world started catching up in the early and mid-​20th century. There is a clear positive correlation between Life expectancy and GDP p/​c, and the historical trend points to these two moving up simultaneously ever since the 19th century. Source of data: Our World in Data, Life Expectancy, Roser et al. (2019).

(d)

(c)

Figure 8.2  Share of world population living in extreme poverty (1820–​2018) (Panel A), Annual working hours per worker (1870–​2017) (Panel B), Homicide rates across western Europe (1300–​2020) (Panel C), and Change in mean male height, 1896–​1996 (Panel D). Source of data: Our World in Data: Hasell et al. (2019) “Poverty”; Herre et al. (2019) “Homicides”; Giattino et al. (2020) “Working Hours”; Roser et al. (2019) “Human Height.”

(b)

(a)

238  Elite Networks (165 cm to 177 cm) over a 100-​year period.2 Even acts of violence, that persistent feature of human DNA, have been significantly reduced in developed capitalist democracies: from an annual rate of 500 murders on 100,000 people in prehistoric times (25% chance of getting killed in a 50-​year life span), to 50 murders on 100,000 people in the Middle Ages (2.5% chance of getting killed in the same life span), to less than 1 murder on 100,000 people (less than 0.05% chance of getting killed in the same life span). We are richer, taller, live longer, eat more and better food, have stopped killing ourselves as much, and have a richer social life than at any time before in recorded human history (Figure 8.2). With this in mind, the aforementioned issues of dysfunctional democracies seem miniscule when compared to the level of progress mankind has made only in the second half of the 20th century. Even inequality, one of the central issues facing modern societies, seems to be less odious when taking into account that people are richer and are living better lives. After all, inequality today is arguably lower than it was before the 20th century when the top 10% of the population—​ the kings, the nobles, and the clergy—​held 80% to 90% of all wealth. This was extreme inequality that we no longer observe today, at least not in developed capitalist democracies. Today, in the most unequal developed countries the share of wealth of the top 10% of earners does not pass 50% of total income and 70% of total wealth (with the United States being the worst example, other developed countries have lower or much lower shares of income and wealth concentration at the top). This is still high and certainly problematic, but is on average nowhere near the 90% of wealth concentration in the hands of the powerful ruling elites, a level of inequality at the end of the 19th century in Europe and the United States. The question is still daunting. Why has inequality failed to exhibit the same long-​term trend as declining poverty and rising living standards? Despite the significant progress we have achieved as a human race, we are still witnessing levels of inequality which the majority of society perceives as unjust and simply too high. Typically, the argument made by critics of capitalism is that capitalism as a system is prone to massive inequality. But this argument fails on two basic facts: (1) capitalism is not only present in Anglo-​Saxon countries that have high(er) levels of inequality. It is also present in Scandinavian countries and across continental western Europe with traditionally much lower levels of income and wealth inequality than anywhere else in the world (varying obviously from country to country). Inequality can also be high in socialist countries, both historically (the Soviet Union and various Eastern European countries during socialism3) and today. Figure 8.3 compares top 10% of income share in China, Vietnam, and Venezuela, three countries that are constitutionally defined as socialist,4 to the United States, the United Kingdom, and France, three

Capitalism and Democracy  239 Top 10% share of pretax income for capitalist vs socialist countries, 1978–2019

0.55 0.5 0.45 0.4 0.35 0.3

1978

1983

1988

1993

1998

2003

2008

2013

United States

China

Vietnam (since 1990)

Venezuela (since 2002)

United Kingdom

France

2018

0.25

Figure 8.3  Top 10% share of pretax income for selected capitalist and socialist countries, 1978–​2019. Source of data: World Inequality Database (WID, 2020), https://​wid.world/​.

different capitalist countries of which the United States is a capitalist democracy with the most worrying inequality trends. On average, inequality is worst in the three socialist countries, where top 10% share in income is 41% and 42% in China and Vietnam, respectfully, 48% in Venezuela, while 45% in the United States, 35% in the United Kingdom, and 32% in France. Expanding the sample to include other capitalist and socialist countries, the inequality numbers do not improve at all for socialist countries. To be fair, top income inequality in Venezuela has a declining trend, while China and Vietnam have stabilized their top 10% share of wealth over the past decade, although still over 40%, larger than in any European country. Inequality is therefore not contingent on a particular economic system. Second (2), inequality, particularly as a consequence of lack of opportunity for the masses, was higher or as high before capitalism as the levels of inequality we observe today in capitalist democracies. The long historical overview in Chapter 2 painted this picture very diligently. Inequality existed before capitalism and it persists in capitalism. Capitalism is not its root cause. A standard debate between advocates and critics of capitalism usually revolves around the former citing the great successes we enjoyed as a society ever since the Industrial Revolution, while the latter typically cite high levels of inequality, corporatism, and the issue of climate change. These things are, however, incomparable. Declining poverty and rising living standards are not driven by the same forces as inequality or corporatism. Poverty and living standards were affected by

240  Elite Networks the rise of capitalism after the Industrial Revolution, but we failed to see any rapid improvements until after World War II when societies started moving toward more inclusive capitalism and more accountable democratic governance. There is no doubt that capitalism has brought incredible success in lifting people out of poverty and rising living standards, being responsible for all of the aforementioned long-​run trends over the past 70 years, sometimes achieved even without democratic accountability (e.g., China, Singapore). But capitalist democracies could not solve the inequality issue. In fact, this is where the critics of capitalism, Marxists in particular, have a point. Capitalism in its early post–​Industrial Revolution age, in the beginning of the 19th century, gave impetus to violence power incentives. The exploitation of workers, even child labor, led to huge levels of inequality where the masses did not see any improvement in living standards. On the contrary, they were worse off. Their living standards were terrible; working in the fields was replaced by working in factories or mines for longer hours and without any consideration for health and safety. People were an expendable asset.5 In monetary terms they were still poor, their purchasing power immensely low, while the elites who held the factors of production were richer than ever. The problem, however, was in the distribution over the allocation of ownership over the means of production. Early capitalism evolved from mercantilism, where the ruling elites determined the ownership over land and, consequently, capital. As a matter of historical customs, one could not own a large company or be granted monopoly rights over exports or imports of a certain good without having the backing of state power. The relationship was highly reciprocal, as suggested by the Tammany Hall practice of merging politics and business into one, cited in the introduction. These were the early stages of capitalism, all marred with examples of exploitation and conglomerates growing on state power and proximity of its key actors to the state. Marx was correct in recognizing the unjustness of the early capitalist system, but he failed to understand the driving forces behind such a system. This was nothing new in the wheels of history. This was a powerful force that adapted to a new paradigm of societal development. Early capitalism thus treated people no better than feudalism or mercantilism. The masses were an expendable resource; cheap labor whose role was to serve the rulers, the nobility, and, in the 19th century, the owners of capital. Social mobility was extremely low. Opportunity to advance in life was only possible if a person was part of an elite network. Even the most intelligent scientist or competent discoverer could not prosper without help of a wealthy patron. Early capitalism in the 18th and 19th centuries simply continued the same practices of feudalist times, driven by the most vicious examples of the violence power principle. This time however, one thing was different. An incentive to change.

Capitalism and Democracy  241

8.2.  Persistent Incentive to Change The Industrial Revolution changed the paradigm of how people think about the future. In Malthusian times people were primarily focused on survival, on achieving subsistence. The Industrial Revolution brought a condition that was unknown to societies before that—​rapid progress, or as economists more popularly call it—​rapid economic growth, the central pillar of capitalism. This was a time where, within a single generation, people could see the effects of economic growth unfolding before their very eyes, witnessing progress like no other generation ever before them. Every single innovation that came as a consequence of the Industrial Revolution—​the steam engine, the locomotion and the railway system, steamships, machines, chemicals, cement, gas lighting, and many other innovations that improved everyday life—​all of these were a testament to rapid economic growth. People were stunned by new innovations. Many saw this era of growth as an opportunity to achieve life beyond subsistence and advance in the world, a motivation that did not exist in Malthusian times other than through conquest or elite network connections. Furthermore, there was a change in working patterns. As agricultural jobs were replaced by industry jobs, working conditions changed and societies adapted. The first consequence was rapid urbanization as people no longer worked in the fields but needed to be located closer to their workplaces—​the factories in big cities. This led to the rise of the so-​called industrial proletariat, the dawn of the modern-​day workforce, but it also signaled a long switch from the family unit to other forms of social organizations, including the market and the state.6 There was also the introduction of universal time in order to define working hours in factories. This may seem like a minor change, but the concept of tracking time changed the way people thought about work and how they made their plans. Hence the most important change: the way people started thinking about the future after witnessing progress around them. They started thinking about the world they were leaving for future generations. Economic growth enabled planning for the future because there was finally a future to look forward to. This is where money and credit took societies to new highs, and cemented the basic capitalist model for the ages.7 In Malthusian times the world was a zero-​ sum game as there was no consistent economic growth. Societies certainly did develop, and innovations happened, but development was very slow, long-​term economic growth was miniscule, and the lasting effects of significant innovations or societal changes were almost never witnessed within a single generation. In the preindustrial age, economic incentives were severely constrained. Getting a loan to start a business was difficult but possible—​e.g., merchants in Italian city-​states used loans since the 11th century to finance their ventures8—​ however, credit creation was severely constrained due to two reasons: (1) lack of

242  Elite Networks trust (lat. credo) and no possibility of planning for the future, and (2) Christian and Muslim religions declaring money lending and charging interest as usury, limiting it to Jewish lenders who were allowed to lend money to non-​Jews.9 Their capacity to do so on a massive scale was obviously limited, particularly since the Catholic Church used the usury allegations to prosecute Jews. Any significant access to finance was therefore limited to elite network members who were the only ones who could provide collateral to justify a loan. Common people with good ideas lacked the means to do that, nor did they have time for creativity and innovation. They were busy achieving subsistence. It was a negative feedback loop: the poor majority had a very limited capacity to access finance in order to change their living standards through exploring new opportunities and opening businesses. No new businesses and no new opportunities meant they remained in feudal poverty, having their already low wages being further depleted through taxation. Rulers and governments thus had an upper limit on tax revenues, prompting them to choose a different strategy for enrichment based on the ideas of mercantilism and conquests of foreign lands: maximizing the intake of gold bullion from abroad was the most efficient way to increase the size of the pie. Economies did not grow endogenously through innovation. They grew through conquest and violence, as did the wealth of the elites. The Industrial Revolution changed this by opening doors to credit expansion and the emergence of capitalism. Rapid economic progress and improving living standards from one generation to the next established trust in a better future. As more people started planning and saving for a future that was becoming less uncertain, the newly attained level of trust increased incentives for credit expansion. Credit was increasingly becoming available to the rising class of merchants, progressively more so than ever before in history. It was also, obviously, more available to the nobility and the rulers who used it to finance wars and new conquests. In addition to credit expansion, capitalism brought new virtues, new ethics of having production being reinvested into more production, ensuring that the current levels of growth persist. It was a true revolution in business. Production kept growing on good foundations—​products were hitting markets at more affordable prices and early consumerism was born. As businesses grew, they hired more people. As they continued to grow, they paid more money to their employees and hired even more of them. All this extra money, plus the availability of credit, made the common folk capable of buying stuff too. Stuff they could never have owned before: houses, vehicles, better food, better clothes, better appliances. The consumerist mentality bounded well with the idea of constant growth and the mass availability of credit. Its end result was an empowered population, first economically and soon politically.

Capitalism and Democracy  243

8.3.  The Gradual Shift to Democracy Thus, we arrive to the most important consequence of the Industrial Revolution. The shift toward democratization. A political consequence of constant economic growth, rising incentives to innovate, and major shifts in patterns of behavior were social upheavals and the many revolutions throughout the next 200 years. People started demanding representation and equality. The gradual shift to democracy had commenced. In England this shift was indeed gradual. It was a 150-​year-​long transition driven by ongoing reform and expansion of the franchise on four occasions (three Reform Acts in the 19th century and the Representation of the Peoples Act in 1918) to gradually grant more and more people voting rights. The process started with the Glorious Revolution of 1688, when Parliament was granted more power than the king, and ended after World War I, with the extension of the franchise first to all men over the age of 21 and then 10 years later to all women over the same age. Interestingly, the English Glorious Revolution is an example of discord within elite networks, an occurrence that happened often throughout history and was typically a catalyst for social change. The nobility rebelled against the king, demanding greater constraints and lesser power from the monarch, and greater representation for themselves. Their success back then initiated a long process of social changes that started delivering greater representation to wealthier men during the 19th century, culminating with universal suffrage in the 20th century. In France the changes were more abrupt. After the French Revolution in 1789 and a decade-​long power vacuum during which the fabric of society was radically changed through a series of long-​lasting social reforms, Napoleon continued the legacy through his European conquests. Even after his resounding defeat the legacy remained. In every territory they conquered, following the ideals enacted after the Revolution, the French brought massive institutional changes that were aimed at repealing the old feudalist regime and set the stage for further expansion of capitalism and, eventually, democracy. They introduced the Code Civil, abolished guilds, initiated land reforms, abolished slavery, and emancipated religious minorities (in most cases the Jews). Even in what was a very short-​lived reign they managed to make drastic changes10 that motivated further revolutions across Europe in subsequent decades, culminating in the famous Revolutions of 1848, also known as the Spring of Nations. These revolutions, led by a temporary coalition between the rising bourgeoisie and the impoverished workers, simultaneously took place in almost all major European empires of the time: France, the Netherlands, the Austro-​Hungarian Empire, the German Confederation states, Italian states,

244  Elite Networks Denmark, Belgium, Poland, etc. Historians describe the 1848 Revolutions as a mixed success at best, but the stage was set for even greater demands for social change. The world, however, had to go through two world wars to finally get there. With the rise of democracies, particularly during the 20th century, capitalism went through its final transition. From an unjust system that exploited labor and encouraged violence power incentives that perpetuated the forces of wealth accumulation, it gradually evolved into a much fairer system. The impoverished workers first organized into unions and soon after into political parties to represent their interests. They successfully enacted labor laws and regulations protecting their rights and banning child labor, thus irreversibly changing the way owners of capital treated their employees. Over time people started demanding greater access to education and information, which increased the role and importance of media in society, not to mention the rapid development of science and various elements of social security (health, pension, unemployment), all of which positively affected economic growth. As the population was getting wealthier, the demand for all these factors—​quality education, health care, information, science—​kept increasing, thus creating a virtuous cycle of development. Citizen empowerment brought about demands for greater participation in collective decision-​making, thus increasing the legitimacy of governments and rendering a positive effect on social stability and the growth of democratic capital. All of these changes happened within evolved capitalist systems.11 It was possible under democratic capitalism to fight for and achieve greater human rights. The evolution from Malthusian times of subsistence into modern-​day democratic capitalism had to go through the early stages of injustice and inequality. The ugly early capitalism was evolving from an odious system to begin with. It cannot bear the burden of forces that preceded it. But it did manage to subdue those forces and gradually, over 200 years, ensure that people live in much richer, fairer, and equal societies. Obviously, democratic capitalism didn’t solve all problems. Many of them still persist even in countries with a long-​lasting democratic tradition, as shown repeatedly throughout the book. We criticize democracy as a system of government so often that we tend to forget the incredible accomplishments achieved in our quest for democracy and its long-​run manifestation. We tend to forget the fact that capitalist democracies have a tendency to change for the better. No other economic system in human history had this feature. Often along the way they may result in suboptimal outcomes and social injustice, even result in wars and revolutions, but the story ever since the Industrial Revolution has been one of persistent progress and an attempt to change the system for the better.

Capitalism and Democracy  245

8.4.  Democracies as a Trial-​and-​Error Process The democratic element in a society carries a strong incentive to change and improve itself. Although we can often see democracy’s deficits in the short run—​characterized by things like political gridlock, corruption, bureaucratic inefficiencies, interest group state capture, vote buying, gerrymandering—​over the long run a democracy undergoes persistent trial and error before arriving at prosperous and just social outcomes. Democracies have a multitude of hidden strengths, against which an authoritarian regime ultimately often fails. Democracies, as well as capitalism itself, strive on the idea of competition. Competition in the form of selection of different alternatives implies more creative solutions when it comes to dealing with challenges. Certainly, this sometimes implies poor judgments being made and dubious outcomes arising as a result, but as the market system, democracy too relies on trial and error. Thinking of the nature of our society, we all thrive on this system. It teaches us not to repeat the same mistakes of the past. A good example is the underlying idea behind the European project—​never again to allow a war on the continent. Thus far the idea has been quite successful (at least among EU countries12), however even with such noble ideas it takes time for them to work and prove their resilience to outside shocks and disturbances. This is why democracies appear to be more fragile than they actually are—​ they are characterized by a complex decision-​making process, which need not always yield the best possible outcomes immediately. Where one can see a sign of weakness (gridlock, slow responsiveness, negative political selection), this may just be a short-​term response of the system to some earlier made errors. When too many of these errors pile up, they congest the system (corruption, cronyism, vote-​buying, the sovereign debt crisis) and signal to the electorate that things need to be changed. It will take time before the electorate recognizes the correct set of ideas and people to solve the piled-​up errors, but at one point they will succeed in doing so. Or at least, within a democracy, voters will be given this option. In the process other errors will surely be made, but due to even closer scrutiny and transparency from the electorate the process will inevitably end in success. As it did many times before despite striking opposition and doomsday prophets. The key to a successful democracy reached by a series of trial-​and-​error processes in Western history is to erect institutions that will limit political power. Democracy mustn’t turn into a tyranny of the majority. This is why political freedoms are crucial and why institutions that enable transparency and scrutiny are essential in ensuring a long-​run survival of democracy. Particularly for a country new to the idea. Many new democracies get preoccupied with elections and fail to design institutions that prevent some of democracy’s main failures. They fail to establish a good constitutional system with an emphasis on the rule of

246  Elite Networks law. This is a typical reason for early failure of democratic consolidation. Their trial-​and-​error process is arguably going to be much longer, particularly if new democracies quickly descend back into authoritarianism. When long-​lasting democracies exhibit signs of failure their trial-​and-​error process is quicker and such countries will resolve their structural problems relatively faster, despite the perception of difficulty to change a congested system. Which brings us back to the initial point: it is much easier to change a faulty system when you have the possibility to do so. In a democracy people have an option to change an unjust and corrupt system, and they often seize that right. The great victories for liberty during the civil rights movement or the emancipation of women, both of which signaled an end to a long period of unfair and unequal societies, testify to this argument. It took more than one hundred years from Lincoln’s abolition of slavery for the African Americans in the US South to gain equal rights. In the mean time they suffered great inequality and almost zero social mobility, not to mention the lack of some basic human rights. Even today, while certainly better off than in the 1960s, African Americans still yearn for greater economic equality and feel their human rights aren’t being respected. All of this is happening in a country that was supposed to be a prime example of an accomplished and fair democracy. This simply shows that those periods were times of “error,” and it took a long time for them to get fixed. Democracy, just like an economy, always converges toward an optimal equilibrium but is never quite there. It presents a persistent notion of positive change. Along the way, during the convergence process, there will always be errors and concerns. Sometimes the trial-​and-​error process will seem to last too long, but eventually it always triumphs. It will make democratic countries look weak, but it is arguably the biggest advantage democracies have over any other system of governance, which is precisely why they will always be successful in fixing their problems.

8.5.  Socialism’s Rapid Industrialization with a Lack of Trial-​ and-​Error in the Underdeveloped Periphery Thus far we have only focused on the democratic capacity building in the countries of the so-​called core,13 where capitalism first emerged and produced the aforementioned incentives for democracies. What about the underdeveloped periphery where, instead of capitalist democracies, socialism took root and molded societal development in an entirely different fashion? In periphery countries, which adapted socialism in the 20th century, democratization came at a very late stage, after the process of rapid industrialization and urbanization, not alongside it as in the core countries. How successful was socialism in delivering the same

Capitalism and Democracy  247 socioeconomic outcomes (in terms of the breakout from the Malthusian trap), existing in parallel with capitalist democracies in the 20th century? Judging the core and periphery countries by the same standards would be wrong. During the early onset of capitalism in the core and its political consequences throughout the 19th century, countries of the periphery were in a lower stage of development. The economic historian Alexander Gerschenkron refers to periphery countries as being characterized by economic backwardness, meaning that such countries go through different stages of catch-​up development than the economically developed countries. He primarily compares Russia, an example of backwardness, and Britain, the frontrunner of economic development. He finds that when backward countries embark on industrialization we witness higher rates of economic growth, a more active role of the government to stimulate growth, greater emphasis on capital-​intensive rather than labor-​intensive production, importing technology from the more developed economies instead of innovating to produce their own, and lower living standards during this rapid process.14 This is similar to the economic theory of convergence. Countries with lower levels of initial development (lower GDP per capita) grow at much higher rates in order to catch up to the more developed countries. This is what happened in almost every peripheral country under socialism: periods of rapid industrialization and urbanization, very often driven by the state to promote quick progress, which led to higher initial levels of economic growth, until the economic model ran out of steam for more structural reasons, namely lack of innovation to sustain progress. Countries in the periphery did not therefore experience a gradual shift from imperialistic mercantilism to a capitalist democracy. Their path was to move from a mainly agricultural economy with very little industry or infrastructure, with no colonial exploits to benefit from (many were colonies themselves), to a phase of rapid industrialization. They never went through the innovation phase of rapid progress that shifted the mentality of the population, nor did they gradually offer incentives for common people to build their fortunes and, over time, demand political change. Their rapid growth in the 30 post–​World War II years (from the 1950s to the 1980s) fast-​tracked 200 years of gradual progress that happened in the core. By the time the economic model ran out of steam, the empowered population demanded changes, and the revolutionary wave of the 1980s demanding democracy had begun. First in Poland, and by the end of the decade in almost every Eastern European country: Romania, Hungary, Czechoslovakia, East Germany, Bulgaria, and Yugoslavia, where the breakup was followed by a series of wars in the newly founded Croatia and Bosnia and Herzegovina. The Soviet Union reached its own dissolution in 1991, and the era of transition from socialism to capitalism had officially begun.

248  Elite Networks Note that this scenario concerns countries of Eastern Europe and the former Soviet Bloc, the most emblematic examples of socialist development in the context of a peripheral country. Peripheral countries in other parts of the world (Africa, Latin America, the Middle East, and parts of Asia) had a colonial heritage to break out of, where socialism was a particularly attractive alternative to the imperialism that continuously exploited them. They thus experienced a different path of development, resulting mostly in two extreme alternatives: a weak state with no capacity to organize a society, or an authoritarian state with little incentives for economic progress. Given that a number of these countries are either nondemocratic or have low state capacity, and given that a lot of them are still agrarian societies, there is no final outcome of the transition to which we can make the comparison. I will therefore limit the rest of the analysis only to Eastern Europe and its path to capitalist democracies. The closest to their experience were Latin American countries, particularly with respect to very similar postcommunist economic outcomes, however given their complex colonial origin and different historical patterns of development, a better comparison to the Western European core is the Eastern European periphery.

8.5.1.  Condensing 200 Years of Gradual Progress into 30 Years of Rapid Growth How did socialism deliver rapid industrialization and what consequences did this render for the peripheral countries of Eastern Europe? While early capitalism and early industrialization were in full swing in the 19th-​century European core, the periphery was still mainly an agrarian society, falling behind rapidly as it managed to elude the innovations of the Industrial Revolution. The Russian Empire, the Ottoman Empire, and partially even Spain and the Austro-​Hungarian Empire had ruling elites that saw a major threat from progress of the Industrial Revolution and actively worked on suppressing it and preventing rapid industrialization.15 With economic backwardness the countries of the periphery stayed entrapped in monarchist systems until World War I. A good proxy for development of those early industrial times is the extent of the railway network and population size of major cities. The shift from agriculture into industry significantly expanded the cities. As more factories were being built, usually around city centers, this attracted more and more people from the countryside, looking for new industrial and manufacturing jobs, to live in cities. Hence urbanization always followed industrialization. The railway network served a dual purpose in this rapid development; it provided a fast link between the factories in big cities and the country’s major sea or river ports, and it was an

Capitalism and Democracy  249 important connection of internal markets. The people and the economy were moving fast and the railroad network was their bloodstream. Figure 8.4 shows that by 1880, Britain, the forefront of industrial development, had a deep network of railroads, while many peripheral countries either had no railroads at all or had only a few bigger cities connected this way. Even among the core countries it is visibly obvious where development was faster, with France, Germany, Switzerland, and the Benelux countries following closely behind Britain. The famous Orient Express, established in 1883 with a goal of connecting Paris to Istanbul through Vienna, had to use a 14-​hour ferry ride from Varna in Bulgaria across the Black Sea to reach Istanbul. The line was completed in 1889 to go through Belgrade, although passengers still had to take carriages in parts of Bulgaria to transfer to other trains before reaching their final destination. By 1875 London already had a population of 4.2 million people, Paris had 2.2 million, and Berlin 1 million. Moscow had 600 thousand and Istanbul 870 thousand, while the smaller dominions throughout the periphery had their biggest cities at less than 100 thousand people. It was a clear lagging behind for the periphery. It is thus no wonder that the idea of socialism, calling for a new world order, took root much easier in peripheral economies, entrapped in agrarian societies and authoritarianism. It failed to develop in countries of the core even though socialist ideas emerged from the core (England, France, Germany) as a response to the ugly early stages of capitalism. It also took roots in the outside-​European periphery where its ideals were accepted as a way of breaking up the imperialist exploitation from the core. Interestingly, this was far from the scenario envisioned by Marx and Engels in the Communist Manifesto. They considered the peripheral countries of Eastern Europe too backward to ever accept the revolutionary ideals of social progress. Not to mention the countries outside of Europe and the colonies. Their goal was clear: a proletarian revolution was supposed to take place in Germany because of their early lag behind Britain and France in the 19th century.16 But socialism became particularly attractive in underdeveloped countries, and for a good reason—​it offered a way to catch up with the developed countries of the core, particularly given the Russian experience. Russia was the first country in the world to embrace communism, instituting it violently through the Bolshevik October Revolution in 1917. Notwithstanding other attempts at communist revolutions in Hungary, Germany, Spain, and outside of Europe during the interwar period, it was not until after World War II that socialism became widely adopted across Eastern Europe and the rest of the developing world. The Soviet Union was the central power and it held the utmost authority over interpreting the Marxist-​Leninist doctrine, particularly over countries in Eastern Europe, with the notable exception of Yugoslavia after 1948 and the Tito-​Stalin split.

250  Elite Networks

Figure 8.4  The railway network in Europe in 1880, offering a visual representation of the difference between the most developed societies (Britain, France, Benelux countries, Switzerland, Germany) from the rest. Note how even the Austro-​ Hungarian Empire, Spain, and Italy failed to develop the same infrastructure network by that time, while countries even further on the periphery had almost no railway links at all. Source: Morillas-​Torne (2012).

Politically, the communist doctrine recognized an authoritarian system that was necessary to fast-​track industrial development and shift the underdeveloped agricultural countries into modern industrial strongholds, perfectly capable of competing with the Western core. Or at least, that was the

Capitalism and Democracy  251 narrative. Killing and imprisoning dissidents, landowners, the aristocracy, certain members of the old elites, or anyone who opposed the doctrine, was justified as purging the country of the enemies of progress and the legacy of the revolution. This ideological justification is particularly interesting when considering contemporary politically socialist countries like China, which uses the same justification to promote what is in essence capitalist economic growth (hence the reason why China’s model is political capitalism,17 not socialism). Nevertheless, the authoritarian political system was successful in achieving robust economic growth via a rapid, state-​managed transition from agricultural jobs into industrial jobs. Sometimes this led to disasters, like in China during Mao’s Great Leap Forward, where people were given unrealistic production targets and were forced to tear down parts of their houses to meet the metal export quotas, or faced starvation to meet the agricultural quotas. The Soviet Union also underwent episodes of famine and scarcity in the interwar period, however as they completed their switch to an industrial economy in the 1950s, stronger economic growth was an obvious consequence. Eastern European countries embarked on the same path from the early 1950s when the shift began, although to a varying level of success determined by the prewar levels of development, and with differences in how the regime was established. In some countries the revolutionary movement was endogenous, developing internally while rising to power, while in others it was imposed exogenously, primarily through Soviet occupation in the immediate aftermath of the war. The consequence was that the new socialist regime was considered as imposed by a foreign power rather than authentic in many Eastern European countries, which created a strong demotivating factor within the population. Countries that developed their socialist governance authentically had a much stronger internal justification and greater support from the majority of the population. This affected the postwar development given the different motivational factors in rebuilding the country. Table 8.1 shows the cumulative GDP per capita growth in the 30-​year period (1950 to 1980) of rapid growth in the periphery compared to the current European core countries and the so-​called Eurozone periphery countries, followed by a 10-​year decline and stagnation (1981–​1990) in the final decade of socialism (only Czechoslovakia was marginally better in 1990 than it was in 1980), and then the next 20-​year period, where GDP per capita first went down in the early stages of transition, recovering only in the late 1990s and early 2000s for the 10 new EU member states (those entering the EU in 2004). I show GDP per capita rather than pure GDP growth, as the per capita values are a better indicator of how economic development benefited the population.

4.50%

4.56%

Poland

Czechoslovakia

Hungary

Yugoslavia

Croatia*

Slovenia*

Serbia*

Romania

Bulgaria

Average (1–​8)

3

4

5

6

6a

6b

6c

7

8

6.10%

6.30%

6.40%

6.20%

5.20%

3.20%

3.30%

2.90%

3.30%

East Germany**

2

2.80%

Soviet Union

Av. annual GDP per capita growth 1950–​1980

1

Country

–​0.48%

–​0.70%

0.30%

–​0.90%

–​1.50%

–​0.90%

–​1.00%

0.30%

1.40%

–​1.10%

–​1.90%

0.70%

Av. annual GDP p/​c growth 1981–​1990

3.11%

2.70%

5.70%

1.80%

1.90%

2.10%

–​

3.50%

3.2%***

4.80%

–​

2.3%***

Av. annual GDP p/​c growth 1991–​2010

3,123

2,632

1,046

2,032

3,934

2,738

2,276

3,953

4,736

3,900

2,794

4,529

GDP p/​c in 1950 (USD)

10,973

9,634

5,482

10,724

21,151

14,193

10,037

10,052

12,336

9,149

7,695

10,245

GDP p/​c in 1980 (USD)

10,303

8,922

5,596

9,754

18,093

12,948

9,000

10,296

14,178

8,150

5,402

10,989

GDP p/​c in 1990 (USD)

19,050

14,686

16,377

11,964

26,001

19,511

–​

20,036

25,922***

20,609

–​

16,345***

GDP p/​c in 2010 (USD)

Table 8.1  Cumulative growth of GDP per capita from 1950 to 1980, from 1981 to 1990, and from 1991 to 2018. GDP per capita is in US dollars, 2011 prices. Three groups of countries are compared: the old periphery including the Soviet Union and the entire Eastern Bloc (1–​8), the countries of the current European periphery, defined as such during the Eurozone sovereign debt crisis (9–​13), and countries of the current European core (14–​21). There was no data for individual countries of the Soviet Union, but there is separate data for the countries of former Yugoslavia (for which the opening year is 1952*). The estimates for East Germany are for the years 1950, 1973, and 1990 (**). The estimates for the Soviet Union and Czechoslovakia after 1990 are estimates based on their former member country’s rates of per capita growth (***). This was not done for Yugoslavia since some of its countries are already included in the Table. Source of data: the Maddison Historical Statistics (2021)

3.10%

4.48%

Portugal

Spain

Ireland

Average (9–​13)

11

12

13

2.90%

3.33%

West Germany

Netherlands

Belgium

Sweden

Austria

Switzerland

Average (14–​21)

16

17

18

19

20

21

4.50%

2.70%

3.30%

3.10%

4.40%

3.60%

France

15

2.10%

United Kingdom

14

4.80%

4.70%

5.30%

Greece

10

4.50%

Italy

9

1.86%

2.30%

2.10%

1.70%

1.80%

1.60%

1.20%

1.80%

2.40%

2.58%

3.30%

3.20%

3.10%

1.10%

2.20%

2.16%

2.60%

2.00%

2.20%

1.60%

2.40%

2.50%

1.30%

1.40%

2.72%

4.90%

2.60%

2.00%

2.60%

1.50%

8,996

11,541

5,907

10,742

8,706

9,558

6,186

8,266

11,061

4,185

5,504

3,464

3,325

3,052

5,582

22,743

23,060

21,932

23,809

23,060

23,438

22,497

23,537

20,612

15,141

13,614

14,008

12,822

14,300

20,959

27,986

34,250

26,930

28,068

27,412

27,515

25,391

28,129

26,189

19,509

18,838

19,215

17,526

15,964

26,003

41,706

57,219

40,288

42,635

37,739

43,812

41,110

36,087

34,754

33,431

48,624

31,786

25,463

26,517

34,766

254  Elite Networks Notice a few interesting facts. First, the average annual growth of GDP per capita for socialist countries in the 30-​year period of rapid growth was almost the same as those of the Eurozone periphery (around 4.5%), but somewhat larger compared to the growth of the European core (4.5% compared to 3.3%). This is, as we mentioned, normal and expected according to the economic convergence theory—​countries at a lower level of development converge faster. In 1950, the European core was three times richer on average than the socialist periphery, while in 1980 it was only twice as rich—​hence these countries managed to slightly close the gap. The Eurozone periphery was also closing the gap, and rapidly so, continuing to do the same in the next 30 years, when the socialist periphery entered into its relative stagnation and could no longer keep up. Notice the differences between West and East Germany. In 1950 West Germany was 2.2 times richer than East Germany. In 1980 it was almost 3 times richer, while in 1990, after a decade-​long decline of GDP per capita in East Germany, West Germany was 5 times richer. East Germany grew, but it failed to catch up. Second, the final decade of socialism in every country shows that the economic model was clearly exhausted and entered a period of stagnation and regression. While the Western core and periphery kept their GDP per capita growing at 1.8% and 2.4%, respectively, the Eastern Bloc collectively declined by 0.5% and found themselves relatively poorer at the start of transition than 10 years earlier. The early years of transition saw huge declines in both GDP and living standards across the Bloc, with significant increases in unemployment and poverty. However, by the end of the 1990s and early 2000s, most Eastern European countries bounced back and again started catching up with the West, growing more strongly than other European countries (with the exception of Russia and former Yugoslav countries). In 2010 they were back to where they were in the 1980s with respect to the developed nations, with countries like Czechia and Slovenia even surpassing some of the peripheral European countries. Eastern Europe lost a lot of ground over the 15 years from 1980 to the mid-​ 1990s, but most of them managed to bounce back and reverse all that regression. It took them 15 years to replace the lost 15 years. However, many countries failed to fully catch up and get back on the pre-​ 1980s trends, among them many former Yugoslav and Soviet states. This brings us to the third fact—​there were obvious between-​country differences in the socialist periphery, both during socialism and afterward. During the 30 years of socialist growth, Romania, Bulgaria, and Yugoslavia, for example, were growing in GDP per capita faster than the Soviet Union and the rest of the Bloc, whose countries were actually developing slower than most of the Western core and periphery. Countries under strong Soviet influence,

Capitalism and Democracy  255 with foreign-​imposed socialism, clearly had lesser incentives for strong, endogenous growth. However, even among the fast growers, there were obvious differences. Romania, for example, started from a very low level of development, being a truly agrarian economy in the aftermath of the war. Both there and in Bulgaria, the state’s central planning had an immense role in driving industrialization, however this failed to fully translate into better living standards for its citizens. Heavy industry was indeed strong, but there was no consumerism. As the Eastern Bloc copied the West, they could produce the same quality steel or plastic for example, but nowhere near the same quality automobile or TV. This is why both Romania and Bulgaria continued stronger growth after their transition, as did Poland, Hungary, and Czechia: the rise of consumerism further propelled their economies. Yugoslavia was an exception. It was the first country to break away from dominant Soviet influence and engage a different economic model than the rest of the Bloc. It opened up its foreign trade with the West and the rest of the world, driven primarily by the political influence of Tito and the Non-​Aligned Movement, which opened up Yugoslav exports to Africa, Asia, and South America. Domestically, it opened its borders to foreign tourists (something quite rare in Soviet-​dominated countries), and introduced more consumerism to its citizens than any other country of the Eastern Bloc. Its economic model was a combination of strong industrialization and urbanization initially, but by the 1970s it was an export-​led growth, boosted by the participation of its labor force in the West (which enhanced the flow of remittances, increasing domestic savings). The one common pattern for the entire Bloc was the stagnation and decline in the 1980s up until the mid-​1990s, driven by the imminent failure of the socialist economic model and the effects of transition. In Yugoslavia this was exacerbated by a political crisis from within “initially” driven by “Serbian” nationalism and territorial aspiration which caused a devastating war. Countries most hurt by the war—​Croatia, Bosnia and Herzegovina, and Serbia—​were left clearly lagging behind the rest (the per capita data for Croatia and Serbia are actually driven more by a population decline due to emigration than by robust economic growth). Even 30 years after the fall of socialism, these countries have failed to reach their pre–​1980s trend of catching up with the West. A final implication is the fast pace of development. It took the three selected countries in Yugoslavia only 30 years to increase their GDP per capita by 5 times. In comparison it took Britain 150 years, taking 1820 as the reference point, to increase their GDP per capita by a factor of 5. France needed 134 years to do the same, Austria 136 years, Germany 109 years, etc. The progress that happened

256  Elite Networks across three generations in the United Kingdom or France, happened in a single generation across most of Eastern Europe. This might have even been too fast. During such a quick process of industrialization there was little time for the population to adapt to massive changes, to develop a successful system of checks and balances or the necessary institutions and public goods that would suppress the antimodernization legacy of the agrarian society. Furthermore, given that the modernization process was imposed by communist governments, often violently, there was a necessary pushback that shaped the feelings of people who were wronged or mistreated by the system. Tribal allegiance remained important and could not be modernized during a single generation. People born in the 1940s, ’50s, and ’60s grew up in typical agrarian communities. The socialist baby boomers lived through three industrial revolutions within a single lifetime, and despite accepting the benefits of progress, certain aspects of the old mentality are still present. In the core this progress was much more gradual, allowing people to adapt across generations, with every new generation feeling less burdened by an agrarian, tribal heritage. The modernization process, even in countries whose growth was less impressive, was rapid. Infrastructure investments were the biggest sign of progress: more factories, energy grids, telecommunication grids, ports, roads, and railroads were built in those 30 years than at any time before or afterward. The rate of urbanization picked up significantly, as cities expanded and had to build more residential buildings and supporting infrastructure like schools, kindergartens, and hospitals. A good measure of industrialization is the decline in the share of agricultural population for the socialist countries, which was around 50% on average in 1950 (and over 60% of the workforce), only to end up at around 10% by the end of the 1980s. In three decades, the shift from an agricultural to an industrial society was achieved. However, keep in mind that the same trends of rapid infrastructure expansion also happened in most nonsocialist European countries at the time, particularly those of the current Eurozone periphery, whose growth was also impressive. Heavy industrialization and rapid development of new infrastructure objects was indeed the main trend of the second half of the 20th century. The main difference between the West and the East was that socialist countries started from a much lower initial level of accumulated capital and hence their progress seemed more impressive. Socialist nondemocracies thus managed to achieve the same level of progress as did capitalist democracies, at least in those 30 years. The socialist experiment at the time was thus considered a great historical success. In the 1960s and 1970s Western economists were predicting that it is only a matter of time before the Soviet Union overtakes the United States by size of GDP. The developed West in the 1970s underwent a period of stagflation,

Capitalism and Democracy  257 while the socialist East was still reaping rewards from its accelerated industrialization and urbanization. Without the benefit of hindsight, in the 1970s one would think that the socialist model had no boundaries. But then the music stopped. The socialist growth model ran out of steam in the 1980s. Most socialist countries started incurring large foreign debts to keep the momentum. Inflation was another problem and currency devaluations a constant worry; standard issues of emerging market societies embarking on the “original sin” of not being able to borrow money in their domestic currencies making them more vulnerable to outside economic shocks—​specifically the oil crises of the mid-​1970s and 1980s. The latter in particular signaled the beginning of the end of the socialist economic model as domestic savings and consumption declined and economies entered a decade-​long stagnation. While the Western core started embarking into the Third Industrial Revolution (the IT revolution), and as globalization picked up pace, the rigid central-​planning system simply could not pick up. Socialist governments of Eastern Europe did not know how to lead a transition from an industrial to a postindustrial society the same way they managed to spearhead the shift from an agrarian to an industrial economy. This was true for one basic reason: technological progress and innovation cannot be planned and managed. Planning the shift toward industrialization was possible given that socialist governments copied the processes that occurred previously, over a much longer time period, in England or Germany. Central planning systems can be relatively good at copying but are ineffective when it comes to creativity as they lack the internal incentive and capacity to innovate. As the socialist economic model in the periphery collapsed, the transition toward democratic capitalism began. However, this transition was merely figurative, and once again brought the violence power principle into the fore. The transition was, in most cases, led by the former socialist political elites, all of which with proven failures to successfully manage the previous economic crises. They condensed power and shifted it into new party formations, only formally respecting the newly established democratic standards (such as multiparty elections and new constitutions), while at the same time enabling cronyist control of resources and dubious privatization schemes which favored the politically chosen well-​organized minority (mostly composed from the old communist elite networks). The oligarchs created during this process contributed to a rise of inequality as wealth was concentrated in the hands of a politically selected elite network. The consequences, besides rising inequalities, were increases in unemployment, emigration, and a decline of GDP and living standards. This is hardly surprising, at least according to the research done by Albertus and Menaldo, who show that democracies are created “by the elites for the elites.” They find that during times of democratic transitions the former regime elites

258  Elite Networks determine the rules of the new democracy, rules that primarily serve to promote their interests, rather than the interests of the median voter. They design electoral rules, write the constitutions, and shape all other institutions, such as the justice system, to serve their benefits and keep them in power. The authors find that two-​thirds of democracies were created by the old elites, where these elites could even determine the timing of when a democracy can be introduced, leveraging it for their benefit. This obviously increases inequality given that the institutions of a new system benefit the old elites. Democratization is thus a deliberate process designed by the elites, rather than a random event that arises from popular uprising.18 Hardly anyone gives power up voluntarily. They strive to preserve it and when faced with undeniable and unmitigable trends, they adapt, negotiate concessions, and build a position that protects them. This argument can serve as a useful explanation of why democracies have failed to eliminate inequality—​it was never part of their design nor their goal. Democracies have, as a consequence of capitalism, succeeded in alleviating poverty, but have failed to account for the problem of elite power, the root cause of inequality. Even when democratic societies enter a developed phase of inclusiveness and prosperity (when they enter and remain in the “Narrow Corridor” maintaining a stable balance between the power of the state and the power of society19), there is no guarantee they will succeed in mitigating the violence power incentives that propagate the concentration of wealth and power within elite networks. However, as argued earlier, despite these initial issues which all early democracies go through (recall the Tammany Hall example from 19th-​century New York mentioned in the Introduction), their trial-​and-​error process does eventually right the wrongs and helps build a more just and prosperous system. I am therefore optimistic in the capacity of capitalist democracies to solve the inequality issue without resorting to violence. As the final chapter will show, in order to do this, another evolution is necessary—​the empowerment of the society with a goal to constrain the power of the elites. We must start with reducing centralized power and shift that power back to the citizens and the community. *** The 20th-​century socialist model managed to deliver rapid industrialization to the periphery, which was still mostly trapped in an agrarian economic system after the end of World War II. However, due to a lack of the very basic trial-​and-​ error process, due to a lack of capacity to innovate and adapt to the new trends of globalization and the IT revolution, it reached its inevitable end after the period of rapid industrialization. The socialist model was not immune to elite networks, and it too failed to curtail their incentives for power and wealth accumulation. This made it vulnerable in its transition phase to capitalist democracy as the old

Capitalism and Democracy  259 socialist elites spearheaded the process in most countries, resulting in an unsuccessful transition marred by cronyism and corruption. The socialist model failed in solving the violence power principle—​if anything it enhanced it—​meaning that it failed in reducing elite network power, the cornerstone incentive for unequal wealth accumulation. The reason for this failure was primarily its inability to innovate and use the trial-​and-​error process that capitalist democracies have proven to be particularly good at. Both systems can end in congestion, economic crises, and deliver erroneous and malign outcomes, but when democracies face such situations, if their institutions are strong enough, they can learn, adapt, and innovate themselves out of the crisis. When socialism faced its first major crisis, because of its lack of capacity to innovate, it was destined to fail. We must therefore ask the next logical question: if capitalist democracies have a tendency for innovation and change from within, driven sometimes spontaneously and sometimes by intelligent design, what would be an ideal set of policies or reforms to drive such a gradual change that would tackle the source of inequality? Before we answer this question in the final chapter, let us consider the standard proposals for curbing inequality in the next chapter: taxation and redistribution.

9

The Pitfalls of Political Power Expanding the Scope of Government to Reduce Inequality

Politicians and diapers must be changed often, and for the same reason —​Mark Twain

9.1.  Progressive Taxation In light of the historical lessons presented in the previous chapter, where democratization gradually evolves and over a longer period of time creates different incentives that improve social outcomes through benevolent governments, we have to ask ourselves is there a way to use the encompassing power of governments to solve the problem of inequality and lack of opportunity? Higher income taxation and redistribution are often advocated as the main policy choices to combat the issue of rising income inequality. The problem, however, is that such policies are focused on alleviating the consequences—​ unequal incomes—​but not the causes—​proximity to political power. Their goal in reducing incentives for inequality is likely to be ineffective. The main reason why advocates of redistribution and taxation support such measures is because they see unequal incomes primarily as a consequence of market failures. If the market fails to achieve a distribution of income that society deems fair—​implying that too many are being left out of the benefits from economic growth, or that some groups are simply taking too much economic gain for themselves (measured as the share of income among the highest income earners versus everyone else)—​the government should step in and tax those with higher incomes, while redistributing these funds to those at the poorer end of society, not just in terms of transfers but also in public goods such as better health and education opportunities. In addition, higher taxes on the rich would create disincentives for very high wages at the top, thus further lowering the market-​generated inequality.1 The central problem with this reasoning is that inequality is only to some extent a consequence of market failures. Political failures, in the form of abuse of political power through elite networks, are a crucially overlooked source of Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0010

The Pitfalls of Political Power  261 abnormal earnings among the top 1% and 0.1%, just as they were the most important source of wealth concentration among the elites prior to the 19th and 20th centuries. As shown in Chapter 4, political connections of top executives in the United States and the United Kingdom are an important explanatory factor of higher salaries and total compensations. These two countries were the forerunners in rising inequality trends over the past decades, driven primarily by the growth in incomes of top income earners. In each case, all other market, nonmarket, and networking variables explained very little of the variation in top incomes. However, distinguishing the market effect from the political effect is difficult to do using official statistics of tax returns (which is how inequality is being measured, today more accurately than ever before). Although we can distinguish between different types of earnings—​for example rents, dividends, or other forms of capital gains, and we can even see if one’s income is entirely generated from inheritance or whether rents are driven by political decisions—​we cannot see why a top executive’s salary is higher. We can only speculate and cite reasons like tax cuts, or the combination of globalization and the skills premium, standard examples of market determinants (i.e., demand) for top executives. Which brings us back to the issue of progressive taxation as a policy with an aim to lower incentives for high salaries at the top. The standard justification for introducing progressive taxation can be visually inferred from Figures 9.1 and 9.2. Top marginal tax rates hit their highest ever levels across many developed economies in the period from the 1930s to the 1970s. Top rates went up as high as 98% in the United Kingdom during the 1940s, and over 90% in the United States in the 1950s, albeit slightly lower in Germany and France during the same period when they reached 75% and 66% respectfully, shown in Figure 9.1. During the same years, inequality (measured as the share of income held by the top decile earners) in all these countries was at its historical lows, going down to 31% on average from as high as 50% in early 1900s, as shown in Figure 9.2. In the 1980s, the income tax system became much less progressive, and inequality shot up once more. Top rates were halved and stood at 30% in the United States and 40% in the United Kingdom, however they remained relatively unchanged in Germany and France in the 1980s, only going down to less than 50% in the 2000s (when inequality started increasing in Germany). Similar trends happened in other European countries. The West, the United Kingdom and the United States especially, never returned to the high-​income tax rates of the mid-​ 20th century (although France had a brief experiment in 2013 and 2014, which is examined in the next section). Economic research also finds a link between tax policies and inequality since the 1980s. In Chapter 3 I mentioned a countrywide study of OECD economies

262  Elite Networks Top income tax rates for selected developed countries, 1900–2020 100% 90% 80% 70% 60% 50% 40% 30% 20% 10%

19 0 19 0 0 19 5 1 19 0 1 19 5 2 19 0 2 19 5 3 19 0 3 19 5 4 19 0 4 19 5 5 19 0 55 19 6 19 0 6 19 5 7 19 0 7 19 5 8 19 0 8 19 5 9 19 0 9 20 5 0 20 0 0 20 5 1 20 0 15

0%

US

UK

Germany

France

Average

Figure 9.1  Top income tax rates for the US, UK, Germany, and France from 1900 to 2020. Source: Piketty (2020), data available at: http://​pike​tty.pse.ens.fr/​fr/​ideol​ogy.

Top decile (10%) share in total income for selected developed countries, 1900–2020 55.0% 50.0% 45.0% 40.0% 35.0% 30.0%

19

0 19 0 0 19 5 1 19 0 1 19 5 2 19 0 2 19 5 3 19 0 3 19 5 4 19 0 4 19 5 5 19 0 5 19 5 60 19 6 19 5 7 19 0 7 19 5 80 19 8 19 5 9 19 0 9 20 5 0 20 0 0 20 5 1 20 0 15

25.0%

US

UK

Germany

France

Figure 9.2  Top decline share in total income for the US, UK, Germany, and France, from 1900 to 2020. Source of data: World Inequality Database, https://​wid.world/​.

The Pitfalls of Political Power  263 conducted by Piketty, Saez, and Stantcheva, who find that the reduction of top marginal tax rates increased the bargaining power of CEOs with respect to their boards, thus increasing their salaries and other benefits.2 Furthermore, they find that the top 1% of earners take advantage of tax evasion opportunities, meaning that an optimal policy should be aimed at eliminating any possibility of tax avoidance. In a study that surveyed the literature on tax policies and their impact on inequality by Saez, Slemrod, and Giertz, it was found that tax evasion significantly increases for those with higher incomes, as does their effort to avoid taxes with respect to higher tax rates.3 They cite the episode of the Reagan tax cuts in the United States, after which a considerable number of wealthy individuals reported a one-​time increase in taxable income, meaning that those people brought back their wealth from abroad to take advantage of the tax amnesty clause. When tax evasion loopholes are closed, taxable income becomes insensitive to tax rates, meaning that one can, theoretically, get more government revenues by increasing top tax rates without reducing work effort. It is the same argument Piketty uses in his books to advocate for a more progressive income tax and a global wealth tax.4 This theoretically justifiable conclusion might, however, be very difficult to implement in practice, as we will touch on in the following sections.

9.1.1.  The Unprecedented 30-​Year Period That Brought Down Inequality Before doing that, let’s get back to the trends from Figures 9.1 and 9.2, specifically focusing on the historical implication of high top income tax rates during and after the 1940s. Can these postwar trends really suggest a causal implication between progressive taxation and income inequality? In other words, did income inequality go down mostly because of progressive income taxation?5 Or were there other, much deeper and more important forces at play? For one, the biggest declines in both income and wealth inequality occurred during a 30-​year period that saw two world wars, between which there was a pandemic that killed even more people than the first war, revolutions, the Great Depression, and episodes of hyperinflation that erased a lot of wealth for many. The rapid decline of inequality during those times is much more likely a consequence of the wars and their aftermath than of the social reforms (franchise extension, the welfare state) that were introduced not with a goal to reduce inequality, but as a reaction to the wars and economic depressions, hoping to prevent such events from ever happening again. Immediate consequences of World War I in the West included the full extension of the franchise to the population demanding greater representation, huge

264  Elite Networks public debts, and war reparations that were being repaid first by galloping inflation and soon by hyperinflation, particularly in Weimar Germany, and a corresponding increase in taxation and introduction of income taxes to additionally pay for the aftermath of the war. Even the rise of the Spanish flu pandemic can to some extent be attributed to the war, given a population with weakened immunity becoming fertile ground for the rapid spread of the virus. In Russia, the direct outcome of the war was the dissolution of its empire through a revolution, and the rise of communism and central planning. The Russian brand of communism would carry on for over 70 years, achieving particular momentum and prominence after World War II, providing at the time a powerful alternative to the Western model of capitalist democracies. The Great Depression resulted in the first big battle of ideas between the then dominant doctrine of laissez-​faire capitalism and the central planning alternative offered by Soviet Russia. President Roosevelt’s New Deal—​introducing social security and unemployment relief, worker protections, major financial and regulatory reforms aimed at consumer protection, public work projects, and of course, new redistributive taxes—​was exactly the type of reform package that shows how societies were adapting to the major social pressures of the time. Preventing another economic depression was the central issue of those turbulent times. Until soon after the Depression, when World War II began, and rendered an even stronger effect on the distribution of incomes and wealth. Destruction of property aside, the biggest negative effect on wealth and incomes was the aftermath of the War. A wave of nationalizations and expropriations of private property spread across Europe, not only in newly established socialist countries but in capitalist democracies like France and the United Kingdom.6 Almost entire industries were nationalized; from banking, to gas and electricity, the coal industry, and the automobile industry.7 Furthermore, public debt was an even bigger issue than after World War I. Countries once again started printing money to pay it off, thus again causing inflation to rise, with efforts like price controls and food rationing applied to prevent it. This added to the misery of the population in the first four to five years immediately after the war, particularly in Germany and France. Personal savings were virtually wiped out. To prevent another hyperinflation episode and pay off the war-​induced public debt, progressive taxes were increased and other forms of taxation introduced. In the battle of ideas the pendulum swung away from laissez-​faire capitalism toward a mixed economy, hoping to get the best out of both a strong social security system and a market economy. Nowhere was this more obvious than in West Germany, whose first postwar government pledged to a social-​market economy (Soziale Marktwirtschaft), a third way approach between the Anglo-​Saxon laissez-​faire capitalism and the central planning system they had during national socialism. It was a market economy with a strong welfare state, the embodiment

The Pitfalls of Political Power  265 of European capitalism in the decades to come. Introducing a strong welfare state happened in almost every developed postwar economy. Atlee’s government in the United Kingdom introduced the NHS, providing public health care, in addition to a number of other reforms that cemented the welfare state. The United States expanded on Roosevelt’s reforms during the 1950s, culminating in Lyndon Johnson’s Great Society in 1965 introducing Medicare, Medicaid, and federal education funding. Societies changed, reforms were introduced, and the democratic trial-​and-​error process delivered. Wealth and income concentration at the top were clearly subdued, as so many factors contributed to their decline. Progressive taxation was a mere drop in the bucket among all these paradigm-​shifting events. Events that coincided and occurred over only three decades—​two major, unprecedented wars, a pandemic, hyperinflation, economic depressions, revolutions—​are the most blatant personification of the forces of occasional destruction. All four of Scheidel’s “Four Horsemen” happened in that relatively short time span. Inequality obviously went down. The social reforms instituted afterward were initiated to prevent these events from ever occurring again. They were not targeted to combat inequality. Their role was to rebuild their societies and prevent economic depressions and the outbreak of another devastating war. Inequality was an afterthought. Claiming that the same set of ambitious reforms would have occurred had it not been for the wars or the depression is simply not rooted in fact. The wars and the depression were obviously not exogenous events, they were driven by internal pressures and were a typical example of a self-​reinforcing negative cycle of the forces of occasional destruction. A number of political, economic, and social pressures combined and delivered such outcomes, where each one merely exacerbated the others. Inequality was hardly the only cause (or even a major cause) of such events, and it was certainly not the main concern during the aftermath of the war when the structural reforms took place. After such unprecedented events it took a while before the forces of wealth accumulation picked up pace once again. Another shift in the pendulum of the battle of ideas in the 1980s certainly helped the forces of wealth accumulation (as was shown in Chapter 3), but it would be wrong to constrain the cause of such forces to a mere policy change or a set of policy changes. Incentives for wealth accumulation, gained through access to political power, are persistent.

9.1.2.  The Artificial Decrease of Inequality Which brings us back to the main issue of progressive taxation and its inability to deal with the causes of inequality. In addition to a doubtful impact of higher top tax rates on the considerable decline of mid-​20th-​century inequality, there is also

266  Elite Networks a structural problem in why they tend to be less effective in a more globalized world. Capital today is more flexible and travels faster than ever. Incidences of tax evasion are quite common, as they are highly correlated with the complexity of a country’s tax code. This might be a deliberate consequence of the political process, given the effort of lobbyists in designing tax legislation to make the system more opaque and hence more prone to loopholes. Because of this a higher tax on top incomes could only lower inequality artificially by incentivizing wealthy individuals to hide their true earnings and wealth overseas. I refer to this as an artificial decrease of inequality given that the official income tax statistics would show lower top incomes, thus lowering the shares of top decile and top centile earnings in total earnings, but only because a part of those earnings would be taken out of the country to an offshore tax haven. In other words, real inequality would be hidden, not solved. A study of tax evasion using the “Panama Papers” and “Swiss leaks” from HSBC for three Scandinavian countries, Norway, Sweden, and Denmark, unsurprisingly confirms that tax evasion increases with wealth. In particular, in these countries tax evasion is about 3% of all personal taxes, but among the wealthiest 0.01% it is close to 30%. When this is taken into account, the inequality numbers in Scandinavia go up significantly.8 This research highlights the importance of going beyond income tax data to fully capture the extent of inequality in a country. As the authors state, even in countries where tax compliance is high, as in Scandinavia, the full extent of one’s tax avoidance cannot be captured by the available data,9 meaning that inequality is likely to be underestimated, particularly in countries where social trust is lower and corruption is higher than in Scandinavia. Given that the very definition of elite network members are people in the top income brackets who (mis)use their power for private gain, it is highly likely that tax evasion is higher the more powerful someone’s position in the elite network hierarchy. This means that raising taxes on these individuals will hardly solve the actual inequality issue. It will only hide it. Closing loopholes and strengthening international tax compliance, despite being the theoretically optimal policy, is almost impossible to enact. Offshore countries have strong incentives to attract wealth this way (whatever effect this might have on their local economy). This is why the solution must be found elsewhere, beyond a set of stand-​alone policy proposals. There is a good recent quasi natural experiment that testifies how the mechanism of an artificial decrease of inequality works. It concerns President Hollande’s introduction of the 75% top marginal tax rate on annual earnings greater than €1 million in France in 2012. Hollande campaigned on this tax in 2011,10 and announced its implementation soon after he won office in May 2012. The French high courts approved the tax in 2013 (given that it was ruled

The Pitfalls of Political Power  267 unconstitutional a year earlier), and the tax had been applied to fiscal years 2013 and 2014, ending abruptly in January 2015 after it was proven woefully ineffective.11 French billionaires were outraged by the announcement of the tax.12 Many declared they would be leaving the country, the most prominent examples being the richest man in France, Bernard Arnault, the CEO of Louis Vuitton, who took out Belgian nationality, and the famous actor Gerard Depardieu, who got Russian citizenship.13 Other less prominent billionaires either shifted their wealth abroad or started buying real estate in London14 (a trend that would significantly increase demand for London’s high-​end real estate in the years to come). The tax itself failed to produce the desired effect. It halved the government’s tax revenues in 2013, not just from income tax but also in corporate tax revenues and VAT revenues.15 It did, however, result in a slight decrease in top 10% and top 1% income inequality, as shown in Figure 9.3. Top 10% share in total income fell from 32.2% in 2011 to 30.9% in 2013, while top 1% share in total income fell from 10.4% to 8.9% in the same period. When the top tax rates were reduced to 45% in 2015, top 10% and 1% inequality again shot up (to 31.9% share of total income for the top 10%, and back to 9.8% for the top 1%). Why did inequality rebound so quickly after the top rate was scrapped? To understand this effect, we must examine the estimated tax evasion in France during the same period. Figure 9.4 shows the total offshore wealth of French citizens Income inequality as share of income, France 1970–2019 40% 75% top rate announced

35% 30%

Rate lowered to 45%

25% 20% 15% 10%

0%

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

5%

Top 10% share

Bottom 50% share

Top 1% share

Figure 9.3  Income inequality in France measured as the share of income for the top 10%, top 1% and bottom 50% share. Source of data: World Inequality Database, https://​wid.world/​.

268  Elite Networks Tax evasion in France: estimated offshore wealth, 2001–2016 400 350

Billions of euro

300 250 200 150 100 50 0

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Figure 9.4  Estimated offshore wealth in France, 2001–​2016. Source: European Commission (2019) “Estimating International Tax Evasion by Individuals,” Taxation Papers, Working paper no. 76, European Commission, September 2019, page 67.

from 2001 to 2016, estimated officially by the European Commission. There is an interesting trend of increasing offshore wealth that started in 2011, expanded in 2012, and culminated in 2014. In that three-​year period, offshore wealth of French citizens increased by 79% cumulative (from €169bn to €342bn). The trend was reversed after 2015, when the 75% top tax rate was scrapped. Furthermore, the French government lost an estimated €51bn of potential tax revenues due to evasion. The four-​year period that revolved around Hollande’s tax hike saw a particularly strong increase in lost revenue due to tax evasion on original income, as shown in Figure 9.5. As did offshore wealth, the revenue loss from evasion started decreasing in 2015. These tax evasion trends, coinciding with the period of highest tax uncertainty in France, suggest that the numbers verify the sentiment of French millionaires and billionaires at the time. It also suggests that with offshore wealth being as high as €360bn (13% of French GDP), amounting to a €50bn loss in tax revenues (5% of annual revenues), capital is indeed very flexible and can hardly be contained by methods of economic policy alone. There could, of course, be other reasons why evasion occurred, and this short quasi-​experiment is not enough to make a causal implication on whether higher top marginal tax rates induce tax evasion of the rich. Anecdotal evidence is never enough to make such a conclusion, however economic research across many countries does confirm that higher tax rates typically result in higher tax evasion. This was confirmed in the United States for capital gains tax and for income

The Pitfalls of Political Power  269 30

Estimated loss in tax revenue due to evasion in France, 2004–2016

Billions of euros

25 20 15 10 5 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Capital income tax evasion Tax evasion on original income

Inheritance and wealth tax evasion Total tax evasion

Figure 9.5  Estimated losses in tax revenue due to different versions of tax evasion in France, 2004–​2016. Source: “Estimating International Tax Evasion by Individuals,” Taxation Papers, Working paper no. 76, European Commission, September 2019, page 192.

tax,16 in China for high tariffs on imports,17 in Switzerland and Italy for income tax,18 and so on. Each study finds a strong negative effect of tax increases on noncompliance, and in many cases the studies also report a loss of tax revenues after higher taxes were introduced. Just like the effect in France after the 75% top marginal rate. Another thing this short quasi-​experiment showed is that increasing the top marginal income tax rate fails to address inequality in real terms. The main problem with models calculating the impact of progressive taxation is that they assume everything would stay the same and that the wealthy would not shift or otherwise hide their incomes after a high marginal income tax rate is introduced. The anticipated gains from high top marginal rates would evaporate if the wealthy simply shifted their earnings overseas, meaning that this revenue could not be used to finance health, education, climate change, or poverty programs, as so eloquently advocated. There would be no such excess revenue from taxation to redistribute, and the only consequence of high marginal tax rates would be to encourage cheating and tax evasion. Higher income taxes would certainly lower (hide) inequality observed in the earnings data, as they would lower the share of top decile and top centile earnings. Measured inequality would therefore go down, but real inequality,

270  Elite Networks real differences in opportunities, wealth, and earnings between the politically connected rent-​seekers and the politically unconnected majority would remain high and possibly even increase if more redistributive power is granted to those who generate top 1% income inequality in the first place. As this book has repeatedly shown, real differences in incomes and opportunities arise from proximity to power, from the elite network effect that makes sure that any regulatory or legislative changes benefit the elite. When the tax code is deliberately complex and opaque this is a direct product of lobbying and legal efforts by the wealthy to legally hide their income abroad. A typical counterargument from advocates of progressive taxation is to introduce stricter laws that can prevent tax evasion globally. The impossibility of international policy coordination on this issue aside, even the domestic legislative process, unfortunately, does not work so smoothly and easily. Every draft of legislation gets watered down significantly before it is voted into law, usually to include many loopholes. Loopholes, counter to popular perception, are not a product of systemic glitches or someone’s error in judgment. They are deliberate and a typical outcome of the political decision-​making process. Any person who ever worked as a legislator is familiar with this. Various interest groups fight to promote their interests in the political arena, but in this battle between competing interests those with the quickest proximity to power tend to get the best outcomes. This is why it is difficult to change gun laws in the United States, or why demand for higher taxation of the rich, even with popular support, rarely gets passed in the legislative assemblies.19 Going back to the spectrum of interest groups delivered in Figure 1.9 in Chapter 1, the elite network members—​the new supermanager elite, clustered in the upper right part of the diagram having high salience and high power—​will hardly ever get punished by the political process. Therefore, the only way to solve the elite network impact on the political decision-​making process—​ the key driver of inequality—​is to change the process: reduce political power, and reduce our dependency on politicians to make such decisions. By increasing taxation and advocating for a greater role of the state to solve redistribution issues and impose additional regulations to prevent adverse political outcomes like collusion, we are inadvertently advocating handing greater power to politicians. And by doing so we are exacerbating negative selection into politics and are increasing systemic fragility. Which brings us to the next argument of why it is futile to depend on political power in order to reduce inequality.

9.2.  Adverse Selection into Politics If a career in politics guarantees access to power, it will undoubtedly attract power-​hungry individuals with questionable moral views and almost entirely

The Pitfalls of Political Power  271 selfish motives. Psychological research has uncovered that an increasing number of political leaders exhibits psychopathic or sociopathic tendencies (incidentally, the same applies to corporate leaders).20 Granting even more incentives to such individuals to obtain power seems absurd and counterproductive to the very idea of trying to achieve fair social outcomes. Politics obviously also attracts genuinely honest individuals with a strong moral compass (something we like to define as “true leadership”), however even a remote probability of having a sociopath in power necessitates trying to find a solution to prevent such an outcome, or even better, never to grant a sociopath any significant power to begin with. Particularly if the probability of choosing a morally corruptible individual is (considerably) higher than choosing an honest individual. Think of it in terms of a battle between competent and incompetent individuals in politics and public service in general. If you are an incompetent yet ambitious individual, loyalty and allegiance to your party and your leader are the only things you can use to advance in your career. Competent individuals do not have to show such blind loyalty, they hope to advance and achieve their ambitions based on merit: their skills and knowledge. However, in a contest between a competent and incompetent individual, even within-​party, the incompetent individual has everything to lose from a defeat, whereas the competent one can always choose an alternative path. This creates strong incentives for incompetent but ambitious individuals to win under all costs, which morally corrupts them during their path to power. By the time they reach office they are burdened with debts to people who helped get them there (unless they get rid of them immediately, which is a strategy directly pulled from an autocrat’s handbook). Selecting an incompetent or morally corrupt leader seems to be a higher probability outcome than selecting an honest incorruptible leader. Even when such an honest, competent leader is chosen, there is still the issue of selection of all the people around them. Changing one person does not change tens of thousands of individuals underneath, all of whom were selected under the same principles of political selection: ambitious and morally dubious incompetence beats morally superior competence in the majority of cases simply because incompetence has much more to lose and hence is more likely to try and win under all costs. It is the violence power principle in action. There are, of course, positive examples of countries where the civil service is made up of mostly honest and competent individuals, where its employees perform their jobs with a sense of pride and duty and seldom hold back progress with unnecessary burdens or corruption. In such countries, politicians get drawn from the same pool of honest and competent individuals and are driven by the motivation to serve their countries, not to usurp power for private gain. These are countries that, according to Acemoglu and Robinson, learned how to

272  Elite Networks contain the power of their Leviathan (existing within the “Narrow Corridor” between weak states that deliver anarchy, and strong states that deliver authoritarianism),21 countries that built institutions that constrain political power, and immediately punish every politician that tries to break the boundaries, even for benign affairs (like buying chocolate and diapers with the government credit card). These are also countries where trust in public institutions is high and in which people do not mind paying higher taxes if they know these taxes will deliver the public goods that benefit their communities. The best example are Scandinavian countries, where most people believe in the fairness of the process and know that any politician that tries to command too much power or misappropriate the vast public resources at their disposal will quickly be punished (the aforementioned credit card example is of a former Swedish minister). Such a mechanism exists in countries where governing institutions are depersonalized, where the question of who holds power does not infringe in any way the institutions that preserve the “Narrow Corridor” balance. Scandinavian elites also avoid paying their taxes, as shown earlier. But they do not usurp political power to protect them. At least not as much to harm social progress.

9.2.1.  Personalization of Governing Institutions The personalization of governing institutions is probably the biggest difference between well-​functioning democracies and poorly functioning democracies. In countries with depersonalized institutions societies are governed by clear rules that everyone respects. These rules are a product of a long trial-​and-​error democratic process and are widely accepted by everyone in society to be fair. As societies evolve the rules change, again to the general acceptance and desire of the society (civil rights movements are good examples). Most importantly, in such countries rules do not change by the will of a single person or narrow group holding power. The power-​holders merely adopt the changing norms if demanded by the majority. Switzerland is a perfect case-​in-​point. Policies change when people vote for such changes in referendums. Politicians in power are merely executors of the peoples’ will. They are respected, they are drawn from a pool of honest and competent individuals, and yet they have very low incentives of ever usurping their political power for private gain. Mostly because their political power itself is constrained by the rules of the game which they cannot change. In such countries elite networks are likely to be less effective simply because individuals at the position of power do not carry the biggest decision-​making weight. Proximity to power does not give a person the leverage to alter legislation to serve narrow interests.

The Pitfalls of Political Power  273 On the other hand, many corrupt democracies suffer from the issue of personalized institutions. This is best described by North, Wallis, and Weingast in their theory of violence and social orders. Countries characterized as limited access orders are those in which well-​organized ruling elites can manipulate the economy by generating privileges based on the personalization of governing institutions. In such systems all political and economic outcomes are merely a consequence of interpersonal relationships among agents within the ruling elite networks. When the institutions of a system are depersonalized it is much harder to create clientelist relationships. This is the case of an open access order, in which the foundation of intrapersonal interactions is a well-​defined, depersonalized legal framework, and not politically generated privileges.22 When institutions are personalized, people start demanding too much from politicians and think that the person behind the institution is the institution. This is dangerous ground for populists to exploit, given that they thrive in environments where they can personify the government and build a cult of personality around themselves. But the problem runs deeper than that, and it often ends up backfiring against the populists. When a government is personified by the person or a group of people in charge, we start demanding too much from those leaders. We place too high of a burden on a centralized system which is by definition unable to solve as many problems as we think they are apt to solve. In such situations our natural inclination is to expand the powers our leaders can use to solve our problems. We are looking for the messiah, the one leader that can save us. We are assigning idealistic characteristics we would like our leaders to have in order to be competent enough to solve all of our and, by extension, the society’s problems. And we keep fooling ourselves that it is just a matter of electing the right person for the job. But this rarely happens. A much more likely outcome for a leader with such power is to misuse it for personal gain, to enjoy it without facing much accountability. We never find the “right” person, because in societies with weak and personalized institutions, the adverse selection into politics effect is only amplified and keeps delivering incompetent power-​hungry individuals. Another consequence of personalized institutions is increased incentives for corruption, not just in national-​level governments but throughout society. When the system fails to solve a given issue—​which happens quite often—​people seek to befriend or reach individuals making the decisions. The decision-​making position on any level therefore grants the individual even greater power. In corrupt countries this is true whether the issue at hand is a court trial where interested parties seek to find a connection to a judge, or getting a building permit, where one seeks connections to public officials making the calls, or a medical procedure, where it is key to be in proximity to the doctor, or at least be able to bribe them. When institutions are personalized, the system fails to deliver optimal

274  Elite Networks outcomes, and when the system fails, people turn to individuals making the decisions to solve their problems. It is a vicious cycle where it is often hard to pinpoint the direction of the causal relationship. Countries with high levels of inequality will typically be countries where institutions are personalized, where selection into politics is skewed toward power-​hungry individuals, and where corruption tends to be higher. Personalization of institutions is a breeding ground for elite networks. One does not befriend systems or institutions, but people. Being close to the right person making the decisions implies getting favorable outcomes. Personalized institutions have no room in successful democracies. And personalized institutions have no room in countries that want to lower inequality. The United States is a specific case here. It does have a strong institutional system that prevents a number of malign undemocratic outcomes, while many of its institutions are indeed depersonalized. However, in the case of the Trump presidency we witnessed how fragile even these strong institutions can be. The United States did not descend into autocracy, nor did the Trump administration manage to disrupt the justice system or change the way American democracy works. Nonetheless, many elements of democracies were under threat, the reason for which was too much power concentrated in the hands of a single person who completely personalized the institution of the presidency. Regardless of the strong constitutional constraints that every president has, even this is sometimes not enough given the pure magnitude of decision-​ making powers given to the president. Particularly when it comes to foreign policy where US presidents have been virtually unconstrained (the only constraint was the size of their domestic budget subject to Congressional approval). The United States is without doubt a strong democracy. Its institutions work. But given that these institutions can be stretched by a single person holding power—​and hence personalizing the institution of the presidency—​perhaps time has come to constrain the power of the executive branch in the United States, and demand less from politics. Which bring us to the necessity for lowering centralized political power and lowering the scope of activities for national-​level politicians. A necessary condition for this to work are stronger incentives to prevent misuse of power in terms of greater transparency and accountability. There is a need for reform aimed not at the consequences but at the root causes of malign outcomes in our societies. Behind them is always too much extractive political power. This is not an inherent argument against government, redistributive taxation, public services, or the necessary public goods in situations where markets fail to provide an optimal social outcome. This is an argument first and foremost against the people behind the government; their incentives and motivations. This is an argument aimed at lowering political power.

The Pitfalls of Political Power  275

9.3.  The Importance of Trust The problem with personalized institutions, in addition to skewing selection into politics, is that it also erodes trust in a society. When people seek individuals instead of institutions to solve their problems, they have no trust in the system. As emphasized before, this is a vicious cycle with no clear direction of causality, but the underlying factor that drives people to seek outside-​institutional solutions is lack of trust. On the other hand, countries with depersonalized institutions imply a population that trusts the system to work and deliver the public goods the people desire. In such countries paying taxes is not met with incentives for evasion, but with a genuine desire to give back to the community. Interestingly, lack of interpersonal trust seems to be a factor that binds most countries with the issue of high inequality, regardless of their levels of development or their economic system (with the exception of China and Saudi Arabia, where interpersonal trust is high for other, mostly historic and religious reasons—​see Figures 9.6 and 9.7). Figure 9.7 in particular shows an interesting correlation between interpersonal trust (the question of whether or not most people in a country should be trusted) and the Gini coefficient. The correlation is negative and shows that countries with higher levels of interpersonal trust typically have lower levels of income inequality. This does not imply a causal relationship between trust and inequality, simply because there might be a common unobservable factor that affects both these variables. It is, nevertheless, interesting to note that countries with higher levels of interpersonal trust also trust their governments more and that this trust seems to be quite consistent over time (like in Scandinavia or China). In the United States, trust in government (measured as % of people who trust their government) dropped from 77% in 1964 down to only 19% in 2015. The decline has been sharp in the mid-​1960s and ’70s (due to the Vietnam War), rebounded in the early ’80s, had a sharp decline in 1992, rebounded in the late ’90s and has been in a steady decline since 2001 (when it was at 48%), again as a consequence of the Afghanistan and Iraq wars and possibly the 2008 financial crisis.23 Interpersonal trust hasn’t fallen so sharply nor has it been so volatile, but the long-​term trend is still negative, from 45% in 1972 to 30% in 2014. It might not be a direct cause of inequality, but rebuilding trust—​especially trust in public institutions—​is certainly an important factor in strengthening the democratic trial-​and-​error process. Another reason why trust is important is its role in perpetuating social progress. Sociopolitical systems in which people don’t trust each other and don’t trust their public institutions are breeding grounds for populism and conspiracy theories. This can best be seen in rapidly declining levels of trust in experts and scientists, particularly noticeable in developed countries of the West during

276  Elite Networks

Figure 9.6  World map of interpersonal trust, based on the World Values Survey and how many people agree with the statement that “most people can be trusted.” Source: Ortiz-​Ospina and Roser (2022). 0.65 0.6 0.55

Gini coefficient

0.5 0.45 0.4 0.35 0.3 0.25 0.2

0

10

20

30

40

50

60

70

80

Trust in others (World Values Survey)

Figure 9.7  Correlation between trust in others from the World Values Survey (2022) (x-​axis) and the Gini coefficient from the World Bank (2020) (y-​axis). Source of data: Ortiz-​Ospina and Roser (2022): “Trust,” https://​our​worl​dind​ata.org/​trust.

The Pitfalls of Political Power  277 the COVID pandemic (although the distrust in experts was a trend that started years, even decades before the pandemic). Everyone, the government especially, is portrayed as having a secret agenda, which raises suspicion and increases the plausibility of off-​the-​shelf conspiracy theories about things that were never before being brought into question—​things like vaccinations, communication networks, or even evolution. There are plenty of social, political, and even economic factors that have contributed to the spread of conspiracy theories and fake news (the Internet and social media being just some of them). The reason for their persistence has been a steady erosion of trust in governments and experts who can easily dismantle every conspiracy theory and keep it on the fringes of society. But when there is no trust in public institutions, conspiracy theories move away from the fringes and into the mainstream. By doing so they can deliver bad social and political outcomes, something we might call an error in the democratic trial-​and-​error process. These periods of error can last a long time and result in extreme political polarization and distrust of anyone who disagrees with one’s opinion. It leads to mass creation of cognitive bubbles (echo chambers) where you only trust people inside the same bubble, while carrying massive distrust toward everyone outside of it, toward people who do not share your worldview. Continued lack of trust is therefore a certain path toward dysfunctional societies. The economic logic of such societies is a zero-​sum game where one’s gain must equal another person’s loss, preventing any incentives for robust economic growth and development. Such societies stay embedded within a self-​ perpetuating bad equilibrium where the only way to break the cycle is a long, protracted process of rebuilding trust through the democratic mechanism. Given that the democratic process itself is extremely fragile, especially in dysfunctional societies, it very often fails to deliver the change in trust, but this is only because it is being used in the wrong way. We will see in the next chapter which types of reforms are crucial to rebuild trust among people, within communities, and for public institutions. Only with higher levels of trust can we start rebuilding dysfunctional societies. Which is why the set of policy proposals presented in the final chapter lists rebuilding trust as the key necessary condition for re-​empowering communities and lowering elite network incentives. If we wish to reduce incentives for inequality, we must forget about solutions that aim to expand political power, or solutions that focus only on inequality’s consequences. Solving this deeply entrenched issue requires a long process that falls outside the scope of a single set of policy changes. It rests on the democratic trial-​and-​error process, it rests on a long road to rebuilding societal trust, and it rests on lowering the magnitude of political power. Hopefully this chapter has shown there are no magic bullets. The process is tedious, and must be met with patience and open-​mindedness.

10

The Three Levers Give me a lever long enough and a fulcrum on which to place it, and I shall move the world. —​Archimedes Men are moved by two levers only: fear and self-​interest. —​Napoleon Bonaparte

10.1.  Moving beyond Standard Approaches: Introducing the Three Levers A set of reforms that seeks to reduce inequality needs to take aim at the main factor that has historically been responsible for higher concentration of wealth among the elites—​political power. This is no easy task, and there is certainly no one-​size-​fits-​all solution. Elite concentration of wealth driven by political power has been a persistent pulling factor throughout human history. It is not something that can be transformed by a one-​off policy change. Changing incentives for elite wealth concentration needs to be designed with second-​and third-​order consequences taken into account. A reform implemented today should aspire to change behavioral patterns of individuals and politicians (first-​and second-​order effects) which would ultimately reduce incentives to misuse political power, lower the allure of elite networks, and hence reduce inequality (third-​order effects). Take the idea of full transparency of budgetary spending and, ideally, decision-​ making behind budget allocation decisions, one of the proposals examined under the First Lever. Full transparency implies that every single receipt a government-​ owned entity spends money on is made publicly available for all to see: major procurement contracts, including the annexes of such contracts where no information is withheld, wages of public sector employees, and all small expenditures such as lunches, dinners, or travel expenses. The consequences would be to discourage all involved in the budget allocation process from cheating and misallocating resources (first-​order effect), which would reduce incentives to misuse public office for private gains (second-​order effect), which would improve selection into politics and increase political accountability (third-​order Elite Networks. Vuk Vuković, Oxford University Press. © Oxford University Press 2024. DOI: 10.1093/​oso/​9780197774229.003.0011

The Three Levers  279 effect). This is by no means straightforward to do, nor does it guarantee a successful outcome, given that a policy change rests on changes in expected behavior in the future. To make the change stick it needs to be gradually introduced while simultaneously encouraging civic participation, namely monitoring of political decision-​making by the media and interested citizens. One important point that needs to be addressed here is the issue of complexity when designing public policies. Complexity renders real costs to the policymaking process, costs which are seldom discussed when proposals are made. Many of the laws and regulations enacted today are unnecessarily long and detailed, burdened with a goal of providing an encompassing solution to each issue. This detailed regulatory approach, because of the exponentially expanding complexity of the society it aims to regulate, typically creates unwanted effects in the form of loopholes, either done by design or by accident, hidden in the vastness of the mere size of each legislative proposal (such as the 850-​page-​long Dodd-​Frank financial regulation bill, or the 1,200-​page-​long Waxman-​Markey climate bill). The United States and the EU have witnessed the rise of an army of regulators and bureaucrats over the past half century. And while their motivations for improving the system are certainly genuine, they often produce adverse effects that increase informational asymmetry benefiting those with access to it. Elite networks, because of their topocratic nature, are the first group to benefit from asymmetric information. Exploiting loopholes in the tax code is a prime example. The inequality literature consistently shows how the wealthy (many of whom are elite network members) are the only ones with the ability to hire legal counsel to dwell through the intricacies of complex tax codes in order to reduce their own tax burden. The rest of society lacks this advantage and is stuck in navigating the added complexity of the regulatory requirement. The same applies to big firms as opposed to smaller firms. An increased regulatory burden takes a much higher toll on small businesses than it does on big business, usually the prime target of said regulations. Considering that a large part of inequality can be explained by between-​firm differences, adding complexity to the regulatory process hardly helps solve the inequality issue. If anything, it exacerbates it. The EU General Data Protection Regulation (GDPR) mandate is a case in point. It was motivated with a very specific issue: to prevent Big Tech firms like Google or Facebook from misusing their users’ data. The initial goal of the legislation was to dampen the business models of online advertisers who used personal data of its users to predict their behavioral and spending patterns and sell them targeted products (from goods and services to political candidates). However, instead of putting even a dent in Big Tech business models, the legislation only added a layer of unnecessary burdens on every single firm with a webpage in how it handles its client’s data. But it had an additional unwanted effect that was much

280  Elite Networks more damaging: it lowered transparency of the public sector in some EU countries. What was usually under the jurisdiction of the “right of access to information” is now hidden by bureaucrats who refuse to provide information on how a certain decision was made (say over allocating a procurement contract or on nepotistic hiring practices), calling on GDPR to shield themselves from further legal inquires. GDPR became a cloak of invisibility for bureaucrats and politicians. In low-​trust and high-​corruption environments, it is rendering a direct negative effect on transparency and the rule of law. So much for thinking about complexity and second-​and third-​order effects while enacting a piece of legislation. Regulatory complexity cannot cope with societal complexity, no matter how detailed it becomes. An effort to fix an issue by adding layers upon layers of regulatory documents only adds to the inefficiency and seldom solves any issue. Not to mention the thousands of hours needed to comply with the regulatory requirements. The goal of effective policymaking should be to introduce clear and simple rules that limit power, and reduce complexity of the regulatory code. Therefore, the policy proposals made here, aimed at curbing elite network power, must not fall into the trap of adding unnecessary complexity. They must aim at relaxing legal and regulatory procedures and replacing them with simple rule-​based policies. The presented set of policy reforms, coalesced in the form of Three powerful Levers aimed at lowering political power and increasing civic engagement and trust, is designed with second-​and third-​order effects in mind, and with an aim to reduce regulatory complexity in its solutions, never to add to them. The goal is to use the trial-​and-​error mechanism of the democratic process through simple enforceable rules to lower incentives of elite network wealth and power accumulation and consequentially lower income and wealth inequality. By now it should be obvious to the reader that all proposed reforms will seek to rebuild trust in public institutions and rely on the trial-​and-​error democratic mechanism. We can therefore define Three Levers that, if applied simultaneously, would gradually reduce incentives for elite network wealth and power accumulation by exploiting the trial-​and-​error mechanism and rebuilding social trust. I use the term “lever” as an apt analogy. Archimedes required a single lever to “move the world,” Napoleon relied on two—​“fear and self-​interest,” but in order to help democracies overcome elite networks and lower inequality, we need three. All three rest on Archimedes’s metaphor of why a lever is useful, and at least one uses some variant of what Napoleon had in mind. The Three Levers are: 1. Reducing political power—​introducing greater constraints on public officials; introducing full budgetary transparency to increase political accountability; lowering scope of centralized government decision-​making activities

The Three Levers  281 2. Re-​empowering the citizens—​paying out gross salaries to people who use part of their tax allotments to directly determine how they are to be spent; introducing mandatory participatory budgeting for community public good provision; giving citizens an option of a credible imminent threat against transgressing office-​holders 3. Re-​ empowering the community—​ greater decision-​ making given to communities over the distribution of local public goods Note that the Levers reinforce and rely on each other. Greater community engagement implies lower centralized political power and the ability of citizens to make choices over the allocation of their tax money into local public goods. Greater citizen empowerment in terms of greater power over the allocation of tax payments implies a stronger community in which they can express their interests. Lower political power implies greater community power and greater power given to citizens to hold politicians accountable. Finally, and most importantly, do keep in mind that the given set of proposed policies was not a product of my explicit research efforts, nor were they tested experimentally and thus proven to work. They are a subjective extrapolation of some of the arguments made in the book. If political power is the issue, we need to understand how to reduce it and make the political process work to our benefit. The reform proposals are purposely designed to be provocative, hopefully stimulating the debate to take another direction, away from the usual set of policy proposals. Let’s explain each in greater detail.

10.2.  First Lever: Reducing Political Power When it comes to reducing political power, trust in public institutions is once again the crucial issue to address. Three reforms in particular can serve as a mechanism to improve trust, improve political accountability, and change selection into politics—​all three of these are particularly important in corruption-​ engulfed countries and dysfunctional societies: 1. Deliver full transparency of budgetary spending and decision-​making in order to uncover all potentially fraudulent deals between politicians, interest groups, and connected firms; 2. Encourage and promote free media and investigative journalism; 3. Introduce term limits for political office and strict conditions that prevent anyone charged with corruption to run for public office.

282  Elite Networks In addition to rebuilding trust and introducing institutional safeguards we must simultaneously implement solutions that reduce discretionary political power. The following proposals can deliver the biggest long-​run benefit in that respect: 4. Reduce discretionary decision-​ making power from office-​ holding politicians. Impose rules as credible threats that imply immediate punishment for politicians (key performance indicators for politics). 5. Reduce the scope of activities office-​holding politicians have the power to influence. The point of each of these reforms is depersonalization of institutions. In order for the system to work it should not matter who holds political office. Political power needs to be decentralized, where public good allocation decisions are collective instead of individual, thus preventing any individual (or a narrow group) from usurping power. The institutional changes proposed under the First Lever seek to be the first step toward turning politicians and office-​holders into true public servants; nothing more than executioners of the people’s will, with an important role in the process of public good allocation.

10.2.1.  Full Transparency of Budgetary Spending and Decision-​Making As shown in the previous chapter, the majority of countries in the world suffer from lack of interpersonal trust and, by extension, trust in public institutions. This can generate unwanted consequences and reduce any impetus for social change, thus condemning a society to a self-​perpetuating negative cycle of increasing polarization and decreasing trust. There is no easy way of breaking this up, and there is certainly no unique recipe, otherwise many societies would have already applied it. In order to rebuild trust, there are only microsteps: small bottom-​up changes that gradually restore confidence within a society. Trust, like confidence or reputation, once lost, is very difficult to reinstate. Marginal changes and efforts must be made to improve it. The first such marginal step, a bottom-​up type of reform, is full budgetary transparency. In itself it is a very simple idea: through the use of basic digital technology—​a website with a search engine that makes it easy to access information from a spreadsheet—​give the public direct insight into every single transaction, every single receipt made by a public entity, whether it is a local government, a national government, every ministry, every public agency and regulatory office, every public firm (e.g., a utility company or a similar natural monopoly under public ownership). Full and complete transparency implies every single

The Three Levers  283 transaction; from small payments (lunches and dinners, travel expenses), to major procurement contracts, to gross salaries of all public employees. It also includes all details of transactions: dates, amount, details of the person or firm that received funds (name and a unique personal identifier), description and justification of the expense—​in other words, everything that goes through an accounting software of a public entity should be available for the public to see. The government, acting like a true servant to its people, should have nothing to hide in front of them, at least in terms of how it spends the taxpayers’ money. The first immediate impact of introducing full transparency is to lower the motivation for corruption and increase political accountability. In corrupt countries a digitalized easy-​to-​use overview of all government transactions and decisions is a powerful tool to prevent fraud in procurement allocations or one-​sided changes in urban planning laws (the two biggest sources of corruption). Furthermore, it can monitor all recipients of government donations and subsidies in an effort to prevent political vote buying, while reducing the possibility of nepotism or employment based on political party membership. Transparency also reduces inefficient spending practices, as the bureaucrats in charge of administrating the decisions are fully aware they are being monitored and scrutinized. This might happen frequently on a national level where there is greater media oversight, but on a local level or among public firms, such detailed scrutiny is almost nonexistent. Finally, having a user-​friendly online tool increases civic participation and encourages citizens toward more active cooperation with the local government in solving the problems of their community. This might seem like a somewhat radical idea, but a system of budgetary transparency already exists, at least on some national or supranational levels. In the United States, every single transaction from a $4.8 trillion federal budget is freely available for anyone to use, search and download through a simple online search engine.1 Even supposedly secret transactions of acquiring combat weapons and military equipment by the Department of Defense (DoD) are all listed. For example, the search engine shows that the DoD awarded over 2.6 million contracts in 2019 worth over $250bn. Each contract is easily accessible and one can browse through its details (such as the type of weapon and how many of them were bought) as well as the details of each company that received the funds, and how much money they have received from the government altogether. Similarly, the European Commission keeps a website called the Financial Transparency System2 that tracks funds that were distributed via various EU grants from the European Commission. A person can search through each Member State and see all organizations and individuals who received an EU grant from the Commission since 2007, including the amount, the purpose, and individual and firm details. However, these are only grants given to Member

284  Elite Networks States, and do not list all expenditures of the Commission, something that would be a welcomed step forward. These are positive examples of national or supranational budgetary transparency, and although they are a commendable effort in the right direction, they are not enough to rebuild trust in public institutions, particularly in corrupt countries. The approach to full transparency must be broad-​based, follow a very specific standard of not hiding any important information, and potentially start on a local level from which it can be scaled up nationally. Furthermore, in addition to the exact amounts and receipts of the money spent, the public should also be able to look behind the reasoning of every single budgetary decision. Transparency in decision-​making is the next logical step for countries that already have some form of budgetary transparency. This implies justifying every single expense to the public, from small payments and official trips, to procurement contracts and decisions over changes in urban planning laws. It could be set up on the same website tracking individual expenses, where alongside each expense or government decision there is a detailed justification and explanation that illustrates how and why such a decision was taken. For example, for every procurement process the public should be able to know why an offer was accepted and why other offers were rejected. Typically, such decisions are hidden under a cloak of bureaucratic procedures and poorly written laws. Having the public be aware of such practices is the first step toward improving the efficiency and fairness of the process. The same goes for decisions on urban planning laws, when the government decides to change agricultural land into construction land—​the public should be informed as to which individual or company stands to benefit from the changes. This would make it easy for the public or the media to keep track of all connections between politicians and those who benefit from their decisions. Needless to say, such transparent procedures will drastically reduce the motivation for anyone to request specific favors from politicians, as the outcomes of such favors will be publicly accessible for everyone to see. Full transparency also reduces asymmetric information surrounding government decisions and access to information. It is a great example of an easy-​to-​implement, easy-​to-​follow, and relatively cheap solution that helps start the process of rebuilding trust in public institutions and increasing civic participation for better government accountability.

10.2.2.  Encourage and Promote Free, Independent Media In order for transparency to yield maximum impact we need engaged citizens and we need someone to keep track of all potentially fraudulent deals or

The Three Levers  285 decisions. In other words, we need free, independent media, and in particular, investigative journalism. What is the optimal way to encourage free media? On one hand, we can have a trustworthy, independent public media outlet like the BBC in the United Kingdom. However, in most countries, especially ones with high levels of corruption, public broadcasters tend to be prone to political capture—​they serve the party in power and deliver biased news. In corrupt countries and countries where public institutions are captured under the control of one-​party rule, public broadcasters are often the worst sources of fake news and propaganda. Such instances of media capture by governments have been well documented in economic research,3 and proven to have a strong positive association with inequality.4 On the other hand, we can have private media outlets, but they too can end up captured and promote personal or political interests of its owners or otherwise distribute fake news and undermine the role of free media in society. Examples here are ample. The most prominent is the case of Rupert Murdoch’s media companies in the United Kingdom, culminating in the phone hacking scandal and the Leveson Inquiry. Another is Berlusconi’s private media empire in Italy which was very successful in hiding any evidence of the prime minister’s wrongdoing (Berlusconi also held firm control over the public television RAI, capturing almost the entire media scene in Italy at one point). Examples of private media companies spreading fake news are even greater, best epitomized in tabloid newspapers which reach a large following and have immense power in determining the political landscape and shaping public opinion. Politically biased broadcasters may even influence elections, shown empirically on the example of Fox News and the 2000 Bush vs. Gore election in the United States.5 However, in most democratic countries there are at least some media outlets that are, in fact, independent, led by real investigative journalists, and help uncover various cases of fraud and political scandals. Their discoveries and investigations are often the only way the public would ever find out about a political or corporate transgression. It is often the case that investigative journalism provides enough evidence and public pressure to start criminal proceedings from the authorities. The authorities seldom react to anonymous tips, but once a scandal becomes public and is present in the media, the authorities have to react. Free media can truly be considered a cornerstone of democracy. It is a public good in its own right. The problem is that such independent outlets are very often on the verge of staying solvent. Particularly if they get slammed by lawsuits, which can either silence them, reduce their sharpness, or push them into counterattack, which further depletes their poor finances. Their societal role is priceless, and yet, they

286  Elite Networks are very fragile and often on the verge of being broken up, typically by politicians and powerful individuals (the elite networks) they investigate. How does one therefore encourage free and independent media? One way would be through citizen empowerment, where people would choose to donate a fraction of their donation allotment to free and independent media outlets of their choosing. The donation allotment mechanism will be explained through the Second Lever, but the basic idea is to give citizens direct power to choose who they support with their tax contributions. Whoever receives such support would still have to be competitive in maintaining high standards in order to attract donations. This is especially important in the digital age where information is abundant, however truthful information is scarce and often hidden behind paywalls. Transparency is important here as well. The media outlets that receive money from individuals through tax contributions need to be very transparent about how they spend it. An alternative would be to have a public broadcaster or newspaper devoid of any political interference. Achieving this separation from political influence is relatively easy in countries with high levels of public trust, but they don’t have the problem with political power to begin with. It is the low-​trust politically captured countries where such an idea is difficult to enact. However, given that free media is indeed a type of public good that strengthens a democracy, one way to ensure lack of political interference is to pass a law that follows a nondiscretionary set of rules on how to distribute funding to media outlets based on their influence (in terms of readership reach or citizens’ donation allotments), the share of investigative articles/​cover stories reported per month, the share of articles/​coverage that led to open criminal or legal investigations, the share of educational articles/​coverage per month, and so on. This could be defined in an easy-​to-​follow points-​based system where different media outlets compete with the quality of their reports and their outcomes. For example, if a political or corporate affair is uncovered, the outlet that uncovers it is rewarded most points, however every other outlet that covers the same story also gets points (obviously much lower than the primary source), in order to encourage the story about a scandal to spread. This may be obvious motivation in countries with free media, but in countries with captured media when a scandal breaks out, the captured media is often forbidden from covering it. A big problem in enacting this type of solution is that it requires passing a law, which is again dependent on having at least initial political support. It is difficult to expect authorities in corrupt countries to empower the media which can hold them accountable. In some cases, this might be superimposed (e.g., in the European Union), but it again requires navigating through domestic institutions which might be politically captured to begin with. Bottom line is that in countries which already have free and independent media, it can continue serving its function of promoting full transparency and

The Three Levers  287 political accountability. They can continue encouraging civic engagement. In countries where the media is captured, empowering the citizens would have to be the first step before liberating the media.

10.2.3.  Term Limits and Strict Conditions for Holding Office The next major reform aimed at rebuilding trust in politicians and public officials is to introduce strict conditions for those willing to run for public office. Specifically, to impose term limits for every executive and legislative office, and to set up a rule that immediately terminates a mandate if a corruption scandal is uncovered. The goals are to promptly punish transgressors, prevent a single person from holding power too long, and ultimately improve selection into politics. Term limits are a standard practice is some democracies, especially for high positions of power. Their basic idea is to provide a nondiscretionary rule that limits one’s ability to be in power for very long periods of time. The institution is not a modern phenomenon. It has its roots in ancient Athens and Rome, and in Renaissance Venice, Florence, and Ragusa (the Dubrovnik Republic), where term limits were imposed on certain elected officials.6 However, even in developed democracies not all positions of power seem to be term limited. Local government office-​holders are rarely constrained in how long they may run for office, and even though their power is limited to their city, county, or municipality, in that local area their relative power is quite high. For example, frauds in procurement contracts or biased changes in urban planning laws are often administered at a local, not national level, where local mayors have too much discretionary power to determine outcomes. Such mayors can easily build minimum winning coalitions that do not require a large number of people, keeping them in power for long periods of time. Chapter 6 introduced the concept of the selectorate theory,7 according to which politicians, in the absence of accountability and oversight, may stay in power for long periods of time if they build a coalition of supporters that benefit directly from them staying in power. This so-​called minimum winning coalition in democracies implies various vote-​buying practices and using the help of key members of groups like unions or religious or ethnic organizations to secure voters. In the corporate sector, companies tied to the person in power have a strong incentive to donate to campaigns or otherwise encourage votes for the office-​holder. Even in developed democracies, small winning coalitions are more likely on a local level of governance where oversight, transparency, and accountability are much lower. Areas in which politicians repeatedly win due to their small but powerful winning coalitions are more likely to have higher corruption, higher

288  Elite Networks levels of inequality, and lower living standards, and to be less developed than areas where a large winning coalition is necessary to win or stay in power. As expected, small winning coalition areas keep their corrupt politicians in power for long periods of time. Term limits are one way to solve this problem. They are designed for that very purpose, to prevent corrupt politicians with powerful winning coalitions from staying in power for too long. Extending term limits to all levels of government, especially local, is the first step toward breaking up local elite networks. The proposition is to introduce a mandatory yet simple and easy to enforce rule that allows only two terms in executive office, typically eight years, after which the office holder is no longer allowed to hold the same office, but may move up to higher office, again for a two-​term limit only. The rule is to be adopted at all levels of government and all positions of elected public office. This includes terms in legislative assemblies as well. Parliamentary status is, however, a bit difficult to confine to a single term, given that governments in many countries don’t always survive a full term, prompting early elections. For parliament and legislative office, the proposition is then to allow three full terms, with a maximum of 12 years, after which the existing member is either replaced by their elected substitute (if the 12th year happens to fall in the middle of a term) or is unable to participate in the next election for Parliament. They can, however, run for higher or lower executive office, provided that they never held other elected office before. The end goal of the term limit imposition is to enable better selection into politics. To increase variation in the number of candidates who get to hold power and change them often. This is particularly important in corrupt countries where selection into politics is dominated by dishonest individuals. If their time in office is limited, the attractiveness of holding office is lowered. This is obviously only the first step. A policy that can strengthen the term limit imposition and further improve selection into politics is another rule-​based constraint that immediately terminates a mandate in office for a person that has been officially charged with instances of corruption. The emphasis here is on the indictment being official, meaning that it has to come as a result of a police investigation for which a court order has been issued. If the courts accept an indictment for a public official, and a politician needs to be placed under arrest, his or her mandate is immediately terminated, and an early election is called to fill a vacancy (if it is a public office that requires an election). Bear in mind that filing a corruption charge or opening an investigation is not enough, nor is a media exposé of a corruption scandal. These could easily be manipulated by opposition candidates or anyone who wishes to hurt the politician in power. Not until the court confirms the indictment can the termination clause be activated. The proposal is therefore to change the election law with a

The Three Levers  289 specific article stating that any politician who is officially indicted on corruption charges is immediately terminated from their position in office, and is forbidden from running for elected office until the charges are officially dropped. After the court hearings are over, if the politician proves their innocence in front of the court, they are free to run for office again. A problem in corrupt countries is that the legal system is often captured as well. However, even if a potentially corrupt judge releases the indicted politician, he was still ousted from office and has to run again to get back into power. His seat is not immediately restored after his court proceedings are finished. Another potential issue with this proposal is the presumption of innocence. Every individual is innocent until proven guilty in front of the court of law. However, very often individuals who stand trial and are accused of a criminal act spend time in jail (unless they make bail) during the court hearings. Is their constitutional right infringed on given that they are innocent during the trial, until proven guilty? Is a bail bond unconstitutional toward those that cannot afford to pay it? Many EU countries have provisions that clearly prohibit public officials accused of or under suspicion of corrupt acts.8 A person holding public office, if officially indicted of misusing public office for private gain, has lost a moral right to hold that office, until they clear their name in the court of law. In the end, this is a purely political decision, not a legal one. If within a society some politicians never get punished by the voters for their transgressions (because they’ve built powerful coalitions of interest that keep them in office), institutional rules should be imposed to prevent such undesired outcomes. Political capture should be prevented by every possible means. Once again, this might seem like a frivolous suggestion in well-​functioning democracies, given that politicians caught in relatively benign scandals resign immediately, not to mention when huge political corruption is uncovered. A well-​ investigated media story is often enough to draw a resignation of a politician, no matter how high-​up the position. However, in many corrupt democracies and anocracies, this policy is essential. In such countries not even direct criminal investigations or jail time can throw politicians out of office. They either enjoy national-​level protection from more powerful politicians or they managed to capture their electorate and used elite network power to permanently preserve their positions, enjoying uncontested power for 20 or even 30 years. This is why institutional rules must be enacted to discourage such outcomes and deliver immediate punishment automatically when a politician is indicted. *** To sum up, this first set of policies is more relevant to countries with higher levels of corruption, which lack transparency, have a problem of corrupt politicians staying in office for too long, and have a serious problem with media capture.

290  Elite Networks In low-​corruption, open access democracies the media is already free and independent, transparency is much higher, and politicians do get punished when caught in scandals. There is still room for progress obviously, as even in advanced democracies the local government can sometimes lack the same level of accountability and scrutiny as is the national standard. Hence full transparency and term limits for every level of public office are certainly necessary. However, a more applicable set of policies for open-​access democracies is to introduce incentives that depersonalize public institutions and lower political power. Let us start with lowering the scope of discretionary decisions.

10.2.4.  Reduce Discretionary Decision-​Making Power and Introduce Rule-​Based Politics One point of representative democracies is to elect politicians to make decisions over the allocation of public goods on our behalf. Aggregating all individual preferences into collective decisions efficiently and fairly is, in fact, impossible.9 Representative democracies are thus the next best thing—​trying to aggregate individual preferences using elected representatives who are given the power to express these preferences for us. Problems arise when too many decisions are bestowed on politicians and when our expectations exceed the capabilities of elected representatives. This is where the regulatory complexity kicks in, creating incentives for mitigating the rules, instead of simplifying them. As more and more complex decisions are required, the decision-​making process grinds to a halt, and room for manipulation is opened. The process becomes too cumbersome, so people seek shortcuts to get quick and favorable decisions in their favor. Discretionary decisions made by individual office-​holders are such shortcuts. Some decisions are meant to be discretionary. It shortens the process and enables a quicker resolution. But it also gives too much power to those who make the decisions. Too much discretionary decision-​making implies that individuals personalize institutions. If you want a favorable decision to be made by a court of law, you seek favors from the judge making the decision. The judge is not the law, he or she only upholds it. If you need a building permit, and the process of getting one takes two years, you look to find a public official to bypass the process and get it done quickly. If you are caught cheating on your tax returns, you look to pull strings with the tax officials so that they let you off the hook. When the system is woefully inefficient or unjust, you seek your own version of justice and efficiency. Individuals become susceptible to corruption simply because too much decision-​making power is bestowed on them. An institution should make decisions, an institution should issue permits, and an institution should uphold the law, in all cases following a clear set of widely acceptable rules.

The Three Levers  291 Institutions are obviously made of individuals, each with their own set of biases and ideologies, and each prone to cognitive illusions that cloud perception. As consumers we are so easily influenced by simple marketing tricks that skew us toward a particular product. Even before the rise of online algorithms we got persuaded by price anchoring, different framing, and visual distractions. As voters we are so easily persuaded by political propaganda, fake news, political ads, and ever more increasingly, our social media bubbles. We make most of our decisions, whether day-​to-​day or life changing, influenced by confirmation bias, hindsight bias, availability heuristics, overconfidence, and various delusions. We are motivated by greed and self-​interest, we follow crowd hubris, and are blindsided by our ignorance of basic statistics.10 Public officials, themselves human beings, are susceptible to the exact same delusions and manipulations as everyone else. Even, and perhaps especially, when making important decisions on public policy. When the stakes are high there are multitudes of individuals attempting to persuade officials to move in one direction or another, playing very often to their egos, something a person in position of power is particularly sensitive to. If certain market failures exist because we as individuals are prone to manipulation, then political failures arise for the exact same reason. Cognitive manipulations of public officials are, in fact, even worse than consumer manipulations, for two reasons: (1) these are people we entrust to make important decisions on our behalf, and (2) power clouds judgment more strongly than any cognitive delusion or product ad. For these reasons the institutions on which we depend to make fair and impartial decisions on our behalf should not be subject to personal cognitive biases or manipulations, but should instead follow clear and mandated rules. Rules that can obviously be adapted and changed but only as a result of the change in societal preferences mandated by the majority of citizens, and not subject to the whim of an individual politician who wishes to bend the rules and laws to their liking. The legal system is the best example of this. When it works well, when it follows a clear set of rules, even if such rules are based on common law practice, it delivers fair outcomes and people place their faith in it. When the legal system is broken, when it is personalized and subject to manipulation and bribery of its officials, it cannot, by definition, deliver fair and impartial outcomes. A successful, depersonalized legal system is often the crucial difference between corrupt and incorrupt countries. It can be a major source of inequality and wealth and power accumulation if it serves the elite networks, and if the justices themselves are members of such networks. Depersonalization of institutions might sound like advocacy for an impersonal Kafkaesque or Orwellian dystopia where individuals are coerced into submission

292  Elite Networks by depersonalized bureaucratic processes. Far from it. Depersonalization implies more emphasis toward building a fair, institutional, rule-​based approach to political decision-​making and less emphasis on discretionary political power, held by both bureaucrats and by politicians. It is the negation of the Orwellian dystopia, where too much power is granted to politics and corporations. The main goal here is to constrain holders of power, not citizens. To do this, democracies need to gradually build a system that promotes rule-​ based solutions and rule-​based credible threats, and practice less discretion. Just like an efficient and fair legal system, political decisions should be guided by the same principles. A decision is made by an individual, but the individual is following a predefined rule-​based approach that cannot be bent to promote partial interests. Each judge certainly has discretionary power, but no judge operating within a fair justice system makes decisions outside their legal boundaries. A rule-​based approach to politics is envisioned as a system of automatic circuit-​breakers that are activated when a certain outcome is achieved. It is a system of credible threats that pushes politicians to respect the wishes of their citizens in having macroeconomic stability and efficient public good provision. It is not a system of rules that defines all roles in a society. It is set of automatic constraints for those in power, not its citizens. For example, one such rule that binds politicians could be Warren Buffet’s impromptu proposal of keeping the budget deficit under control. Buffet suggested that if the budget deficit goes over 3% of GDP, every sitting Congressman should be ineligible to run for office again. A similar proposal was sometimes advocated for central bank governors: if a set of key monetary policy targets (CPI inflation among them) is broken, the governor should be dismissed. A real-​world example of such a strong constraint was the 2012 US fiscal cliff provision, an imposed rule that would have automatically been invoked had the Obama administration failed to reach an agreement with a Republican-​led Congress over the fiscal reform package. If no decision was to be made by December 31, a series of deep budget cuts and big tax hikes would have been imposed which would have automatically closed the deficit but also sent the economy in a deep recession. The outcome of another uncertain bargaining game, the Brexit negotiations, hung by a thread for over three years with the impeding threat of a no-​deal Brexit. This was a strong credible threat that forced the United Kingdom and the EU to, eventually, reach a deal. These kinds of rules operate as a pure game of deterrence: when there is a credible threat, people will be forced to cooperate to avoid it, whether in nuclear wars, or fiscal cliffs. The rule-​based approach advocated here is more inclined to follow Buffet’s proposal or the central bank constraint, where certain outcomes can trigger immediate punishment of politicians and office-​holders. Having such rules, which everyone agrees are fair and which have the legitimacy to trigger immediate

The Three Levers  293 punishment, are the best possible incentive for politicians not to misuse their power and always be focused on delivering the common good. These rules are designed to reduce the necessity for overregulating. They serve to promote a simple and easily enforceable institutional constraint, very transparent in its conditions, making public officials highly accountable toward their electorates. I will not advocate overly specific rules, as these must be decided by individual countries based on their history and voter preferences (some voters care more about inflation for example, others about unemployment), but I will make a few suggestions as to what this might imply: • Unemployment: if it goes up over an agreed-​on amount the entire parliament11 and executive government branch are dissolved and ineligible to run for office again. This might be manipulated by politicians employing too many people in the public sector, so additional provisions are necessary such as defining the relative upper limit of the private-​to-​public sector employment ratio. Also, it is necessary to control for extenuating circumstances of unexpected spikes of unemployment, like the one that happened during the COVID pandemic. In such cases, the executive and legislative branches have 6 or 12 months to get the situation under control before the rule is triggered. It is easy to see how motivated politicians would be to solve this problem immediately to preserve their seats. • Inflation and monetary policy: if inflation goes up too high for too long (above an agreed-​on upper level like 3% for over two years), the central bank governor and the board are all terminated from their posts. Some countries, small open economies in particular, have no power over inflation, so their monetary policy goals should adapt accordingly (e.g., exchange rate targeting, or nominal GDP targeting). The rule can also be relaxed to allow the governor more time to get inflation (or any chosen target) under control before the mechanism is exercised. It can also include other extenuating circumstances like the situation in the labor market—​if unemployment is low, and inflation remains elevated, this might buy more time for the governor. • Deficits, debts, and fiscal policy: the “Buffet rule” would suffice here, with a few modifications. If the budget deficit or public debt go up above a predetermined level, the government and parliament have 12 months to set it in order. If not, they are all forced to resign, call a new election, and are forbidden to run for office again. This might seem excessive, but with such a credible threat in place it is unlikely it will ever be invoked. Politicians will have very strong incentives (fear and self-​interest) to prevent it from happening. The US fiscal cliff and a “hard” Brexit are examples of credible threats that never materialized. Office-​holders made sure of it. In terms of

294  Elite Networks fiscal policy, there could be a strict provision of an upper limit to the total societal tax burden, where no tax can be increased or new tax introduced if it increases the overall tax burden to citizens and firms. A tax can be introduced only if it immediately replaces another tax, without increasing the overall tax burden (a progressive tax can be introduced under such a provision, but if a society chooses to tax the rich more, they must lower the burden of something else). This is similar to having regulatory sunset clauses, or cases in which a new regulation cannot be introduced without removing at least two old regulations. Furthermore, in line with the tax practices in Switzerland’s direct democracy, a budget law can prohibit the government from imposing new or higher taxes without the approval of its citizens. This may be determined directly via referendums (held online, to reduce costs), which would make an important part of the social contract between the government and its people. • Other pressing issues like inequality, climate change, or even health and education outcomes over the long run can also be subject to their own set of rules that trigger severe punishment of politicians if broken. This is more difficult to implement than fiscal or monetary policies, and is particularly more difficult to hold politicians accountable (who is to say which government is more responsible for climate change issues or rising inequality?), but having long-​term oriented governments is crucial to promote socially desirable outcomes. The proposal, in this case, is to punish governments immediately only if they digress from expected, predefined trends in health, education, or climate outcomes (e.g., population not getting healthier, PISA scores getting worse, CO2 emissions increasing, etc.) Think of these rules as key performance indicators (KPIs) for politics. There is very little we can hold politicians accountable for. During election campaigns many broken promises are quickly forgotten, and voters fail to punish politicians for their transgressions, very often because of poor choices they are faced with. When given clear KPIs, for which automatic punishment is applied, we can have institutional rules to do the punishment for us. If asymmetric information is a problem many voters are facing due to policy complexity and lack of accountability, clear and enforceable institutional rules are the best way to solve such a problem. Last but not least, what about congestion? When there are simply too many rules to follow, and when some rules are in potential conflict with each other? This could bring gridlock, and politicians would be afraid to push a decision because a certain rule is preventing them. To prevent congestion and gridlock arising from too many complex rules that might even be at odds with each other, we must turn to the final reform proposal under the First Lever; to reduce the scope of centralized decision-​making.

The Three Levers  295

10.2.5.  Reduce the Scope of Centralized Government Reducing the scope of government is not the same as reducing its size, even if one can act as a catalyst for the other. Reducing scope of centralized government, in the context of the theory and evidence presented in this book, implies shifting decision-​making over public good allocation from central governments to local communities. Not local governments, but local communities, as shall be elaborated in greater detail under the Third Lever. The main idea here is to lower centralization and reduce the scope of activities and decisions handled by national-​level politicians, for all the reasons elaborated in the previous sections, without giving any more powers or responsibilities to local politicians. This reform is not a call for more autonomy to local governments, as this only exacerbates lack of transparency and lack of accountability. The shift must happen on a community level, engaging more people in local public good provision. In order to reduce centralized political power, we must constrain national-​ level politics mostly to keeping overall macroeconomic stability and delivering national-​level public goods like safety, judicial independence, public infrastructure, and the like. Local public good provision should be left entirely on the community and its citizens to decide. The central government can act as an aggregator of funds (through collecting taxes), it can—​with full transparency—​ redistribute these funds to communities, but cannot decide on how communities should disburse them. The central government, guided by the aforementioned clear and simple institutional rules, would retain the ability to impose strong constraints on upholding judicial independence and punishment for local-​level transgressions. It is a system of checks and balances between the central government and the community, where the main role of the central government is to design incentives to make communities more efficient in public good provision, rather than doing it itself. One must not make a mistake of confusing the act of reducing the scope of political power with reducing the scope of beneficial regulation and public good provision. The argument presented here certainly does not advocate a return to times when governments were small (measured by the size of their revenues relative to GDP), as this is simply not feasible—​the wheels of history do not turn backward. A welfare state cannot and should not be revoked just for the sake of keeping government size small. This is hardly the point. Keep in mind, however, that even in times when governments were small in scope and in size, such as the 19th century times of robber barons in the United States, they still played an important role in granting monopoly rights and licenses. A robber baron couldn’t amass his fortunes without having (usually buying) the backing of political power. During the same period, many European countries still had powerful states led by monarchs, the proximity to which determined one’s riches.

296  Elite Networks Therefore, decreasing political power and the scope of activities handled by national-​level politicians does not mean reducing the power of beneficial regulations, in itself a crucial public good. Providing the necessary public goods like health care, education, roads and public infrastructure, public safety, an independent legal system, maintaining cultural heritage, and financial and corporate regulation that prevents monopolies and the concentration of power, are and should remain the focus of governments. The decisions in allocating these resources, however, should be given to local communities where and whenever possible, while maintaining the credible threat of punishment in case of transgressions. For example, education spending, health care spending, spending on arts and culture, local-​level public safety, and local infrastructure projects are good examples of public goods where the majority can be delivered by the community and for the community. Running kindergartens, schools, and hospitals should be under direct jurisdiction of a community, financed directly by the taxes and donation allotments of people living in these communities. Local public safety, services like the police or fire departments should be under jurisdiction of communities where collective action problems are easier to overcome. People living in a community with high levels of social capital have strong incentives to have the best schools for their children and the best health care for their elderly. Social capital on the other hand is dependent on rebuilding interpersonal trust and trust in public institutions, the key factors the First Lever is targeted to improve. Therefore, shifting centralized power, reducing discretionary decision-​ making, and introducing full transparency and accountability are all designed to build trust in order to allow communities to deliver optimal social outcomes. The other two Levers will explain exactly how empowered communities and citizens can achieve this. Some central government activities cannot be allocated to community-​level decision-​making. Interstate transportation and power grids (gas, electricity, water) are one example, science and higher education are another. Also important are beneficial regulations like antitrust laws that prevent concentration of corporate power and unfair market advantages for huge corporations. There will therefore always be demand for centralized governments, as we have no better system for collective preference aggregation for specific types of public goods. However, when turning to centralized government solutions we hope to draw more efficiency simply by reducing the scope of activities and policy complexity that public officials have to handle. Limiting them to providing national-​level public goods only and imposing strict rules under the credible threat of being thrown out of office are best hopes of leveling their power and reducing the impact and motivation for elite networks.

The Three Levers  297 The same logic is applied to local government officials. Decentralization does not imply granting more responsibilities to local governments, as this would defeat the very purpose of lowering political power. As Chapter 6 has shown, it is much easier to exploit political connections and engage in corruption on a local level where there is lack of oversight and lack of transparency. Local politicians, then, should be given the same institutional rule-​based constraints as national-​ level politicians. They should be granted even less power as their role turns into a facilitator of community-​level decisions. They will be evaluated based on their operational merit, which is key for local governments. Ideology plays little importance when solving local issues, but operational prowess is crucial. The reforms of the First Lever are obviously focused on politicians. But they represent only the first part of elite networks. What about wealth accumulation and the second part of the network—​corporate bosses, the 20th-​century nobility? Unjust wealth accumulation is and always was generated and enabled through the political process. Throughout human history, as detailed in Chapter 2, the oligarchy surrounding a leader would benefit from the mere proximity to unhinged political power. The lesser the scope of activities given to the politician, and the lesser the power of the politician, the lesser the probability one exploits political power for private gain.

10.2.6.  Minor Reforms That Could Strengthen the First Lever In addition to the aforementioned major structural reforms that would establish an effective lever of reduced political power, there are several minor reforms that could be introduced to additionally strengthen the First Lever. However, this set of minor reforms would be applied on a case-​by-​case basis rather than universally. Different countries have different experiences and different electoral and political systems, so proposing an encompassing electoral reform would be pointless. This section suggests a framework within which to think about further electoral reforms that might work better in some countries than in others. Such reforms would most likely arise eventually through the democratic process once the Three Levers are implemented. This is why they are, at this stage, considered to be marginal improvements, but are nevertheless interesting to consider. 1. Introduce a 100% tax rate on campaign spending (direct campaign spending or PACs in the United States) for all donations above a certain amount, e.g., $1,000 in the United States, €1,000 in the EU, etc. Any donation above hence needs to be doubled to account for the tax levy.

298  Elite Networks This reform would discourage large sums of money being spent in political campaigns by large donors and big organizations. It would also lower the amount of money in politics and turn funds away from glossy campaigns and their well-​oiled PR machines and shift it, in relative terms, toward true grassroots movements that better represent opinions of concerned citizens. The tax limit to the donation size should be at or above the mean monthly salary in a country so as to encourage small donors who wish to support their party or candidate. There is obviously scope for misuse where one wealthy person can donate more than, say, $1,000 by distributing it to his/​her friends and family members. Still, finding 10,000 people in order to make a $10,000,000 donation would be much more difficult to do. They are obviously free to spend much more than $1,000, however that payment renders a 100% tax on whatever they donate. Another scope of misuse is to funnel money unofficially and illegally to candidates, an issue that is particularly problematic in corrupt countries. Preventing this type of misuse would require a powerful anticorruption watchdog that would compare the average money spent in campaigns (by counting the costs of posters, events, TV and other ads) to how much the candidate reported they have spent and received. This would be an extension of the full transparency proposal. As for corruption and bribes given directly to candidates for office, the very idea of lowering political power and the scope of politically determined activities has a goal of reducing the demand for political favors in the first place. It does not eliminate political corruption, but it makes it less effective. 2. Lower informational asymmetry for lobbying by increasing transparency. The idea is similar to the aforementioned full budgetary transparency where the public gets to see which pieces of lobbying-​funded research went into the decision-​making process behind every newly enacted law or regulation. Most people have a misconception of how lobbying works. The public image is that this is money been used to strongarm politicians into doing what the lobbyists want, or money spent on various fundraising events. In reality, most funds spent on lobbying are for salaries of the lobbyists themselves who provide crucial pieces of information for legislators. They organize lunches, dinners, and policy events where politicians and their staffers are invited to hear what experts have to say on a certain issue. Money is also spent on experts to deliver conclusions that someone in the industry welcomes. This is not an act of an evil organization paying experts and scientists to lie, as it is often perceived. It is simply a matter of finding the right experts who already agree with the industry view and then paying them to help out and draft a policy paper, or invite them on

The Three Levers  299 a panel to discuss a topic they are comfortable with. No one is paid to lie. Bending statistics—​sometimes. The point is that lobbying solves an informational asymmetry problem for politicians. Legislators in parliament rely on either their staffers or, as we have learned in the book, on their close network of friends for information. Politicians are typically overworked in various committees, spend a lot of time talking with voters or on media appearances, and spend an obscene amount of time fundraising for the next election. They have little time left to read long policy papers and make informed judgements, so they rely on their staffers for this information (who get it from a variety of sources, the most important of which are lobbyists), or directly on lobbyists through a variety of events where they interact. Lobbyists are therefore rarely perceived—​at least by politicians—​as unwelcomed salesmen pushing their agenda, but rather as beneficial actors helping make better and more informed decisions. Lowering their influence through bans is impossible. Their influence can be lowered by lowering political power in general, but also by imposing full transparency over the process through which they influence policymakers. More precisely this would imply that every legislator, when making their vote on a piece of legislation, should make public their entire decision-​making process: which policy papers they’ve used, which opinions they have taken into account, and which ideas they have been exposed to. To a majority of people this information would be irrelevant, but to a group that stands to gain or lose a lot from a piece of legislation, knowing the process is crucial, as they can use it to pressure a legislator to take into account arguments of both sides before making the final decision. This would make the legislative process much fairer and would expose how successful lobby groups actually are in moving political decisions. Public scrutiny is the best remedy to anything done in the shadows of government buildings or at invitation-​only gatherings. Finally, many often advocated political reforms typically include discouraging gerrymandering in order to make democracies more representative, call for more prudent regulatory oversight bodies in many different areas under government control, or propose to improve selection into civil service, making it more apt to help politicians make the right decisions that would favor the many not the few. While these are certainly noteworthy efforts, all of which I would wholeheartedly support, I see them as only marginal improvements over the current state of affairs where political power is too high. Once political institutions are depersonalized, and political power is reduced, such reforms will arise spontaneously from the democratic process itself. That, at least, is the goal. Lowering political power is closely connected to encouraging greater civic engagement and participation among its citizens, operating through their communities. To ensure this will happen, we need two more Levers.

300  Elite Networks

10.3.  Second Lever: Re-​Empowering the Citizens Engaging citizens is the most essential lever in activating the democratic trial-​ and-​error process. There is nothing more important than civic engagement in holding politicians and public officials accountable. After all, politicians are mere agents of the people in the principal–​agent relationship. The very root of the word “minister” implies an act of helping, taking care of, or looking after people, not ruling over them. A prime minister is merely the first among equals, also at the service of all citizens. Similar for the use of the word “state secretary” (as used in US and UK public service), which implies assistance and offering help. The very meaning of the words we assign to the executive cabinet members means they are here to help people, not rule over them. Voters hire politicians to do a job for them, and should have many mechanisms at their disposal for punishing politicians. The Second Lever defines some of these mechanisms. The following set of proposals would work best in empowering citizens: 1. Paying gross salaries to people who can take control of their tax payments. Each citizen would have a tax donation allotment through which they could choose who to donate directly. The goal is to encourage citizens to directly participate and support what they stand for. 2. Introducing mandatory participatory budgeting for community public good provision, as a way of extending and strengthening citizen participation. 3. Punishment referendums: giving citizens impeachment powers to directly punish transgressing office-​holders, introducing a credible threat against political power. Let’s explain each of these with an emphasis on how they help in teaching citizens about the real impact of direct democracy.

10.3.1.  Give Citizens the Ability to Take Control of Their Tax Payments This reform proposal aims to encourage citizens to care about where their tax money is going and how it is being spent. It is the first step toward increased civic engagement. Paying out a gross salary, without automatically deducting the income tax, and then encouraging them to distribute a part of that salary equivalent to their tax payment might sound revolutionary, but if implemented it could lead to several positive outcomes. First, it will help perpetuate the perception that tax revenues are not the government’s money that can be spent as those in

The Three Levers  301 power desire. It is the peoples’ money and they have the right to say where and how it should be spent. Second, having people engaged directly in how they want their money being spent will certainly increase the levels of civic participation and help people understand their own rights with respect to their governments. Finally, having the choice to directly support what they stand for will make sure the money is really going to where it is most needed—​to health care, education, safety, infrastructure, charity, or various civil society groups. It is direct representation at its finest. Having said that, such a reform might be difficult to introduce and implement. An often-​cited reason against it is that people will be late or otherwise negligent with their payments. That argument is baseless given that people are obliged to regularly make payments on their bills and various other communal services or local taxes. These are not taken automatically, but are billed and paid later. Not paying your bills means losing the service. Not paying your taxes means jail time and big fines. People will certainly have a motivation to pay their income taxes on time, every month, just like they pay all their other bills. It should be noted here that pension, health, and unemployment contributions could still be deducted automatically, as these represent individualized insurance policies that people benefit from when the event occurs, i.e., when a person gets sick, losses a job, or retires. This proposal is thus limited to the ability to have control over income tax payments, to have the power to decide where they should be spent. How exactly would it work? Each person receives their monthly salary from which income taxes are usually automatically deducted, or has to pay their income taxes at the end of a fiscal year (in case of self-​employed, rentiers, business owners, etc.), with practices varying from country to country. When those payments are to be made, either personally or via the employer, each individual is given the right to choose from a list of government programs or institutions they consider more important to them, locally or nationally. They are hence given a tax donation allotment that they may choose how to distribute. This is done online or through a government app, where people sign-​in via their social security numbers and unique codes given on each pay slip or each tax return they have to file. Within the app they first start with a choice between most general sectors (e.g., health, education, defense, transport, housing, etc.), and after making their initial choices they go into greater detail and award the specific programs within each major program. This needs to be very user-​friendly and almost gamified to attract users, making the process simple and straightforward to use while still delivering the empowerment effect. However, a large part of every national budget is more or less fixed, given that social security contributions are directly used to fund the pension system or health care programs and are often not enough to bear the entire burden, so revenues from other sources have to be used to fill up the gap. This would imply

302  Elite Networks that every individual could only choose a part of their tax donation allotment (say, 50%) to distribute as they please, or could, for example, be asked to choose at least N out of M (where 2N≤M) institutions or programs listed from each main sector. When designing this the focus is always on simplicity, the process cannot become too complex for an average citizen to use. Furthermore, a part of their tax payment can also be allocated to civil society or charity organizations, or as mentioned under the First Lever, to support local or national media that uncovers political or corporate scandals. For example, 5% or 10% of each person’s tax allotment can go toward supporting these groups, which would reduce their dependence on government money and orient them toward their core work—​doing charity and raising awareness on important issues. It would introduce incentives for these organizations to become even better at their jobs, where they could attract even more money from engaged citizens who divert part of tax revenues directly to these organizations. A more lenient approach would be to introduce tax deductions for individual donations to charity, media, or civil society groups, similar to the tax deduction incentive firms have to donate money. This system could also be set up online, in the exact same way as the tax allotment process—​simple, approachable, and gamified. One question that obviously arises here is, how can we make sure that taxpayer money actually goes to where the people want it to go? Given that the allotment decisions are to be made online, we could use a form of blockchain to record all transactions and see exactly where they end up, i.e., at which public institution. A blockchain is by definition resistant to altering its data entries, as data in a single block cannot be modified without changing all the subsequent blocks on the entire decentralized network. It also makes it possible for anyone to monitor where the money is going and how is it being spent. It therefore strengthens the full transparency proposal of the First Lever, particularly if both of these are to be implemented simultaneously, thus significantly improving trust in the system, and making sure that citizens’ desires are respected. Individual tax allotments will, of course, be anonymized to protect privacy of people’s income statements. There are obviously drawbacks. People might not have the time or the inclination to determine who they send their tax allotments to and would prefer to leave it to someone else, presumably the government. This option will remain a possibility. A person that does not wish to individually determine where their tax money should go is free to opt out and state specifically (within the online app) that he or she has no intention of determining the allocation. This decision can always be reversed, by asking the government to grant back the right if requested. This way people are always given a choice and a possibility to participate more directly in the allocation process. Furthermore, there is an issue of some important services being overlooked and defunded if people do not consider them worthy or if they simply begin to

The Three Levers  303 favor some services over others. For example, it is safe to assume the majority will prefer more money going to health care and education, and less toward relatively unknown regulatory agencies, or to the public administration which has a role that is difficult to measure or value. This is why only a part of the budget is to be allocated through direct citizen engagement. Income tax revenues for OECD countries are on average 24% of all budget revenues, social security contributions (health and pension insurance) about 26% of revenues, and VAT revenues about 30%. The United States, not having VAT, is a slight outlier here as the share of income tax revenues in the federal budget is at 50%, while social contributions in the form of payroll taxes are 36%. However, on a state level in the United States, sales and property taxes contribute more than income taxes. Therefore, even though income tax revenues can be a significant contributor to the budget, if only a part of them (say, half) is allocated directly, this need not at all impede important yet overlooked public services, and it can still deliver the crucial incentive to kick-​start greater citizen engagement over the allocation of and control over taxpayer money. Having all this in mind, this type of proposal is a game-​changer and should be introduced gradually. The best way is to start on a local level, rather than national, to help people get used to the idea, to familiarize themselves with the budget allocation process and how to choose optimal projects. Call it a beta version of the reform. They thus choose only between various local projects and how they want their local tax allotments distributed. The second reform proposal, budgetary participation, builds on this idea and starts locally, focusing on the allocation of funds for important community-​level projects (decisions on what to build, what to fix, etc.) As people gradually increase their civic engagement on a local level, we can expect the same desires to be translated to the national level. It may be a bit more complicated as national-​level governments handle much more duties and responsibilities than local level governments, but some form of direct citizen participation in the public good allocation process must be made available. Finally, an important immediate consequence of such a reform (the first-​order effect), is to make an even bigger part of the budget nondiscretionary and fixed, so that public officials are forced to allocate resources per the citizens’ wishes. This further reduces the scope for government inefficiency or corruption and strengthens the full transparency proposal of the First Lever.

10.3.2.  Mandatory Participatory Budgeting for Community Public Good Provision This reform proposal is merely a continuation of the previous one, with the idea of strengthening and reinforcing it, particularly if applied gradually, starting

304  Elite Networks with community-​level participation. The burden once again falls on technology to enable people to directly express which public goods they wish to be provided. Using the exact same process as described above—​an online app where people sign in with their unique information—​we can introduce direct democracy on a community level where people can choose to allocate public goods that immediately help their well-​being. This proposal is nothing new and is already being implemented in a number of cities around the world. It has proven to be particularly efficient at the local level of government, although it does require a learning curve for the citizens to get accustomed to using the feature. Once they get accustomed to directly participating in the creation of their budgets, they get acquainted with the process and play a very active role in their community development. Some people are obviously more active than others, but the very idea of something like this existing significantly improves trust among all actors. It also educates the citizens of the standard economic concepts of opportunity cost and scarcity. Budgets are limited and only so many projects can be handled at a given time. Also, financing one big project means less money available for a number of smaller ones, and vice versa. The implementation is even more straightforward than the previous proposal. Citizens get a simple survey accessible through an online app where they are offered a set of projects they may rank according to importance, thus choosing which projects are to be funded. For example, for the upcoming fiscal year a set of N projects of roughly equal size and cost are given, where citizens have to choose at least 2. This can be done in groups of projects; for example, 5 expensive projects, from which 2 are chosen, 10 medium-​priced projects from which 4 are chosen, and 20 low-​cost projects from which 8 are chosen. People will have to choose between a set of minor projects such as whether to fix a road pothole, a streetlight, or broken benches in the public park, or will face a choice between larger-​scale projects, such as a new town swimming pool, a new local sports stadium, or a new power plant. The vote would use the Borda count approach, where citizens assign points to each project expressing the strength of their preference. For example, for 5 given projects, each citizen distributes from 1 to 5 points, where 5 points are given to the most preferred, and 1 point to the least preferred. When the points are summed up the two projects with the highest score get funded. A Borda count approach of allocating preferences is chosen as it has the highest probability of delivering a Condorcet winner12 in a pool of multiple choices. To administer, each household is given a letter (or email) alongside its bill for communal services or local taxes, where the rules are clearly specified and each household is given a unique code to use while submitting their votes. The letter tracks the size of each household where bigger households get a higher weight,

The Three Levers  305 i.e., a higher value of each individual vote. People can then choose how to express their preferences: this can be done immediately on the letter they receive which is then mailed by post (by assumption elderly citizens will most likely vote in this fashion), via logging into a web app where people use their unique code, or directly from their phones by scanning the QR codes from the letter/​ email. Anonymity and protection of privacy must be guaranteed, which is easy to implement from a technical standpoint.13 Just like with the previous proposal, people who wish to opt out are free to do so. Having direct democracy this way might not always result in projects that are socially optimal. Which is why, for example, health care or education spending could be handled separately. The choice should not be between a new school or a town swimming pool, but how to distribute the within-​health and within-​ education budgets: what to prioritize and why. Nevertheless, even this might not deliver the most socially optimal policy, which brings us back to the democratic trial-​and-​error process. Even if, at first, projects are chosen that do not deliver the greatest value to the community, in the second or third trial, these errors will be fixed. The community will eventually converge toward an optimal social equilibrium and deliver the necessary public goods. The whole point is not having a central authority force a project onto a community. It should be left to the citizens themselves to decide. Even if they sometimes make wrong decisions. Wrong decisions are useful as they can help citizens see the errors they’ve made and correct them. On a community level, errors are very easy to spot. The main goal of this proposal is to offer citizens an option to choose exactly the type of projects they want, thus respecting their will and creating a positive and true perception that citizens do actually have control over budgetary redistribution. This is another example of an irreversible reform. Once citizens get used to having direct participation and control over budgets, it becomes very difficult to regress to the old way of nontransparent, nonaccountable, and discretionary decision-​making local governments.

10.3.3.  Granting Impeachment Powers to the Citizens for Direct Punishment of Officials Ancient Athenians used an effective method of controlling political power called ostracizing. Each year the citizens of Athens with a right to vote (adult men only; women, slaves, and foreigners living in the city were not allowed to vote) could nominate one citizen who they assumed was becoming too powerful or too dangerous and vote to exile him from the city for 10 years. Each citizen would engrave a name on a piece of pottery (ostrakon), the names would

306  Elite Networks be counted by the highest officials, and the person with the most votes would get exiled (ostracized), but would not lose his citizenship or his property in the city. They could come back after 10 years and resume participating in public life. Even though this was a well-​established institutional rule, the citizens of Athens seldom used this power. Over the course of the rule being in place (it is believed it was only used in the 5th century BC), less than 15 people in total were ever ostracized, meaning that their transgressions must have been significant to deserve such an outcome. This could have been done to any citizen of Athens, not just the governing politicians, anyone who was thought to be a threat to democracy. Reintroducing the same rule today would be impossible. For one, it is an infringement of basic human rights, something that ancient Athenians did not possess as an institutional safeguard. It is easy to imagine governments, especially anocratic or dictatorial regimes, using propaganda to punish a scapegoat, or prosecuting their political opponents (or within-​party threats) and hence eliminating them as competition (dictators and wannabe-​dictators already do this, without the institution of ostracism or having to ask the people what they think). Second, the legal system is designed to punish transgressions of all kinds. It is complex and often slow for a good reason; to avoid manipulation and ensure just outcomes.14 Punishment for breaking the law is not a power bestowed on the masses, and for a good reason—​it can quickly and easily turn into mob rule. However, having elections every four years as the only method for citizens to enact punishment against power holders is clearly not enough. Elections rarely act as a referendum on government performance during the previous four years. First of all, voters are myopic and only take into account what happened a few months before the election, meaning that the government is held accountable only for things that happened recently. Second, in many high-​corruption countries, governments engage in clientelist practices and buy votes, usually through local power brokers (as shown in Chapter 6). Third, elections offer a number of choices, where people typically choose the least bad option (“the lesser of two evils”), and may be stuck with a bad government simply because the alternatives are even worse. Elections therefore lack the necessary credible threat against political power, especially if politicians operate within powerful small winning coalitions. Citizens need an additional constitutional force which they may use in between elections to remove individuals they see as unfit or too powerful for their position. I propose an option to hold a punishment referendum against any in-​power elected official where people may vote directly, via a referendum, to punish the official outside of regular elections. The punishment is enacted in the form a ban against running for public office again.

The Three Levers  307 Think of this as the power of impeachment given to the people, instead of the Parliament. Parliaments in every country have the power of calling a vote of no confidence against a government or any single government official. But if the party in power holds an absolute majority in Parliament, it is difficult, if not impossible, to successfully enact this punishment. Giving impeachment powers directly to the people via referendums is a compelling credible threat that acts as a constraint against any transgression of power. Elections are few and far between. Citizens need to be able to hold officials accountable more often. A punishment referendum can be a compelling deterrent against politicians, so we need strong institutional safeguards and rules to prevent potential abuse. First, voting needs to follow the same rules as any other referendum. In order to call it, at least 10% of registered voters need to organize a petition. The vote does not have to be an annual event as it was in Athens, it can only be triggered if there is enough will from the people to hold it. Second, at least 50% of registered voters need to turn out to make the results binding. Third, a punishment referendum can be called for a single person, or as a vote on up to five individuals. This depends on the petition demanding the referendum. If there is only one person on the ballot, he or she is punished with at least 50% plus one vote. If there is more than one person, then the person with the most votes gets punished. Fourth, the referendum can only be held against those holding office, executive, legislative, or judicial. It cannot be applied against any citizen or against opposition politicians, only those who need a constraint over their position of power. Fifth, there can be only one punishment referendum in a single year, and there should be at least 12 months between the two processes. This is to prevent procedural abuse. Finally, the punishment implied by the referendum needs to be strictly imposed and defined constitutionally. The proposal is to ban the person from running from public office again. The point of having such a rule is to give citizens an option of a credible threat against those holding power. Like nuclear games of deterrence, the threat of retaliation prevents the other side from using their attack. In a standard Nash equilibrium,15 it is hard to expect that citizens will use their credible threat too often, but it delivers an important signal to any dishonest politician who wishes to engage in corruption, or to any politician with autocratic or dictatorial tendencies that the punishment for such transgression can be severe. The idea is to make public officials afraid of the people, to turn them into true public servants, and thus change incentives for entry into politics to include only honest and morally incorruptible individuals. Very often losing elections is not enough of a credible threat, because winning elections requires forming a minimum winning coalition of interests to keep a person or group of people in power. These referendums merely add another dimension of political accountability and control, and grant a powerful credible threat in the hands of the people.

308  Elite Networks

10.4.  Third Lever: Re-​Empowering the Community Why emphasize the community? What makes it so important in improving people’s trust in each other and their trust in public institutions? According to a number sociological and economic research studies, the community delivers great value to its members. It offers them meaning through participation in various local organizations and associations like parent-​teacher associations, school boards, neighborhood oversights, religious groups, volunteer fire departments, charity organizations, community events and gatherings, and the like. The community anchors individuals, gives them a sense of control and self-​ determination, and provides an opportunity to directly engage in public good provision, thus encouraging them to make a difference in the communities they care about. Research has found that growing up in better communities carries a significant impact on a child’s economic and social outcomes later in life (such as college attendance and earnings).16 Most importantly a healthy and vibrant community is a crucial characteristic of a well-​functioning democracy. Local communities can also be much better at solving the collective action or common pool resource problem. Through continued cooperative behavior they can successfully internalize externalities and discourage free riders. This was best proven by Elinor Ostrom’s Nobel prize–​winning work using the cases of local fisheries, groundwater basins, and irrigation systems.17 A community offers full transparency and decentralized decision-​making where every member sees how much another member is contributing. Members encourage and help each other while building strong social capital and interpersonal trust. In such environments the incentives for free riding are extremely low, meaning that collective action is achievable and can be quite efficient. This is true for any close-​ knit community, whether defined geographically or through common interest. In his book The Third Pillar, Raghuram Rajan delivers a persuasive argument on the importance of the community in breaking up concentration of political and economic power. Rajan stresses several positive roles of a community. It teaches younger individuals how to internalize community norms and values, how to promote community interest before self-​interest, and it even supports education for its members. It helps build binding social relationships where neighbors help each other in time of distress and can even act as lenders of last resort.18 Finally, communities encourage favors and resolve conflicts and disputes between members, which reduces the necessity for the legal system to intervene. A community therefore reduces the need for both judges and regulators.19 It does not eliminate either (nor should it), but it reduces the demand for both, making them more efficient and less costly. Rajan, however, also finds that in the persistent battle between states and markets, between protectionism and globalization, it is the community—​dubbed the third pillar of this

The Three Levers  309 delicate balance—​that has been in persistent decline; social, economic, and political.20 Among the first to notice this trend in the United States was Robert Putnam in his seminal book Bowling Alone, where he details how Americans are becoming increasingly detached from their communities. He paints a bleak picture of American society where social capital has steadily been depleting ever since the 1960s. Civic engagement—​illustrated through a decline of league bowling even though individual bowling has increased, and measured directly through membership in various organizations like religious groups, labor unions, parent-​ teacher associations, veteran organizations, and of course political engagement—​ has been in persistent decline.21 One consequence of this is depleted democratic capital, which limits the power of democracy’s trial and error mechanism. But an even more important consequence might be something that was documented in Case and Deaton’s book Deaths of Despair: middle-​aged non-​Hispanic whites in the United States, lacking a college education and typically living in remote rural areas, are increasingly dying from despair-​related incidences: suicide, opioid abuse, and alcohol poisoning.22 Community disengagement, along with failing US health care, are major reasons behind such adverse social outcomes hitting a large fraction of Americans. The lower educated white working class has faced particularly hard times over the past 20 years, not just in the United States but across the developed OECD economies. Well-​paid manufacturing jobs being replaced by lower-​paid service jobs offering much lower job security was the key economic consequence of technological progress and globalization in the developed West. These trends carry remarkable resemblance to the postcommunist transitions in eastern Europe23 in the 1990s. The less educated working class, living mostly in rural areas, lost their jobs in traditional industries, became long-​term unemployed or were faced with lower-​paid service jobs, and saw greater instances of depression, anxiety, and social apathy. As a consequence, the political spectrum in eastern Europe switched completely: losers of transition—​lower educated working classes with low incomes, higher instances of unemployment, and living closer to the poverty line started voting for right-​wing parties, whereas the winners of transition, living in low unemployment areas, with higher incomes and greater opportunities (the new bourgeoisie) started voting for left-​wing parties.24 This sounds very familiar to the contemporary political landscape of many Western democracies, the United States and United Kingdom in particular. Low-​ income working-​class voters are turning to the Right, whereas high-​income upper-​class voters are voting for the Left. The working classes in particular are feeling detached from the society’s elites. They distrust them, as much as they distrust politicians in general, and blame them for negative social outcomes

310  Elite Networks and their own failures.25 Nothing showcases this better than the reactions of our societies to the COVID-​19 pandemic. In the United States and across many European countries the pandemic has been remarkably politicized. Societies are so divisive that even a natural occurrence like a virus is used to sow political divisions and spread conspiracy theories. Distrust against science and experts mirrored the distrust against governments and society’s elites, making the pandemic seem as yet another deliberate failure of establishment elites. The anger and anxiety felt by voters searching for antiestablishment candidates (like Trump) or antiestablishment outcomes (like Brexit) is fueling even greater discord and throwing citizens into echo chambers (social bubbles) where it is becoming increasingly difficult to understand a person you disagree with politically. The social media is perpetuating such echo chambers by serving us only the information we agree with, thus further insulating us from other peoples’ experiences. Evidence of this can be seen in Figure 10.1, which shows the map of the 2020 US election results between Biden (blue) and Trump (red) by county, where each county is visualized according to its population size. Notice the immense urban–​ rural political divide. In traditional Red States, where Trump won easily, counties with denser urban populations voted Biden (e.g., Houston, Dallas, Austin, and San Antonio in Texas, New Orleans in Louisiana, Nashville and Memphis in Tennessee, etc.) In traditional Blue States, where Biden won easily (like New York, California, Illinois, or Washington), the large urban areas all voted for Biden, but the small rural areas all voted for Trump. This pattern is strong across the country, and is a testament to the urban–​rural political divide in US politics that has been present for the past few decades. The striking thing about this divide is how it emphasizes geographical echo chambers. Case and Deaton’s deaths of despair are most frequent among middle-​ aged low-​educated whites typically living in rural areas. Their voting patterns suggest they are looking for a way out, and populism is a particularly attractive remedy. But as they get increasingly detached from urban voters—​and vice versa, as urban voters get detached from rural experiences—​we are left with a deeply divided country, both parts of which are wrapped in their own bubbles, now even geographically completely detached from each other. Add to this the different news sources used by different communities and the bubbles become obvious. No wonder that US political polarization as measured by Congress voting patterns is at an all-​time high.26 However, unlike their urban bubble counterparts, the rural bubble inhabitants are increasingly living in dysfunctional communities characterized by higher levels of distrust, higher levels of social apathy, and in general a zero-​sum mentality. This makes them easily swayed by zero-​sum populism, not just in the United States but across Europe as well.

The Three Levers  311

Figure 10.1  Map of 2020 US election results by county. Red circles are counties won by Trump (Rep), blue circles are counties won by Biden (Dem). The size of a circle depicts the population size of the county. Source: New York Times, https://​www.nyti​ mes.com/​inte​ract​ive/​2020/​11/​03/​us/​electi​ons/​resu​lts-​presid​ent.html.

Spotting a dysfunctional community is easy. It is typically a place with higher levels of crime (theft, murder, rape), poor public services, corruption, and perverse incentives that propagate the self-​defeating negative social cycle. As Case and Deaton pointed out, these are also places with higher instances of middle-​ aged white deaths related to drug overdose, alcohol poisoning, and suicide. Such communities carry a strong distrust against the state and are in general bad or depressing places to live in. However, despite their distrust of governments and poor public services there is still a greater demand for government intervention. As Rajan27 puts it: The faint hope that the government will dig a latrine, pave a road, or discipline school teachers can prevent the local population from organizing to do so. In frontier towns in the United States, the community raised a barn or built a road itself knowing there was no one else who would do it. In dysfunctional communities where the government is closer, the misplaced expectation that the ghost of the inefficient government will eventually appear and do the job crowds out what little private incentive there is.

This finding has been shown to be true on a cross-​national level as well: in countries with lower levels of social capital—​approximated by interpersonal trust—​people

312  Elite Networks support more government intervention in the economy.28 In other words, even when people know their governments are corrupt and inefficient, they still prefer greater regulation and government intervention, simply because they don’t trust private incentives or community engagement to provide solutions. This makes sense if one thinks about the interrelation of a country’s formal and informal institutions. If formal institutions are weak and inefficient, meaning that if the rules in society are poorly defined and offer too much scope for corruption and perverse incentives, then informal institutions will adjust accordingly and this will affect the accumulation of social capital (people start losing trust in the system and each other). The relationship works in both ways; if the “rules of the game” are defined to be inclusive, then the accumulation of democratic capital will eventually lead to a change in customs, norms, and trust in a society (i.e., our informal institutions). Moving this down to a community level, in functional communities that deliver the right incentives for their members, people invest heavily in their social capital. This rewards them through greater incentives for cooperation, reduces the necessity for greater regulation, while government intervention is limited to providing the key public goods that the community desires. In dysfunctional communities, reverse forces are in play. People fail to invest in social capital, they shun cooperation with their neighbors and demand greater regulation. Trust is extremely low as are economic opportunities. The local government is likely to be corrupt and public good provision subpar. In such communities or societies not even supranational authority can be too helpful as it creates the same dependency and the same crowding out effect. A good example here is the relationship between the European Union and its more corrupt Member States. The EU has no credible threat of punishing corruption in its Member States. It does extend pressure on candidate countries to lower corruption and increase institutional safeguards, but once a country becomes a member, the EU has little power to prevent corruption and kleptocratic practices of a country’s political class.29 It continues, however, diverting resources to infrastructure projects in many of its more corrupt Member States (under undoubtedly noble goals of fostering development), only to see a large chuck of such projects get red flagged for procurement corruption.30 The EU’s institutional bodies that are given the task to prevent corruption (like the European Anti-​Fraud Office, OLAF) are relying on each Member State’s domestic institutions to aid them in their investigations, which they obviously fail to do if such institutions are themselves locally politically captured. Corruption goes unpunished, infrastructural projects funded by EU money get overpaid, while local politicians use the excess funds to divide them between themselves and their cronies (rewarding their winning coalitions), thus strengthening their local rule, much to the dissatisfaction of their societies. The negative vicious cycle keeps spinning as greater government

The Three Levers  313 inefficiency and corruption increase social apathy and discourage civic and political engagements, trapping voters in the same social echo chambers as we have seen in the United States. It is getting increasingly difficult to break up dysfunctional communities, even in developed societies.

10.4.1.  Decentralize and Depersonalize Decision-​Making over Allocation of Public Funds How do we prevent a community falling into the vicious cycle? Or even more importantly how do we pull dysfunctional communities out of their abyss in order to ensure that delivering more power to community-​level decision-​making will actually take effect? As emphasized earlier, rebuilding social trust is such communities is the key to kick-​start the democratic trial-​and-​error mechanism. This is becoming the number one priority in rejuvenating communities and in the long run lowering incentives for power accumulation. Where to start? Community re-​empowerment is a natural extension of the first two levers and is rooted in the idea of delivering more decision-​making power to community self-​government, down to a neighborhood level. The point is to have decisions decentralized and depersonalized, meaning that allocation over public goods should not be contingent on individual power, but on collective power. When more power is given to communities, this helps them in their economic reinvigoration by attracting more individuals to live there, and sets the stage for the rise of dynamism and innovation. It is important to emphasize that this does not mean simply relocating power, taking it away from national-​level politicians only to give it to local-​level politicians. That amounts to nothing as local-​level politics most often exists in the shadows, away from media attention, and deprived of any accountability. The idea is to give more decision-​making power to the community itself, not its representatives or its local leaders. The local government needs to only be a service for the community, to provide the public goods that the community needs, and that it collectively enforces. Being the leader of a community cannot be a function that allows any discretionary spending decisions given to the leader. On the contrary, all spending and redistribution decisions need to be handled by the community itself, where everyone participates, whether it is a street, a neighborhood, a small town, or a village. Size is key here, as the smaller the community, the more likely it will be able to solve the collective action problem in line with Elinor Ostrom’s common-​pool solution. Power is in the hands of a small and easily accountable collective, not the individual. What would this look like? Imagine a neighborhood where all local issues—​ decisions on where to build a road; fix a street lamp; build a park; build a new

314  Elite Networks kindergarten, school, or small hospital; organize an event; or grant a building license for a new house, apartment complex, business, or other major investment—​ are decided solely by people living in that neighborhood. Neighborhoods don’t have large budgets, nor should they. But they have the people’s decisions on where to place their tax money (Second Lever), and they have the power to order the city officials (or higher government officials in case of bigger infrastructural projects) to build them what they voted for via participatory budgeting (again, Second Lever). As explained under the Second Lever, budgets are limited, and voters would not always get what they want, but they would have choices over which projects are of higher priority for their community. Each community would get assigned a fixed amount (proportional to its size and tax contributions) not to spend it themselves, but to decide how to spend it, on which projects. The community does not handle any funds. It only participates in decision-​making over the allocation of such funds. All decisions coming from local communities—​ villages, streets, neighborhoods, no matter how they’re organized—​are binding by law, and higher-​level officials have an obligation to fulfill them. Higher-​level officials are thus nothing more than executors of the people’s will. They have no power over what gets built on a local level. They don’t even pick the contractor. Their only job is operational: to administer the decision, write up the public tender, conduct it, make payments, and make sure the legal and regulatory requirements are respected. As described under the First Lever, governments, whether national or local, truly become servants to the people. They still maintain the rule of law, the justice system, and all necessary public services, but the people have a much greater say over how funding is allocated, including for things like wages of public sector servants. This would imply that, for example, teachers’ unions no longer bargain over wages with governments for the entire country. Wages of school or kindergarten teachers, doctors in public hospitals, police officers or fireman, local judges, or local public sector employees are negotiated within each community directly. The people decide who they value more. Communities will soon come to realize the laws of supply and demand governing such wages. If there is no good doctor or teacher in a community, they might want to attract one by offering higher salaries. If the public office doesn’t attract good enough candidates, wages in the public sector need to be increased. Only wages of public sector employees on a national level (people working in ministries or national-​level authorities like the central bank, regulatory agencies, etc.) are decided on more centralized levels, which is where employees are free to join institutional-​level unions if they require greater bargaining power. Universities, for example, can function as their own community. Many, in fact,

The Three Levers  315 already do. One more level is to hand them decisions over the allocation of all public goods in their community, while maintaining the provision that members do not have direct access or the ability to spend these funds, just the power to make the allocation decision. Empowering the communities this way might seem like a radical change. This depends on how one looks at it. A re-​empowered community does not imply a return to tribal societies where the nation-​state is constrained and impotent. Its point is to engage citizens into local-​level allocation decisions over public goods, where they can actually make a difference in their community. It does not imply that central governments should stop offering their key public services (like public safety, health care, education, infrastructure, defense), but it does imply that the provision of such services will be considered much fairer and more representative of what the voters really want. This is far from radical; it is democracy at its finest. One thing that might discourage decentralization of power and community-​ level decision-​making is the geographical clustering of capital and powerful people in big cities. Figure 1.2 in Chapter 1 shows this—​the majority of wealth and power are created and maintained in a few locations in each country. The network effect is strongest in such areas, due to the mere proximity between politicians, corporate executives, and other powerful individuals. A full decentralization and disruption of elite networks would imply dissolving such clusters and discouraging the network homophily effect in order to prevent clustering of corporate and political power. It would basically imply that we force the most powerful companies, individuals, and politicians to be randomly scattered across the country in order to reduce their exposure to each other and make them more focused on developing the local community they find themselves in. This, however, is impossible to achieve as it defies the laws of physics when it comes to how networks behave (recall the examples from Chapters 4 and 5), not to mention that it is woefully unconstitutional and illiberal. People have every right to choose where they live. While a few wealthy individuals do relocate, the main clusters are still formed in places like Washington, DC; Silicon Valley; or Manhattan in the United States (and its close proximity, like business owners living in Connecticut, or Silicon Valley “tech bros” living in San Francisco); London, Brussels, Berlin, or Paris in Europe; Hong Kong, Singapore, or Shanghai in East Asia; Dubai and Doha in the Middle East, etc., or are perpetuated in places like Davos. Washington, DC, is a good example of the network clustering effect—​the capital became the headquarters of all major interest groups and lobby organizations, and all individuals working in or around politics. The huge US political industry is condensed in DC, so much so that the city itself underwent remarkable gentrification due to a lot of wealthier individuals working in

316  Elite Networks the industry starting to move into the city (rather than its suburbs), thus completely rejuvenating once high-​crime neighborhoods. One consequence of the COVID-​19 pandemic—​very possibly another critical juncture in human history—​that could help increase community empowerment and possibly lower the clustering effect in big cities, is the trend to work from home. People from big cities, working from home, were increasingly moving to smaller towns during the lockdowns. A good sign of this was the rise of room lettings in smaller communities around big cities, and a significant decline of room lettings within big cities during lockdowns across the United States and many European countries. After the pandemic ended, however, life in big cities was quickly restored, as they still offer a multitude of benefits, but some fraction of individuals might choose to stay living in smaller communities, working remotely. Some companies are already encouraging this, which is lowering their fixed costs. Some are even considering offering monetary incentives to their employees to move to other places. Think of the potential this can bring to rural areas and small towns. They could flourish as individuals with higher incomes drive greater demand for goods and services in such communities. Working remotely outside the big city furthermore lowers commuting time and helps save the environment by reducing emissions. The majority will probably not choose to do so, but it doesn’t take a lot of people to start changing local communities. Finally, clustering in big cities around nodes of power does not imply that decision-​making over public good allocation in such cities cannot be organized on a neighborhood or street level. The units will be smaller, but the collective action solution is the same. As for powerful individuals which form clusters around themselves, their limit to power is contingent on the limit of power of politicians, the First Lever, but also by decentralization of decision-​making. If a corporation wishes to build a factory or a plant in a community, it needs to get approval from all its members, not from a single politician who they might be tempted to bribe or donate to their campaign. It will be much more difficult to foster a negative impact on a community (in terms of negative externalities from pollution for example), when decision-​making power is decentralized. Depersonalization and decentralization are thus crucial in order for community re-​empowerment to work. They are crucial for democracies to deliver its greatest benefit.

10.5.  Second-​and Third-​Order Effects of the Three Levers The proposals under each lever are all designed with second-​and third-​order effects in mind, and they are all conceptualized to support and reinforce each other. What, therefore, can we expect after gradually implementing each? Which societal changes are most likely to occur?

The Three Levers  317 The first desired long-​term effect of the Three Levers is to increase both interpersonal trust and trust in public institutions. In the gradual process of achieving this effect, incentives for elite network creation should be reduced, and hence inequality is expected to decline. All this is supposed to be achieved by a combination of lowering political power, depersonalizing government institutions, increasing accountability and transparency, improving selection into politics, and diverting more centralized decision-​making over public good allocation to local communities. This is obviously easier said than done and depends on a gradual implementation process that slowly integrates second-​and third-​order effects. There is no easy and quick path to dismantling elite networks. Having the citizens participate and enact more control and accountability over its public institutions is a process that needs to be learned and adjusted to within a democratic setting. It will take time before the implementation of each reform proposal encourages more citizens to actively engage by monitoring politicians and bureaucrats and discouraging them from trying to embezzle public funds. It will take time before citizens get used to directly allocating their tax allotments and choosing local public goods directly. And it will obviously take time before institutions get depersonalized, and centralization of power is reduced from national and local politics. After this first-​order effect is achieved, the second-​order effect is to expect office-​holders to start being more careful and accountable. After making an even bigger part of the budget nondiscretionary and fixed, public officials are forced to allocate more resources per the citizens’ wishes. This further reduces the scope for government inefficiency or corruption and strengthens the full transparency proposal of the First Lever. The very idea of granting more power to citizens in the allocation process and controlling this allocation through an online decentralized mechanism further increases transparency and discourages incentives for cheating, fraud, and political discretion. Greater citizen participation also encourages free media, while free media encourages citizen participation. Their combined strength ensures that politicians and corporations are called to greater scrutiny and held accountable, while rebuilding trust in public institutions and among each other. Add to all this the credible threat of citizens being able to severely punish office-​holding transgressors, and incentives for entering politics are immediately different. When there is little room for embezzlement and corruption, when their power is severely limited to the role of true public servants, dishonest power-​ hungry individuals have less incentive to participate. They shift selection into politics to a pool of honest individuals with a sense of public duty who can truly perform the role of public servants (by being benevolent leaders and executors of people’s will) instead of current dissolute rulers. Similarly, less centralized decision-​making for politicians lowers incentives for corporate bosses to form

318  Elite Networks elite networks to influence politicians, as the costs of doing so significantly outweigh the benefits. If being in close proximity to an office-​holding politician does not help you get a favorable decision from a government, given that such a decision depends on institutional processes and not individual will, there is no incentive to form strong elite networks that help skew incomes of top executives. Over time inequality should thus go down. These are the desired third-​order effects. Table 10.1 summarizes the full mechanism and expected long-​term effects. On a community level, greater decision-​making has a goal of greater civic engagement and participation in democratic decision-​making as its second-​order effect, thus strengthening the Second Lever, citizen empowerment. Today, in many communities and societies, a lot of individuals feel underrepresented, either due to their income, race, gender, nationality, or otherwise. Many feel left out of social progress and seek radical changes hoping this might change their position in society. The Second and Third Levers directly aim to help underrepresented groups become more engaged and more likely to be able to change their social outcomes by changing their communities. Changing communities via greater civic engagement is much easier and significantly more likely than changing national-​level politics. Community engagement is the first step to changing national-​level outcomes. The second-​order effect is thus to teach people about democracy. Teach them how to interact with others in their proximate community, particularly those they disagree with. Teach them about the very process of how decisions are being made, who is to be held accountable, have them participate in raising issues, giving opinions, and making actual contributions. Once people get accustomed to this, once they start seeing social changes they’ve instigated, this will encourage them to further increase their social engagement, thus raising both social trust and democratic capital in a society. The third-​order effect is, finally, to improve trust and build democratic capital. It is the cornerstone change that enables dysfunctional societies and communities to break the negative cycle of underdevelopment. This will no doubt be a process of trial and error, where many errors will be made, many bad decisions reached, and many negative outcomes witnessed. But with greater interpersonal trust, built through the process of community engagement over collective action decision-​making, members of a community will continue down that path regardless of a few wrong choices. And the final outcome will without doubt be beneficial to the community and its members. *** The Three Levers elaborated in this chapter do, however, leave a few open issues. First, politicians have a legitimate interest to get elected, and there is usually a divergence between their interests and the dispersed interests of the citizens. What

–  Full transparency – ​Free media – ​Term limits – ​Rules instead of discretion – ​Reduce scope of centralized activities

– ​Tax allotment to citizens – ​Participatory budgeting – ​Citizen impeachment (credible threat)

– ​Full decentralization of public good allocation

Lever 1: Reduce political power

Lever 2: Empower citizens

Lever 3: Empower the community

Reforms

– ​Greater inclusion of underrepresented groups in society – ​More power to the community, not local government

– ​Greater citizen control over the budget process – ​Direct engagement in public good allocation

– ​Greater transparency and accountability – ​Better use of technology to prevent corruption – ​Rule-​based politics

First-​order effects

Table 10.1  Long-​term effects of the reforms proposed under the Three Levers

– ​Changing dysfunctional communities

Third-​order effects

– ​Greater civic engagement – ​Discouraging dishonest politicians – ​Improved trust and – ​Turning politicians into greater democratic public servants capital – ​Depersonalization of – ​Better selection into governing institutions politics – ​Lower incentives for elite network formation – ​Full citizen engagement – ​Lower inequality on local-​level issues

Second-​order effects

320  Elite Networks is the optimal way of resolving this conflict when the motivation of a politician is much stronger than something we might refer to as public interest? Are the superimposed rules that I mention in the First and Second Lever enough? How can we expect politicians to even impose such rules that would reduce their own power? Is there a group of citizens powerful enough to impose the implementation of such rules, following the long democratic trial-​and-​error process? How long should we be forced to wait for the crucial societal reforms to take place? In other words, how can we disempower politicians with minimum resistance? The second set of issues concerns the question Friedrich Hayek asks in the Road to Serfdom: “why do the worst get on top”?31 His response, based on the experience of the rise of Nazism and fascism in 20th-​century Europe, was that the ruthless and unscrupulous spared no effort to reach the heights of power in societies where hierarchical orders are handed too much power. If selection into politics is on average adverse (as elaborated in Chapter 9), how can we expect such individuals to ever agree to constrain their own power? The issue is therefore whether the aforementioned gradual reforms are enough to prevent the worst from rising to the top. After all, some dictators and war criminals came to power via democratic elections (after which they’ve abolished democracy, or made it pro forma). Can these reforms be robust enough to prevent another potential autocratic or anocratic ruler from reversing them? Rules and constitutions are often changed under immense populist pressure. To make sure that these rules become irreversible, we as a society need to learn more about democracy. This is why community-​level decision-​making is important—​it teaches people about democratic decision-​making. It teaches people about the advantages and disadvantages of representative and direct democracies, and which is better for which type of public good allocation. Preventing the worse from coming to the top and breaking down political pressure to lower their power is not something we can expect to happen overnight, or through a violent revolution. It is a gradual, even painful, process that might last for decades before we can see its first results. Given the constraints that we face from too much political power and the effect it renders on inequality, skewed opportunity, and wealth concentration among elites, there is no alternative to a gradual democratic trial-​and-​error that reduces political power to eliminate all of its malign outcomes.

Afterword Throughout the book I refer to elites and elite networks as something inherently negative. Societies need elites. They have always existed and always will exist. They have shown a remarkable level of persistence throughout history, while its modern descendants have played positive roles in developing societies, often even as agents of progressive change. The problem is not with the existence of elites, it is about the way these elites may capture corporate and political power and misuse it to prevent opportunities for everyone else, and hence cause an increase in wealth concentration and inequality. Elites are just like politicians; good ones certainly do exist, but the problem arises when the bad ones—​with a far stronger motivation for usurping power—​ take over and skew the distribution of resources and opportunities within a society. This book has been about explaining how this mechanism works and eventually finding solutions to constrain such behavior. It adds the missing element in the explanations of long-​term inequality: the capture and misuse of political power. If there is one thing, one line that should come out as the ultimate conclusion of this entire study it is exactly that. Political power is the root cause of inequality, through the mechanism of elite networks. Fixing this issue requires taking aim at a different target: the cause, not the consequence.

Notes Introduction 1. The report, classified as “secret,” was published on WikiLeaks, and is available in full here: https://​wikile​aks.org/​plusd/​cab​les/​08TUN​IS67​9_​a.html. 2. Rijkers, Freund, and Nucifora (2014). 3. The Christian Democracy party dominated Italian politics for 50 years, winning power in every single election since 1946 until 1992, except for one 3-​year mandate by the Social Democrat Benito Craxi, himself a victim of the Manu pulite investigation that got him convicted for corruption. 4. Bloomberg Businessweek, June 2018: https://​www.bloomb​erg.com/​news/​featu​res/​ 2017-​06-​08/​no-​one-​has-​ever-​made-​a-​cor​rupt​ion-​mach​ine-​like-​this-​one. 5. Riordan (1995). 6. Ackerman (2005). 7. Riordan (1995). 8. Bueno De Mesquita and Smith (2011). 9. “Bell’s Rizzo sentenced to 12 years in prison,” Los Angeles Times, April 4, 2014: https://​ www.lati​mes.com/​local/​la-​me-​0417-​rizzo-​pri​son-​20140​417-​story.html. 10. Abramoff (2012). 11. Fukuyama (2014). 12. Olson (1971). 13. Among others, Atkinson (2008, 2015); Atkinson and Piketty (2007, 2010); Atkinson, Piketty, and Saez (2011); Banerjee and Duflo (2019); Milanović (2016); Piketty (2014); Piketty and Saez (2003), etc. 14. Piketty (2014). 15. Such as Milanović (2016), Scheidel (2017), or Piketty (2020). 16. Shiller (2013). 17. Piketty (2014) does the most accurate description of this transition from capital-​ based to income-​based inequality, stating that the 20th century marked a switch from a “society of rentiers” to a “society of supermanagers.” 18. Locke (2003[1690]), p. 46. 19. The argument emphasizing the crucial role of free trade in economic development has been eloquently elaborated since Adam Smith (1776) and David Ricardo (1817). 20. Obviously, there are examples of societies achieving some form of economic development without free trade, but all such examples are short-​lived and their relative progress was entirely unsustainable (like the example of the Soviet Union during Stalin). Read more about this in Acemoglu and Robinson’s (2012) Why Nations Fail. 21. McCloskey (2012).

324 Notes 22. Mokyr (2016) and Ferguson (2017). 23. Chapters 5 and 7 develop this in greater depth by defining between-​firm differences that depend on type of industry and within-​industry competition. There are rent-​ seeking firms, which build their position on the political market, and there are customer-​seeking firms, which build their position on the market for customers. Rent-​seeking firms are the ones seeking a nonmarket way—​political protection—​ of capturing a stronger market position. Their relationship with politics is the main focus of this book. 24. North, Wallis, and Weingast (2010). 25. Acemoglu and Robinson (2019). 26. Ibid., Chapter 2. 27. Rajan (2019). 28. See Diamond (1997) or Cameron and Neal (2003). 29. Olson (2000). 30. Scheidel (2017). 31. Acemoglu and Robinson (2012). 32. Clark (2011). 33. Milanović, Lindert, and Williamson (2011). 34. According to Clark’s (2014) research on social mobility based on rare elite surnames. 35. Piketty (2014). 36. Clark (2014). 37. Atkinson, Piketty, and Saez (2011). 38. Bakija, Cole, and Heim (2012). 39. See Hacker and Pierson (2011) and Bonica et al. (2013). 40. Kuznets (1955). 41. Milanović (2016), Chapter 2. 42. Scheidel (2017), pp. 5–​9. 43. Milanović (2016) also develops the argument that seemingly exogenous factors like wars and revolutions are actually driven by internal structural factors that are triggered by unsustainably high levels of inequality. Inequality does not go down itself but it sets in motion the events that lower it. This, however, is difficult to prove and can fall into the standard trap of reverse causality. 44. See more in Piketty and Saez (2003), Milanović (2016, Chapter 2), and Scheidel (2017, Chapters 12 & 13). 45. Clark (2014). 46. Olson (1971), Chapters 1 & 2. 47. Olson (1982), Chapter 3.

Chapter 1 1. As defined by Tullock (1967), Stigler (1971), and Krueger (1974). 2. Dabla-​Norris and Wade (2001).

Notes  325

3. Stiglitz (2012). 4. Furman and Orszag (2015). 5. Song et al. (2019) and Autor et al. (2020). 6. Lindsey and Teles (2017). 7. Scheidel (2017). 8. Piketty (2014). 9. Similar to Schoenman’s (2014) description of network creation in postcommunist societies. 10. Becker (1968). 11. Empirically shown by several studies including: Rundquist, Storm, and Peters (1977); Kurer (2001); and Winters and Weitz-​Shapiro (2013). 12. Transparency International (2021). 13. Weber (1978 [1922]). 14. Pareto (1935). 15. Mosca (1939). 16. Michels (1962 [1915]). 17. For example, Mills (1956), Endruweit (1984), Dye (2001), and Gilens (2012). 18. I touch on the democratic trial-​and-​error mechanism as a way to curtail elite network power, as shown in the final part of the book, in Chapters 8 and 10. 19. Acemoglu and Robinson (2006). 20. Smith (1904 [1776]). 21. Ansell and Samuels (2014). 22. Ferguson (2017) makes a compelling argument to that end, detailing the spontaneous nature of forming elite networks throughout history, while dismissing the conspiracy accusations even within organizations like the Illuminati (Chapters 1 & 10) or the Freemasons (Chapter 20), both of which certainly made contributions in their time, but their roles since have been vastly exaggerated. 23. In network theory these are called high-​degree nodes, where the degree of a node determines the total amount of other nodes connected to it (Jackson, 2010). Therefore, the scope of power for one node is determined by the sum of all other nodes it is connected to. 24. Borondo et al. (2014). 25. Barbasi (2003) describes superhubs as highly connected leaders whose connections exert influence over most other hubs and nodes. Navidi (2017) applies the same term to financial networks in her book Superhubs. 26. The source of the data is a private company BoardEx which collects this type of data. Access to the database has been granted via the Wharton Research Data Service (WRDS). See Chapter 4 for more details. 27. As, for example, used on LinkedIn. I ignore second-​and third-​level connections to superhubs, given that these networks would be too large and would all look the same—​as one huge circle. 28. Such as preventing the Commodity Futures Trading Commission (CFTC) from gaining oversight on over-​ the-​ counter (OTC) derivatives. Alongside Fed Chairman Alan Greenspan, Rubin publicly shot down the proposal of the CFTC

326 Notes then-​chairperson Brooksley Born. Despite sounding early warnings over the danger of derivatives, Born was forced to resign in 1999 after Congress accepted Rubin and Greenspan’s proposals over hers. See Goodman (2008). 29. Nassim Taleb (2012, 2018) used this episode to coin the phrase “Bob Rubin trade,” describing a decade of hidden risks being transferred from banks to the taxpayers, at the end of which bankers like Rubin got away with huge payoffs, while the losses had to be carried by the taxpayers. It was an asymmetric trade at its worst: the downside is protected by public money, while a select few get to keep the entire upside. “Heads he wins, tails he shouts ‘Black Swan’ ” (Taleb, 2018, p. 13). 30. Cohan (2012). 31. Becker and Morgenson (2009) and Stewart (2009). The nine banks were: Bank of America, JP Morgan Chase, Citigroup, Wells Fargo, Goldman Sachs, Morgan Stanley, Bank of NY Mellon, State Street, and US Bancorp. 32. Their second-​order ties were deliberately cut off to avoid congestion in the graph. 33. Meaning that these individuals are connected to each other despite also being connected to Paulson through two different channels. 34. Acemoglu et al. (2016). 35. Moyer (2016). 36. Frank (2015). 37. Together with Vanguard and State Street, they (the Big Three) constitute the largest shareholder in 88% of all S&P500 companies. See Fichtner, Heemskerk, and Garcia-​ Bernardo (2017). 38. See, for example, Mills (1956), Guttsman (1965), Useem (1984), Ruostetsaari (2006), Maclean et al. (2010), and Griffiths et al. (2014).

Chapter 2

1. Lenski (1966). 2. Milanović, Lindert, and Williamson (2011). 3. Trigger (2003), p. 142. 4. Scheidel (2017) refers to it as the “Great Disequalization,” while Harari (2011) calls it the “biggest fraud in history” because it turned the Homo sapiens from free, egalitarian foragers into servants of agricultural cycles, and introduced hierarchies, and by extension disparities, in the distribution of wealth. 5. Marean (2014). 6. Brown et al. (2012). 7. Marean (2015) and Harari (2011). 8. Diamond (1997). 9. Harari (2011), p. 94. 10. Kohler et al. (2017). 11. Diamond (1997) diligently describes the geographical conditions that enabled this early development.

Notes  327 12. Bogaard, Fochesato, and Bowles (2019). 13. Ibid. 14. Kohler et al. (2017). 15. Ibid. 16. Olson (2000), pp. 6–​12. 17. It wasn’t the only social order, as Stasavage (2020) shows with a multitude of examples of early democracies all over the world in ancient times, but these were limited to small-​scale environments. When the first empires arose, the violence power principle took over and established a dominant hierarchical order wherever it penetrated. 18. Carneiro (1970) 19. Harari (2011), Chapter 6. 20. Ibid., p. 124. 21. Ibid., Chapter 8. 22. See for example: Abul-​Magd (2002) or Stein (1994). 23. Scheidel (2017), p. 43. 24. Gregory Clark in his 2011 book A Farewell to Alms, Chapter 2, delivers a very accurate economic portrayal of the Malthusian model that kept societies virtually constrained since the Agricultural Revolution until the 1800s. 25. Scheidel (2017), p. 63–​70. 26. Clark (2014), Chapter 9. 27. Scheidel (2017), pp. 71–​79. 28. Milanović (2016), p. 68. 29. Frankopan (2015). 30. This is perfectly encapsulated in Machiavelli’s (1999[1532]) The Prince, a 16th-​ century political treatise, paints a powerful image of the political structures that existed during his time, in particular showing how the papacy and the Church were very often at the very center of the political dominance hierarchy. 31. Clark (2014). 32. Frankopan (2015), Chapters 6 & 7. 33. Piketty (2014), Chapter 10. 34. Acemoglu and Robinson (2006). 35. Ansell and Samuels (2014). 36. Wallerstein’s core, see Chapter 8. 37. Ferguson (2017). 38. It should be noted that even though the “robber barons” were clearly motivated by the violence power principle while amassing their huge fortunes, later in life they’ve embraced the gift exchange principle. Their philanthropic efforts, particularly in the United States, enabled building thousands of schools, hospitals, colleges, museums, libraries, charities, cultural institutions, and the like. 39. Scheidel (2017). 40. Note also how the share of wealth of the top 1% and 10% was lower in the United States than it was in European countries in the 19th century. This is hardly surprising given that Europe (the Old World) featured highly entrenched historical elites while the United States was a New World country where elite networks were yet to

328 Notes develop to their full extent. However, by the end of the 20th century their roles were reversed: European elites held a lesser share of total wealth than did American elites. 41. Piketty (2014), pp. 276–​278. 42. Ibid., pp. 407–​408. 43. Bakija, Cole, and Heim (2012). 44. Atkinson, Piketty, and Saez (2011).

Chapter 3

1. See Blanchard (1997) and Fidrmuc (2003). 2. See Milanović (1998) and Tucker (2006). 3. Miller, Grodeland, and Koshechkina (2001). 4. Przeworski (1991) and Tucker (2006). 5. Atkinson, Piketty, and Saez (2011). 6. This was most obviously portrayed in the entertainment industry: African American TV shows started appearing on network television already in the 1970s but were gaining particular popularity in the 1980s and 1990s (Soul Train started in 1971, Fat Albert a year later, followed by classics such as Good Times, The Jeffersons, The Cosby Show, Family Matters, Fresh Prince of Bel-​Air, and of course the Oprah Winfrey Show starting in 1986). African American actors started gaining prominence in Hollywood; the stereotypical black character in the 1950s was to some extent related to crime (or slavery), while the stereotypical black character in the 1980s was the chief of police—​a huge difference in perception. Popular songs—​arguably one of the biggest embodiments was Donna Summer’s 1983 hit “She Works Hard for the Money”—​ushered a new era in women’s rights as women were finally being perceived as earners and potentially corporate executives—​a significant difference over the 1950s and 1960s perception of women as stay-​at-​home moms. 7. Meltzer and Richard (1981). 8. Downs (1957). 9. Sirowy and Inkeles (1990). 10. Muller (1988), Simpson (1990), Burkhart (1997), and Huber et al. (2006). 11. Rodrik (1999) and Lee (2005). 12. Scheve and Stasavage (2009) and Acemoglu et al. (2019). 13. Kuznets (1955). 14. Olson (1982). 15. The data was taken from the IMF database based on a working paper by Mauro et al. (2013). 16. Tanzi and Schuknecht (2000), pp. 9–​13. 17. Pigou (1932 [1920]), Samuelson (1983 [1947], 1954), and Arrow (1953). 18. Musgrave (1959). 19. Tanzi and Schuknecht (2000), pp. 32–​45. 20. Friedman (1962) and Buchanan and Tullock (1962).

Notes  329 21. Tanzi and Schuknecht (2000), Chapter II. 22. See Mauro et al. (2013) for the most accurate depiction of historical debt trends. There have been many hypotheses as to why debt levels have risen so substantively over the past decades. Alesina and Perotti (1994) survey the literature and present six hypotheses to account for the rise in public debt, all of which are politically motivated. These include the classical political business cycle model, the intergenerational distribution argument, the role of political institutions a country has, its distributional conflicts, its geographically dispersed interests, and even that parties may use public debt as a strategic variable. Persson and Tabellini (2000) add to this political instability and the common-​pool problem (building on Olson, 1982) in the battle of groups over redistribution, where powerful groups internalize benefits from public goods, while the costs are imposed on taxpayers via either higher taxes or overborrowing. 23. Tanzi and Schuknecht (2000), p. 46, and Mauro et al. (2013). 24. Tanzi and Schuknecht (2000), pp. 26–​27. 25. See Deaton’s (2013) The Great Escape for an overview of these trends. 26. Atkinson (2008), p. 5. 27. See Atkinson (2008), pp. 23–​27, for a more detailed explanation. 28. Atkinson (2008), pp. 33–​37. 29. Okun (1975), p. 66. 30. To borrow the terms from Acemoglu, Robinson, and Verdier (2017). 31. Atkinson et al. (2017). 32. Data on top income earner distribution is taken from Piketty and Saez (2003), and Atkinson, Piketty, and Saez (2011), while the indicators of wealth inequality and poverty are taken from Atkinson and Piketty (2007, 2010). 33. Keep in mind that the periods of lowest reported inequality in the United States, the 1950s and 1960s, allowed huge parts of the population to remain disenfranchised from the benefits of economic and productivity growth. 34. Atkinson, Piketty, and Saez (2011). 35. See, e.g., Atkinson (2008) or Autor (2014). 36. Case and Deaton (2020). Another factor that is often overlooked in the story of stagnating real wages, at least in the United States, is the rising share of noncash benefits in the incomes of lower-​and middle-​income workers. Mainly, health insurance premiums and pensions savings. Such benefits will not be reported as taxable income in the official statistics; however, they should show up in gross wages data. Therefore, the data on total income inequality could be slightly biased upward, but the data on earnings dispersion should remain robust to this. 37. Atkinson (2015). 38. Deaton (2013). 39. Gabaix and Landier (2008). 40. Furman and Orszag (2015). 41. See Kahneman (2011), Chapter 20. A similar study was done by De Bondt and Thaler (1985). 42. Bertrand and Mullainathan (2001).

330 Notes 43. Piketty, Saez, and Stantcheva (2014). 44. Banerjee and Duflo (2019). 45. Shleifer (2012). 46. Fukuyama (2014). 47. Song et al. (2019). 48. Autor et al. (2020). 49. Philippon and Reshef (2013). 50. Ibid. 51. Bell and Van Reenen (2013). 52. Godechot (2012). 53. Greenwood and Scharfstein (2013). 54. Claessens and Perotti (2007). 55. See Albertus and Menaldo (2018) for more on how the old regime elites capture their newly created democracies. 56. Stiglitz (2012), p. 32. 57. Furman and Orszag (2015). Their empirical finding is further supported by Song et al. (2019) and Autor et al. (2020). 58. Lindsey and Teles (2017). 59. Svejnar and Bagchi (2015). 60. A similar finding was confirmed by Gupta et al. (2002), Glaeser et al. (2003), and You and Khagram (2005). 61. Gehlbach (2006); Hochberg et al. (2009); Mian, Sufi, and Trebbi (2010); and de Figueiredo and Richter (2014). 62. Igan, Mishra, and Tressel (2012) and Blau, Brough, and Thomas (2013). 63. Vukovic (2021); Duchin and Sosyura (2012); Blau, Brough, and Thomas (2013); and Calomiris and Khan (2015). 64. Evans (1996), Kelleher and Yackee (2009), and de Figueiredo and Silverman (2006, 2007) for the United States; Alt et al. (1999) and Helland (2008) for Norway; or even to avoid paying taxes (Richter et al. 2009 for the United States, and Schone et al. 2013 for France). 65. Deaton (2013), pp. 180–​181. 66. For comparison, the child poverty rate in the United States is 21%, the African American poverty rate is 26%, for Hispanics 24%, for Asians 12%, and for whites 10% (all data for 2014). The overall poverty rate for 2015 in the United States was 13.5%. Source: US Census Bureau (2016). 67. Fortune Magazine (2005), “Fortune’s ‘Power 25.’ The 25 Most Effective Interest Groups.” 68. Center for Responsive Politics (2016). 69. Arguably in countries with inclusive institutions and a well-​functioning constitutional order, when such practices get uncovered, the actors get punished. 70. Bonica et al. (2013). 71. Bermeo (2009). 72. Hacker and Pierson (2011). 73. McChesney (1997).

Notes  331 74. Bartels (2016). 75. Gilens (2012). 76. Moss (2008). 77. Persson and Tabellini (2009).

Chapter 4 1. Wharton Research Data Service (WRDS) (2017). The BoardEx dataset has been used in academic research thus far, mostly focusing on network effects, mergers and acquisitions, executive compensation, and corporate performance of CEOs (e.g., El-​ Khatib, Fogel, and Jandik 2015; De Cesari and Ozkan 2015; Kim, Kogut, and Yang 2015; Schmidt 2015; etc.) 2. The UK dataset included more private, unlisted firms than the US case. 3. Relevance is determined either by geographical proximity (e.g., State representatives or city mayors) or by key positions in government or the legislature (ministers, members of key parliamentary committees, etc.). 4. BoardEx (2017) Data Dictionary. 5. The POLCON variable is disentangled by its two main components, previous government experience and same organization membership. The conclusions are identical: no matter how we define political connections, those who are connected in each case have a much larger average network size. 6. The relative size of each individual’s network declines exponentially. In the United States more than half have less than 1,000 connections, and 90% have less than 4,000 connections. There are only a few extremes on the upper tail, but these do not affect the overall findings. 7. Topocratic networks are defined by Borondo et al. (2014). 8. See Atkinson et al. (2017). 9. The trends for absolute values of earnings and salaries for both countries point to the same suggestion as their logarithmic versions in Figures 4.7 and 4.8. 10. Also with microlevel data there is always a potential problem with measurement error. If there is measurement error in the data the additive random errors of the main explanatory variables could bias the estimated coefficient toward zero. 11. Angrist and Krueger (2001). 12. Fisman and Svensson (2007) and Desai and Olofsgard (2011). 13. Calculated as 100 * exp (0.119 ) − 1. 14. This shouldn’t be surprising given the way I define the instrument with an emphasis on reducing the measurement error. Angrist and Krueger (2001) suggest that in cases where there the explanatory variable is likely to be biased due to measurement error, the estimated coefficients of standard OLS regressions would be biased toward zero. In this case this means that the non-​IV estimates are likely to be lower than when correcting for potential measurement error using industry-​level averages of the explanatory variable.

332 Notes 15. For example, Fisman (2001); Faccio (2006); Faccio, Masulis, and McConnell (2006); Bellettini, Ceroni, and Parolo (2013); etc. Chapter 7 presents these findings in greater depth. 16. Eggers and Hainmueller (2009). 17. Ding et al. (2015). 18. Wu et al. (2018). 19. Alonso and Simon (2023). 20. Kong et al. (2023). 21. Meyer et al. (2018). 22. Alregab (2015).

Chapter 5 1. He was hardly the only one. Bloomberg’s reporters listed 64 former congressmen or senators in 2015 sitting on boards of companies they used to regulate. See Green, Kochkodin, and Migliozzi (2016). 2. The tendency of highly connected individuals (high-​degree nodes) to be connected with other highly connected individuals. Also known as network homophily; attraction of similar individuals. See Jackson (2010). 3. According to Borondo et al. (2014). 4. For example, a new law being passed, a regulatory opportunity, having privileged access to government tenders, etc. 5. This is known as Dunbar’s (1992) number: the cognitive limit to the number of people one can maintain successful social relationships with. 6. Jackson (2010). 7. Ibid. 8. The distribution is a pure Poisson, where the majority of individuals have low levels of connections (mostly family and a few friends), while a very small number of people are highly connected. 9. Navidi (2017). 10. Ferguson (2000). 11. Ferguson (2017), Chapters 24 & 25. 12. The project features maps of various historical connections, from the networks of academic economists and researchers, to social networks of citizens in 18th-​century France, to the privilege networks of powerful families or monarchs. The maps and the data are available here: https://​histe​con.fas.harv​ard.edu/​visu​aliz​ing/​index.html. 13. The project was led by Montserrat Cachero Vinuesa, and is available here: https://​ histe​con.fas.harv​ard.edu/​visu​aliz​ing/​pri​vile​ges/​index.html. The full research is available in Cachero and Rodriguez-​Modrono (2022). 14. Charles V was a Habsburg born in Burgundy, and the Welsers rose to prominence as the bankers of the Habsburgs. Hence the connection to the monarch. 15. Rothschild (2012).

Notes  333 16. Fisman (2001), Faccio (2006), and Desai and Olofsgard (2011). 17. Desai and Olofsgard (2011). 18. The empirical findings in Tanzi and Davoodi (1998), Faccio et al. (2006), Bellettini et al. (2013), Garcia-​Santana et al. (2020), or Gamberoni et al. (2016) confirm these insights for a number of different countries. 19. Tullock (1967) and Krueger (1974). 20. Bandiera, Prat, and Valletti (2009); Fazekas, Toth, and King (2016); Fazekas and Kocsis (2017). 21. Song et al. (2019). 22. This finding is also supported by Card et al. (2013) for Germany, Hakanson et al. (2021) and Akerman et al. (2013) for Sweden, Mueller et al. (2017) for the United Kingdom, and Alvarez et al. (2018) for Brazil. Other papers that confirm the between-​ firm effect in the United States are Davis and Haltiwagner (1991), Barth et al. (2016), and Furman and Orszag (2015). 23. This would be similar to the conclusions of Autor et al. (2020) on the effect of “superstar firms” as the drivers of compensations at the top. 24. Note that the corresponding benefits do not drop. Convicted white-​collar criminals tend to keep most of their material possessions. 25. Note that point M represents an unstable equilibrium as agents have an incentive to move to the right. They realize that one cannot stay below point M for too long and they expect total benefits to start exceeding total costs after initial time and effort have been spent. 26. I am referring here to the motivation of gift-​ giving, not its administration. Inheritance is usually regulated by a legal contract that is necessary to keep all parties satisfied. 27. Many inheritors of family business fortunes often squander their inherited wealth. If they lack the competitive edge of their parents or grandparents, an equalization force should level the playing field. However, as shown by Clark (2014) in The Son Also Rises, family fortunes can persist for centuries. The problem here is that most such century-​long fortunes were hardly a product of voluntary exchange and were much more likely a product of coercion, violence, or proximity to political power. 28. This change in rhetoric is most diligently portrayed in McCloskey’s (2012) Bourgeois Virtues, while the impact of the change in culture is best explained by Mokyr (2016).

Chapter 6 1. A similar and often interchangeably used term in this context is patronage, where the rewards from being part of a politician’s close network of supporters usually include being given jobs in the public sector, exclusive government contracts, or other forms of protectionism. The literature recognizes patronage as a particular form of clientelism (see Piattoni 2001, Root and Nellis 2000, or Robinson and Verdier 2013). 2. Stokes et al. (2013).

334 Notes 3. For a definition, see Shleifer and Vishny (1993), Mauro (1995), Knack and Keefer (1995), and Rose-​Ackerman (1978). 4. Gardiner (2002). 5. Shleifer and Vishny (1998). 6. This cynical conclusion is the foundation of the political agency literature; for examples, see Brennan and Buchanan (1980), Ferejohn (1986), and Besley (2006). 7. There is vast literature covering these issues: Coate and Morris (1995), McChesney (1997), Grossman and Helpman (2002), Gilens (2012), and Gilens and Page (2014). 8. North, Wallis, and Weingeist (2010) call such societies limited access orders. 9. Buchanan and Tullock (1962) and Brennan and Buchanan (1980). 10. Bueno De Mesquita et al. (2005). 11. Besley (2004) and Casselli and Morrelli (2004) both follow that assumption. 12. A good overview of the logic and theory is given in Besley (2006). 13. The standard term limit assumption in Besley and Case (1995a, 1995b); Alt, Bueno De Mesquita, and Rose (2011); and Ferraz and Finan (2011). 14. Austen-​Smith and Banks (1989), Banks and Sundaram (1993), Besley and Case (1995a), Rogoff (1990), Persson and Tabellini (2000), and Besley and Smart (2007). 15. For the United States, see Rundquist, Storm, and Peters (1977), who use a survey experiment to confirm that voters trade-​off corruption for other values in a candidate, while Peters and Welch (1980) find that US Congressmen are less likely to be punished after being charged for corruption. Almost identical findings were confirmed in Italy (Chang and Golden 2004; Golden 2004), Greece (Dobratz and Whitfield 1992; Patrikios and Karyotis 2008), and Japan (Reed 1999; Nyblade and Reed 2008). 16. See Ferraz and Finan (2011) and de Figueiredo, Hidalgo, and Kasahara (2012) for Brazil, Dimock and Jacobson (1995) for the effect of the House banking scandal on US voters in 1992, or a repeated study of US Congressmen from 1982 to 1990 by Welch and Hibbing (1997). The literature examining this relationship in developing and emerging countries found evidence both in support of the idea that corruption does not get punished at the polls (Manzetti and Wilson 2007; Chang and Kerr 2017), and against it (Klasnja 2015; Klasnja, Tucker, and Deegan-​Krause 2016). 17. See Rundquist, Storm, and Peters (1977), Kurer (2001), and Winters and Weitz-​ Shapiro (2013) for a detailed overview of the hypotheses. 18. Downs (1957). 19. Ferraz and Finan (2011) famously show that this mechanism works in Brazil—​an independent audit on political corruption closed the informational asymmetry and the voters punished the corrupt politicians. 20. Podobnik, Vuković, and Stanley (2015). 21. Verdier (1995) and Root and Nellis (2000). 22. Shleifer and Vishny (1998) and Besley (2006). 23. McChesney (1997) and Robinson and Verdier (2013). 24. Bueno De Mesquita et al. (2005). 25. I show this on the example of corruption in Croatian local government in Vuković (2019). 26. Nyblade and Reed (2008). 27. Vuković (2020).

Notes  335 28. Although no research has directly linked corruption to re-​election, procurement-​ based corruption was found to be high in a number of Eastern European countries, see Fazekas and Kocsis (2017). 29. Vuković (2019).

Chapter 7 1. Phelps (2013). 2. Acemoglu and Johnson (2023). 3. Typically, the political economy literature separates firms into rent-​seekers and profit-​seekers, however, I deliberately use the term “customer-​seeker” given that such firms achieve their profits by satisfying and pandering to customers, not politicians. 4. Monopolies cannot exist without support and encouragement from the government. The aforementioned examples of historical monopolies were clear cases of a government granting monopoly rent-​extraction rights in colonies to connected executives who in turn were allowed to amass vast private fortunes. They were free to set high prices and achieve huge profits primarily due to their realization of a nonmarket position bestowed to them by political power. 5. Facebook hired former UK deputy prime minister Nick Clegg to be their head of global affairs, Apple named former US vice president Al Gore to their board of directors, while Google, Amazon, and Microsoft all hired senior US government officials over the past decade. 6. The top 5 Big Tech companies (Amazon, Apple, Facebook, Google, Microsoft) spent more than $60m on lobbying in both 2019 and 2020. Source: https://​www.cnbc.com/​ 2021/​01/​22/​faceb​ook-​spent-​more-​on-​lobby​ing-​than-​any-​other-​big-​tech-​comp​any-​ in-​2020.html. 7. Stigler (1971). 8. Peltzman (1976). 9. Becker (1983). 10. Tullock (1967) and Krueger (1974). 11. McChesney (1997). 12. Beck, Hoskins, and Connolly (1992). 13. Shleifer and Vishny (1994, 1998). 14. Olson (1982). 15. Bellettini, Ceroni, and Prarolo (2013). 16. Fisman (2001). 17. Faccio (2006). 18. Goldman, Rocholl, and So (2009). 19. Eggers and Hainmueller (2009). 20. Jayachandran (2006) and Blanes i Vidal, Draca, and Fons-​Rosen (2012). 21. See Faccio et al. (2006), Sapienza (2004), Dinc (2005), and Khwaja and Mian (2005). 22. Fan et al. (2007).

336 Notes 23. Gehlbach, Sonin, and Zhuravskaya (2010). 24. Earle and Gehlbach (2015). 25. Schoenman (2014). 26. Bertrand, Bombardini, and Trebbi (2014) and Blanes i Vidal, Draca, and Fons-​Rosen (2012). 27. Correia (2014). 28. Vitali, Glattfelder, and Battiston (2011). 29. Their networking power, exhibited through high rates of network homophily and clustering, is well documented, on a case-​by-​case basis, in Navidi (2017). 30. Kwak (2014). 31. Becker and Morgenson (2009) and Stewart (2009). 32. Walsh (2009). 33. Each of them wrote books about this, and they even wrote a book together to explain and justify their decisions: Bernanke, Geithner, and Paulson (2019). 34. Vuković (2021). 35. Acemoglu et al. (2016). 36. Querubin and Snyder (2013). 37. Bradley, Gebrekidan, and McCann (2020). 38. Vuković (2021). 39. The Troubled Asset Relief Program (TARP) was drafted already in October and provided the immediate necessary liquidity to the industry. It directly injected a total of $621 billion, with the banks getting $372 billion (the rest went to the government mortgage giants, Fannie Mae and Freddie Mac, the automobile industry, and AIG). At the same time, the lobbying and campaign spending of the finance industry was on historical highs during 2008 and 2009. Firms that got bailouts spent $114m in lobbying and campaign spending in 2008, and an additional $80m on lobbying in 2009. 40. Abnormal lobbying is defined as the difference between actual lobbying that happened in 2008 and 2009, and the counterfactual estimate of lobbying—​its precrisis moving average trend that suggests how much money would have been spent on lobbying had it not been for the crisis. The point of using abnormal lobbying is to eliminate the impact of precrisis aggregate lobbying and focus only on what transpired in the immediate aftermath of the crisis. 41. Any established correlation of political connections with the allocation of bailouts still suffers from omitted variable bias, does not address the self-​selection problem of connected firms, and can have potential reverse causality. For example, greater exposure to risk of bankruptcy could have drawn both greater lobbying and donations, as well as the bailout allocation. 42. Vuković (2021).

Part III 1. To use the notation from Acemoglu and Robinson (2019), these are countries of a “Shackled Leviathan” who are well within the “Narrow Corridor,” with the “Red Queen Effect” fully in place.

Notes  337

Chapter 8

1. Roser, Ortiz-​Ospina, and Ritchie (2019). 2. Roser, Appel, and Ritchie (2019). 3. See Milanović (1991) and Bergson (1984). 4. One can hardly define China’s economic system as socialism (in fact, it is a pure example of political capitalism, as defined in Milanović’s [2019] Capitalism, Alone), but with a socialist government China may care more about a fair redistribution of income and wealth that a typical capitalist country. Vietnam is a similar case of political capitalism. 5. There is no better evidence of this than in the popular literature. The works of Charles Dickens, Jane Austen, Honoré de Balzac, or Émile Zola paint powerful images of terrible living standards of workers in 19th-​century England and France. This was the violence power principle at its worst. 6. Harari (2010) and Rajan (2019) describe this process in greater depth. 7. Ferguson’s (2008) Ascent of Money makes the most convincing argument of this case. 8. Acemoglu and Robinson (2019), Chapter 5. 9. Ferguson (2008), Chapter 1. 10. Acemoglu et al. (2011) show that areas which accepted these changes after the French retracted, were set on a course of positive growth and development over the next century. In areas that repealed the French reforms and reintroduced the old regime, institutions resisted positive change and had worse economic outcomes. 11. Not exclusively, obviously, as socialist countries also delivered major reforms that produced rapid industrialization (socialism was accepted primarily in less developed countries), and a form of the welfare state, to a varying level of success. The last part of the chapter covers their development. 12. The continental exceptions include the Russian invasion of Ukraine in 2014 and particularly in 2022 (an event that prompted Ukraine to apply for both NATO and EU membership in the subsequent months), as well as the failure of the European Community in 1991 to prevent the rise and aggression of Serbian nationalism on Croatia and Bosnia and Herzegovina, leading to a devastating five-​year war in the Balkans (Glaurdić, 2011), and again in 1998 in Kosovo. 13. The core and the periphery are concepts developed by the economic historian Immanuel Wallerstein in his 1974 classic The Modern World-​System I: Capitalist Agriculture and the Origins of the European World-​Economy in the Sixteenth Century. He presents an initial version of the world-​systems theory that separates countries in the early stages of capitalism into the core (British Empire, France, Germany/​Prussia, Netherlands, Belgium, Austria, Spain, Portugal, Italy), the semiperiphery (China, India, Argentina, Brazil, Mexico, Iran), and the periphery (everyone else, most notably Russia, Eastern Europe, the rest of Latin America and Asia, and Africa). 14. Gerschenkron (1962). 15. There were exceptions however, but they preceded the most rapid era of industrial development. Emperor Joseph II of the Habsburg monarchy and the Holy Roman Empire and Empress Catherine the Great of the Russian Empire were both lauded as Enlightenment monarchs of the 18th century, initiating a series of reforms. However,

338 Notes their successors had a much lower impetus for change and actively prevented the progress of the Industrial Revolution. See Acemoglu and Robinson (2012), Chapter 8. 16. Marx and Engels (2004 [1848]). 17. Milanović (2019). 18. Albertus and Menaldo (2018). 19. Triggering the “Red Queen Effect,” Acemoglu and Robinson (2019), Chapter 2.

Chapter 9 1. An excellent overview of all the arguments in favor of such proposals is given in Banerjee and Duflo (2019), Chapter 7. 2. Piketty, Saez, and Stantcheva (2014). 3. Saez, Slemrod, and Giertz (2012). 4. Piketty (2014, 2020). 5. A similar case is made for a decline of wealth inequality in the 20th century, given that progressive rates were also applied to the inheritance tax. However, high top marginal inheritance rates were only introduced in the United States and the United Kingdom (up to 70% and 80%), even though wealth inequality went down across the developed world, mostly as a result of the structural shifts in property ownership, as discussed in Chapter 2. 6. Germany and Italy made the majority of their nationalizations before the war, also contributing to declining inequality. Italy during Mussolini’s rule at one point had three-​quarters of its economy nationalized. In Germany nationalizations continued after the war, particularly in the socialist East Germany, where the majority of firms were placed under public ownership. West Germany took a slightly different approach. 7. The most famous case-​in-​point was the nationalization of the car company Renault in France, whose owner Louis Renault was accused of alleged collaborationism with the Nazis (which was later disproved), arrested, and had all of his factories nationalized. It is easy to see how, in that particular case, wealth inequality would go down. 8. Alstadsæter, Johannesen, and Zucman (2019). 9. Ibid, pp. 2098–​2099. 10. He became the frontrunner in the race against incumbent President Sarkozy in May 2011, after the current frontrunner Dominique Strauss Khan was arrested on suspicion of sexual assault in New York. 11. “France Forced to Drop 75% Supertax after Meagre Returns,” The Guardian, December 2014, https://​www.theg​uard​ian.com/​world/​2014/​dec/​31/​fra​nce-​drops-​ 75perc​ent-​super​tax. 12. “Indigestion for ‘Les Riches’ in a Plan for Higher Taxes,” New York Times, August 2012, https://​www.nyti​mes.com/​2012/​08/​08/​busin​ess/​glo​bal/​fran​ces-​les-​ric​hes-​ vow-​to-​leave-​if-​75-​tax-​rate-​is-​pas​sed.html. 13. “After Depardieu, Belgian Plan by France’s Richest Man Fuels 75% Tax Row,” The Guardian, January 2013. https://​www.theg​uard​ian.com/​busin​ess/​2013/​jan/​24/​belg​ ian-​move-​bern​ard-​arna​ult-​fra​nce-​tax.

Notes  339 14. “Wealthy French Take Their Assets to London,” Der Spiegel, May 2012, https://​www. spie​gel.de/​intern​atio​nal/​eur​ope/​weal​thy-​fre​nch-​flee-​to-​lon​don-​amid-​fear-​of-​holla​ nde-​a-​833​814.html. 15. “France in 14 Billion Euro Tax Black Hole,” BBC News, May 2014, https://​www.bbc. com/​news/​busin​ess-​27602​312. 16. Poterba (1987), Clotfeler (1983), and Crane and Nourzad (1990). 17. Fisman and Wei (2004). 18. Pommerehne and Weck-​Hannemann (1996) and Chiarini, Marzano, and Schneider (2011). 19. Gilens (2012) shows that in American politics affluent voters carry a much higher weight in policy decisions, and that low-​and middle-​income voters never get the policies they want when their preferences diverge from those of the affluent voters. 20. See, e.g., Schouten and Silver’s (2012) book Almost a Psychopath or Eddy’s (2019) Why We Elect Narcissists and Sociopaths—​And How We Can Stop. 21. Acemoglu and Robinson (2019). 22. North, Wallis, and Weingeist (2010). 23. Ortiz-​Ospina and Roser (2016), “Trust” (Public trust in government, United States, 1958 to 2015).

Chapter 10 1. Administered via an official government website, USAspending.gov, https://​www. usas​pend​ing.gov/​#/​sea​rch. 2. The EC Financial Transparency System: https://​ec.eur​opa.eu/​bud​get/​fts/​index​_​ en.htm. 3. Besley and Prat (2006). 4. Petrova (2008). 5. DellaVigna and Kaplan (2007). 6. The Dubrovnik Republic arguably took it furthest. Even though it was an aristocratic republic, ruled by the noble classes, it recognized a division of power (two houses of its parliament, an executive branch, and even a supervisory body that ensured protection of the law), and had term limits for its head of state—​the rector—​who was subject to a monthly change to prevent any possibility of usurpation of power. 7. From Bueno De Mesquita et al. (2005). 8. A brief overview can be found here: https://​www.europ​arl.eur​opa.eu/​workin​gpap​ ers/​juri/​101/​def​ault​_​en.htm. 9. According to Arrow’s (1951) impossibility theorem. 10. The literature on this field of behavioral economics is vast and expanding. I recommend a few titles: Kahneman’s (2011) Thinking, Fast and Slow, Ariely’s (2008) Predictably Irrational, Akerlof and Shiller’s (2015) Phishing for Phools, and Thaler’s (2016) Misbehaving. 11. The entire parliament can be held accountable in presidential systems where parliaments can be the cause of gridlock and where bipartisan solutions are necessary

340 Notes to resolve important issues. In parliamentary systems, this provision would only apply to governing coalitions (given that the opposition often has no real power), unless the inaction is triggered by the opposition itself (if, for example, one governing coalition party switches ranks and tries to exploit the rule). 12. A Condorcet winner is a preference aggregation mechanism that is considered the fairest, as it implies choosing an option that beats all others in a pairwise vote. The Borda count voting mechanism is much more likely to deliver a Condorcet winner than, for example, majority voting, especially as the number of alternatives increases. See more in Black (1958). 13. Many developed countries use electronic and online voting for political office, so problems of privacy, hacking, and anonymity are easily solvable. 14. If the legal system itself is inefficient or corrupt, the solution is not to take away its powers, but to subject them to greater transparency and accountability, and by all means, remove any political influence over it. 15. Nash (1950). 16. Chetty, Hendren, and Katz (2016). 17. Ostrom (1990). 18. Think of the 1946 classic film It’s a Wonderful Life, starring James Stewart as George Bailey, and how the entire community comes to rescue him in his time of need, after he spent his entire life helping the community. This is fiction, but it paints a clear picture of how community life was perceived in the United States. 19. This is a reference to Shleifer’s book The Failure of Judges and the Rise of Regulators, where the argument revolves around the systemic failure of the courts to handle disputes which increases the demand for regulation. 20. Rajan (2019). 21. Putnam (2000). 22. Case and Deaton (2020). 23. For reference, see Tucker (2006). 24. Tavits and Letki (2014). 25. This too testifies to how much we as a society depend on political leaders and blame them for all the faults in our own lives. Too much power is given to governments and we expect too much in return. 26. Andris et al. (2015). 27. Rajan (2019). 28. Aghion et al. (2010). 29. This was exposed to a dramatic extent during the Eurozone sovereign debt crisis. Eventually, international markets and other outside pressures led to downfalls of two governments, replacing them with technocratic prime ministers in the most unstable countries in 2013, Italy and Greece. 30. According to the European Commission’s (2014) “EU Anti-​Corruption Report”, procurement corruption is the biggest source of corruption in EU Member States, reaching the size of the entire annual EU budget. The report recognized too much discretionary power given to politicians and bureaucrats as the biggest motivation for procurement corruption. See more at https://​eur-​lex.eur​opa.eu/​legal-​cont​ent/​EN/​ ALL/​?uri=​celex%3A5201​4DC0​038. 31. Title of Chapter 10 in Hayek’s (2007 [1944]) The Road to Serfdom.

References Abramoff, J. 2012. “I Know the Congressional Culture of Corruption.” The Atlantic, July 24, 2012. Available at: https://​www.thea​tlan​tic.com/​polit​ics/​arch​ive/​2012/​07/​i-​know-​ the-​congre​ssio​nal-​cult​ure-​of-​cor​rupt​ion/​260​081/​ (accessed September 2017). Abul-​Magd, A.Y. 2002. “Wealth Distribution in an Ancient Egyptian Society.” Physical Review E, 66: 057104. Acemoglu, D., Cantoni, D., Johnson, S., and Robinson, J. 2011. “The Consequences of Radical Reform: The French Revolution.” American Economic Review, 101(7): 3286–​3307. Acemoglu, D., and Johnson, S. 2023. Power and Progress: Our Thousand-​Year Struggle over Technology and Prosperity. London: Basic Books. Acemoglu, D., Johnson, S., Kermani, A., Kwak, J., and Mitton, T. 2016. “The Value of Connections in Turbulent Times: Evidence from the United States.” Journal of Financial Economics, 121: 368–​391. Acemoglu, D., Naidu, S., Restrepo, P., and Robinson, J. 2019. “Democracy Does Cause Growth.” Journal of Political Economy, 127(1): 47–​100. Acemoglu, D., and Robinson, J. 2006. Economic Origins of Dictatorship and Democracy. Cambridge: Cambridge University Press. Acemoglu, D., and Robinson, J. 2012. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. London: Crown Publishers. Acemoglu, D., and Robinson, J. 2019. The Narrow Corridor: States, Societies, and the Fate of Liberty. New York: Penguin Books. Acemoglu, D., Robinson, J., and Verdier, T. 2017. “Asymmetric Growth and Institutions in an Interdependent World.” Journal of Political Economy, 125(5): 1245–​1305. Ackerman, K.D. 2005. Boss Tweed: The Rise and Fall of the Corrupt Politician Who Conceived the Soul of Modern New York. New York: Da Capo Press. Aghion, P., Algan, Y., Cahuc, P., and Shleifer, A. 2010. “Regulation and Distrust.” Quarterly Journal of Economics, 125(3): 1015–​1049. Akerlof, G., and Shiller, R. 2015. Phishing for Phools: The Economics of Manipulation and Deception. Princeton: Princeton University Press. Akerman, A., Helpman, E., Itskhoki, O., Muendler M.A., and Redding, S. 2013. “Sources of Wage Inequality.” American Economic Review, 103(3): 214–​219. Albertus, M., and Menaldo, V. 2018. Authoritarianism and the Elite Origins of Democracy. Cambridge: Cambridge University Press. Alesina, A., and Perotti, R. 1994. “The Political Economy of Growth: A Critical Survey of the Recent Literature.” World Bank Economic Review, 8(3): 351–​371. Alfani, G., and Ammannati, F. 2017. “Long‐Term Trends in Economic Inequality: The Case of the Florentine State, c. 1300–​1800.” Economic History Review, 70(4): 1072–​1102. Alonso, M., and Simon, B. 2023. “Elite CEOs: Political Connections and Pay in China.” SSRN, July 27, 2023. Available at: https://​ssrn.com/​abstr​act=​4363​367. Alregab, H. 2015. “CEO Pay: The Role of Performance, Governance and Political Connection.” Doctoral dissertation, University of Leicester, School of Management,

342 References 2015. Available at: https://​leices​ter.figsh​are.com/​artic​les/​the​sis/​CEO_​pay_​the_​role_​ of_​performance_​governan​ce_​a​nd_​p​olit​ical​_​con​nect​ion/​10157​642. Alstadsæter, A., Johannesen, N., and Zucman, G. 2019. “Tax Evasion and Inequality.” American Economic Review, 109(6): 2073–​2103. Alt, J., Bueno de Mesquita, E., and Rose, S. 2011. “Disentangling Accountability and Competence in Elections: Evidence from US Term Limits.” Journal of Politics, 73(1): 171–​186. Alt, J., Carlsen, F., Heum, P., and Johansen, K. 1999. “Asset Specificity and the Political Behavior of Firms: Lobbying for Subsidies in Norway.” International Organization, 53(1): 99–​116. Alvarez, J., Benguria, F., Engbom, N., and Moser, C. 2018. “Firms and the Decline in Earnings Inequality in Brazil.” American Economic Journal: Macroeconomics, 10(1): 149–​189. Andris, C., Lee, D., Hamilton, M., Martino, M., Gunning, C., and Selden, J. 2015. “The Rise of Partisanship and Super-​Cooperators in the US House of Representatives,” PLoS ONE, 10(4): e0123507. Angrist, J., and Krueger, A. 2001. “Instrumental Variables and the Search for Identification: From Supply and Demand to Natural Experiments.” Journal of Economic Perspectives, 15 (4): 69–​85. Ansell, B., and Samuels, D. 2014. Inequality and Democratization: An Elite-​Competition Approach. Cambridge: Cambridge University Press. Ariely, D. 2008. Predictably Irrational: The Hidden Forces That Shape Our Decisions. New York: HarperCollins Publishers. Arrow, K. 1951. Social Choice and Individual Values. New York: Wiley & Sons. Atkinson, A.B. 2008. The Changing Distribution of Earnings in OECD Countries. Oxford: Oxford University Press. Atkinson, A.B. 2015. Inequality: What Can Be Done? Cambridge, MA: Harvard University Press. Atkinson, A.B., Hasell, J., Morelli, S., and Roser, M. 2017. The Chartbook of Economic Inequality. Institute for New Economic Thinking, Oxford Martin School. Available at: https://​www.chartb​ooko​feco​nomi​cine​qual​ity.com/​. Atkinson, A.B., and Piketty, T. (eds.). 2007. Top Incomes over the Twentieth Century: A Contrast between European and English-​Speaking Countries. Oxford: Oxford University Press. Atkinson, A.B., and Piketty, T. (eds.). 2010. Top Incomes: A Global Perspective. Oxford: Oxford University Press. Atkinson, A.B., Piketty, T., and Saez, E. 2011. “Top Incomes in the Long Run of History.” Journal of Economic Literature, 49(1): 3–​71. Austen-​Smith, D., and Banks, J. 1989. “Electoral Accountability and Incumbency.” In P.C. Ordeshook (ed.), Models of Strategic Choice in Politics. Ann Arbor: University of Michigan Press, pp.121–​148. Autor, D. 2014. “Skills, Education, and the Rise of Earnings Inequality among the ‘Other 99 Percent.’” Science, 344(6186): 843–​851. Autor, D., Dorn, D., Katz, L., Patterson, C., and Van Reenen, J. 2020. “The Fall of the Labor Share and the Rise of Superstar Firms.” Quarterly Journal of Economics, 135(2): 645–​709. Bakija, J., Cole, A., and Heim, B. 2012. “Jobs and Income Growth of Top Earners and the Causes of Changing Income Inequality: Evidence from U.S. Tax Return Data.” Working paper, Department of Economics, Williams College.

References  343 Bandiera, O., Prat, A., and Valletti, T. 2009. “Active and Passive Waste in Government Spending: Evidence from a Policy Experiment.” American Economic Review, 99(4): 1278–​1308. Banerjee, A., and Duflo, E. 2019. Good Economics for Hard Times: Better Answers to Our Biggest Problems. New York: Penguin Random House. Banks, J., and Sundaram, R. 1993. “Adverse Selection and Moral Hazard in a Repeated Elections Model.” In W. Barnett, M. Hinich, and N. Schofield (eds.), Political Economy: Institutions, Information, Competition and Representation. New York: Cambridge University Press, pp. 295–​311. Barbasi, A.L. 2003. Linked: How Everything Is Connected to Everything Else and What It Means for Business, Science, and Everyday Life. New York: Basic Books. Bartels, L.M. 2016. Unequal Democracy: The Political Economy of the New Gilded Age. Second Edition. Princeton: Princeton University Press. Barth, E., Bryson, A., Davis, J., and Freeman, R. 2016. “It’s Where You Work: Increases in the Dispersion of Earnings across Establishments and Individuals in the United States.” Journal of Labor Economics, 34(S2): 67–​97. Beck, R., Hoskins, C., and Connolly, J.M. 1992. “Rent Extraction through Political Extortion: An Empirical Examination.” Journal of Legal Studies, 21: 217–​224. Becker, G. 1968. “Crime and Punishment: An Economic Approach.” Journal of Political Economy, 76(2): 169–​217. Becker, G. 1983. “A Theory of Competition among Pressure Groups for Political Influence.” Quarterly Journal of Economics, 98: 371–​400. Becker, J., and Morgenson, G. 2009. “Geithner, Member and Overseer of Finance Club.” New York Times, April 26, 2009. Available at: https://​www.nyti​mes.com/​2009/​04/​27/​ busin​ess/​27g​eith​ner.html (accessed March 2017). Bell, B.D., and Van Reenen, J. 2013. “Extreme Wage Inequality: Pay at the Very Top.” American Economic Review, 103(3): 153–​57. Bellettini, G., Ceroni, C.B., and Prarolo, G. 2013. “Political Persistence and Economic Growth.” European Journal of Political Economy, 31: 165–​179. Bergson, A. 1984. “Income Inequality under Soviet Socialism.” Journal of Economic Literature, 22: 1052–​1099. Bermeo, N. 2009. “Does Electoral Democracy Boost Economic Equality?” Journal of Democracy, 20(4): 21–​35. Bernanke, B., Geithner, T., and Paulson, H. 2019. Firefighting: The Financial Crisis and Its Lessons. New York: Penguin Random House. Bertrand, M., Bombardini, M., and Trebbi, F. 2014. “Is It Whom You Know or What You Know? An Empirical Assessment of the Lobbying Process.” American Economic Review, 104(12): 3885–​3920. Bertrand, M., and Mullainathan, S. 2001. “Are CEOs Rewarded for Luck? The Ones without Principals Are.” Quarterly Journal of Economics, 116(3): 371–​400. Besley, T. 2004. “Paying Politicians: Theory and Evidence.” Journal of the European Economic Association, 2(2–​3): 193–​215. Besley, T. 2006. Principled Agents? The Political Economy of Good Government. New York: Oxford University Press. Besley, T., and Case, A. 1995a. “Incumbent Behavior: Vote-​Seeking, Tax-​Setting and Yardstick Competition.” American Economic Review, 85(1): 25–​45. Besley, T., and Case, A. 1995b. “Does Political Accountability Affect Economic Policy Choices? Evidence from Gubernatorial Term Limits.” Quarterly Journal of Economics, 110(3): 769–​798.

344 References Besley, T., and Prat, A. 2006. “Handcuffs for the Grabbing Hand? Media Capture and Government Accountability.” American Economic Review, 96(3): 720–​736. Besley, T., and Smart, M. 2007. “Fiscal Restraint and Voter Welfare.” Journal of Public Economics, 91: 755–​773. Black, D. 1958. The Theory of Committees and Elections. Cambridge: Cambridge University Press. Blanchard, O. 1997. The Economics of Post-​Communist Transition. Oxford: Oxford University Press. Blanes i Vidal, J., Draca M., and Fons-​Rosen, C. 2012. “Revolving Door Lobbyists.” American Economic Review, 102(7): 3731–​48. Blau, B., Brough, T., and Thomas, D.W. 2013. “Corporate Lobbying, Political Connections, and the Bailout of Banks.” Journal of Banking and Finance, 37(8): 3007–​3017. BoardEx. 2017. Data Dictionary. BoardEx Ltd, New York, 2017. Available at: http://​meta​ lib.ie.edu/​ayuda/​Var​ios/​BoardE​xWRD​SDat​aDic​tion​ary.pdf (accessed November 12, 2017). Bogaard, A., Fochesato, M., and Bowles, S. 2019. “The Farming-​Inequality Nexus: New Insights from Ancient Western Eurasia.” Antiquity, 93(371): 1129–​1143. Bonica, A., McCarty, N., Poole, K.T., and Rosenthal, H. 2013. “Why Hasn’t Democracy Slowed Rising Inequality?” Journal of Economic Perspectives, 27(3): 103–​124. Borondo, J., Borondo, F., Rodriguez-​Sickert, C., and Hidalgo, C.A. 2014. “To Each According to Its Degree: The Meritocracy and Topocracy of Embedded Markets.” Scientific Reports, 4, Article no. 3784. Bradley, J., Gebrekidan, S., and McCann, A. 2020. “Waste, Negligence, and Cronyism: Inside Britain’s Pandemic Spending.” New York Times, December 17, 2020. Available at: https://​www.nyti​mes.com/​inte​ract​ive/​2020/​12/​17/​world/​eur​ope/​brit​ain-​ covid-​contra​cts.html (accessed December 2020). Brennan, G., and Buchanan, J. 1980. The Power to Tax: Analytical Foundations of a Fiscal Constitution. Cambridge: Cambridge University Press. Brown, K.S., et al. 2012. “An Early and Enduring Advanced Technology Originating 71,000 Years Ago in South Africa.” Nature, 491: 590–​593. Buchanan, J., and Tullock, G. 1962. The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press. Bueno De Mesquita, B., and Smith, A. 2011. The Dictator’s Handbook: Why Bad Behavior Is Almost Always Good Politics. New York: Public Affairs. Bueno De Mesquita, B., Smith, A., Siverson, R., and Morrow J. 2005. The Logic of Political Survival. Cambridge, MA: MIT Press. Burkhart, R. 1997. “Comparative Democracy and Income Distribution: Shape and Direction of the Causal Arrow.” Journal of Politics, 59(1): 148–​164. Cachero, M., and Rodriguez-​Modrono, P. 2022. “An Empire of Networks. The Political Economy of the Habsburgs in the Caribbean (1492–​1556).” Journal of Historical Network Research, 7: 181–​215. Calomiris, C.W., and Khan, U. 2015. “An Assessment of TARP Assistance to Financial Institutions.” Journal of Economic Perspectives, 29(2): 53–​80. Cameron, L., and Neal, R. 2003. A Concise Economic History of the World: From Paleolithic Times to the Present. Oxford: Oxford University Press. Card, D., Heining, J., and Kline, P. 2013. “Workplace Heterogeneity and the Rise of West German Wage Inequality.” Quarterly Journal of Economics, 128(3): 967–​1015. Carneiro, R. 1970. “A Theory of the Origin of the State.” Science, 169: 733–​738.

References  345 Case, A., and Deaton, A. 2020. Deaths of Despair and the Future of Capitalism. Princeton: Princeton University Press. Caselli, F., and Morelli, M. 2004. “Bad Politicians.” Journal of Public Economics, 88: 759–​782. Center for Responsive Politics. 2016. Lobbying Spending Database. Washington, DC: AARP, Available at: https://​www.open​secr​ets.org/​orgs/​summ​ary.php?id=​D00​ 0023​726 (accessed September 2017). Chang, E., and Golden, M. 2004. “Does Corruption Pay? The Survival of Politicians Charged with Malfeasance in the Postwar Italian Chamber of Deputies.” Unpublished paper, Michigan State University and the University of California at Los Angeles. Chang, E., and Kerr, N. 2017. “An Insider Outsider Theory of Popular Tolerance for Corrupt Politicians.” Governance, 30(1): 67–​84. Chetty, R., Hendren, N., and Katz, L. 2016. “The Effects of Exposure to Better Neighborhoods on Children: New Evidence from the Moving to the Opportunity Experiment.” American Economic Review, 106(4): 855–​902. Chiarini, B., Marzano, E., and Schneider, F. 2011. “Tax Rates and Tax Evasion: An Empirical Analysis of the Long-​Run Aspects in Italy.” European Journal of Law and Economics, 35: 273–​293. Claessens, S., and Perotti, E. 2007. “Finance and Inequality: Channels and Evidence.” Journal of Comparative Economics, 35:748–​773. Clark, G. 2011. A Farewell to Alms: A Brief Economic History of the World. Princeton: Princeton University Press. Clark, G. 2014. The Son Also Rises: Surnames and the History of Social Mobility. Princeton: Princeton University Press. Clotfeler, C. 1983. “Tax Evasion and Tax Rates: An Analysis of Individual Returns.” Review of Economics and Statistics, 65(3): 363–​373. Coate, S., and Morris, S. 1995. “On the Form of Transfers to Special Interests.” Journal of Political Economy, 103(1995): 1210–​1235. Cohan, B. 2012. “Rethinking Robert Rubin.” Bloomberg Businessweek, September 30, 2012. Available at: https://​www.bloomb​erg.com/​news/​artic​les/​2012-​09-​30/​ret​hink​ ing-​rob​ert-​rubin#p5 (accessed February 2021). Correia, M. 2014. “Political Connections and SEC Enforcement.” Journal of Accounting and Economics, 57 (2–​3): 241–​262. Crane, S., and Nourzad, F. 1990. “Tax Rates and Tax Evasion: Evidence from California Amnesty Data.” National Tax Journal, 43(2): 189–​199. Dabla-​Norris, E., and Wade, P. 2001. “Rent-​Seeking and Endogenous Income Inequality.” IMF Working Paper 01/​15, Washington, DC. Darwin, C. 1871. The Descent of Man, and Selection in Relation to Sex. London: John Murray. Davis, S., and Haltiwanger, J. 1991. “Wage Dispersion between and within U.S. Manufacturing Plants, 1963–​86.” Brookings Papers: Microeconomics, (1991): 115–​200. De Bondt, W.F.M., and Thaler, R. 1985. “Does the Stock Market Overreact?” Journal of Finance, 40(3): 793–​805. De Cesari, A., and Ozkan, N. 2015. “Executive Incentives and Payout Policy: Empirical Evidence from Europe.” Journal of Banking and Finance, 55(C): 70–​91. de Figueiredo, J., Hidalgo, D., and Kasahara, Y. 2012. “When Do Voters Punish Corrupt Politicians? Experimental Evidence from Brazil.” Unpublished manuscript, University of California, Berkeley.

346 References de Figueiredo, J., and Richter, B. 2014. “Advancing the Empirical Research on Lobbying.” Annual Review of Political Science, 17(1): 163–​185. de Figueiredo, J.M., and Silverman, B.S. 2006. “Academic Earmarks and the Returns to Lobbying.” Journal of Law Economics, 49(2): 597–​625. de Figueiredo, J.M., and Silverman, B.S. 2007. “How Do We (Want to) Fund Science? Politics, Lobbying, and Academic Earmarks.” In R. Ehrenberg (ed.), Science and the University. Madison: University of Wisconsin Press. Deaton, A. 2013. The Great Escape: Health, Wealth, and the Origins of Inequality. Princeton: Princeton University Press. DellaVigna, S., and Kaplan, E. 2007. “The Fox News Effect: Media Bias and Voting.” Quarterly Journal of Economics, 122(3): 1187–​1234. Desai, R., and Olofsgard, A. 2011. “The Costs of Political Influence: Firm-​Level Evidence from Developing Countries.” Quarterly Journal of Political Science, 6(2): 137–​178. Diamond, J. 1997. Guns, Germs, and Steel: The Fates of Human Societies. New York: W.W. Norton. Dimock, M., and Jacobson, G. 1995. “Checks and Choices: The House Bank Scandal’s Impact on Voters in 1992.” Journal of Politics, 57(4): 1143–​1159. Dinc, S. 2005. “Politicians and Banks: Political Influences on Government-​Owned Banks in Emerging Markets.” Journal of Financial Economics, 77(2): 453–​479. Ding, S., Jia, C., Wilson, C., and Wu, Z. 2015. “Political Connections and Agency Conflicts: The Roles of Owner and Manager Political Influence on Executive Compensation.” Review of Quantitative Finance and Accounting, 45: 407–​434. Dobratz, B., and Whitfield, S. 1992. “Does Scandal Influence Voters’ Party Preference? The Case of Greece during the Papandreou Era.” European Sociology Review, 8(2): 167–​180. Downs, A. 1957. An Economic Theory of Democracy. New York: Addison-​Wesley. Duchin, R., and Sosyura, D. 2012. “The Politics of Government Investment.” Journal of Financial Economics, 106: 24–​48. Dunbar, R. 1992. “Neocortex Size as a Constraint on Group Size in Primates.” Journal of Human Evolution, 22(6): 469–​493. Dye, T. 2001. Top Down Policymaking. Washington, DC: CQ Press. Earle, J.S., and Gehlbach, S. 2015. “The Productivity Consequences of Political Turnover: Firm-​Level Evidence from Ukraine’s Orange Revolution.” American Journal of Political Science, 59(3): 708–​723. Eddy, B. 2019. Why We Elect Narcissists and Sociopaths: And How We Can Stop. Oakland, CA: Berrett-​Koehler Publishers. Eggers, A.C., and Hainmueller, J. 2009. “MPs for Sale? Returns to Office in Postwar British Politics.” American Political Science Review, 103(4): 1–​21. El-​Khatib, R., Fogel, K., and Jandik, T. 2015. “CEO Network Centrality and Merger Performance.” Journal of Financial Economics, 116(2): 349–​382. Endruweit, G. 1984. “Elites as Agents of Economic Development.” Intereconomics, 19: 32–​38. European Commission, 2014. “EU Anti-​ Corruption Report.” Report from the Commission to the Council and the European Parliament. Available at: https://​ eur-​lex.eur​opa.eu/​legal-​cont​ent/​EN/​ALL/​?uri=​celex%3A5201​4DC0​038 (accessed November 2014). European Commission, 2019. “Estimating International Tax Evasion by Individuals.” Taxation Papers, Working paper no. 76. Available at: https://​ec.eur​opa.eu/​taxat​ ion_​cust​oms/​sites/​taxat​ion/​files/​2019-​taxat​ion-​pap​ers-​76.pdf (accessed April 2020).

References  347 Evans, D. 1996. “Before the Roll Call: Interest Group Lobbying and Public Policy Outcomes in House Committees.” Political Research Quarterly, 49(2): 287–​304. Faccio, M. 2006. “Politically Connected Firms.” American Economic Review, 96(1): 369–​386. Faccio, M., Masulis, R., and McConnell, J. 2006. “Political Connections and Corporate Bailouts.” Journal of Finance, 61(6): 2597–​2635. Fan, J.P.H., Wong, T.J., and Zhang, T. 2007. “Politically Connected CEOs, Corporate Governance and Post-​Ipo Performance of China’s Newly Partially Privatized Firms.” Journal of Financial Economics, 84: 330–​357. Fazekas, M., and Kocsis, G. 2017. “Uncovering High-​Level Corruption: Cross-​National Objective Corruption Risk Indicators Using Public Procurement Data.” British Journal of Political Science, 50(1): 155–​164. Fazekas, M., Toth, I., and King, L. 2016. “An Objective Corruption Risk Index Using Public Procurement Data.” European Journal on Criminal Policy and Research, 22 (3): 369–​397. Ferejohn, J. 1986. “Incumbent Performance and Electoral Control.” Public Choice, 50: 5–​25. Ferguson, N. 2000. The House of Rothschild, Volume 2: The World’s Banker: 1849–​1999. New York: Penguin Books. Ferguson, N. 2008. The Ascent of Money: A Financial History of the World. New York: Penguin Books. Ferguson, N. 2017. The Square and the Tower: Networks, Hierarchies and the Struggle for Global Power. New York: Penguin Books. Ferraz, C., and Finan, F. 2011. “Electoral Accountability and Corruption: Evidence from the Audits of Local Governments.” American Economic Review, 101(4): 1274–​1311. Fichtner, J., Heemskerk, E., and Garcia-​Bernardo, J. 2017. “Hidden Power of the Big Three? Passive Index Funds, Re-​Concentration of Corporate Ownership, and New Financial Risk.” Business and Politics, 19(2): 298–​326. Fidrmuc, J. 2003. “Economic Reform, Democracy and Growth during Post-​Communist Transition.” European Journal of Political Economy, 19(3): 583–​604. Fisman, R. 2001. “Estimating the Value of Political Connections.” American Economic Review, 91(4): 1095–​1102. Fisman, R., and Svensson, J. 2007. “Are Corruption and Taxation Really Harmful to Growth? Firm Level Evidence.” Journal of Development Economics, 83(1): 63–​75. Fisman, R., and Wei, S. 2004. “Tax Rates and Tax Evasion: Evidence from ‘Missing Imports’ in China.” Journal of Political Economy, 112(2): 471–​496. Frank, R. 2015. “The ‘Manager CEOs’ Who Became Billionaires.” CNBC, June 3, 2015. Available at: https://​www.cnbc.com/​2015/​06/​03/​jpmor​gan-​ceo-​jamie-​dimon-​beco​ mes-​a-​bill​iona​ire.html (accessed February 2021). Frankopan, P. 2015. The Silk Roads: A New History of the World. London: Bloomsbury. Friedman, M. 1962. Capitalism and Freedom. Chicago: University of Chicago Press. Fukuyama, F. 2014. Political Order and Political Decay: From the Industrial Revolution to the Globalization of Democracy. New York: Farrar, Straus and Giroux. Furman, J., and Orszag, P. 2015. “A Firm-​Level Perspective on the Role of Rents in the Rise in Inequality.” Presentation at “A Just Society” Centennial Event in Honor of Joseph Stiglitz, Columbia University, October 16, 2015. Gabaix, X., and Landier, A. 2008. “Why Has CEO Pay Increased So Much?” Quarterly Journal of Economics, 123(1): 49–​100.

348 References Gamberoni, E., Giordano, C., and Lopez-​Garcia, P. 2016. “Capital and Labour (Mis) Allocation in the Euro Area: Some Stylized Facts and Determinants.” ECB Working Paper No. 1981, European Central Bank. Garcia-​Santana, M., Moral-​Benito, E., Pijoan, M., and Ramos, R. 2020. “Growing like Spain: 1995–​2007.” International Economic Review, 61(1): 383–​416. Gardiner, J.A. 2002. “Defining Corruption.” In A. Heidenheimer, and M. Johnston (eds.), Political Corruption: Concepts and Contexts. Third edition. New Brunswick: Transaction Publishers, pp. 25–​40. Gehlbach, S. 2006. “The Consequences of Collective Action: An Incomplete-​Contracts Approach.” American Journal of Political Science, 50(3): 802–​823. Gehlbach, S., Sonin, K., and Zhuravskaya, E. 2010. “Businessmen Candidates.” American Journal of Political Science, 54(3): 718–​736. Gerschenkron, A. 1962. Economic Backwardness in Historical Perspective: A Book of Essays. Cambridge, MA: Belknap, Harvard University Press. Giattino, C., Ortiz-​ Ospina, E., and Roser, M. 2020 [2013]. “Working Hours.” OurWorldInData.org. Retrieved from: https://​our​worl​dind​ata.org/​work​ing-​hours (accessed October 2022). Gilens, M. 2012. Affluence and Influence: Economic Inequality and Political Power in America. Princeton: Princeton University Press. Gilens, M., and Page, B. 2014. “Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens.” Perspectives on Politics, 12(3): 564–​581. Glaeser, E., Scheinkman, J., and Shleifer, A. 2003. “The Injustice of Inequality.” Journal of Monetary Economics, 50(1): 199–​222. Glaurdić, J. 2011. The Hour of Europe: Western Powers and the Breakup of Yugoslavia. New Haven: Yale University Press. Godechot, O. 2012. “Is Finance Responsible for the Rise in Wage Inequality in France?” Socio-​Economic Review, 10(3): 447–​470. Golden, M. 2004. “International Economic Sources of Regime Change: How European Economic Integration Undermined Italy’s Postwar Party System.” Comparative Political Studies, 37(10): 1238–​1274. Goldman, E., Rocholl, J., and So, J. 2009. “Do Politically Connected Boards Affect Firm Value?” Review of Financial Studies, 22(6): 2331–​2360. Goodman, P.S. 2008. “Taking Hard New Look at a Greenspan Legacy.” New York Times, October 8, 2008. Available at: https://​www.nyti​mes.com/​2008/​10/​09/​busin​ess/​econ​ omy/​09gr​eens​pan.html (accessed February 2021). Green, J., Kochkodin, B., and Migliozzi, B. 2016. “The Lucrative Road from Congress to Corporate Boards.” Bloomberg, October 18, 2016. Available at: https://​www.bloomb​ erg.com/​graph​ics/​2016-​congre​ssio​nal-​board-​pay/​. Greenwood, R., and Scharfstein, D. 2013. “The Growth of Finance.” Journal of Economic Perspectives, 27(2): 3–​28. Griffiths, D., Lambert, P., and Bihagen, E. 2014. “Measuring the Potential Power Elite in the UK and Sweden.” European Societies, 16(5): 742–​762. Grossman, G.M., and Helpman, E. 2002. Special Interest Politics. Cambridge, MA: MIT Press. Gupta, S., Davoodi, H., and Alonso-​Terme, R. 2002. “Does Corruption Affect Income Inequality and Poverty?” Economics of Governance, 3(1): 23–​45. Guttsman, W.C. 1965. The British Political Elite. London: MacGibbon and Kee.

References  349 Hacker, J., and Pierson, P. 2011. Winner-​Take-​All Politics: How Washington Made the Rich Richer–​And Turned Its Back on the Middle Class. New York: Simon & Schuster. Håkanson, C., Lindqvist, E., and Vlachos, J. 2021. “Firms and Skills: The Evolution of Worker Sorting.” Journal of Human Resources, 56(2): 512–​538. Harari, Y.N. 2011. Sapiens: A Brief History of Mankind. London: Vintage. Harvard Centre for History and Economics. 2020. “Visualizing Historical Networks: Inner Life of Empires & The Economy of Privilege.” Harvard University, Cambridge, MA. Available at: https://​histe​con.fas.harv​ard.edu/​visu​aliz​ing/​index.html (accessed December 2020). Hasell, J., Roser, M., Ortiz-​Ospina, E., and Arriaga, P. 2019. “Poverty.” OurWorldInData. org. Retrieved from: https://​our​worl​dind​ata.org/​pove​rty (accessed October 2022). Hayek, F.A. 2007 [1944]. The Road to Serfdom. Chicago: University of Chicago Press. Helland, L. 2008. “Lobbying with Conflicting Interests: Norwegian Local-​ Central Relations.” European Journal of Political Research, 47(2): 184–​205. Herre, B. 2021. “Regimes of the World.” OurWorldInData.org. Retrieved from: https://​ our​worl​dind​ata.org/​regi​mes-​of-​the-​world-​data (accessed October 2022). Herre, B., Spooner, F., and Roser, M. 2019. “Homicides.” OurWorldInData.org. Retrieved from: https://​our​worl​dind​ata.org/​homici​des (accessed October 2022). Hochberg, Y.V., Sapienza, P., and Vissing-​Jørgensen, A. 2009. “A Lobbying Approach to Evaluating the Sarbanes-​ Oxley Act of 2002.” Journal of Accounting Research, 47(2): 519–​583. Huber, E., Nielsen, F., Pribble, J., and Stephens, J. 2006. “Politics and Inequality in Latin America and the Caribbean.” American Sociological Review, 71(6): 943–​963. Igan, D., Mishra, P., and Tressel, T. 2012. “A Fistful of Dollars: Lobbying and the Financial Crisis.” NBER Macroeconomics Annual, 26(1): 195–​230. ILOSTAT. 2020. “Statistics on Labor Productivity.” Geneva: International Labor Organization. Available at: https://​ilos​tat.ilo.org/​top​ics/​lab​our-​produ​ctiv​ity/​ (accessed February 2021). International Monetary Fund (IMF). 2013. Dataset. “Public Finances in Modern History.” Washington, DC: International Monetary Fund. Available at: https://​www.imf.org/​ exter​nal/​np/​fad/​his​tdb/​ (accessed January 2017). Jackson, M.O. 2010. Social and Economic Networks. Princeton: Princeton University Press. Jayachandran, S. 2006. “The Jeffords Effect.” Journal of Law Economics, 49(2): 397–​425. Kahneman, D. 2011. Thinking, Fast and Slow. New York: Farrar Straus Giroux. Kelleher, C.A., and Yackee, S.W. 2009. “A Political Consequence of Contracting: Organized Interests and State Agency Decision Making.” Journal of Public Administration Research Theory, 19(3): 579–​602. Khwaja, A.I., and Mian, A. 2005. “Do Lenders Favor Politically Connected Firms? Rent Provision in an Emerging Financial Market.” Quarterly Journal of Economics 120(4): 1371–​1411. Kim, J.W., Kogut, B., and Yang, J.S. 2015. “Executive Compensation, Fat Cats, and Best Athletes.” American Sociological Review, 80(2): 299–​328. Klašnja, M. 2015. “Corruption and the Incumbency Disadvantage: Theory and Evidence.” Journal of Politics, 77 (4): 928–​942. Klašnja, M., Tucker, J., and Deegan-​ Krause, K. 2016. “Pocketbook vs. Sociotropic Corruption Voting.” British Journal of Political Science, 46(1): 67–​94. Knack, S., and Keefer, P. 1995. “Institutions and Economic Performance: Cross-​Country Tests Using Alternative Institutional Measures.” Economics and Politics, 7(3): 207–​227.

350 References Kohler, T.A. et al. 2017. “Greater Post-​Neolithic Wealth Disparities in Eurasia than in North America and Mesoamerica.” Nature, 551: 619–​622. Kong, G., Huang, J., and Ma, G. 2023. “Anti-​ Corruption and within-​ Firm Pay Gap: Evidence from China.” Pacific-​Basin Finance Journal, 79: 102041. Krueger, A. 1974. “The Political Economy of the Rent Seeking Society.” American Economic Review, 64: 291–​303. Kurer, O. 2001. “Why Do Voters Support Corrupt Politicians?” In A.K. Jain (ed.), The Political Economy of Corruption. London: Routledge, pp. 60–​84. Kuznets, S. 1955. “Economic Growth and Income Inequality.” American Economic Review, 45(March): 1–​28. Kwak, J. 2014. “Cultural Capture and the Financial Crisis.” In D. Carpenter, and D. Moss (eds.), Preventing Regulatory Capture: Special Interest Influence and How to Limit It. Cambridge: Cambridge University Press, pp. 71–​98. Lee, C.S. 2005. “Income Inequality, Democracy, and Public Sector Size.” American Sociological Review, 70(1):158–​181. Lenski, G. 1966. Power and Privilege: A Theory of Social Stratification. New York: McGraw-​Hill. Lindsey, B., and Teles, S. 2017. The Captured Economy: How the Powerful Enrich Themselves, Slow Down Growth, and Increase Inequality. New York: Oxford University Press. Locke, J. 2003 [1690]. Two Treatises of Government. Second Treatise. New Haven: Yale University Press. Machiavelli, N. 1999 [1532]. The Prince. London: Penguin Classics. Maclean, M., Harvey, C., and Chia, R. 2010. “Dominant Corporate Agents and the Power Elite in France and Britain.” Organization Studies, 31(3): 327–​348. Maddison Historical Statistics. 2021. Maddison Project Database 2020. Available at: https://​www.rug.nl/​ggdc/​histor​ical​deve​lopm​ent/​maddi​son/​relea​ses/​maddi​son-​ proj​ect-​datab​ase-​2020 (accessed February 2021). Manzetti, L., and Wilson, C. 2007. “Why Do Corrupt Governments Maintain Public Support?” Comparative Political Studies, 40(8): 949–​970. Marean, C.W. 2014. “The Origins and Significance of Coastal Resource Use in Africa and Western Eurasia.” Journal of Human Evolution, 77: 17–​40. Marean, C.W. 2015. “The Most Invasive Species of All.” Scientific American, 313(2): 23–​29, August 2015. Marx, K., and Engels, F. 2004 [1848]. Manifesto of the Communist Party. Marxist Internet Archive. Available at: https://​www.marxi​sts.org/​arch​ive/​marx/​works/​1848/​commun​ ist-​manife​sto/​ (accessed February 2021). Mauro, P. 1995. “Corruption and Growth.” Quarterly Journal of Economics, 110(3): 681–​712. Mauro, P., Romeu, R., Binder, A., and Zaman, A. 2013. “A Modern History of Fiscal Prudence and Profligacy.” IMF Working paper WP/​13/​5, Washington, DC. McChesney, F.S. 1997. Money for Nothing: Politicians, Rent Extraction, and Political Extortion. Cambridge, MA: Harvard University Press. McCloskey, D. 2012. Bourgeois Dignity: Why Economics Can’t Explain the Modern World. Chicago: University of Chicago Press. Meltzer, A., and Richard, S. 1981. “A Rational Theory of the Size of Government.” Journal of Political Economy, 89(5): 914–​927. Meyer, R., Höllerer, M., and Leixnering, S. 2018. “A Question of Value(S): Political Connectedness and Executive Compensation in Public Sector Organizations.” International Public Management Journal, 21(3): 477–​500.

References  351 Mian, A., Sufi, A., and Trebbi, F. 2010. “The Political Economy of the US Mortgage Default Crisis.” American Economic Review, 100(5): 1967–​98. Michels, R. 1962 [1915]. A Sociological Study of the Oligarchical Tendencies of Modern Democracy. New York: Crowell-​Collier. Milanović, B. 1991. “Poverty in Eastern Europe in the Years of Crisis, 1978 to 1987: Poland, Hungary, and Yugoslavia.” World Bank Economic Review, 5(2): 187–​205. Milanović, B. 1998. Income, Inequality, and Poverty during the Transition from Planned to Market Economy. World Bank regional and sectoral studies, Washington, DC: World Bank Group. Milanović, B. 2016. Global Inequality: A New Approach for the Age of Globalization. Cambridge, MA: Belknap, Harvard University Press. Milanović, B. 2019. Capitalism, Alone: The Future of the System That Rules the World. Cambridge, MA: Belknap, Harvard University Press. Milanović, B., Lindert, P., and Williamson, J. 2011. “Pre-​Industrial Inequality.” Economic Journal, 121(551): 255–​272. Miller, W.L., Grodeland, A.B., and Koshechkina, T.Y. 2001. A Culture of Corruption? Coping with Government in Post-​Communist Europe. Budapest: Central European University Press. Mills, C.W. 1956. The Power Elite. Oxford: Oxford University Press. Mokyr, R. 2016. A Culture of Growth: The Origins of the Modern Economy. Princeton: Princeton University Press. Morillas-​Torne, M. 2012. “Creation of a Geo-​Spatial Database to Analyse Railways in Europe (1830–​2010). A Historical GIS Approach.” Journal of Geographic Information System, 4(2): 176–​187. Mosca, G. 1939. The Ruling Class. New York: McGraw Hill. Moss, A.L. 2008. Selling Out America’s Democracy. Westport, CT: Praeger Publishers. Moyer, L. 2016. “JPMorgan Backs Geithner’s Investments with Warburg Pincus.” New York Times, February 8, 2016. Available at: https://​www.nyti​mes.com/​2016/​02/​ 09/​busin​ess/​dealb​ook/​jpmor​gan-​backs-​geithn​ers-​inve​stme​nts-​with-​warb​urg-​pin​cus. html (accessed February 2021). Mueller, H., Ouimet, P., and Simintzi, E. 2017. “Wage Inequality and Firm Growth.” American Economic Review, 107(5): 379–​83. Muller, E. 1988. “Democracy, Economic Development, and Income Inequality.” American Sociological Review, 53(1): 50–​68. Musgrave, R. 1959. The Theory of Public Finance: A Study in Public Economy. New York: McGraw-​Hill. Nash, J. 1950. “Equilibrium Points in n-​Person Games.” Proceedings of the National Academy of Sciences, 36(1): 48–​49. Navidi, S. 2017. Superhubs: How the Financial Elite and Their Networks Rule Our World. London: Nicholas Brealey Publishing. North, D.C., Wallis, J.J., and Weingeist, B.R. 2010. Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History. Cambridge: Cambridge University Press. Nyblade, B., and Reed, S. 2008. “Who Cheats? Who Loots? Political Competition and Corruption in Japan,1947–​ 1993.” American Journal of Political Science, 52(4): 926–​941. Okun, A. 1975. Equality and Efficiency: The Big Tradeoff. Washington, DC: The Brookings Institution.

352 References Olson, M. 1971 [1965]. The Logic of Collective Action. Cambridge, MA: Harvard University Press. Olson, M. 1982. The Rise and Decline of Nations. New Haven: Yale University Press. Olson, M. 2000. Power and Prosperity: Outgrowing Communist and Capitalist Dictatorships. New York: Basic Books. Ortiz-​Ospina, E., and Roser, M. 2022 [2016]. “Trust.” OurWorldInData.org. Retrieved from: https://​our​worl​dind​ata.org/​trust (accessed October 2022). Ostrom, E. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press. Pareto, V. 1935. The Mind and Society. New York: Harcourt, Brace and Company. Patrikios, S., and Karyotis, G. 2008. “The Greek Parliamentary Election of 2007.” Electoral Studies, 27(2): 356–​359. Peltzman, S. 1976. “Toward a More General Theory of Regulation.” Journal of Law and Economics, 19: 211–​240. Persson, T., and Tabellini, G. 2000. Political Economics: Explaining Economic Policy. Cambridge, MA: MIT Press. Persson, T., and Tabellini, G. 2009. “Democratic Capital: The Nexus of Political and Economic Change.” American Economic Journal: Macroeconomics, 1(2): 88–​126. Peters, J., and Welch, S. 1980. “The Effects of Charges of Corruption on Voting Behavior in Congressional Elections.” American Political Science Review, 74 (3): 697–​708. Petrova, M. 2008. “Inequality and Media Capture.” Journal of Public Economics, 92(1–​ 2): 183–​212. Phelps, E. 2013. Mass Flourishing: How Grassroots Innovation Created Jobs, Challenge, and Change. Princeton: Princeton University Press. Philippon, T., and Reshef, A. 2013. “An International Look at the Growth of Modern Finance.” Journal of Economic Perspectives, 27(2): 73–​96. Piattoni, S. 2001. “Clientelism in Historical and Comparative Perspective.” In S. Piattoni (ed.), Clientelism, Interests, and Democratic Representation. Cambridge: Cambridge University Press, pp. 1–​30. Pigou, A. 1932 [1920]. The Economics of Welfare. Fourth edition. London: Macmillan. Piketty, T. 2014. Capital in the Twenty-​ First Century. Cambridge, MA: Harvard University Press. Piketty, T. 2020. Capital and Ideology. Cambridge: Harvard University Press. Piketty, T., and Saez, E. 2003. “Income Inequality in the United States, 1913–​1998.” Quarterly Journal of Economics, 118(1): 1–​41. Piketty, T., Saez, E., and Stantcheva, S. 2014. “Optimal Taxation of Top Labor Incomes: A Tale of Three Elasticities.” American Economic Journal: Economic Policy, 6(1): 230–​71. Podobnik, B., Vuković, V., and Stanley, E. 2015. “Does the Wage Gap between Private and Public Sectors Encourage Political Corruption?” PLoS ONE, 10(10): e0141211. Polity IV Project, 2018. “Political Regime Characteristics and Transitions, 1800–​2013.” Center for Systemic Peace. Available at: https://​www.system​icpe​ace.org/​pol​ity/​poli​ty4. htm (accessed September 2018). Pommerehne, W., and Weck-​Hannemann, H. 1996. “Tax Rates, Tax Administration and Income Tax Evasion in Switzerland.” Public Choice, 88: 161–​170. Poterba, J. 1987. “Tax Evasion and Capital Gains Taxation.” American Economic Review, 77(2): 234–​239. Przeworski, A. 1991. Democracy and the Market: Political and Economic Reforms in Eastern Europe and Latin America. Cambridge: Cambridge University Press.

References  353 Putnam, R. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. Querubin, P., and Snyder, J. 2013. “The Control of Politicians in Normal Times and Times of Crisis: Wealth Accumulation by U.S. Congressmen, 1850–​1880.” Quarterly Journal of Political Science, 8(4): 409–​450. Rajan, R. 2019. The Third Pillar: How Markets and the State Leave the Community Behind. New York: Penguin Random House. Reed, S. 1999. “Punishing Corruption: The Response of the Japanese Electorate to Scandals.” In O. Feldman (ed.), Political Psychology in Japan: Behind the Nails Which Sometimes Stick Out (and Get Hammered Down). New York: Nova Science. Ricardo, D. 1817. On the Principles of Political Economy and Taxation. London: John Murray. Richter, B.K., Samphantharak, K., and Timmons, J.F. 2009. “Lobbying and Taxes.” American Journal of Political Science, 53(4): 893–​909. Rijkers, B., Freund, C., and Nucifora, A. 2014. “All in the Family: State Capture in Tunisia.” Policy Research Working paper WPS6810. World Bank, Washington, DC. Riordan, W.L. 1995. Plunkitt of Tammany Hall: A Series of Very Plain Talks on Very Practical Politics. New York: Signet Classics. Robinson, J., and Verdier, T. 2013. “The Political Economy of Clientelism.” Scandinavian Journal of Economics, 115(2): 260–​291. Rodrik, D. 1999. “Where Did All the Growth Go? External Shocks, Social Conflict, and Growth Collapses.” Journal of Economic Growth, 4(1999): 385–​412. Rogoff, K. 1990. “Equilibrium Political Budget Cycles.” American Economic Review, 80(1): 21–​36. Root, H.L., and Nellis, N. 2000. “The Compulsion of Patronage: Political Sources of Information Asymmetry and Risk in Developing Country Economies.” In Bueno De Mesquita, and H. L. Root (eds.), Governing for Prosperity. New Haven: Yale University Press, pp. 85–​105. Rose-​Ackerman, S. 1978. Corruption: A Study in Political Economy. New York: Academic Press. Roser, M., Appel, C., and Ritchie, H. 2019 [2013]. “Human Height.” OurWorldInData.org. Retrieved from: https://​our​worl​dind​ata.org/​human-​hei​ght (accessed October 2022). Roser, M., Ortiz-​ Ospina, E., and Ritchie, H. 2019 [2013]. “Life Expectancy.” OurWorldInData.org. Retrieved from: https://​our​worl​dind​ata.org/​life-​exp​ecta​ncy (accessed October 2022). Rothschild, E. 2012. The Inner Life of Empires: An Eighteen-​Century History. Princeton: Princeton University Press. Rundquist, B.S., Strom, G.S., and Peters, J.G. 1977. “Corrupt Politicians and Their Electoral Support: Some Experimental Observations.” American Political Science Review, 71(3): 954–​963. Ruostetsaari, I. 2006. “Elites and Democracy: Are They Compatible?.” In F. Engelstad, and T. Gulbrandsen (eds.), Comparative Studies of Social and Political Elites. Comparative Social Research, Vol. 23. Bingley: Emerald Group Publishing Limited, pp. 265–​274. Saez, E., Slemrod, J., and Giertz, S.H. 2012. “The Elasticity of Taxable Income with Respect to Marginal Tax Rates: A Critical Review.” Journal of Economic Literature, 50(1): 3–​50. Samuelson, P. 1954. “The Pure Theory of Public Expenditure.” Review of Economics and Statistics, 36(4): 387–​389.

354 References Samuelson, P. 1983 [1947]. Foundations of Economic Analysis. Cambridge, MA: Harvard University Press. Sapienza, P. 2004. “The Effects of Government Ownership on Bank Lending.” Journal of Financial Economics, 72(2): 357–​384. Scheidel, W. 2017. The Great Leveler: Violence and the History of Inequality from the Stone Age to the Twenty-​First Century. Princeton: Princeton University Press. Scheve, K., and Stasavage, D. 2009. “Institutions, Partisanship, and Inequality in the Long Run.” World Politics, 61(2): 215–​253. Schmidt, B. 2015. “Costs and Benefits of Friendly Boards during Mergers and Acquisitions.” Journal of Financial Economics, 117(2): 424–​447. Schoenman, R. 2014. Networks and Institutions in Europe’s Emerging Markets. Cambridge: Cambridge University Press. Schone, K., Koch, E., and Baumont, C. 2013. “Modeling Local Growth Control Decisions in a Multi-​City Case: Do Spatial Interactions and Lobbying Efforts Matter?” Public Choice, 154(1–​2): 95–​117. Schouten, R., and Silver, J. 2012. Almost a Psychopath: Do I (or Does Someone I Know) Have a Problem with Manipulation and Lack of Empathy? Center City, MN: Hazelden Publishing. Shiller, R. 2013. “The Best, Brightest, and Least Productive?” Project Syndicate, September 20, 2013. Available at: https://​www.proj​ect-​syndic​ate.org/​com​ment​ary/​the-​rent-​seek​ ing-​prob​lem-​in-​conte​mpor​ary-​fina​nce-​by-​rob​ert-​j-​-​shil​ler-​2013-​09. Shleifer, A. (ed.). 2012. The Failure of Judges and the Rise of Regulators. Cambridge: MIT Press. Shleifer, A., and Vishny, R. 1993. “Corruption.” Quarterly Journal of Economics, 108(3): 599–​617. Shleifer, A., and Vishny, R. 1994. “Politicians and Firms.” Quarterly Journal of Economics, 109: 995–​1025. Shleifer, A., and Vishny, R. 1998. The Grabbing Hand. Cambridge, MA: Harvard University Press. Simpson, M. 1990. “Political Rights and Income Inequality: A Cross-​National Test.” American Sociological Review, 55(5): 682–​693. Sirowy, L., and Inkeles, A. 1990. “The Effects of Democracy on Economic Growth and Inequality: A Review.” Studies In Comparative International Development, 25(1990): 126–​157. Smith, A., 1904 [1776]. An Inquiry into the Nature and Causes of the Wealth of Nations. Fifth edition. London: Methuen & Co. Song, J., Price, D., Guvenen, F., Bloom, N., and Wachter, T. 2019. “Firming Up Inequality.” Quarterly Journal of Economics, 134(1): 1–​50. St Louis Federal Reserves, FRED database. 2021. Corporate Profits after Tax (without IVA and CCAdj)/​Gross Domestic Product St Louis Federal Reserves. Available at: https://​fred. stl​ouis​fed.org/​graph/​?g=​1Pik (accessed February 2021). Stasavage, D. 2020. The Decline and Rise of Democracy. A Global History from Antiquity to Today. Princeton: Princeton University Press. Stein, G. 1994. “Economy, Ritual, and Power in Ubaid Mesopotamia.” In G. Stein, and M. S. Rothman (eds.), Chiefdoms and Early States in the Near East: The Organizational Dynamics of Complexity. Madison, WI: Prehistory Press.

References  355 Stewart, J.B. 2009. “Eight Days: The Battle to Save the American Financial System.” New Yorker, September 21, 2009. Available at: https://​www.newyor​ker.com/​magaz​ine/​ 2009/​09/​21/​eight-​days. Stigler, G. 1971. “The Theory of Economic Regulation.” Bell Journal of Economics and Management Science, 2(Spring): 3–​21. Stiglitz, J. 2012. The Price of Inequality: How Today’s Divided Society Endangers Our Future. New York: WW Norton. Stokes, S., Dunning, T., Nazareno, M., and Brusco, V. 2013. Brokers, Voters, and Clientelism: The Puzzle of Distributive Politics. Cambridge: Cambridge University Press. Sussman, N. 2006. “Income Inequality in Paris in the Heyday of the Commercial Revolution.” Unpublished manuscript, Hebrew University. Svejnar, J., and Bagchi, S. 2015. “Does Wealth Inequality Matter for Growth? The Effect of Billionaire Wealth, Income Distribution, and Poverty.” Journal of Comparative Economics, 43(3): 505–​530. Taleb, N.N. 2012. Antifragile: Things That Gain from Disorder. New York: Random House. Taleb, N.N. 2018. Skin in the Game: Hidden Asymmetries in Daily Life. New York: Random House. Tanzi, V., and Davoodi, H. 1998. “Corruption, Public Investment and Growth.” In H. Shibata et al. (eds.), The Welfare State, Public Investment, and Growth. Tokyo: Springer, 1998, pp. 41–​60. Tanzi, V., and Schuknecht, L. 2000. Public Spending in the 20th Century: A Global Perspective. Cambridge: Cambridge University Press. Tavits, M., and Letki, N. 2014. “When Left Is Right: Party Ideology and Policy in Post-​ Communist Europe.” American Political Science Review, 103(4): 555–​569. Thaler, R. 2016. Misbehaving: The Making of Behavioral Economics. New York: W.W. Norton. Transparency International, 2021. “Corruption by Country.” Berlin: Transparency International. Available at: https://​www.trans​pare​ncy.org/​en/​cpi/​2020/​index/​table (accessed February 2021). Trigger, B. 2003. Understanding Early Civilizations: A Comparative Study. New York: Cambridge University Press. Tucker, J. 2006. Regional Economic Voting: Russia, Poland, Hungary, Slovakia and Russia, 1990–​99. Cambridge: Cambridge University Press. Tullock, G. 1967. “The Welfare Costs of Tariffs, Monopolies and Theft.” Western Economic Journal, 5(3): 224–​232. US Census Bureau. 2016. Income and Poverty in the United States: 2015. Report Number: P60-​256, September 13, 2016, Washington, DC. Useem, M. 1984. The Inner Circle. New York: Oxford University Press. Verdier, D. 1995. “The Politics of Public Aid to Private Industry: The Role of Policy Networks.” Comparative Political Studies, 28(1): 3–​42. Vitali, S., Glattfelder, J., and Battiston, S. 2011. “The Network of Global Corporate Control.” PLoS ONE, 6(10): e25995. Vuković, V. 2019. “Political Economy of Local Government in Croatia: Winning Coalitions, Corruption, and Taxes.” Public Sector Economics, 41(4): 387–​420. Vuković, V. 2020. “Corruption and Re-​Election. How Much Can Politicians Steal before Getting Punished?” Journal of Comparative Economics, 48(1): 124–​143.

356 References Vuković, V. 2021. “The Politics of Bailouts: Estimating the Causal Effect of Political Connections on Corporate Bailouts during the 2008–​2009 US Financial Crisis.” Public Choice, 189: 213–​238. Wallerstein, I. 1974. The Modern World-​System I: Capitalist Agriculture and the Origins of the European World-​Economy in the Sixteenth Century. New York: Academic Press. Walsh, M.W. 2009. “A.I.G. Lists Banks It Paid with U.S. Bailout Funds.” New York Times, March 19, 2009. Available at: https://​www.nyti​mes.com/​2009/​03/​16/​busin​ess/​16res​ cue.html (accessed March 2017). Weber, M. 1978 [1922]. Economy and Society. Oakland: University of California Press. Welch, S., and Hibbing, J. 1997. “The Effects of Charges of Corruption on Voting Behavior in Congressional Elections, 1982–​1990.” Journal of Politics, 59(1): 226–​239. Wharton Research Data Service (WRDS). 2017. BoardEx Database—​North America. The Wharton School, University of Pennsylvania, PA. Available at: https://​wrds​web.whar​ton. upenn.edu/​wrds/​ (accessed continuously between September 2017 and February 2018). Winters, M.S., and Weitz-​Shapiro, R. 2013. “Lacking Information or Condoning Corruption: When Do Voters Support Corrupt Politicians?” Comparative Politics, 45(4): 418–​436. World Bank. 2020. Gini Index (World Bank Estimate). Development Research Group. World Bank, Washington, DC. Available at: https://​data.worldb​ank.org/​indica​tor/​ SI.POV.GINI (accessed November 2016). World Inequality Database (WID). 2020. Top Decile Share of Pre-​Tax Income for US, UK, France, Germany, China, Venezuela, Vietnam. Available at: https://​wid.world/​ (accessed September to December 2020). Wu, H., Li, S., Ying, S.X., and Chen, X. 2018. “Politically Connected CEOs, Firm Performance, and CEO Pay.” Journal of Business Research, 91: 169–​180. You, J.S., and Khagram, S. 2005. “A Comparative Study of Inequality and Corruption.” American Sociological Review, 70(1): 136–​157.

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–​53) may, on occasion, appear on only one of those pages. accountability, xiii, 202, 203, 273, 274, 287–​88, 289–​90, 294, 295, 296 political, 278–​79, 280, 281, 283, 286–​87, 307 accumulation of power, 25, 29–​30, 40–​41, 232, 280, 291, 313 Acemoglu, Daron, 16, 18–​19, 215, 271–​72 Acemoglu and Robinson, 16, 18–​19, 271–​72 narrow corridor (NC), 16, 258, 271–​72 Agricultural Revolution, vii, 17–​18, 21–​22, 71, 72, 73, 75, 81–​82, 189–​90 ancient times, 12, 77–​78, 79, 82–​83, 85–​86, 91–​92, 191 Ansell, Ben, xv, 43 assortativity, 9, 55, 58–​59, 169, 177, 179 positive, 44, 172, 175 Atkinson, Tony, 20–​21, 115–​22 Austro-​Hungarian Empire, 243–​44, 248, 249 autocracies, 3–​4, 40–​41, 42, 77–​78, 101–​2, 106–​ 7, 190, 203, 274 Autor, David, 36, 123–​24 bailouts, 110–​11, 224–​25, 226, 227, 228–​29 allocation, 31, 65–​66, 226, 227–​28, 229–​30 Bankman-​Fried, Sam, 5–​6 Bank of America, 51, 57, 135–​36 Bank of England, 55–​57, 224 banks, 4, 37–​38, 47, 54–​55, 127, 171, 224–​ 25, 229–​30 big, 49, 51, 55–​57, 124, 179, 224–​25 bargaining power, 60, 62, 122, 123, 261–​ 63, 314–​15 Bartels, Larry, 129 Becker, Gary, 38, 218–​19 behavior, xiii, 7, 13–​14, 33, 63, 65–​66, 80, 171–​ 72, 198–​99, 204 corrupt, 197, 206–​7, 212 predatory, 14, 194 Bernanke, Ben, 49, 50–​51, 53–​55, 57, 135–​36, 178, 224–​25 betweenness centrality, 46, 53, 55, 58–​59, 174, 175, 176–​77, 179, 180, 182

Bezos, Jeff, 58–​59, 96, 135–​36 Biden, Joe, 310 Big Tech, xiii, 36, 123–​24, 218, 224, 279–​80 billionaires, 57, 126, 165, 267, 268 BlackRock, 58–​59, 222 Blankfein, Lloyd, 57, 58–​59, 135–​36 BoardEx database, xvi, 132–​33, 134–​35, 136–​37, 139–​40, 173–​74 board members, 6, 96, 136, 165, 173, 185 boards, 5–​6, 36, 49, 133–​35, 138, 154–​55, 157–​ 58, 159–​60, 218, 220–​21 Buchanan, James, 198 Bueno De Mesquita, Bruce, 198, 202 Buffet, Warren, 132, 135–​36, 292 Bush, George, 223, 285 campaign donations, 7–​8, 21, 30, 34–​35, 62, 65–​66, 132, 171, 173, 202–​3, 226 capital, 9, 11, 18–​20, 22, 83–​84, 90–​92, 93, 94–​ 95, 96–​97, 240, 315–​16 capital gains, 20–​21, 97, 120–​22, 139–​40, 261 capitalism, 30, 31–​32, 90, 91–​92, 99–​100, 101–​2, 231, 238–​40, 241, 242, 243–​44, 246–​47 early, 9–​10, 240, 244, 248 laissez-​faire, 264–​65 capitalism and democracy, 235–​59 capitalist democracies, 31–​32, 230, 233, 238–​40, 244, 246–​48, 256–​57, 258–​59, 263–​64 Catholic Church, 85–​86, 193, 241–​42 CEOs, 55–​57, 58–​59, 96, 97, 122–​23, 132, 133–​ 34, 135–​36, 164–​65, 173, 217 bank, 57, 178, 226 connected, 165, 217–​18, 220–​21 corporate, 36, 66, 135–​36, 175, 176–​77 former, 50–​51, 57 CFR (Council on Foreign Relations), 48–​49, 53, 55–​57, 58–​59 Chartbook of economic inequality, 115, 117–​20 China, 20, 40–​41, 44–​46, 73–​74, 84, 89–​90, 220–​21, 238–​40, 250–​51, 275 Citigroup, 47–​48, 51–​52, 53, 57, 223, 224

358 Index citizens, vii, 32–​33, 129–​30, 255, 281, 286, 292, 293–​94, 299–​301, 302, 304, 305–​7, 317, 318–​20 civic engagement, 32, 286–​87, 299–​300, 303, 309, 318 Clark, Gregory, 19, 20, 23, 84, 89 clientelism, vii, 196–​99 Clinton, Bill, 47–​48, 51, 57, 223 Clinton administration, 48–​49, 223 clustering, vii, 9, 44–​46, 59, 61, 174–​77, 178–​79, 185, 315, 316 level of, 175–​77 clusters, 9, 43–​44, 46, 47–​49, 51–​52, 53, 172, 173–​74, 175–​76 collective action, 25, 108, 218–​19, 308, 318 collective action problem, 7, 59–​60, 63, 296, 313 collusion, vii, 2–​5, 36–​37, 61, 166, 218–​21 communism, 43, 101–​2, 112–​13, 249, 263–​64 communities, 32–​33, 78, 281, 295, 296, 305, 308–​9, 310, 311, 312–​13, 314–​15, 316, 318–​20 dysfunctional, 310–​11, 312–​13 engagement, 311–​12, 318 functional, 232, 312 local, 152, 295, 296, 308, 314, 315, 317 concentration of power, 9, 16, 17, 71–​72, 84, 108, 169, 258, 296 Congress, 65–​66, 127–​28, 137, 226–​27, 292 Congressmen, 5, 57, 137–​38, 223, 227 connected executives, 30, 49–​50, 66, 69–​70, 133, 138–​39, 145, 164, 217 connected firms, 7–​8, 138–​39, 184, 205, 206, 220–​22, 225–​26, 227–​28, 229–​30 connected individuals, 15–​16, 29, 35–​36, 44–​46, 138, 145, 152, 172, 173, 184, 185 connections, 3–​4, 5–​6, 27, 34–​36, 44–​46, 59, 64, 97–​98, 134–​36, 137–​38, 152, 155–​56, 171, 173–​74, 175–​76, 179–​80, 273–​74 corporate, 164, 227 deep, 49, 215, 218 direct, 166, 173, 227 firm, 153–​54, 158, 162 high number of, 46 individual, 28–​29, 44, 133, 134–​35 industry-​level, 154, 158–​59 low levels of, 135, 175 social, 51, 174, 225, 229–​30 conquest, 17–​18, 83–​84, 86, 241, 242 constraints, 7, 23–​24, 72, 84, 85, 97, 98, 124–​25, 243, 274, 307 institutional, 288, 292–​93, 297 strong, 274, 292, 295 corporate boards, 46, 164, 169–​70, 171, 195

corporate elites, vii, 23–​24, 34–​46, 61, 124–​25, 171 corporate executives, xi–​xii, 2–​3, 7–​8, 36, 46, 47–​48, 49, 96–​98, 164, 165, 166, 173, 230 corporations, i, 15, 38, 169, 213, 222, 291–​92, 316, 317 corporatism, 213–​14, 230, 231, 239–​40 corruption, vii, 39–​41, 65, 195–​201, 204–​6, 207–​12, 273–​74, 283, 312–​13, 317–​18 charges, 4–​5, 205–​6, 288–​89 local, 209, 210 political, 4–​5, 196, 200, 289, 298 punish, 39–​40, 199, 200, 207 Council on Foreign Relations. See CFR countries, corrupt, 216, 217–​18, 273–​74, 283, 284, 285, 286, 288–​89, 298 credible threat, 32–​33, 282, 292, 293–​94, 296, 300, 307, 317, 318–​20 crisis, 47, 49, 51, 110–​11, 112–​14, 140–​45, 193, 224–​26, 227, 229–​30 Croatia, xv, 31, 40–​41, 205, 206, 209, 210, 212, 251, 255 customer-​seeking firms, vii, 9, 31, 169–​70, 184, 185, 216–​17 Davos, 55–​57, 172, 315–​16 Deaton, Angus, 121–​22, 127–​28, 309, 310, 311 degree centrality, 21, 29–​30, 46, 97–​98, 134–​35, 153, 173, 174–​76 democracies, vii, 30, 40, 41–​42, 60, 61–​62, 99–​100, 101–​2, 104–​8, 128, 130, 200, 203, 242–​47, 257–​59 advanced, 7, 289–​90 consolidated, 101–​2 corrupt, 273, 289 developed, 110–​11, 117–​18, 130, 171, 205, 217–​18, 287–​88 direct, 293–​94, 300, 303–​4, 305 new, 245–​46, 257–​58 representative, 196, 290 rich, 40–​41, 116, 120–​21 successful, 245–​46, 274 democracy and inequality, vii, 99–​131 democratic capital, 7, 33, 207, 233, 244, 311–​ 12, 318 democratic capitalism, 31–​32, 193, 232, 244, 257 democratization, 43, 69, 90, 91–​92, 107–​8, 243, 246–​47, 257–​58, 260 developed countries, 114, 124–​25, 165, 199–​ 200, 238, 247, 249, 275–​77 Diamond, Jared, 73–​74 dictators, 2, 60, 61, 62–​63, 98–​99, 202–​3, 306, 320

Index  359 Dimon, Jamie, 55–​59, 135–​36 dispersion of earnings, xi–​xii, 116–​17, 122, 177, 185, 217 distribution of income and wealth, 11–​12, 18–​ 19, 89, 169, 264 distribution of wealth, 17, 18, 29–​30, 71, 82, 89, 91, 98, 174, 179 Downsian model, 105–​6, 200 earnings, 115–​17, 122–​24, 136, 140–​53, 157, 158–​59, 162, 163, 166–​67, 185, 265–​66 abnormal, 9–​10, 21, 29–​30, 35–​36, 164, 260–​61 higher, 97–​98, 124, 160–​61, 164–​65 inequality, 119–​20, 164 Eastern Europe, 101–​2, 130, 207, 248, 249, 254, 255–​56, 257, 309 East India Company, 9–​10, 182, 215 economic growth, 23–​24, 25, 124, 126–​27, 220, 241, 243, 244, 247, 250–​51 effect of political connections, 42, 162, 163, 164–​65 elections, 7–​8, 38, 101–​2, 188, 197, 206, 210, 226–​27, 245–​46, 288, 306–​7 elite network behavior, 36–​37, 39, 42, 59–​60 elite network formation, xiii, 9–​10, 30–​31, 32, 33, 59, 189 elite networks, vii, 5–​8, 26–​28, 40–​42, 44–​46, 61–​62, 64–​67, 83–​85, 86–​90, 93–​98, 132–​ 33, 169–​94, 211–​12 and inequality, 74–​75, 76–​77, 78–​79, 80–​81, 82, 84, 86, 88–​89, 90, 92–​93, 94–​95, 96–​ 97, 98–​99 members, 6, 7–​8, 11, 49–​50, 51, 53, 62, 63, 66, 132–​33, 266 membership, vii, 28, 38, 42, 59, 65, 166–​67, 179, 186, 191–​92, 194 new, 29–​30, 93, 96, 97 power, 25, 130, 232, 289 relationships, 15, 49–​50, 65, 169–​70, 182, 216–​17, 219–​20, 222–​23 superhubs, vii, 177–​79, 180, 221–​22 theory, vii, 16–​17, 23, 31, 34, 36, 172–​85, 205, 206–​7 elite power, i, 1–​33, 90–​100, 130, 192–​93, 258 elites, xii–​xiii, xiv, 3–​4, 16–​17, 19, 20, 23–​24, 25–​26, 41–​42, 43–​44, 66–​167, 172, 257–​58 elite theory, vii, 41, 42 empires, 9–​10, 18–​19, 23, 79–​80, 82, 84, 85, 89–​ 90, 95–​96, 180–​82, 215 empirical findings, xv, 30–​31, 46, 66, 184, 197, 218 England, 55–​57, 86, 87–​88, 89, 90, 91–​94, 182, 224, 243

Enlightenment, xii–​xiii, 14–​15, 20, 88–​89, 193 equality, 12, 29, 31–​32, 231, 243 Europe, 85–​86, 87–​88, 89–​90, 91, 93–​94, 95–​96, 108–​9, 111–​12, 192, 193, 235–​38, 249 European Commission, 267–​69, 283–​84 exchange, voluntary, xii–​xiii, 12, 17, 83–​84, 169, 189–​91, 194 executive earnings, 139–​40, 145, 152, 153, 154–​ 56, 157, 158, 160–​62, 163, 164 executives, 51, 132–​33, 136, 137–​38, 139, 140, 145, 157, 162, 166–​67, 169–​70, 184–​ 85, 217 executive salaries, vii, 42, 140–​52, 163, 164–​67 Ferguson, Niall, 92, 179 finance industry, 37–​38, 124, 125, 127, 222, 226 financial crisis, 6, 59, 65–​66, 110–​11, 119–​21, 222, 224–​25, 226, 229–​30 Fink, Larry, 58–​59 firms, vii, 9–​10, 34–​36, 65–​66, 133–​38, 152, 173–​74, 184–​85, 206–​7, 212–​30 biggest, 123–​24, 185 large, 122, 216 nonconnected, 139, 184, 220–​21 rent-​seeking, 65, 125–​26, 184–​85, 197, 215–​17 superstar, 125–​26 First Lever, vii, 281–​99, 302, 303, 314, 316, 317 forces of occasional destruction, xiii–​xiv, 23, 24, 25–​26, 72, 85, 88–​89, 90, 93, 265 forces of wealth accumulation, 31–​32, 169, 190–​ 91, 192–​93, 244, 265 forces of wealth concentration, xiii, 23–​26, 69–​ 70, 72, 85, 87–​90, 93–​94, 97 France, 39–​40, 91–​92, 93–​94, 95–​96, 108–​9, 115–​16, 238–​39, 243–​44, 249, 251, 255–​56, 261–​63, 264, 266–​69 Frankopan, Peter, xvi, 79 fraudulent procurement contracts, 31, 64, 65, 195, 197, 204 free exchange principle, 13, 92–​93, 96–​97, 190, 191, 193, 194 French Revolution, 95–​96, 243–​44 Friedman, Milton, 112 Fukuyama, Francis, 7, 123 Furman, Jason, 36, 126 Geithner, Timothy, 49, 50–​53, 57, 58–​59, 178, 223, 224–​25 appointment, 51, 225 connections, 51–​52, 53, 57 Germany, 87–​89, 108–​9, 111–​12, 115, 117–​19, 249, 255–​56, 257, 261–​63

360 Index Gerschenkron, Alexander, 247 gift exchange principle, 12, 13, 59, 189, 191 Gilens, Martin, 129–​30 Gini coefficients, 76, 88–​89, 106–​7, 117–​ 18, 275–​77 Goldman Sachs, 2–​3, 47–​48, 49, 50–​51, 53, 57, 135–​36, 171, 223–​24 governing institutions, vii, 16, 32, 197, 272–​ 73, 318–​20 government contracts, 15–​16, 30, 34–​35, 65, 132, 215–​16, 217 government intervention, 111, 112, 311–​12 governments, 1, 2–​3, 15, 37–​38, 64, 65–​66, 108, 110–​13, 114–​15, 136, 191, 215–​16, 217–​18, 274, 275–​77, 288, 293–​94, 295–​96, 306, 311–​12 Greenspan, Alan, 224 groups, 7, 25, 37–​38, 39–​41, 42–​44, 59–​61, 62–​ 64, 73, 108–​9, 127–​28, 202 narrow, 44–​46, 62, 120, 129–​30, 169, 171–​72, 204, 272, 282 organized, 7, 40–​41, 59–​60 religious, 28–​29, 30, 66, 132–​33, 308, 309 small, 4–​5, 9, 80–​81, 176–​77 underrepresented, 318–​20 Harari, Yuval, 74, 78–​79, 80 Harvard Center for History and Economics, 180 Hayek, Friedrich, 112, 320 hierarchical orders, vii, 77–​81, 320 hierarchies, social, xii–​xiii, 20–​21, 97, 98, 105 Holmes, Elizabeth, 5–​6 Homo sapiens, 13–​14, 72, 73–​74 human history, xi, 14–​15, 16, 64–​65, 66–​67, 69, 97–​99, 190, 191–​92, 193 human rights, 31–​32, 101–​2, 189–​90, 191–​92, 231, 244, 246, 306 hunter-​gatherers, vii, 72, 73–​74, 75–​77, 81–​82 IMF (International Monetary Fund), 50–​51, 53, 54–​55, 108–​9, 179 incentives, xii, 11–​12, 15–​16, 32, 42–​43, 77–​78, 184, 198–​99, 201–​2, 231, 247–​48, 260, 274–​75, 277, 278–​79, 289–​90, 317–​18 for elite network formation, xiv, 32, 33, 317 strong, 14, 25, 38, 43, 44, 79, 197, 198, 232–​ 33, 293–​94, 296 income and wealth inequality, 71, 99–​100, 104, 105–​6, 238–​39, 263 income distribution, xi, 7–​8, 11–​12, 18–​19, 20–​ 21, 29, 57, 69–​70, 89, 105–​6 income inequality, i, 27, 66, 115–​16, 117–​20, 128–​ 30, 139–​40, 165, 166, 263, 267–​68, 269–​70

incomes, 7–​8, 11–​12, 18–​21, 29, 34–​35, 36, 83, 94–​96, 105–​6, 116–​17, 118–​21, 125, 215–​ 16, 260–​61, 267–​69, 270 capital-​based, 95, 96–​97 executive, 124, 139–​45 higher, 36–​37, 260, 263, 309, 316 income taxes, 263–​64, 267, 268–​69, 300–​ 1, 302–​3 Industrial Revolution, 13–​15, 22–​24, 29–​30, 90–​ 94, 193, 230, 231, 239–​40, 241, 242, 248 inequality, vii, 7–​8, 11–​12, 21–​24, 25–​26, 34–​37, 66–​73, 87–​89, 98–​131, 230–​32, 238–​40, 260–​63, 265–​66 between-​firm, 36, 126, 185 in democracies, vii, 106–​7, 115–​20 economic, xiii, 44–​46, 82, 115, 117–​20 high, 37, 59, 117–​18, 128, 232, 275 higher, 27, 37, 69–​70, 76–​77, 106–​7, 117–​18, 125, 203 literature, 7–​8, 29–​30, 34, 185, 279 lower, xiv, 22–​23, 32, 83, 104–​6, 108, 265–​66, 274, 280 rising, 7–​8, 14, 27, 29–​30, 97, 104, 126–​27, 128, 129, 130 trends, 22–​23, 25–​26, 91, 105, 116–​17, 132–​ 33, 238–​39 informational asymmetries, 44–​46, 123, 174, 200, 201, 279 inheritance, 11, 13, 19–​20, 76, 91–​92, 94–​96, 191, 194, 261 innovations, disruptive, 10–​11, 92–​93, 215, 231 institutions, 12–​13, 16, 39–​41, 130, 171–​72, 189–​90, 212, 257–​58, 271–​72, 273–​74, 290–​92, 301–​2 personalized, 273–​74, 275 strong, 7, 101, 274 interest groups, 25, 37–​38, 40–​41, 42, 63–​64, 108, 127, 129–​30, 177, 218–​19 powerful, 3–​4, 28–​29, 127–​28, 176 special, 25, 28–​29, 128–​29 state capture, 7, 104, 108, 214, 245 International Monetary Fund. See IMF interpersonal trust, 232, 275–​77, 308, 311–​12, 317, 318 Japan, 20, 39–​40, 44–​46, 89, 104, 108–​9, 111–​12, 115–​16, 117–​19, 165 jobs, 2–​3, 94–​95, 96, 121, 122, 135–​36, 204, 222, 223, 271–​72, 273, 301, 302 agricultural, 241, 251 executive, 137–​38 industrial, 251 labor-​intensive, 213–​14

Index  361 low-​skilled, 121 previous, 134–​35, 156 Johnstone family, 180–​82 JP Morgan, 9–​10, 51, 52–​53, 57, 135–​36, 223 judges, 4, 180–​82, 273–​74, 290, 292, 308–​9 corrupt, 288–​89 justice system, 206, 257–​58, 274, 314 Kahneman, Daniel, 122–​23 Keynesian, 111, 112 key performance indicators (KPIs), 32–​33, 282, 294 kings, 19–​20, 61, 85–​87, 91, 93, 95–​97, 98–​99, 180, 243 Kissinger, Henry, 5–​6, 53, 195 Krueger, Anne, 50–​51, 54–​55 Kuznets, Simon, 22 Kuznets theory, 22, 107–​8 curve, 232 first wave, 22 second wave, 22 waves, 22–​23 labor, 11, 18–​19, 20, 22, 84, 93, 94–​96, 192, 213 cheap, 104–​5, 240 exploited, 244 market, 29, 293 slave, 18, 83–​84 unions, 40–​41, 60, 62, 121–​22, 176, 309 land, 18, 19–​20, 75–​76, 80–​81, 82, 83–​84, 85–​ 86, 91–​92, 93, 94–​95, 97–​98 agricultural, 284 ownership, 29–​30, 76, 91–​92, 94–​95, 96–​97 owners of, 11, 19–​20, 69–​70, 90 Latin America, 3, 40, 101–​2, 112, 124–​25, 130, 248 law, 4–​5, 31–​32, 171–​72, 231, 270, 279–​80, 286, 289, 290, 291, 314, 315–​16 leaders, 4, 64, 78, 79, 81, 98, 202–​3, 271, 273, 297, 313 political, 79, 198, 270–​71 legislation, 7–​8, 11, 34–​35, 130, 171, 217–​18, 219, 270, 272, 279–​80, 299 favorable, 31, 34–​35, 38, 40–​41, 127, 169–​ 70, 202–​3 legislators, 3, 137, 160, 218–​19, 224–​25, 270, 298–​99 legitimacy, xii, 13–​14, 71, 86, 191–​92, 194, 244, 292–​93 levers, vii, 33, 34, 232–​33, 278–​81, 296, 299–​ 300, 316–​17, 318–​20 Lew, Jack, 51–​53, 57, 135–​36, 223 limited access orders, 16–​17, 273

links, i, 46, 55–​57, 128, 129, 134–​35, 137, 164, 174–​75 direct, 65–​66, 86, 163, 164–​65 random, 174–​75 superhub, 59 living standards, xi, 3–​4, 82, 83, 90, 92–​93, 94–​ 95, 99, 101–​2, 235, 239–​40 lobby groups, 60–​61, 62 lobbying, xvi, 10–​11, 65–​66, 125, 127–​28, 216, 217, 219, 298–​99 activities, 7, 127, 218 efforts, 127–​28, 226 expenditures, 127 finance industry, 171 lobbyists, 37–​38, 62, 63–​64, 219, 220, 221–​22, 265–​66, 298–​99 local governments, xv, 2–​3, 32–​33, 203, 204, 282–​83, 289–​90, 295, 297, 312, 313 Madoff, Bernie, 6, 7 Malthusian, xi, 72, 85, 90, 98–​99 economic model, vii, 81–​85 subsistence, 244 times, 87–​88, 90, 241 trap, 14, 18–​19, 23–​24, 83, 90, 192–​93, 246–​47 managers, 28–​29, 36, 92, 96–​97, 136 society of, vii, 20, 90–​100 market, xiii, 15, 98, 112–​13, 184–​85, 215–​17, 218, 225, 260–​61 dominance, xiii, 10–​11, 15, 220 economy, 111–​12, 264–​65 failures, 260–​61, 291 financial, 20, 96, 97, 178 mechanism, 9, 11, 174, 191 political, 185, 198 power, i, 15, 34–​35, 164, 216–​17 Marx, Karl, 91–​92, 240, 249 Maslow, Abraham, 177–​78, 189 Mauro, Paolo, 108–​9 mayors, 137–​38, 204, 206, 209, 210, 287 corrupt, 5–​6, 203 local, 3, 207–​8, 210, 287 McChesney, Fred, 1, 3–​4, 128–​29, 219 mechanism, 7–​8, 9, 29–​30, 177, 179, 184–​85, 200, 202, 203, 204, 300 democratic trial-​and-​error, 32, 128, 277, 280 violence power, xii–​xiii, 14 well-​defined institutional, 21, 232 media, 6, 96, 200, 244, 278–​79, 284, 285, 286–​ 87, 289–​90, 302 captured, 285, 286 free, 32–​33, 203, 281, 285, 286, 317, 318–​20 independent, vii, 284–​85, 286–​87

362 Index Meltzer, Allan, 105–​6, 128–​29 members, 25, 28, 41–​42, 59–​60, 61–​64, 66, 132–​ 34, 135–​36, 137–​38, 166, 182, 201, 202–​3, 207–​8, 210, 308–​9, 312–​13, 318 active, 53, 136–​37 connected, 176, 201 of elite networks, 62, 82, 185 of parliament, 137, 164, 182, 206 Menaldo, Victor, xv, 257–​58 mercantilism, 85, 240, 242 Middle Ages, vii, 13–​14, 19, 22–​23, 85–​86, 87–​ 88, 89, 189, 191, 193, 235–​38 middle classes, 43, 90, 91–​93, 95, 104, 121, 123, 129–​30, 213–​14 Milanović, Branko, 19, 22, 23, 71–​72 minimum winning coalitions, vii, 195, 202, 205–​12, 287, 307 misuse, 4–​5, 10–​11, 14, 35–​36, 171–​72, 197, 212, 219, 278–​80, 298 of political power, xiii, 36–​37, 232 of power, 7–​8, 11, 15, 35–​37, 166, 274 money, 5–​6, 59–​60, 186, 188, 241–​42, 282–​83, 298–​99, 300–​1, 302–​3 laundering, 1–​2, 5 for nothing, 128–​29, 219 and power, 60–​61, 62–​63 taxpayer, 302–​3 monopolies, xii, 3, 7–​8, 16, 86, 96–​97, 126, 193, 216–​17, 296 natural, 215–​16, 282–​83 monopoly rights, 10–​11, 215, 240, 295 motivation, 169, 189–​90, 192–​93, 194, 201, 226, 230, 283, 284, 286 for corruption, 206, 283 for politicians, vii, 195–​212 municipalities, 37–​38, 206–​10, 287 Murdoch, 58–​59 James, 57 Rupert, 285 Navidi, Sandra, 34, 55–​57, 177 network centrality, 46, 47, 49, 173, 188 network effect, xiv, 9, 21, 66, 97, 157, 160, 162, 163, 166 networks, 44–​46, 47–​49, 51–​53, 55–​57, 58–​59, 97–​99, 133–​36, 171–​75, 176–​78, 180–​82, 187–​89, 201–​2 close, 38, 185, 196, 299 homophily, 9, 30–​31, 169, 176, 178–​79, 182, 315 of influence, 21, 29–​30, 64–​65 larger, 54, 135, 139, 152, 163 political, 98–​99, 177–​78

powerful, 5, 57, 171–​72, 177, 200 superhubs, 29 topocratic, 9, 30–​31, 44–​46, 59, 139, 169, 172, 175 network size, vii, 134–​36, 139, 152, 154–​55, 156, 157–​58, 159, 161, 162 effect, 155, 158 firm, 157–​58, 162–​63 network theory, 30–​31, 46, 169, 184, 189, 203 NGOs, 21, 28–​29, 51, 60, 62, 63–​64, 66, 136–​37, 160, 166 nobility, xii, 7–​8, 9–​10, 19–​20, 85–​86, 89–​90, 91, 182, 240, 242, 243 nobles, 3–​4, 43, 69–​70, 84, 86, 87, 93, 96–​97, 98–​99, 189, 191–​92 nodes, 21, 46, 53, 54–​55, 173–​76, 177–​78, 180, 182, 188, 316 connected, 173, 174, 222 highest-​degree, 175, 177–​78, 188–​89 important, 55–​57, 173, 177 low-​degree, 175, 188 Obama, Barack, 49, 51, 53, 223 Obama administration, 47–​48, 51, 53, 223, 292 office-​holders, 206–​7, 224–​25, 226, 282, 287, 292–​94, 317 officials, elected, vii, 287, 288, 291, 305–​7 Olson, Mancur, 7, 18, 25, 28–​29, 59–​60, 77–​78, 128–​29, 213, 220 Olson’s interest group theory, 28–​29, 41–​42, 108 Olson’s logic of group behavior, 25, 29, 41, 43, 218–​19 open access orders, 16, 17, 273 organizations, 4, 60, 132–​38, 145–​52, 156–​57, 160–​61, 302 alumni, 136–​37, 227 charity, 62, 135–​36, 302, 308 lobby, 63, 315–​16 military, 64, 136–​37 professional, 30, 132–​34, 136–​37, 166 religious, 17–​18, 42, 62, 63–​64, 136–​37 social, 16, 241 Orszag, Peter, 36, 47–​48, 126, 135–​36, 223 owners of capital, xii, 69–​70, 90, 93, 94, 165, 179, 240, 244 Paulson, Henry, 49, 50–​51, 52–​53, 57, 58–​59, 178, 223, 224–​25 network, 50–​51 Peltzman, Sam, 218–​19, 220 Phelps, Edmund, 213 Piketty, Thomas, 19–​21, 23, 36, 37, 91–​92, 93–​ 96, 118–​19, 120–​21, 123, 261–​63

Index  363 POLCON, 136, 145, 152–​53, 154–​56, 159, 161 policies, 33, 47, 129–​30, 198–​99, 224–​25, 259, 260, 261, 288, 289–​90 austerity, 110–​11, 112–​13 changes, 123, 130, 265, 272, 277, 278–​79 fiscal, 112–​13, 293–​94 monetary, 293, 294 optimal, 198–​99, 261–​63, 266, 305 regulatory, 128–​29, 218 policy complexity, 123, 294, 296 political connections, 5–​6, 35–​36, 41–​42, 65–​66, 125–​26, 136, 138–​39, 145–​52, 154–​56, 157–​58, 159–​63, 164–​66, 215–​16, 217–​18, 220–​21, 226–​27, 228–​29 of board members, 165 of corporate executives, 7–​8, 132, 260–​61 direct, 155, 158, 166–​67 firm, 65–​66, 138, 145, 158, 164 and network size, vii, 139, 154–​55 political elites, 37–​38, 41, 92, 130, 197, 202, 212, 257 political influence, 6, 11, 98, 127, 184, 255, 286 political institutions, 26, 206, 299 political orders, 16–​17, 23–​24, 42, 72, 190 political parties, 15, 42, 62, 63–​64, 126–​27, 129, 136–​37, 166, 176, 205–​6 political power, xiii, 3–​4, 9–​10, 11, 15–​16, 18–​ 19, 21, 23–​24, 31–​32, 41–​42, 69–​70, 82–​83, 96–​97, 225–​26, 270–​72, 278, 299 centralized, i, 33, 295 discretionary, 282, 291–​92 extractive, i, 7–​8, 192–​93, 274 lowering, xiv, 31–​33, 194, 232, 274, 280, 281, 297, 298, 299 political process, 10–​11, 16–​17, 34–​36, 125, 128, 130, 196, 219, 265–​66, 270 political protection, xiii, 6, 165, 166, 214, 215, 216–​17 political rents, 34–​35, 184, 198–​99, 201–​2 political survival, vii, 26, 31, 38, 65, 197, 202, 204, 212 political systems, 8, 39, 98–​99, 101–​2, 120, 126–​ 27, 128, 164–​65, 182, 206–​7 politicians, 30–​31, 37–​38, 42, 64–​66, 136–​37, 166–​67, 195, 196–​99, 200–​1, 202–​3, 204–​5, 206–​8, 210, 218–​19, 227–​29, 271–​72, 288–​ 90, 292–​94 bad, 198, 199, 205 corrupt, 65, 199–​200, 287–​88, 289–​90 elected, 7–​8, 69–​70, 169–​70, 180 and firms, 7–​8, 27, 29, 194, 201, 206, 220, 317 former, 5–​6, 53, 54–​55, 218 incumbent, 197–​98, 199–​200

local, 206, 210, 295, 297 office-​holding, 5–​6, 30, 65, 169, 171, 282, 317–​18 politicians in power, xi–​xii, 28–​29, 129–​30, 132, 185, 197, 198–​99, 207, 210, 287–​89 position of power, 3–​4, 43, 44–​46, 55–​57, 59, 169–​70, 171–​72, 222–​23, 225, 226, 229–​30 poverty, xi, 3–​4, 13, 101–​2, 105, 115, 118, 119–​ 20, 233, 235–​38, 239–​40 declining, 238, 239–​40 extreme, 235–​38 reduced, 32, 115 power, 7–​8, 13–​17, 25, 28–​29, 34–​37, 40–​42, 59, 71–​72, 177–​78, 191–​93, 197, 210, 257–​59, 270–​72, 287–​88, 307, 312–​14, 318–​20 corporate, xi, 98–​99, 125–​26, 233, 296 discretionary, 65, 287, 292 holding, xii, 82, 196–​97, 272, 274, 287, 307 relative, 29, 60, 287 rising, 91–​92, 124–​25, 220 structures, 16, 81, 98, 189–​90 privatization schemes, 124–​25, 219–​20, 257 privileged access, 2, 34–​35, 123, 174 privileged position, 11, 15, 98, 184 privileges, xiv, 9, 10–​11, 23–​24, 71–​72, 83, 91–​ 92, 180–​82, 192, 202, 229–​30 procurement contracts, 34–​35, 65, 98, 204, 206–​7, 209, 212, 217–​18, 225–​26, 279–​80 corruption, 207–​10, 312–​13 exclusive, 2–​3, 31, 38, 169–​70, 171, 205–​6 public, 65, 197–​98, 204, 206–​7 suspicious, 26, 207–​9, 210 productivity, 8, 21, 99, 184, 185, 213–​15, 235 growth, xiii, 104–​5 total factor, 214–​15 progress, xii–​xiii, xiv, 12, 104–​5, 214–​15, 230, 231, 247, 248, 250–​51, 255–​57 social, 31–​32, 249, 272, 275–​77, 318 progressive taxation, vii, 32, 260–​70 protection, xiii, 15–​16, 77–​78, 81, 82, 84, 189–​ 90, 215–​16, 217 demand for, 18, 36, 75, 78, 81 proximity to political power, xiii, 13, 18–​19, 21–​ 22, 69–​70, 96, 97, 98, 225–​26 public goods, 30, 32–​33, 198, 199, 204, 205, 296, 303–​4, 312, 313, 314–​15 necessary, xii–​xiii, 3, 190–​91, 274, 296, 305 provision, 281, 292, 295, 300, 308 public officials, 46, 49–​50, 53, 54–​55, 57, 59, 287, 289, 291 public service, 47, 52–​53, 202–​3, 223, 224, 230, 271, 274, 300, 302–​3, 315

364 Index punishment, xiv, 171–​72, 188, 200–​1, 202, 203, 292–​93, 294, 295, 296, 306–​7 Putnam, Robert, 309 Rajan, Raghuram, 308–​9, 311 Reagan, Ronald, 47–​48 administrations, 112–​13 reducing political power, vii, 231, 232–​33, 280, 281 re-​election, 199–​200, 204, 205, 206, 207–​8 reforms, 32–​33, 124–​25, 194, 232, 233, 264–​65, 278, 282–​83, 297, 298, 303, 318–​20 regulations, 30, 31, 128–​29, 130, 132, 218–​19, 230, 279, 295–​96, 311–​12 favorable, xi–​xii, 7–​8, 34–​35, 171, 196, 218–​ 19, 220 rent-​extraction, vii, 3, 11, 18–​19, 31, 85, 195–​ 99, 219 rentiers, society of, vii, 20, 90–​100 rents, 10–​11, 19–​20, 34–​36, 94–​96, 169–​70, 171, 197–​98, 199, 261 extracting, vii, 31, 169–​70, 185, 195–​212 rent-​seekers, xiii, 9, 31, 126, 169–​70, 184–​85, 216–​17, 218 rent-​seeking, vii, 10, 11, 31, 34–​35, 36, 50–​51, 84, 216–​18, 219 Richard, Scott F., 105–​6, 128–​29 robber barons, 4, 7, 9–​10, 295 Robin Hood Foundation, 58–​59, 135–​36 Robinson, James, 16, 18–​19, 271–​72 role of government, 111, 114, 115 roving bandits, 18, 77–​78, 81, 189–​90 Rubin,Robert, 47–​49, 50–​53, 59, 223 appointment, 51 network, 47–​48, 49 rule of law, 31–​32, 101, 231, 279–​80, 314 rules, 79–​80, 89–​90, 130, 189–​90, 212, 257–​58, 272, 288, 290, 291, 292–​93, 294, 305–​6, 307, 318–​20 ruling elites, xiv, 13–​14, 16, 17, 18–​19, 69, 71–​ 72, 77–​79 Saez, Emmanuel, 36, 120–​21, 123, 261–​63 salaries, 97, 122, 124, 139–​52, 154–​55, 156–​58, 159–​60, 161–​62, 163, 164–​65 higher, xi–​xii, 27, 30, 66, 69–​70, 145, 152, 154–​55, 157–​59, 217 Scheidel, Walter, 18–​19, 23, 37, 82, 87–​89, 93, 118–​19, 265 Second Lever, vii, 32–​33, 286, 300, 313–​ 14, 318–​20 selectorate theory, vii, 31, 195, 201–​5, 287 Shiller, Robert, 10

Smith, Adam, 43, 182, 202 social capital, 26, 232, 296, 309, 311–​12 socialism, 23, 90, 91–​92, 111, 238–​39, 246–​48, 249–​51, 254–​55 socialist countries, 31–​32, 98–​99, 101–​2, 238–​ 39, 250–​51, 254, 256–​57, 264 social mobility, 19, 20, 29, 41–​42, 84, 89, 104, 240 social networks, 44, 64, 136, 166–​67, 172, 174–​ 75, 185, 227, 229–​30 informal, i, 7–​8, 34 societies, xiii–​xiv, 11–​12, 13–​14, 16–​17, 25–​26, 28, 29–​30, 37, 41–​43, 85–​86, 88–​89, 105–​6, 190–​91, 192–​94, 241, 272, 277, 318 ancient, 82 developed, 22, 249, 312–​13 dysfunctional, 277, 281, 318 hunter-​gatherer, 18–​19, 72 modern, xi, 23–​24, 231, 238 preindustrial, 19, 23–​24, 71–​72, 83–​84, 90 Soros, George, 55–​57, 135–​36 Soviet Union, 101–​2, 238–​39, 247, 249, 251, 254–​55, 256–​57 special interests, 14, 42, 124–​25, 128–​30, 196, 197 state failures, xiii–​xiv, 23, 87–​88, 99–​100 stationary bandit, 18–​19, 77–​78, 81, 83, 85–​86, 87–​88, 89, 98–​99, 189–​90, 194 Stiglitz, Joseph, 36, 126 superhubs, 5, 34, 46, 47, 55–​57, 173–​74, 177, 224 supermanagers, 8, 21, 29–​30, 37, 66, 69–​ 70, 118–​19 Tammany Hall, 4, 7, 9–​10, 93, 182, 240, 258 Tanzi and Schuknecht, 113–​14 TARP (Troubled Asset Relief Program), vii, 49, 65–​66, 224–​25, 226–​30 Tax allotment, 281, 302, 317, 318–​20 taxation, 128, 232, 242, 259, 263–​64, 269 taxes, 128–​29, 260, 261, 263, 266–​67, 268–​69, 272, 293–​94, 295, 296, 297, 298 tax evasion, 261–​63, 265–​66, 267–​69, 270 tax payments, vii, 281, 300–​1, 302 tax rates, 207–​8, 211–​12, 263, 297 higher, 211–​12, 261, 263, 265–​66, 268–​69 marginal, xiv, 123, 261–​63, 266–​67, 268–​69 tax revenues, 207–​8, 242, 268–​69, 300–​1, 302 technological progress, 8, 18–​19, 75, 121, 123, 214–​15, 257, 309 Third Lever, vii, 32–​33, 295, 308–​16, 318 top executives, 11, 27, 34, 123, 124, 164, 185, 260–​61, 317–​18

Index  365 top income earners, 7–​8, 30, 34, 36–​37, 116, 118–​19, 122, 130, 260–​61 top incomes, 21, 42, 49–​50, 97, 98, 116–​17, 119–​ 20, 122, 124, 139–​40, 145 total earnings, 30, 66, 140–​52, 153, 154–​55, 157, 159, 161–​62, 163 transparency, 245–​46, 278–​80, 283, 284–​85, 286–​87, 289–​90, 295, 296, 317 trial-​and-​error, vii, 246–​59, 280 mechanism, xiv, 26, 31–​32, 280, 313 process, vii, 232, 233, 245–​46, 258–​59, 264–​ 65, 275, 277 Troubled Asset Relief Program. See TARP Trump, Donald, 57, 274, 309–​10 Trump administration, 223, 274 trust, vii, 12, 32, 33, 98, 178–​79, 241–​42, 275–​ 77, 311–​12 declining, 32 in public institutions, 272, 275, 277, 281, 282, 284, 296, 308, 317 rebuilding, xiv, 26, 275, 277, 280, 282, 284, 287, 317 social, 266, 280, 313, 318 Tullock, Gordon, 198, 219 unemployment, 111, 112, 115, 244, 254, 257, 293, 309 unequal distribution of wealth, 66–​67, 69, 75, 190–​91 unions, 60–​61, 62, 63–​64, 121–​22, 202–​3, 244, 287, 314 United Kingdom, 27, 109, 116–​20, 124, 133–​34, 137–​39, 140–​52, 159–​64, 165, 166, 238–​39, 251, 261, 264–​65, 285 United States, 44–​46, 93–​94, 104–​5, 111–​13, 116–​20, 123–​24, 127–​30, 133–​34, 137–​38, 140–​52, 154–​55, 157–​62, 166, 238–​39, 261, 274, 309–​10 use of violence, xii, 13–​14, 16–​17, 18, 25, 81, 194

violence, xi, 13–​15, 16, 18, 73, 77–​78, 80–​81, 83–​84, 189, 190, 191–​92 violence power principle, xii, 12, 13–​14, 15, 16–​ 17, 30–​32, 89–​90, 191–​94 voluntary exchange principle, xii–​xiii, 12, 14–​ 15, 16–​17, 96–​97, 99, 189, 190–​91, 193 voters, 128–​30, 196, 197, 198–​201, 202–​3, 204, 205, 206, 207, 294, 314, 315 median, 105, 106–​7, 257–​58 Vuković, Vuk, i, 1–​2, 34–​35, 71–​72, 101–​2, 132–​ 33, 171–​72, 195, 207–​8, 209–​10 wage premium, 124, 132–​33, 153, 158, 160, 166 wages, 83, 94, 95–​96, 97, 101–​2, 121–​22, 140–​ 52, 154–​55, 156, 166–​67, 314 Wall Street, 5–​6, 47, 48–​49, 52–​53, 223, 224 wealth, 11–​13, 14–​15, 17–​20, 23–​24, 35–​36, 66–​67, 69, 71–​72, 75, 76, 85, 89–​90, 91–​92, 93–​97, 98–​100, 126, 191 wealth accumulation, 14–​15, 17, 31–​32, 85, 190–​91, 192–​93, 230, 232, 233, 265, 297 wealth concentration, xiii, 8, 9, 23–​26, 69–​70, 82–​83, 85, 87–​90, 95–​96, 238 wealth inequality, 2, 11, 19, 87–​88, 95–​96, 98–​ 100, 104, 105–​6, 118, 119–​20 welfare state, 20, 23, 99–​100, 104, 105–​6, 108, 111–​12, 113, 115, 263, 264–​65 Wharton Research Data Service. See WRDS WID. See World Inequality Database winning coalitions, 202–​3, 210, 287–​88, 312–​13 powerful, 203, 212, 287–​88 small, 31, 202, 287–​88, 306 World Economic Forum, 49, 58–​59, 172 World Inequality Database (WID), 238–​39, 261–​63, 267–​68 World War II, 23, 28–​29, 110–​11, 118–​19, 215, 239–​40, 249, 258–​59, 263–​64 WRDS (Wharton Research Data Service), xvi, 133 W/​S ratio, 202, 210, 211–​12