Trust in a Polarized Age [1 ed.] 2020018600, 2020018601, 9780190887223, 9780190887247

Citation preview

Trust in a Polarized Age

Trust in a Polarized Age K EV I N VA L L I E R

1

3 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 2021 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: Vallier, Kevin, author. Title: Trust in a polarized age /​Kevin Vallier. Description: New York, NY : Oxford University Press, [2021] | Includes bibliographical references and index. Identifiers: LCCN 2020018600 (print) | LCCN 2020018601 (ebook) | ISBN 9780190887223 (hardback) | ISBN 9780190887247 (epub) Subjects: LCSH: Political culture—​United States. | Trust—​United States. | Polarization (Social sciences)—​Political aspects—​United States. | Civil society—​United States. Classification: LCC JK1726 .V353 2021 (print) | LCC JK1726 (ebook) | DDC 306.20973—​dc23 LC record available at https://​lccn.loc.gov/​2020018600 LC ebook record available at https://​lccn.loc.gov/​2020018601 1 3 5 7 9 8 6 4 2 Printed by Integrated Books International, United States of America

In Memory of Jerry Gaus

Contents Acknowledgments  About the companion website 

ix xi

Introduction: Trust and Polarization 

1

1. Must Politics Be War Here and Now? 

20

2. Social and Political Trust: Concepts, Causes, and Consequences 

49

3. Civil Society and Freedom of Association 

88

4. The Market Economy 

119

5. The Welfare State 

139

6. Against Egalitarianism 

173

7. Democratic Constitutionalism 

210

8. Elections and Process Democracy 

237

Epilogue: How Is Trust Restored? 

277

Works Cited  Index

281 299

Acknowledgments Trust is a treasure. It allows people to cooperate and form relations of love and friendship that would otherwise not be possible. That is also true for me. I have profited from the relations of love and friendship that made this book possible. The work has taken several forms since I wrote the first draft in 2015, and this is the second of two books that grew out of that original manuscript. Writing two books instead of one has required me to rely even more on family, friends, and my professional relationships. So I am eager to thank everyone who helped me. As is usual with my books, I first thank my wife, Alicia, who made the entire trust project possible by being an incredible wife and mother, and supporting me throughout the writing process. This was especially true during the COVID-​19 pandemic, when finishing the book became harder, and parenting young children harder still. And all the while, she has continued to push me to see the humanity in those I have failed to trust as much as they deserve. I am grateful to my mother as well, whose unyielding support has been a constant throughout my life. I am indebted to everyone who worked for or with Oxford University Press in helping me get this book under contract and ready for publication. I am most grateful to my incredible editor, Lucy Randall, who has been consistently helpful and wise, and Hannah Doyle for her constant support throughout production. I am grateful to Andrew Lister, Christie Hartley, Bob Talisse, and anonymous referees for the book, and to Bob for convincing me to split the original project into two book projects. I thank Kayla Fox for helping me put together the index of the book. I am grateful to Paul Stevenson Photography for the photo of the Hands Across the Divide Statue that is the basis for the cover of the book. And I am grateful to Jessica Pellien and Mark Fortier at Fortier Public Relations for their input on the book’s design and marketing. I am also grateful to Alexander Cohen, who commented on and copyedited the entire book, as well as Mark Laufgraben, who was essential in the page-​proofing process. I thank everyone in my department at Bowling Green State University who helped me move the project forward. I am especially thankful for my graduate students, whose input into the project was invaluable, especially Colin Manning, Ryan Fischbeck, and Ami Palmer, as well as Will Lugar. I am also grateful for help from my colleagues Rishi Joshi, Brandon Warmke, and Michael Weber. But I am most grateful to Margy DeLuca and Sally Pietras, our department secretaries,

x Acknowledgments for helping with a large number of small tasks that allowed me to focus on my writing. The Institute for Humane Studies was generous enough to provide the financial support necessary to free me from some of my teaching duties to finish the book. My thanks to the Institute too. Several people provided me with extensive, page-​by-​page comments on the entire manuscript: Scott Simmons, Eduardo Martinez, and Alexander Cohen. These were wonderful, deeply helpful gifts. Much of the shape of this book arose from comments on the previous volume, Must Politics Be War?, after it was published, especially from Nic Southwood, Simone Chambers, Adam Gjesdal, Jacob Barrett, and Alex Motchoulski. Southwood in particular helped me see the need for an important revision. I also profited from comments at a book manuscript workshop hosting by the Institute for Humane Studies in 2019, and from all who attended and commented on the book, especially Christian Bjørnskov, Brian Kogelmann, Alex Motchoulski, Molly McGrath, Bart Wilson, Alan Thomas, and Micah Schwartzman. I  am also grateful for help from Bill Glod, Jenny Lambe, Josh Grubbs, Gary Uzonyi, Alan Gibson, Irina Soboleva, and Sean Stevens. Much of my work on the empirical side of trust research was supported by the Niskanen Center in conjunction with the Knight Foundation. I am especially grateful to Will Wilkinson for coordinating that opportunity. The white paper on which c­ hapter 2 was based has been widely appreciated thanks to their efforts and can be found at kf.org/​vallier. I received helpful comments from a number of others as well, including Greg Robson, Andreas Bergh, Jacob T.  Levy, Victor Muñiz-​Fraticelli, Eric Schliesser, Jason Brennan, Jonny Anomaly, Nolan McCarty, Jerry Gaus, Ryan Muldoon, Marc Cohen, John Thrasher, Chad Van Schoelandt, Paul Billingham, Lori Watson, Alex Guerrero, Enzo Rossi, Joshua Miller, Alan Reynolds, Paul Weithman, and David Estlund. Since the first volume was published, I  launched a blog, Reconciled (kevinvallier.com/​reconciled), to discuss themes related to the books and other issues. I’m grateful to a host of discussants on the blog and social media for helping to sharpen my ideas. Parts of this book draw on published journal articles. Some of the material in ­chapter 3 is indebted to “Associations in Social Contract Theory: Toward a Pluralist Contractarianism.” Political Studies 68, 486–​ 503, 2020. Chapter  6 draws on “A Moral and Economic Critique of the New Property-​Owning Democrats: On Behalf of a Rawlsian Welfare State.” Philosophical Studies 172, 283–​304, 2015. Chapter 8 shares material with “Process Democracy.” Journal of Moral Philosophy, published online. In August 2020, Jerry Gaus, my friend and teacher, passed away unexpectedly. My intellectual debt to him is incalculable, so I dedicate this book in his memory.

About the companion website A companion website housing the following appendices referenced in the book can be found at www.oup.com/us/trustinapolarizedage. Appendix A. Bias and the Veil Appendix B. Accommodating Radicals in a System of Trust Appendix C. On Tyler Cowen’s Stubborn Attachments and the Trust-Growth Connection Appendix D. On Jason Brennan’s Case for Epistocracy Appendix E. On Alan Thomas’s Arguments for Property-Owning Democracy

Introduction Trust and Polarization

Americans are finding it harder and harder to trust one another. Social trust in the United States, our trust in our fellow citizens, has fallen dramatically.1 In the early 1970s, around half of Americans said that most people can be trusted. Today that figure is less than a third.2 Political trust, trust in government and democracy, has fallen steeply as well.3 Throughout the 1960s, over 70 percent of Americans said they trusted government in Washington always or most of the time. By the early 1990s, that number had fallen below 30 percent, and after a brief rebound in the early 2000s, it has collapsed to 17 percent as of 2019.4 More troublesome still is that Americans reporting no confidence at all in their national government doubled from around 14 percent between 1995 and 2011 to 28 percent in 2017.5 We see a similarly disturbing pattern in partisan distrust. In 2017, around 70  percent of Republicans said they distrusted anyone who voted for Hillary Clinton for president; likewise, around 70  percent of Democrats said they distrusted people who voted for Donald Trump. People not only distrust politicians from other parties, they distrust anyone who votes for the other party, that is, many millions of people.6 In our politically polarized age, we trust each other less simply based on how we vote. Worse, distrust is giving way to darker impulses, like hatred.7 From 1980, when measurement began, some members of each major political party reported that they hated the other one. But they used to be far fewer, numbering between 10 and 20  percent. In 2000, something changed. Now as many as half of the 1 Esteban Ortiz-​Ospina and Max Roser, “Trust,” Our World in Data, graph “Share of People Agreeing with the Statement ‘Most People Can Be Trusted,’ 1993–​2014.” https://​ourworldindata.org/​ trust. 2 The General Social Survey 1972–​2018 puts social trust at around 32% in 2018. See Smith et al. 2018. Though see Inglehart et al. 2020 for comparison. 3 “Public Trust in Government:  1958–​2019.” http://​www.pewresearch.org/​politics/​2019/​04/​11/​ public-​trust-​in-​government-​1958-​2019/​ 4 Ibid. 5 Inglehart et al. 2020. 6 For a discussion of partisan mistrust and findings that back it up, see Westwood et al. 2018, p. 336. 7 Hetherington and Weiler 2018, p. 129, chart “Partisan Hatred for the Other Party.”

2 Introduction members of each party despise the other party. It is unclear how a democracy can remain stable under these conditions. We surely do not trust those we despise, and we certainly would not like to be ruled by them, even temporarily. Worst of all is a Pew poll taken in 2018 that breaks down trust levels by age. The most trusting Americans are those over 65, whereas the least trusting by far are the youngest. In 2018, around 29 percent of Americans over 65 said that most people cannot be trusted, which is sad, but 60 percent of Americans 18 to 29 agree.8 And one thing we know from the trust research is that trust levels tend to harden in early adulthood.9 We face a low-​trust future. Eroding social and political trust comes with grave costs. Falling social trust can undermine democracy, economic growth, economic equality, the rule of law, and the protection of minorities; it can foment tribalism and bigotry, weaken our capacity to form relations of love and friendship with others, and even negatively affect personal psychological well-​being.10 Falling political trust makes it hard to form effective public policy, can threaten democratic government and political stability, and can generate excessive inequalities of power, corruption, and violations of basic rights.11 These phenomena are not merely abstract; they affect us viscerally. Recall your experience watching the hearings on Brett Kavanaugh’s nomination to the Supreme Court. As I’m sure you remember, Christine Blasey Ford claimed that Kavanaugh sexually assaulted her in high school, accusations Kavanaugh fiercely denied. You were surely on one side, absolutely convinced you were correct (He’s lying! No, she’s lying!), and you weren’t alone. Your friends probably agreed with you. Your family too. Now recall how you felt. In that moment, you may have been right, but you were also polarized. And you distrusted not only politicians on the other side but supporters of those politicians. Distrust and polarization occupy us socially and emotionally; this is no mere intellectual problem. We can see other consequences of falling trust as well. Arguably, one recent effect of lower political trust is norm erosion. Democracies can only function if they are supported by norms that serve as “the soft guardrails of democracy preventing day-​to-​day political competition from devolving into a no-​holds-​ barred conflict.”12 We need tolerance and institutional forbearance to make democracy work; otherwise, we cannot keep political losers integrated into the political system. Perhaps the most unusual and troubling way in which President Trump differs from previous presidents is his deliberate erosion of norms. 8 Gramlich 2019, chart “Most Young Adults in U.S. See Others as Selfish, Exploitative, Untrustworthy.” 9 See ­chapter 2. 10 See ­chapter 2. 11 See ­chapter 2. 12 Levitsky and Ziblatt 2018, p. 101.

Introduction  3 Trump’s temporary decision to deny funding to Ukraine to compel it to investigate one of his domestic political rivals, arguably with the intent to boost his chance of re-​election, severely erodes norms that prohibit using state power to interfere with free elections. Such actions suggest that the future will lead “to more departures from unwritten political conventions, and increasing institutional warfare—​in other words, democracy without solid guardrails.”13 Here again we are not musing about an abstract intellectual issue, but encountering a real problem that we can follow on television or the internet every hour of the day. Falling trust creates even deeper problems, including in how we form our deepest beliefs in other domains. Partisan identities increasingly drive the formation of other social identities, such as religious identity.14 Indeed today partisan identities are more stable than religious identities in the United States.15 With nonpolitical identities in decline, partisan identities cause alternative identities to line up in opposition to the opposing partisan group.16 For people who value these alternative identities, the data are troubling. Politics is swallowing everything else.

1. Trust and Polarization: The Distrust and Divergence Hypothesis These problems are cause for concern, but I am particularly worried that falling trust is driving increased political polarization. In common parlance, political “polarization” actually denotes two distinct phenomena: polarization and sorting.17 Polarization means that people are changing their minds on issues or changing or strengthening political group loyalty. Sorting occurs when people with the same views and loyalties associate more often among themselves, say by living closer together, marrying one another more often, sharing social networks, and the like. Polarization means that people change; sorting means that people move. Nolan McCarty distinguishes between issue-​ and affect-​based polarization and sorting. Issue-​based polarization occurs when people adopt increasingly divergent positions on issues. In contrast, affect-​based polarization occurs when people adopt new identities based on their membership in a political group 13 Ibid., p. 208. 14 Margolis 2018, p. 3. 15 Ibid., p. 104. 16 For the social costs of building our identities around politics, especially the important argument that doing so makes democratic politics function more poorly, see Talisse 2019. 17 McCarty 2019.

4 Introduction because they have warm attitudes toward the in-​group and hostility to the out-​group. Issue-​based sorting occurs when people sort into different political groups based on their preexisting stances on issues, whereas affect-​based sorting occurs when people sort into different political groups based on their positive or negative emotional attitudes toward those groups. I will use the term partisan divergence or just divergence to refer to all four phenomena: issue-​based polarization, affect-​based polarization, issue-​based sorting, and affect-​based sorting. McCarty argues that the big story in American politics over the past several decades is considerable partisan divergence.18 He claims, among other things, that “the evidence that the political parties have polarized over the past forty years is quite strong” and that “polarization has coincided with a large drop in the dimensionality of voting in Congress.”19 Most of the divergence is due not to rational disagreement but to affect formation. For the most part, we are emotionally rather than intellectually at odds. But that is nonetheless cause for concern. For our affective divergence is so great today that “partisans exhibit discriminatory behavior against opposing partisans at levels exceeding discrimination based on race.”20 Other scholars agree. Liliana Mason argues that American political parties have sorted into socially distinct groups.21 Today we have fewer conservative Christian Democrats and liberal Republicans. We are losing the kinds of cross-​ cutting identities that reduce intolerance toward out-​groups. That obscures when a political conflict is a legitimate controversy, where both sides have valid rationales for their points of view.22 When our ideological identities sync with our partisan identities, “It becomes harder to understand opponents as reasonable people and easier to feel threatened and angered by them,” which further increases partisan divergence.23 Party members adopt many of their political positions based on what their parties believe.24 What’s worse, sorting can be self-​reinforcing, as can polarization, since “those with more extreme political attitudes [tend] to view others as more extreme: which some call ‘polarization projection.’ ”25 18 Ibid., p. 16. 19 Ibid., p. 49. Here “dimensionality” refers to the dimensions of evaluation along which politicians evaluate bills; one dimension means that a politician only evaluates a bill based on the extent to which it realizes a single value. 20 Ibid., p. 64. For the original study, see Iyengar and Westwood 2015; they use multiple measures of both political and racial bias to come to this admittedly controversial conclusion. The effect is due both to the fall in racism and the rise in divergence. 21 Mason 2018, p. 63. 22 Mutz 2002, p. 850. 23 Mason 2018, p. 90. 24 Gould and Klor 2019, p. 19. So, we shift between political positions a lot, but we have stable core concerns. 25 Mellers, Tetlock, and Arkes 2019, p. 4.

Introduction  5 Scholars disagree about whether ordinary voters have experienced partisan divergence to any significant degree; many seem moderate, but it’s not clear that they’re principled centrists either, since they know little about politics and may care even less. However, when we look at citizens who are engaged and sophisticated, as opposed to those who aren’t, we have clearer evidence of divergence.26 Partisan divergence is widespread, even if it is worst among elites. One worry about divergent politics is that it generates a vicious cycle that drives centrists out of the public sphere; further, rising use of new technologies like social media increase one’s exposure to ideological and partisan content.27 In some societies, more partisan divergence might be good because it would give democratic voters more policy choices, but US institutions tend to produce legislative gridlock during periods of divergence.28 And indeed, US institutions may make divergence worse by giving greater power to already polarized party activists, which they can use to shape party platforms.29 On top of that, divergence is arguably worse when social trust and political trust are low, since the reaction to divergence will be less cooperative and peaceful.30 Scholars disagree about the causes of partisan divergence. Probably many factors are at work, such as southern realignment,31 greater economic inequality, greater racial diversity, and strong party competition at the federal level.32 In particular, increases in income inequality are highly correlated with more divergence, but we don’t know why. This may be because inequality pushes Democrats left, but it moves state legislatures to the right on the whole because moderate Democrats get replaced by Republicans.33 Since the causes of partisan divergence are complex and mysterious, we have few strategies for reform. Electoral reform probably cannot do much to stop or reduce divergence. Campaign finance reform may be promising, but the reforms we have put into practice have had, at best, mixed effects.34 Even worse, bipartisan redistricting, fewer partisan primaries, and reduced corporate spending on elections may not help reduce divergence either, even if they are otherwise good 26 McCarty 2019, p. 67. 27 Ibid., p. 100. 28 Ibid., p. 154. 29 Layman, Carsey, and Horowitz 2006, pp. 96–​7. I thank Eduardo Martinez for this point. 30 This is even worse in light of recent research, which suggests that there are social contexts where both conservatives and liberals see life as a zero-​sum game (for conservatives, when the status quo is challenged, but for liberals when the status quo is upheld), and so people appear to tend to adopt a politics-​as-​war mentality when the relevant cues are present. Thus, partisan divergence, by exacerbating our divisions, may increase our nascent zero-​sum attitudes. See Davidai and Ongis 2019. 31 Southern realignment is the shift of the southeastern US states from being primarily Democratic to primarily Republican starting in the mid-​1960s. 32 McCarty 2019, p. 99. One of the major causes is arguably attempts by figures in the leading parties in the 1950s and 1960s to make the parties more diverse. See Rosenfeld 2018. 33 McCarty 2019, p. 80. 34 Ibid., p. 132.

6 Introduction ideas.35 Some lab-​based research explores how to reduce partisan divergence in small groups, but the work is in its infancy.36 In this book, I argue that there is a close relationship between partisan divergence and trust, in particular, trust at the societal and institutional level. As I explain later in the book, we need to look at the causal relationship between social trust and political trust. Briefly, social trust is that trust which each member of a society has that other members of her society will generally follow publicly recognized moral rules; political trust is that trust which each member of a society has that governmental institutions will follow fair procedures and produce positive results. We will see that both social and political trust are extremely important societal goods and that losing trust has many negative consequences. Social trust and political trust are measured primarily through the use of surveys and questionnaires, though some measurement comes through observing individual decision-​making in structured experiments, and these measures enable us to empirically explore potential causal connections between trust and divergence. I think falling trust and increasing partisan divergence are mutually reinforcing: falling trust leads to more divergence, and divergence leads to falling trust. We have some evidence that falling social trust increases divergence. Hans Pitlik and Martin Rode have found that “trusting people have a lower propensity to express support for extreme policies, leading to a general moderation of preferences in trusting societies,” and that high levels of trust enable more consensus on critical reforms.37 Markus Leibrecht and Pitlik find a link between social trust and actors tending to focus on promoting the public good rather than merely their private interests.38 Social trust improves coordination and cooperation between societal groups, lowering the costs of building agreement on needed economic reforms; agreement among groups raises the credibility of a policy change. Thus, more trusting societies are less likely to have policy change blocked by powerful executives or other political officials.39 Niclas Berggren and Christian Bjørnskov find that social trust facilitates liberalizing economic reforms but makes reforms that reduce market liberalization more difficult.40 High social trust implies that people are often patient, prepared to let a process of reform work itself out while they remain hopeful about its outcome.41 Social trust leads to more consensus in agreeing on reforms



35 36 37 38 39

40 41

Ibid., p. 133. Mellers, Tetlock, and Arkes 2019, p. 7. Pitlik and Rode 2019, p. 4. Leibrecht and Pitlik 2015, p. 249. Ibid., pp. 264–​5. Berggren and Bjørnskov 2017, p. 4. Ibid, p. 7.

Introduction  7 and, in general, makes effective reforms more likely.42 Friderich Neinemann and Banjamin Tanz have also found that trust is conducive to reforms and that falling trust makes it more difficult for societies to cope with challenges from phenomena like demographic change.43 Relatedly, Iasonas Lamprianou and Antonis Ellinas find that falling political trust leads to more support for extreme right-​wing policies, and this is through grievances with political institutions, for example those over large-​scale immigration, rather than through either economic losses or cultural change in itself.44 They found that falling political trust, in conjunction with a lack of support for democracy and a disposition to use political violence, increased support for the far-​right Golden Dawn party in Greece. Similarly, Carl Berning and Conrad Ziller have found that as social trust falls, preferences for radical right-​wing populist parties increase.45 This relationship is largely mediated by anti-​immigrant sentiments.46 Lower social trust means more distrust of immigrants, which means stronger preference for extreme right-​wing populist parties and policies. The result, among other things, is that falling social trust has led to increased support for the right-​wing populist Party for Freedom in the Netherlands.47 There is generally a plausible case that falling social trust has led to more divergence and populism in Germany. East German trust fell greatly under Communism. Now, the former East Germany, which still has lower social and political trust than the rest of Germany, exhibits more divergence, with powerful socialist and populist right-​wing parties: Die Linke on the left and Alternative for Germany on the right. These two parties are much weaker elsewhere in Germany. Thus, two otherwise very similar groups of people—​Germans—​exhibit different trust levels and also exhibit differing degrees of divergence and populist hostility to liberal democratic institutions. In many ways, the American political situation is similar, with falling social trust, rising partisan divergence, and rising populism. Right-​wing populism has taken power through the Republican Party, and left-​wing populist sentiment in the Democratic Party has manifested most clearly in the popular Bernie Sanders campaign. Trust falls, divergence rises, and populism comes in its wake. There are other potential causal channels from trust to partisan divergence as well. For instance, falling social trust often leads to greater hostility to income

42 Ibid., pp. 20–​1. 43 Heinemann and Tanz 2008, p. 182. 44 Lamprianou and Ellinas 2017, p. 45. 45 Berning and Ziller 2017, p. 198. This is true for both an individual’s level of social trust and the level of social trust that prevails in a single neighborhood. Decreases in either increase a preference for right-​wing populist parties. 46 Ibid., p. 201. 47 Ibid., p. 212.

8 Introduction redistribution. This can increase economic inequality, which may aggravate divergence.48 The claim that divergence leads to falling social trust finds less-​compelling support in the data. Indeed, American social trust has only fallen a few points over the last twenty-​five years, while we have seen continuing increases in polarization.49 But there are a number of plausible links that may be being obscured, many of which are substantiated by data that I discuss in c­ hapter 2. Greater divergence may lead people to trust others less because they recognize that others are often members of a political out-​group. Increases in divergence alone may decrease trust because people may be less likely to trust persons who have values distant from their own. As cultures develop social markers for identifying these diverse groups, and as content creators draw our attention to those markers by representing or exaggerating them in news, TV, movies, and social media, we will tend to culturally sort ourselves into different social silos that seldom interact with one another. One could also argue that decreasing political trust and increasing divergence and norm erosion lead to corruption, including visible corruption, which reduces social trust. For example, falling trust in government can lead to populist voting, and populist leaders often engage in nepotism and violate ordinary democratic norms; falling political trust in the United States arguably led to the election of Donald Trump, who has given family members unprecedented positions of power and has challenged democratic norms in ways unheard in most of modern American history. The best arguments against a causal channel from partisan divergence to social distrust are not compelling. For example, Pitlik and Rode suggest that highly trusting societies should lead to preference homogeneity. They claim that societies with many low trusters and many high trusters should therefore manifest more divergence than we see in countries with social trust levels around 50 percent, since such societies are often not very divergent but should be, given that modest levels of social trust are often less politically divergent.50 Yet many other factors may yield political differences among people who trust one another, such as their personal ideals, values, and faith commitments. Trust alone should not be expected to yield complete agreement. Pitlik and Rode also argue that social trust is too difficult to alter for partisan divergence to make much of a dent in it, since social trust levels (as I discuss in ­chapter 2) tend to harden in early adulthood. But that’s compatible with there being a generational effect, where greater divergence leads young people to adopt

48

McCarty 2019, pp. 79–​80. Smith et al. 2018. Inglehart et al. 2020 finds no decrease during this period. 50 Pitlik and Rode 2019, p. 4. 49

Introduction  9 distrusting attitudes when they take on a political identity and encounter those with a different one. So, while divergence may not change the trust levels of most adults, it can exert a gradual effect on trust across generations, and indeed, as I have noted, much of the fall in social trust is concentrated among the youngest. Even if divergence does not reduce social trust, there is good evidence that divergence reduces political trust. Eric Uslaner argues that more legislative divergence leads to less political trust; in general, polarizing politics makes people less trusting.51 Perhaps the main causal channel is that divergence leads to more gridlock, lowering the quality of governance; since quality of governance is the main determinant of political trust, then divergence should decrease political trust.52 One might also argue, along similar lines, that if gridlock also reduces economic growth and economic security, political trust will fall further still. And if divergence leads to more economic inequality, as Uslaner53 and McCarty54 argue, since inequality seems to reduce political trust, then that is yet another source of distrust. Falling political trust also sometimes leads to populist voting,55 and populist politicians often erode democratic norms, undermining political trust further. As Pippa Norris and Ronald Inglehart argue, populist discourse is by nature focused on making the case that democratic elites cannot be trusted, and indeed, populist rhetoric can lower political trust as a result.56 Lower political trust may then engender the erosion of central constitutional norms because the political system is seen as broken and corrupt.57 On these grounds, I defend a distrust-​divergence hypothesis: that social and political distrust and partisan divergence are mutually reinforcing. If my hypothesis is correct, then this book can contribute to disrupting the causal feedback loop between distrust and divergence by examining how diverse societies sustain social and political trust. In particular, I hope to identify trust-​increasing liberal rights practices, where certain basic liberal rights, recognized in constitutional law, exercised regularly by the people, and embodied in public policy, are the principal drivers of trust. If we can identify feasible reforms to protect liberal

51 Uslaner 2015, p. 363. Uslaner uses the term “polarization” but does not distinguish between polarization and sorting, so I use the term “divergence.” 52 Ibid. 53 Ibid. 54 McCarty 2019, pp. 78–​81. 55 Geurkink, Zaslove, and Sluiter 2019. Following this study, I understand populist politics as a stoked antagonism of a movement that regards itself as embodying a “pure” people standing against a “corrupt elite.” 56 Norris and Inglehart 2019, p. 65. 57 McCarty 2019, pp. 154–​5. For a similar argument, see Rosenfeld 2018, pp. 282–​3. On the other hand, polarized/​divergent populations may be more inclined to monitor government for corruption and punish it when it occurs, so long as political competition is high and incumbent officials can run for office; for evidence, see Melki and Pickering 2020.

10 Introduction rights and encourage their exercise, we may be able to arrest the growth of distrust and divergence. This does raise a powerful challenge, however. For if liberal rights practices are so great, then why are we seeing big declines in social and political trust? After all, the United States protects liberal rights better than almost any other regime in human history. So, what’s going on? When it comes to political trust, I think we have a pretty good idea: there’s been a decline in economic growth and some elements of quality of governance relative to the 1950s and 1960s. But the decline in social trust is more mysterious. I’ve suggested that divergence is part of the story, but it is only part: as I point out in ­chapter 2, we just don’t know much about the causes of social trust, despite knowing of a few phenomena that can reduce it. Thus, this book does not argue that we can end our decline in social and political trust with liberal democratic practice but rather that the ordinary tools of liberal-​democratic law and policy are the best hope we have for arresting the decline. Perhaps there are other factors driving trust lower, and liberal democracy can only slow them down. But if liberal democracy is the best we can do, that is a good reason to stick with it through thick and thin. And the book also makes a stronger case, which is that the policy tools of liberal democratic order is a source of trust building, not just preventing collapse.

2. Competing Hypotheses Before we go further, it will be helpful to have a sense of some hypotheses that are alternatives to the distrust-​and-​divergence hypothesis: inherent-​corruption theories, direct theories, and group theories. Inherent-​corruption theories hold that distrust and divergence are the natural result of some central dimension of liberal-​democratic political order. Chantal Mouffe, for example, argues that conflict of various kinds is inevitable in democracy,58 and this leads her to adopt an “agonist” position, which I address in c­ hapter 8. And Patrick Deneen’s recent book, Why Liberalism Failed, famously argues that liberalism has an ingrained tendency toward inequality, social atomization, and, ultimately, institutional decay.59 Liberalism adopts a conception of liberty where people count as free when they have the ability to plan their lives free from the oppressions of the state and social norms in society broadly, but on Deneen’s argument, what members of liberal society get instead is devastating loneliness and meaninglessness in their lives. Mouffe claims democracy sets us at odds; Deneen says that a liberal social ethos and liberal institutions divide and isolate us.

58 59

Mouffe 2005. Deneen 2018.

Introduction  11 Importantly, neither Mouffe nor Deneen calls for the abolition of liberal democracy; indeed, Mouffe thinks we should simply recognize the true conflictual nature of democracy and accept it. And while Deneen seems comfortable with the collapse of liberalism, he does not advocate limiting or abolishing any major liberal rights. Nonetheless, both think there are major social dynamics that lock us into a warlike politics. The distrust-​divergence hypothesis is much more consistent with the data than these inherent-​corruption theories are; indeed, inherent-​corruption theories are seldom defended on empirical grounds of any kind. Rather, these theorists ask us to condemn certain broad institutions, forms, and cultural beliefs based on how they fit into a kind of powerful social narrative. However, the data strongly suggest social instability, division, and conflict are due to a reinforcing cycle of falling trust and increasing partisan divergence, not anything inherent in democracy or liberalism. That means the solution to division does not require learning to live with a warlike politics or waiting for civilizational collapse to replace the old with the new, but rather focusing on using the tools provided by liberal social orders to interrupt the distrust-​divergence cycle. Second, social trust is quite stable in most liberal democracies (and in most nonliberal nondemocracies); the United States alone has seen a steep decline in social trust. That suggests we need a country-​level explanation of falling trust, not one ascribed to generic institutions like democracy or a generic liberal ethos. Finally, the argument of my previous book, Must Politics Be War?, helps answer inherent corruption theories because, as I review subsequently, it shows that liberal democratic societies with a high degree of diversity have no robust propensity to distrust among their members. Thus, we have little reason to think that democracy must yield conflict that goes beyond what is morally permitted within a trusting social order; nor do we have reason to think that liberal institutions undermine trust by undermining social cohesion and increasing social atomization. Direct theories of divergence point to social phenomena that directly cause divergence, unmediated by low trust, or they point to a shared cause of low trust and divergence. As we have seen, there are a number of such theories, such as southern realignment, economic inequality, party engineering, media, and electoral processes such as gerrymandering and partisan primaries. There is evidence that some direct causes arose as party divergence developed in the 1970s and 1980s.60 First, southern realignment, the shift of southern whites from the Democratic Party to the Republican Party, arguably has a number of causes, including the push for racial equality in the 1950s–​1970s. This created stronger



60

McCarty 2019, pp. 99–​100.

12 Introduction party competition and so yielded some polarizing effects. Increasing inequality appears to track partisan divergence quite well; more divergent legislatures hurt moderate Democrats, shift other Democrats left, and enable Republicans to use gridlock to prevent income redistribution that would reduce inequality. Some electoral practices, like first-​past-​the-​post voting,61 also appear to contribute to divergence. However, McCarty thinks that social science does not help many of these hypotheses, such as those that blame gerrymandering and primary elections. Campaign finance may play a role, but “bipartisan redistricting commissions, nonpartisan primaries, and curbs on corporate spending” probably can’t stop divergence.62 A related direct theory takes note of the asymmetry of partisan divergence between Republicans and Democrats, as McCarty notes that Republicans have moved further right than the Democrats have moved left. Here, then, the Republican Party is a direct cause of divergence. But McCarty plausibly argues that the Republican Party is quite weak, historically speaking. It has a great deal of trouble determining its own agenda independently of powerful personalities and megadonors. Nearly all elite Republicans wanted to stop Donald Trump from being the nominee, and yet they failed, quite spectacularly. What’s more, the GOP controlled the entire federal government from 2017 to 2019 and almost no major legislation was passed, not even the Republicans’ own long-​sought projects. So, postulating the Republican Party elite as a prime cause of divergence confronts real difficulties. One might then turn to blame the structure of conservative media, such as Fox News, for causing divergence, but while there is evidence of a “Fox News effect” in increasing vote share for Republicans,63 it is hardly a major explanatory factor: partisan divergence began before Fox News gained its current high level of influence, and Fox News has competitors with different perspectives.64 Some of the polarization on the right has been exacerbated by their closely-​knit media ecosystem, and may well explain why polarization is greater on the right then the left.65 But this process is also arguably an effect of greater polarization and mistrust as much as a cause, since there was clearly both a demand for such media among parts of the general public and elites willing to supply it.

61 First-​past-​the-​post voting means that whoever gets a majority of votes wins an office, as opposed to proportional representation systems where minority groups may get some share in power despite losing the vote. 62 McCarty 2019, pp. 132–​3. 63 Martin and Yurukoglu 2017. See Benkler, Faris, and Roberts 2018 for a detailed account of how the right-​wing media ecosystem as a whole affects voting in American elections. 64 Though admittedly, Fox’s competitors on the left are less successful individually in changing voting patterns and influence. 65 Benkler, Faris, and Roberts 2018, pp. 311–​340.

Introduction  13 From a review of the literature, there are two compelling direct cause theories. The first is southern realignment: that the push for racial equality in the South split the Democratic Party, revitalized the Republican Party in the South, and thus created more competitive parties. The second is economic inequality, which then polarized the now more competitive political parties. These direct theories are only problematic insofar they purport to explain divergence independently of falling trust, or if they explain trust and divergence as a product of the same cause. However, it is plausible that falling trust exacerbates the polarizing effects of these historical changes. This is especially true in the case of income inequality, since falling social trust seems to lead to more economic inequality, as I show in ­chapter 2. Southern realignment is also arguably mediated by trust: the prospects of electoral success in ending segregation and securing equal rights plausibly led to more conflicts between black and white southerners, sustaining, and perhaps even increasing, distrust between races as whites adopted new policies for monitoring and oppressing the black population. The distrust-​divergence hypothesis is therefore not incompatible with the factors the direct theories highlight contributing to divergence. Rather, it holds that trust plays an important role in explaining divergence, including, possibly, by amplifying the power of such proposed direct causes as realignment and inequality. The distrust-​divergence hypothesis helps to enrich and vindicate these direct theories. The third class of partisan divergence theories is group theories. Based on findings from political science, political psychology, evolutionary biology, and sociology, these views postulate that human beings are naturally tribal and biased toward their tribe. The most recent group theory of divergence is Ezra Klein’s Why We’re Polarized.66 Klein reviews a great deal of evidence in favor of our having tribal psychologies that our institutions are somehow ginning up. He points to some institutional causes, like southern realignment. And he describes some promising institutional solutions. However, he is largely unique in stressing individualized de-​biasing techniques like “identity mindfulness,” and he encourages people to identify more with the local culture than with national politics.67 Robert Talisse, in Overdoing Democracy, focuses on much of the same data to argue that we are politically oversaturated: politics occupies too central a space in our lives.68 He encourages people to develop more cross-​cutting identities where politics does not much matter. As he says, we cannot allow politics “to permeate the whole of our collective lives.”69 66 Klein 2020. Klein also uses the term “polarization” in its more common sense. 67 Ibid., pp.  261–​8. Klein also recommends institutional changes, some of which McCarty is skeptical about. 68 Talisse 2019. 69 Ibid., p. 171.

14 Introduction Group theories of divergence have a lot of insight, but they too are incomplete. First, they tell an incomplete story about how our tribal psychologies are activated. Why are we retreating into new groups rather than sticking to our older, bigger group, namely the country as a whole? That is, why do we think we must stick to the red tribe or the blue tribe? Some institutional changes, such as an increase in publicly perceived racial and economic inequality, have helped activate tribal mindsets. But we need an explanation of why political identity has suffused American life so broadly in order to explain the degree of divergence in the United States. Declines in trust have primed our tribal impulses. Falling trust explains why we have retreated into new identities rather than reforming our older national identity. It helps explain why people are adopting more extreme opinions; and if people are mistrusting, it helps to explain sorting as well, as people will prefer to associate with those more like them when trust is relatively low. To see this, imagine a United States with as much social trust as Sweden today: imagine 65 percent of people said most people can be trusted, as opposed to 35 percent. With our psychologies flooded by trust, people would see little need to retreat into new identities; older identities should hold. Similarly, if most people trusted the government to do the right thing, for instance, if 65 percent of people trusted Congress as opposed to 20 percent, our political divisions would be less severe and less tribal. A second difficulty with group theories of divergence is that they tend to overstress the individual, psychological dimensions of divergence, quickly scaling up from bias in the individual brain to large-​scale institutional changes. The mechanism of transmission from individual brains to national-​level politics is often unclear. We are told that people group in tribes and that this makes a big institutional difference, but biased and polarized minds need not be coordinated with one another at such a large level. One plausible transmission mechanism from individually biased minds to national-​level attitude and institutional change is a decline in social and political trust. If falling trust is priming our tribal psychologies and we are retreating into political identities as a result, then we can explain why individual-​level bias leads to national-​level changes, because trust is a kind of collective attitude that coordinates large-​ scale behavior. When it breaks down, it leaves room for certain kinds of neurological tendencies to make differences at the macro level. And in those cases, phenomena like Cass Sunstein’s law of group polarization can take hold: as opposing tribes polarize and deliberate mostly within themselves, polarization gets worse.70



70

Sunstein 2002. Here Sunstein seems to refer to polarization alone, not sorting.

Introduction  15 Direct and group theories of partisan divergence have a lot of insights; they’re not entirely wrong. But they both have a missing piece: falling social and political trust. Inherent corruption theories, however, are mistaken.

3.  The Limitations of a Merely Philosophical Approach In my previous book, Must Politics Be War? (MPBW), I asked whether declines in social and political trust pose a general skeptical challenge to the feasibility of a diverse, yet high-​trust, society. As we become less trusting, it is natural to wonder whether our shared political life is necessarily one of conquest, domination, and control among competing ideological groups. But we might instead hope that our inevitable political disagreements can be resolved in ways that all can regard as morally acceptable, and that the processes that resolve disputes have authority over each person. MPBW argued that we can end a warlike politics by adopting and maintaining liberal institutions, those that protect a wide range of equal liberties for all. Liberal institutions alone furnish the rational basis for trust among diverse perspectives, both liberal and illiberal. Unlike regimes that promote particular sectarian values,71 liberal regimes can be justified to each group in virtue of protecting its members’ ability to live their own lives in their own way. When people comply with the directives of liberal institutions, they thereby manifest a kind of trustworthiness; they show that liberal institutions have their allegiance, which gives others reason to trust them, even if our values and worldviews are incompatible, even antagonistic. But, I argued, we can pursue trust in the wrong way. We should not try to establish trust based on propaganda, manipulation, and threats.72 Rather, we want trust grounded in the fact that others are trustworthy. This is why it is essential that liberal institutions be justifiable for all, or publicly justified. If liberal institutions are justified to all, then each person and group can see good moral reason of their own to comply with those institutions, motivating them to be trustworthy by following those rules from conscience and conviction. As I argued, to end political war, we need trust; to get trust, we need trustworthiness; to get trustworthiness in a diverse society, we need public justification; and liberal institutions alone are publicly justified. Hence my defense of what is now called public reason liberalism, a commitment to liberal institutions based

71 Here I understand values as sectarian if they are only embraced by a particular religious or ideological group, and those values are rejected by other religious or ideological groups. Thus, sectarian groups are not necessarily religious groups. 72 One might argue that this is how China maintains its high levels of national trust. Though there are doubts that the China data are sound. See Uslaner 2002, p. 226.

16 Introduction on the requirement that norms of interpersonal authority and political power are justified to multiple reasonable points of view.73 However, we need to go further than this, beyond mere philosophical argument. Previously I only argued that liberal institutions furnish the rational basis for trust—​that people will see no reason, on balance, to distrust those who are different from them and so reject liberal institutions. As I put it there: reflection stops defection from liberal order. My aim was to meet a difficult philosophical challenge, not the practical problem of distrust and divergence. But this argument was incomplete. First, if the goal of the project is to show how liberal institutions create trust in the right way, it would be helpful to show that liberal institutions create real trust. There is a point in showing how a model society like our own can sustain high levels of social cooperation under diverse conditions; such has been the aim of social contract theories for centuries. But we now have access to data on how social stability and cooperation are sustained, data that earlier social contract theorists did not have, and so we can speak more directly to the problems we currently face. Second, one aim of solving the philosophical challenge is to pave the way to address the related practical challenge of interrupting the cycle of distrust and divergence, but to address the practical challenge, we must now look at the empirical data on which institutions and policies create and sustain social and political trust in the real world. And solving the practical challenge is the greater endeavor.

4.  The Aim of Trust in a Polarized Age The strategy of this book should now be clear. I want to argue that certain key liberal rights create real trust for the right reasons. So, I offer a dual justification for a number of central liberal rights, first making an empirical argument that exercises of each right promote social or political trust or both, then arguing that the right is publicly justified and so promotes trust in the right way. I focus on the rights to freedom of association, private property, adequate social welfare programs, a democratic constitution, and participation in elections. Together, these rights practices should help disrupt the distrust and divergence loop. I initially hoped to defend the liberal rights of speech, press, and religion, as well as federalism and open immigration. But I  met with disappointment. Freedom of speech, press, and religion don’t seem to noticeably increase or decrease social or political trust, and it’s not clear why. The data on trust and federalism are too meager and inconclusive, though perhaps future work will tease



73

I have reviewed the details of public reason views in Vallier 2014a, Vallier 2016, and Vallier 2019.

Introduction  17 out some connections. My biggest concern is that there is some evidence that immigration restrictions preserve political trust among citizens within a particular country.74 Since publicly justifying immigration restrictions to poor and fearful immigrants is rather difficult, there may be a divergence between what sustains trust and what is publicly justified. That means we will have to choose between creating trust and mistreating people, which is unfortunate. But I must set those issues aside.75 Despite these disappointments, some liberal rights practices promote real trust in the right way, and that is strong reason to embrace them, especially in polarized times.

5. Chapter Summary This book has eight chapters. The first chapter restates the problem and solution of MPBW in terms that I think are accessible for a wider audience. If you’ve read the previous book, then, feel free to skip to c­ hapter 2. Chapter 2 reviews the empirical literatures on social and political trust and shows how liberal institutions create social and political trust in the real world. Chapters 3–​8 cover specific rights practices. Chapter 3 argues that freedom of association promotes social and political trust by promoting contact between diverse persons. It then argues that a right to freedom of association is publicly justified because diverse persons will insist on protecting their associations from their government and other citizens, and so will embrace associational rights within a contractarian procedure. Chapters 4–​6 turn to economic rights. Chapter 4 defends private property and the market economy. Private property rights give rise to markets, which promote peaceful interaction and promote a prosperous and impartial economic system, creating social and political trust. And, since markets arise from publicly justified private property rights, markets create trust in the right way. Chapter 5 argues that the practices of assigning welfare rights and distributing welfare benefits, as well as the regularized use of those benefits, promote social and political trust. If the market economy is to be publicly justified, it must be supplemented with social insurance programs, though not with a strong administrative state. Strong administrative states are likely to undermine trust through public corruption and reduced economic growth. Chapter 6 resists more economically egalitarian

74 McLaren 2017. 75 One might worry that the causal effects of these institutions on trust are causally intertwined with the causal effects of the institutions I focus on. Plausibly or no, the trust literature proceeds on the assumption that we can isolate the effects of certain institutions on trust and vice versa.

18 Introduction systems, especially property-​owning democracy and liberal socialism.76 Such regimes are likely to be poor, violent, and corrupt, undermining several central bases for trust. And we are unlikely to be able to create trust simply by compressing economic inequalities. Chapters 7 and 8 address the policy-​formation process as well as rights to vote and otherwise participate in elections. Chapter 7 outlines and defends a policy-​ formation process based on high standards of evidence, high-​quality deliberation, and extensive public input that I call democratic constitutionalism. There’s strong evidence that high-​quality governance and procedural fairness are absolutely essential for political trust. Chapter 8 argues that elections can play an important role in promoting trust, though the electoral process creates social and political trust in some cases but not in others. However, since democratic practice arguably sustains other institutions that directly create and sustain social and political trust, such as the rule of law, economic growth, and economic security, democracy promotes social and political trust at a remove. I then argue that electoral rights are publicly justified on grounds of fairness and instrumental benefits. One criticism of MPBW is that it didn’t have enough to say about the real democratic process of policymaking and elections.77 Chapters 7 and 8 explain how my theory impacts real politics.

6.  Liberal Order Is Our Best Hope for Peace, but Is Peace What We Want? The overall message of this book, like that of the previous volume, is both liberal and conservative. It is liberal in arguing that liberal rights practices have the unique capacity to sustain social and political trust among diverse persons. It is conservative in arguing that distrust and partisan divergence can be addressed through the liberal institutions we have in place. We need modest reforms to our present liberal rights practices. I encounter opposition to my thesis from all over the political spectrum. Egalitarians, conservatives, and libertarians are all prone to think that political life is fundamentally about the good defeating the bad. People in all these camps are committed to seeing politics as a struggle between the good (themselves) and the bad (those of other ideological persuasions). At the risk of giving offense, I am sometimes tempted to think that we want politics to be war because political war is a force that gives us meaning,78 with the attractive feature of avoiding the

76

These systems are defined in c­ hapter 6. Thompson 2019. 78 Hedges 2002. 77

Introduction  19 risks of the direct use of violence. Instead, we let politicians do our dirty work for us. This book, like its predecessor, is meant to criticize partisanship by showing that our desire to destroy one another is not based on the sober recognition of the cold reality of political life. There is little reason to think liberal politics must be war. The real problem is that we want to destroy each other, not that we must. Once these dark desires are recognized, we must fight them. That means always recognizing the humanity of our political opponents, repenting of our desire to treat them as inhuman foes, and joining others in the hard work of maintaining liberal democracy, markets, and social welfare programs. Few think our system is ideal. I certainly don’t. But a liberal democratic peace is our best hope for living well together in a world where we will never agree about what living well entails.

7.  How to Read This Book This book lies at the intersection of philosophy, political science, and economics, and I have written it so that people in those disciplines can appreciate my arguments. I recognize that nonphilosophers can grow weary of philosophical argumentation and may want to focus on the empirical discussions and my engagement with scholarly work outside of analytic political philosophy, such as my discussion of Christopher Achen and Larry Bartels’s work in ­chapter 8. For this reason, I have marked more technical sections and subsections with an asterisk. Nonphilosophers can skip these sections without much loss. I have also placed the empirical data toward the beginning of each chapter whenever I can, again to help the nonphilosophical reader. I think philosophical detail is important, and I make no apology for it. But I want people with different interests to profit from the book.

1

Must Politics Be War Here and Now? While liberal democracy has triumphed over its twentieth-​century competitors, the twenty-​first century has seen liberal democracies challenged by politicians and parties that erode democratic norms, the informal rules that make democracies functional and free. Democratic-​norm erosion is both a cause and an effect of the growing sense that democratic politics is a struggle for domination, a thinly veiled war between political factions trying to conquer one another. Populist leaders tell us that we are at war with “globalist” elites and that citizens are at war with migrants. Progressives and egalitarians insist that we are at war against an ever-​expanding list of bigotries. Consequently, many are losing their capacity to trust our most important social institutions—​and one another. Our increasingly confrontational political attitudes tempt us to think that politics must be war. A politics that transcends institutionalized aggression, one grounded in at least a partial agreement on shared values and moral norms, is no longer feasible for us (if it ever was). I believe that the only way to solve this problem of political war is to identify institutions that can create and sustain high levels of trust among persons with diverse values and commitments, specifically ones that sustain trust in society (social trust) and trust in political institutions (political trust). As I  argued in Must Politics Be War?, we can understand a state of political war as a condition in which a society has low levels of social and political trust but is not in an open civil war.1 In a low-​trust environment, we should expect violence and oppression to rise because even the best of us will face the temptation to strike first to ensure that we are not harmed or controlled by our untrustworthy opponents. Low-​trust societies are tinderboxes, as even ordinary behavior can be seen as dishonest and dangerous. High-​trust environments, where people trust one another to follow a wide array of central moral rules, such as rules that prohibit harm and require keeping promises and avoiding harming others, are quite different. When we trust each other to follow the most urgent and universal moral rules, we have the basis for bringing conflict to a close. If we can trust our fellow citizens to keep their promises and not harm us without cause, we can give peace

1 This is not to identify a state of political war as any social state with low levels of social and political trust, however.

Must Politics Be War Here and Now?  21 a chance. The problem of political war is thereby solved by high levels of social and political trust, and the institutions that sustain them. Sustaining social and political trust in a diverse society is arduous because people with diverse perspectives find it difficult to see one another as trustworthy. We are tempted to see those who reject our values as thereby manifesting a moral or intellectual vice. In this way, we succumb to what I have called the illusion of culpable dissent. We hold others morally responsible for disagreeing with us despite having limited access to their reasons for believing as they do.2 And the greater the distance between our respective values, the greater the temptation to treat dissent as a fault. The resulting distrust may not lead to open war, but it generates destructive conflict, where we co-​opt shared institutions to defeat the “enemy.” In this way, distrust and diversity can lead to a warlike politics. But, as I argued in MPBW, the appearance of culpable dissent is often an illusion. That’s due to the fact of evaluative pluralism: that sincere and informed people can nonculpably disagree about many important matters, including what the good life consists in and what justice requires. Evaluative pluralism occurs because a range of social and cognitive factors, such as our imperfect rationality and limited ability to process information, as well as the complexity and vagueness of many moral judgments, lead people to disparate evaluations of moral and political claims. The sources of honest disagreement are vast, and they afflict us so much that we cannot reliably detect when disagreement is reasonable, given that the reasoning and experiences of others can differ from ours dramatically, without our realizing it. Of course, many of our disagreements are due to one or both parties are being irrational, unreasonable, or simply malicious. Because we tend to be biased in favor of attributing intellectual or moral error to our opponents; recognizing evaluative pluralism is a hard skill to develop. What I have just described is the central problem of MPBW, which I will finish solving in this book. One aspect of my solution is an account of the right reasons to trust. So, from here, I will review my solution to the problem of political war (section 1), develop the central idea of trust for the right reasons (section 2) and review the key concepts I used in MPBW to formulate and resolve the problem of political war (sections 3–​6). I  next outline the relevance of the empirical 2 Alex Motchoulski points out that we might treat those we disagree with most not as morally responsible agents, but as inept patients to be contained or isolated. This may be true in some cases, but in light of P. Kyle Stanford’s work, I take it that we bear reactive attitudes toward what he calls “morality-​based” out-​groups, such that we treat groups who morally disagree with us as the objects of scorn, derision, and even hatred. That is the situation I think we find ourselves in in the United States. Stanford’s work challenges my thesis because we may consider our political out-​group as so beyond the pale that they become inhuman and so become patients to be destroyed, rather than moral agents. But at least for the societies addressed in this book, it looks like our value systems are sufficiently close that our practice of holding accountable isn’t stretched beyond the breaking point. For more, see Stanford 2018.

22  Trust in a Polarized Age literatures on social and political trust to the creation and maintenance of trust for the right reasons (section 7) and review how liberal rights practices create trust for the right reasons (section 8).

1.  The Basic Solution to the Problem of Political War Solving the problem of political war requires that people have a public basis for trusting one another, a touchstone of trustworthiness. In large societies, the chief touchstones of trustworthiness are public compliance with social norms, norms that the public collectively regards as morally binding or appropriate. A  society whose members obey these norms, especially the norms that comprise the economy, the legal system, and the political process, provides a rational basis for its members to trust one another, even though they know others have distinct and incompatible values. Persons who comply with these norms can be seen as trustworthy even when their personal ideals differ, which allows diverse persons to build mutual trust. And with trust, we can avoid the temptation to preemptively engage in institutionalized aggression and conquest, since we will see no need to destroy our opponents. In this way, providing a touchstone of trustworthiness will help ensure that even deep political conflict and contestation will be seen as compatible with a shared moral order. The chief touchstones of trustworthiness in a diverse political order are its rights practices: observable acts of publicly protecting and exercising basic rights. Free speech is a rights practice where individuals exercise their liberty to speak under protection from coercive interference from others and from government. Property holding is a rights practice where citizens use their property in ways that they think best under protection from interference from others and from government. Similarly, receiving and using social insurance can be seen as a welfare rights practice, where government operates social welfare programs, like food stamps and public medical insurance. My argument is that the institutions of an open society, that is, a society with a broad range of liberal rights practices, have the unique power to sustain trust between diverse perspectives, overcome the illusion of culpable dissent, and stop political war. Stopping political war, however, requires something more than sustaining trust, since we can sustain trust inappropriately. We instead want our trust in others to be grounded in good evidence that others are trustworthy,3 not in deception or manipulation. We want to trust others for good reason. Next, we want 3 Here I understand good evidence simply as indicators that persons are in fact trustworthy. I allow that good evidence can be mistaken in principle, so I do not mean to commit myself to either an externalist or an internalist account of how evidence indicates trustworthiness.

Must Politics Be War Here and Now?  23 trustworthy behavior to be driven by moral considerations. We want others to engender our trust based on a free choice to act in concert with their moral values and commitments, not due to violence or threats. To establish trust for the right reasons, then, we must find touchstones of trustworthiness that provide moral reasons to be trustworthy that, in turn, rationally justify trust through the observation of morally motivated trustworthy behavior.4 And for those touchstones to serve that function in a diverse society, they need to appeal to the diverse moral reasons of persons with distinct and incompatible worldviews. That is to say the touchstones of trustworthiness must be justified to each person, or publicly justified. I argued in MPBW that liberal rights alone can be publicly justified and that, therefore, such liberal rights practices uniquely have the capacity to create and sustain trust for the right reasons.5 In this way, liberal rights practices end political war and show that politics can be based in moral peace between persons.6 For a number of reasons, establishing trust on the right basis requires an inquiry into the real-​world causes of trust and trustworthiness. The most basic reason is that institutions can only sustain trust for the right reasons if they in fact sustain trust. If liberal rights practices sustain trust for the right reasons, they must sustain trust in the real world. In this way, the institutions that provide good reason to trust, specifically epistemic reasons to trust and moral reasons to be trustworthy, must also engender stable trust judgments among real people.

2.  Trust for the Right Reasons 2.1.  Trust and Trustworthiness First, I draw on the idea of interpersonal and social trust I employed in MPBW. I understand trust as a relationship in which one person trusts another to do a particular thing, meet a particular standard, or the like—​in technical terms, as a three-​place relation of the form A trusts B to Φ. Trust is more than an expectation of how others will act. We depend on those we trust, so for A to trust B, A has to have a goal that she can only achieve if B Φs, and she must believe that B is willing and able to Φ. Yet trust is more than mere reliance. For A to trust B, A has to think that B is a common participant in their shared moral practices and that B is a morally responsible agent—​that is, the sort of entity that can be held 4 Importantly, though, we must also assume that people are trustworthy because they are prepared to comply for moral reasons, in part or on the whole. This raises an important challenge about how we learn the motives of others, which I address in ­chapter 2, section 7. 5 I did not use the term “trust for the right reasons” in MPBW, but I often spoke of establishing trust “in the right way.” We will see why the new formulation is superior. 6 This thereby constitutes a defense a form of public reason liberalism.

24  Trust in a Polarized Age morally accountable. We can rely on our cars and computers, but we don’t trust them. Trust is restricted to persons we recognize as morally responsible agents. Finally, for A to trust B, A has to think that B Φs based on a certain kind of moral consideration, such as because B bears A goodwill, or because B sees herself as morally required to Φ on A’s behalf, or perhaps because B owes Φing to someone else. If A does not see B as normally acting from broadly moral considerations, say because she believes B Φs merely due to self-​interest, or out of bad will, then trust becomes much more difficult. Social trust is trust scaled up to the societal level. Social trust is trust held by society and placed in society conjunctively. A person socially trusts, therefore, not when she trusts her society as a collective agent, but when she thinks that most morally responsible agents in that society are trustworthy with respect to central moral rules. Thus, one socially trusts when she thinks other moral agents will tend to comply with the publicly recognized moral rules that she needs to rely on in her pursuit of her goals. And to socially trust, one must think others are generally willing and able to do their part to achieve those goals, knowingly or unknowingly, by following publicly recognized rules, and that others do so at least in part based on moral reasons. We socially trust others to follow central moral rules. A moral rule is a kind of social norm distinguished from other kinds of social norm because the normative expectations that back the norm have a moral character.7 We all think that others think we morally ought to follow the norm. Social trust involves trusting others to do what we collectively regard as the right thing to do, which is why we generally trust others to follow social norms rather than personal norms. We also trust people to follow central moral rules without which our main social institutions cannot function. We socially trust others to be honest, not to harm without cause, not to kill the innocent, to keep our promises, to show gratitude, to aid the impoverished, and so on. Social trust is also trust that others will comply with moral rules for moral reasons, at least much of the time, or that moral reasons would motivate trustworthy behavior in the absence of self-​interested considerations. These moral considerations are diverse: they might derive from our personal ideals, faith, or values, but they also might just be things we think are morally true.8 It is hard to trust others when we think they cannot be counted on to act from these basic moral considerations. If you learn everyone in your society is a sociopath, you will find it hard to trust them.

7 For further discussion of moral rules, see MPBW, pp. 31–​35. 8 In this way, moral reasons are varied and numerous, and they contrast with our reasons to violate our ideals, faith, or values, such as reasons of pure self-​interest or animus.

Must Politics Be War Here and Now?  25 The moral considerations must be substantive; if the motive is small or insignificant, our capacity to trust is undermined by the sense that the moral considerations in the trustee are unimportant. The considerations are non-​exclusive in order to allow for other motives, like self-​interest, though it does include compliance from bad will. And finally, the moral considerations need not be sufficient to guarantee compliance. The truster need only the see trustee as normally prepared to comply based on moral considerations, not that she always or even usually does so.9 This leads into an account of social trustworthiness. One is socially trustworthy when she tends to comply with the moral rules that are the object of social trust. Social trustworthiness is essential for social trust because it provides the right kind of reason to trust. Social trust is appropriate only when we reasonably believe, based on observation or testimony, that others are socially trustworthy. Further, to count as socially trustworthy, persons should be understood as usually disposed to comply with central moral rules for moral reasons. More precisely, for a person to count as socially trustworthy, moral reasons must normally motivate her actions, or would normally motivate them in the absence of self-​interested reasons. We must be able to count on people to not always do what is in their mere self-​interest. Now that we know what social trust and social trustworthiness are, we can define social trust for the right reasons in terms of epistemic reasons10 to trust and moral reasons to be trustworthy. The epistemic reasons to socially trust are that others are socially trustworthy; social trustworthiness is what makes trust beliefs rational. Evidence-​based trust is essential because if trust is based on poor evidence, it cannot long survive awareness of this fact and so will be less stable than trust based in good evidence. This is one reason to worry about trust that is produced immorally, say through deception or manipulation.11 The moral reasons to be trustworthy are more complex. These reasons need to meet two closely related conditions:  (i) they must be seen on reflection as 9 Readers of MPBW will notice that I have weakened the requirement on social trust that we trust others only when we think they follow moral rules for moral reasons that would always, or nearly always, be sufficient to motivate compliance with the relevant moral rule. I do this primarily because the argument of the book does not require this stronger condition. For my original account, see MPBW, p. 28. I thank Christopher Eberle for discussion on this point. 10 An epistemic reason to believe X, roughly, is a reason that contributes to making belief X rational. 11 Note that this point is compatible with the presence of a social norm requiring us to trust others when the evidence of their trustworthiness is unclear or insufficient. Perhaps we should default to presuming trustworthiness and only distrust others when there is evidence of untrustworthiness. I do not think that we can default to trust when the evidence of another’s trustworthiness is unclear or insufficient, however, since it is hard to rationally trust someone when it is unclear whether she is trustworthy, since the risk of betrayal or disappointment will be a live option. But it does make sense that we try to trust, that we don’t try to alter our trusting attitudes as they arise until we encounter untrustworthy behavior. I thank Jacob Barrett for encouraging me to be clearer on this point.

26  Trust in a Polarized Age morally appropriate motives, and (ii) they must ground our practice of accountability, such that acting contrary to those reasons is grounds for moral blame and punishment. I turn now to some of the technical details of the reasons that make trustworthiness appropriate, but the machinery has a purpose: to articulate our implicit commitment to sustaining trust appropriately.

2.2.  Intelligible Reasons to be Trustworthy* In MPBW, I called these reasons intelligible reasons.12 An intelligible reason is, first, any reason that an agent will regard as morally justifying her actions on reflection, given what she believes and values. Intelligible reasons, therefore, are not whatever reasons a person says that she presently accepts, but the reasons implied by her beliefs and values by way of valid inference. Intelligible reasons are therefore those that a person is rationally entitled to affirm. Thus intelligible reasons meet condition (i). An intelligible reason must also have a public dimension, in that other members of the trusted person’s society can see the reasons as reasons for her according to her own evaluative standards. That is, people can see on reflection that she has the reason, even if they reject the reason for themselves. I sometimes illustrate this idea by inviting the reader to imagine a Muslim, a Christian, and an atheist discussing public policy together. If the Christian said that she supported a policy based on the Qu’ran, then the atheist would balk. After all, Christians believe what the Bible teaches, not the Qu’ran, and the atheist, despite believing in neither, knows this. Thus, the Christian would have offered an unintelligible reason. But if the Muslim cites the Qu’ran, then the atheist can see that as an intelligible reason for the Muslim, based on the Muslim’s evaluative standards. Intelligible reasons must meet this weak publicity condition in order for others to be able to hold a person appropriately responsible for failing to act on those reasons. If we cannot even see moral reasons as moral for those who offer them, it is hard to trust them based on unintelligible reasons. It is hard to trust a Christian who pursues policies based on the teachings of the Qu’ran, and it is hard to hold someone responsible for a failure to live up to standards that are clearly not her own. An intelligible reason is one we can see another as endorsing by her own lights, such that it makes sense to hold her culpable or responsible for failing to act on the reason.



12

MPBW, pp. 88–​92,  96–​97.

Must Politics Be War Here and Now?  27 This, then, is why intelligible reasons are a good specification of moral reasons to be trustworthy: because they are reasons that the trusted person sees as moral, which trusters can appeal to in order to hold her responsible or decline to do so.

2.3.  Moderately Idealized Reasons to Be Trustworthy* We cannot determine which reasons are intelligible without appealing to some form of idealization. A person can have an intelligible reason even if she does not affirm the reason at present. An intelligible reason is one that an agent is rationally entitled to affirm after some amount of reasoning, which includes the collection of information, and making proper inferences based on that information.13 I ascribe intelligible reasons to persons based on the reasons they would affirm as their own if they were moderately idealized. Moderately idealized agents correspond to real persons, but they have reflected enough to respond to considerations that we would hold them responsible for ignoring. In this way, moderate idealization appeals to standards of information and inference that are not perfect but that are appropriate to our practice of responsibility. This, in turn, supplies the reasons on which trust and trustworthiness may be based, since the practice of trust and trustworthiness is a practice of responsibility. We hold others responsible for failures to be trustworthy, and we trust them less as a result. To clarify, notice that our ordinary practice of holding responsible contains an implicit idealization. We hold others accountable for engaging in conduct they did not know was wrong when they engaged in the action. In those cases, we say that the person we blame and punish should have known better than to act as she did. We appeal to idealization here because we expect persons to appreciate certain considerations that they are not currently aware of. We thus ascribe moral reasons to persons that they do not presently recognize but that we think that they could recognize with some effort. Notice that our ordinary practice of trust and trustworthiness implicitly appeals only to moderate idealization. We do not hold persons responsible for not knowing everything relevant to their decisions or for not engaging in flawless cognition.14 The case for moderate idealization is based on two arguments: an argument that we must idealize some and an argument that we must not idealize too much.15 The most fundamental reason to idealize is that idealization helps to 13 I develop a detailed account of idealization along these lines in Vallier 2014a, chap. 5. But I only connect idealization and trust in MPBW, pp. 97–​107. I discuss moderate idealization in more detail in appendix A. 14 In ­chapter 3 of MPBW, I spend a great deal of effort connecting moderate idealization to our practice of trust and trustworthiness. 15 For more detail, see MPBW, pp. 92–​3, 97–​106.

28  Trust in a Polarized Age identify the right reasons for trust. The reasons acknowledged by real people are not always good reasons, given the great limitations of our cognitive capacities and our generally middling moral character. By idealizing, we can ignore bad reasons and focus on the reasons moral agents will endorse on reflection. These psychologically accessible reasons drive trustworthy behavior and ground our practice of holding responsible. We must also avoid radically idealizing, say by ascribing reasons to persons based on full rationality and full information. Our reasons to be trustworthy would then be those that we would embrace only if we knew all the relevant information and had completed all the relevant reasoning. The principal reason to avoid radical idealization is because the reasons that morally motivate trustworthiness cannot be too far from our ordinary cognitive capacities. We have two types of moral reason to be trustworthy: reasons derived from our own evaluative commitments, and reasons to avoid ostracism, punishment, guilt, and shame from others when held appropriately responsible.16 Moderate idealization will help us identify both types of moral reasons. We idealize, as before, in order to avoid counting bad reasons as good ones. But radically idealized reasoning will not identify moral reasons that motivate ordinary persons to be trustworthy, so it cannot license our practice of moral responsibility. Radically idealized reasons are rationally inaccessible to agents and so cannot motivate ordinary moral agents. Similarly, these reasons cannot legitimize our practice of moral responsibility, since we can only hold persons responsible for not appreciating reasons that can be grasped through normal cognitive feats. I cannot hold my seven-​year-​old son responsible for ignoring the rationale for special relativity, even though there are good reasons for him to believe it. While real people will not always be motivated by the reasoning of moderately idealized agents, due to irrationality and bias, ordinary reflection can uncover the reasons and generate the associated motives.17 Appealing to moderately idealized agents has an added advantage: Because the intelligible reasons affirmed by moderately idealized agents rationalize our 16 Of course, real agents are also afraid of ostracism, punishment, guilt, and shame when others inappropriately hold them responsible, but those do not provide good moral reasons not to comply with a moral rule whose force is grounded in their own values and commitments. We can expect moderately idealized agents to have a bit of courage. 17 To expand on this point, real agents may engage in reasoning that does not measure up to moderate idealization. Thus, when they are presented with the deliverances of their moderately idealized counterpart, they may be hesitant to give up the reasoning they have actually completed. In MPBW, I argued that the standard of moderate idealization arises from our practice of trust and trustworthiness. If the standard of moderate idealization is grounded in our shared moral practice, then we can say to the real reasoner that she ought to reason further by her own lights, and this entitles us, in judging her, to replace her current reasoning with the reasoning of her moderately idealized counterpart. I address these issues in detail in Vallier 2020c.

Must Politics Be War Here and Now?  29 practice of moral responsibility, they can be used as a basis for criticizing real persons who depart from the rules justified by those reasons. We can credibly say that others should have known better than to depart from those rules. Therefore, if liberal order has the unique capacity to justify trust and trustworthiness across diverse perspectives, then we can criticize those who fail to comply with the directives of liberal regimes. We thereby accomplish one of the great goals of the social contract tradition: to show that there are regimes supported by the reason of the public and that have authority for citizens in those regimes. For moderate idealization will give us an account of sufficient, morally motivating reasons to endorse a political regime. And this is one way of morally justifying that regime.

2.4.  Internal, Nonexclusive, Deontic Reasons* The “right reasons” to be trustworthy are not necessarily given by the true moral theory. Instead, the “right reasons” criterion takes no position on which moral standards are correct. “Right reasons” are the reasons that agents take themselves to have. These are reasons internal, therefore, to our practice of trust and trustworthiness, reasons that people affirm as valid and that others see as such. The “right reasons” also do not exclude other reasons that lead people to trust one another. The fact that a liberal right is publicly justified shows that each person has sufficient moral reason of her own to go along with the norm. But it is common for people to have multiple reasons to trust others socially or politically. The only bad reasons to trust are those grounded in inadequate evidence; the only bad reasons to be trustworthy are those that are in tension with one’s moral convictions, like manipulation, threats, and brainwashing. Fortunately, the reasons persons trust and are trustworthy are seldom based solely on these inappropriate factors. Finally, we want the reasons driving trust and trustworthiness to be reasons other than that trust and trustworthiness have good effects. If we only appeal to instrumental reasons, this may tempt some to free ride on the systems of trust they find themselves in because they can garner the social benefits of trust without being trustworthy themselves. But if a person’s reasons to trust and be trustworthy derive from her own moral perspective, such that she sees herself as under a duty or obligation to be trustworthy, then these noninstrumental or deontic reasons will give her powerful motivation to be trustworthy, so powerful that it may combat her temptation to free ride on the trust and trustworthiness of others. We have some reason to think people are often in this situation because we have empirical evidence that people generally see trustworthy

30  Trust in a Polarized Age behavior as morally required, regardless of their diverse commitments and beliefs, so a duty of trustworthiness is a widespread basis for moral approbation or disapprobation.18

2.5.  Social Trust for the Right Reasons We can now provide a formal definition of social trust for the right reasons. Social trust for the right reasons: a society enjoys social trust for the right reasons when its social trust is grounded in adequate evidence available to every member that others are socially trustworthy because each is normally prepared to comply with moral rules from her own intelligible moral reasons.19

This is the central normative notion in this book. I want to establish that liberal rights help to generate trust for the right reasons. This requires a series of two-​ part arguments. In each, I first argue that the empirical literature shows that a certain liberal right creates real trust, then show that the trust is grounded in the right kind of reasons because the right in question is publicly justified. As noted previously, my strict definition of public justification is that it occurs when a moral rule is justified for each person by way of her moderately idealized intelligible reasons; in this way, public justification provides good moral reason to be trustworthy, creating touchstones of trustworthiness that justify and motivate social trust. More simply, when a moral rule is justified to each person, everyone has reason to be trustworthy in a public way, which helps everyone trust one another. We can extend social trust for the right reasons to trust in legal and political systems. So, political trust for the right reasons will obtain when a society trusts a political institution and that trust is grounded in adequate evidence that members of that institution are generally politically trustworthy based on their intelligible reasons. To vindicate my thesis as to each liberal rights practice, I must show that it helps to create or sustain (or both) social or political trust (or both) for the right reasons.

18 Bicchieri, Xiao, and Muldoon 2011. 19 I define intelligible reasons in this chapter, section 2.2. An intelligible reason is one that we can see as a reason for the person who offers it in accord with that person’s own evaluative standards. Also, note here that I do not say that one must be prepared to comply with moral rules from intelligible moral reasons alone, or that her actions must always be motivated by her intelligible moral reasons, in accord with my weakening of the moral motivation requirement for trust in MPBW.

Must Politics Be War Here and Now?  31

3.  Must Politics Be War? The Question Specified Based on the foregoing discussion, I can now provide a precise formulation of the question of whether politics must be war: Are there any moral, legal, and constitutional rules that sustain social trust for the right reasons under conditions of evaluative pluralism?

If some moral, legal, and constitutional rules meet this description, and if those rules and their associated institutions are themselves feasible for us, then a non-​warlike politics should be feasible for us, because trust discourages institutionalized aggression. To end political war, therefore, we need institutions that create and sustain social and political trust for the right reasons. In particular, we need institutions that can sustain a high-​trust equilibrium. When trust falls below high levels, these institutions should help return trust to previous (or even higher) levels of trust through a process of self-​correction. I will not argue that there are institutions that can lift us out of a low-​trust equilibrium. Sometimes trust is so low that escaping a low-​trust equilibrium is impossible or extremely difficult. But, fortunately, many countries have sufficiently high trust levels that creating high-​or higher-​trust equilibria is feasible for them. And in the case of the United States, it is not clear we are in a low-​trust equilibrium. Our trust levels are not especially low, and our trust level seems to be changing. I focus on moral, legal, and constitutional rules because people under conditions of evaluative pluralism come to trust one another by observing compliance with mutually morally acceptable social rules or determining that others have complied through testimony.20 Rule compliance serves as a touchstone of trustworthiness because people with different perspectives can jointly observe compliance with rules they know others have sufficient reason to accept and obey. I appeal to people faced with evaluative pluralism because we are moral agents under conditions of evaluative pluralism like this. So, if we can identify institutions that sustain social trust for the right reasons between such individuals, we will have gone a long way toward showing that it is feasible for our societies to sustain a high-​trust equilibrium under some kinds of legal and political institutions. Finally, I draw on moderate idealization for reasons canvassed in the previous section. 20 It is important that the rules are social norms, since it is publicly held that we each morally should comply with the norm. If the rules are not social norms, then the rule is not backed by morally normative expectations, in which case the joint recognition of normative pressure to comply will be absent, making it much harder to use norm compliance to convince others that one is acting on moral considerations. I thank Scott Simmons for pressing me to clarify this point.

32  Trust in a Polarized Age So, must politics be war? Not if politics is grounded in social rules that sustain trust for the right reasons between moderately idealized moral agents under conditions of evaluative pluralism. We have now at least asked the right question.

4.  The Problem of Political War Framed in Terms of Feasibility* In this section, I want to say a bit more about how to frame my project in terms of the feasibility of trust and liberal order. In MPBW, I asked whether political life is invariably a struggle between factions competing for political hegemony. But the titular question “Must politics be war?” does not explain how the “must” is to be understood. On my view, politics need not be war because there is a possible world where politics is not war.21 But not just any possible world will do. We need a set of possible worlds narrower than the set of logically possible worlds, the states of affairs that contain no logical contradictions—​indeed, narrower than the set of nomologically possible worlds, the states of affairs that obey our laws of physics and the laws of nature generally. Nothing in the laws of logic or of nature entails that politics is war. But that alone does not mean that a non-​warlike politics is available to us. So, I am interested in describing a practically possible world, one that is feasible. I want to show that politics need not be war in the actual world because it is not war in a nearby feasible world.22 In other words, a peaceful politics is a state of affairs that we can reach with significant, but not overwhelming, effort.23 There is a growing literature on the proper conception of feasibility in political philosophy, and I would love to review it here, but I lack the space. Further, Nic Southwood has raised a number of serious issues for all three major approaches in the literature, cost-​based accounts, probability-​based accounts, and possibility-​ based accounts.24 He also points out that conceptions of feasibility haven’t been applied to institutions as often as they have been applied to individual and group agents. And since many institutions aren’t even group agents, it can be hard to apply a notion of individual or agent-​based feasibility to determine what 21 To clarify for readers not grounded in analytic philosophy, a possible world is a world we can imagine without internal contradiction. More precisely, it is a description of a complete states of affairs that contains no contradictory propositions; the actual world is a possible world, and all other possible worlds contain at least one counterfactual proposition in the set of states of affairs that comprise that possible world. 22 The nearness of a possible world or a feasible world is specified by how much the states of affairs in the two compared worlds differ from one another. 23 I take it this statement is general enough that it does not commit me to a particular account of feasibility. 24 Southwood 2018.

Must Politics Be War Here and Now?  33 is institutionally feasible. So, I will here articulate the notion of feasibility that I used in MPBW and use in this book. My aim is not to provide a new definition of feasibility but, rather, to state some conditions that correlate with or indicate whether an arrangement is feasible. I want to describe an institutional arrangement as feasible without committing myself to a particular account of feasibility, so I’m looking for an intuitive heuristic that will help track feasibility. I will understand an institutional arrangement as feasible when it can be sustained by persons acting on reasons that they are morally responsible for attending to and blameworthy for ignoring.25 In brief, feasibility co-​occurs with what we ordinarily think persons are responsible for doing or failing to do. Consider, as an illustration, our practice of holding other motorists responsible for following informal traffic norms like allowing others to merge into oncoming traffic. We expect motorists to notice other vehicles and to care about their drivers’ need to merge into oncoming traffic free from fear for their safety. We then expect others to recognize a moral reason to let other drivers through, namely to help them avoid serious injury, and we blame motorists who ignore that moral reason. In this way, through appropriate disapproval and sanction, it is feasible to maintain the informal norm of allowing motorists to merge into oncoming traffic if the norm can be sustained by moral reasons that persons are ordinarily responsible for attending to and blameworthy for ignoring. The feasibility of following these informal traffic norms tracks what we think motorists are responsible for. I associate feasibility and responsibility because I  think our practice of responsibility contains an implicit, shared conception of feasibility that forms the basis for evaluating ordinary moral behavior. When I hold someone responsible for ignoring a moral reason, I ordinarily presume that the person is capable of attending to the reason and acting accordingly. In short, I presume that complying with the reason is within her power, or feasible for her. It matters that the notion of feasibility is implicit in our moral practice, since this suggests that we pretheoretically share an understanding of feasibility that philosophers can articulate or define in various ways.26 25 Here I will simply set aside the large range of fascinating questions about the claim that “ought implies can,” as I assume some version of the claim is true. If we think others ought to comply with certain reasons, we normally think that they are capable of doing so. This is not because being morally required to do something gives them the power to do it; it is because we do not think people are required to do what they are unable to do, and therefore, if we thought the person was unable to do the thing, we would not think the person was required to do it. 26 I’m also attracted to the view that feasibility is grounded in facts about responsibility. An arrangement is feasible if and only if we are responsible for bringing about that arrangement or preventing it. But perhaps a conception of feasibility should explain why we are responsible for bringing about an arrangement or preventing it, and not the other way around. So I set aside whether feasibility can be made sense of in terms of responsibility judgments. Going further, we presume that there are answers to questions about who is responsible and what moral norms apply. An interesting consequence of this presumption is that even if it is currently infeasible to establish a non-​warlike

34  Trust in a Polarized Age We have a recipe for determining when political peace is feasible. Political peace is feasible when it can be brought about by persons acting on moral reasons that they hold each other responsible for attending to and blameworthy for ignoring. Peace will emerge when people respond to the moral reasons they have the ordinary capacity to grasp and be motivated by. This will yield trustworthy behavior, and, in turn, trust. Since feasibility is tied to what we are responsible for, and we standardly assume that abiding by our responsibilities is within our power, we have an attractive notion of feasibility that we can use to describe the problem of political war. We must now address whether political peace is feasible in the aggregate. After all, even if it is feasible for each person to respond to the reasons that we would hold her responsible for ignoring, it may not be feasible for everyone to respond to these reasons in concert. As Southwood notes, “Even if it is feasible for every individual member to do her part in a joint activity, it is a further question whether it is feasible for members of the group to work together in ways that are required for successful group action.”27 Fortunately, my account of political peace is not grounded in a group action. Peace is an emergent property of interactions among diverse persons when they trust and are trusted. So, we do not need to explain the feasibility of group action. But we must still address whether political peace can feasibly emerge from social interactions in which we trust and are trusted. I answer this challenge at length in c­ hapter 2, where I explain how social scientists measure social and political trust. Insofar as political peace is the product of emergent trust, the existence of emergent trust should help establish the feasibility of political peace because peace arrives just when there is sufficient emergent trust. That’s the best way to show that a state of affairs is feasible—​to show that it is so.28 Thus, we can say, at least provisionally, that political peace is feasible based on (i) our holding others responsible for being untrustworthy and (ii) the scientific measurement of real trust. In what way is it feasible for real-​world institutions to sustain political peace? How do institutional rules and compliance with those rules make political peace feasible for us? The short answer is that we have established political peace when diverse persons embedded within institutions lack a robust propensity to distrust one another in that they can maintain trust without a constant drag on politics, our presuming there are answers to the problem may make moral peace more feasible over time. I thank Alex Motchoulski for this insight. 27 Southwood 2018, p. 4. 28 Though in some cases, the actual need not be feasible if, say, the state of affairs was reached through a fluke. See Southwood and Wiens 2016. I’m grateful to Jacob Barrett and Nic Southwood for this point.

Must Politics Be War Here and Now?  35 their trust levels.29 An institution makes peace feasible when those bound by the institution’s constitutive norms lack a robust propensity to fall victim to the illusion of culpable dissent because they grasp moral reasons to act in ways that cause political peace to emerge, namely by trusting and being trustworthy. In this way, liberal rights practices solve the problem of political war when people in societies with those practices lack a robust propensity to distrust those with diverse and conflicting values and beliefs. A liberal politics need not be war because liberal institutions have no ingrained tendency to undermine social and political trust.

5.  Public Justification, Legal Systems, and Constitutions* To explain why politics need not be war, we need a test to determine which legal and political rules will be endorsed by moderately idealized agents. The central idea here is public justification, the normative idea at the heart of public reason liberalism. To properly sustain trust and trustworthiness, the moral rules that are the object of trust and trustworthiness must be publicly justified, or justified to each person based on her intelligible reasons. Thus, the object of public justification is a moral rule. The reasons in which public justification consists, its currency, are intelligible reasons drawn from moderate idealization. Consequently, moderately idealized agents are the subjects of public justification. I have previously called these agents members of the public. Importantly, it is not enough for people simply to comply with the rule. They must internalize it as well, by habituating themselves to the rule sufficiently that they are psychologically motivated to comply with the rule and experience the appropriate reactive attitudes when they or others violate the rule.30 With accounts of the subject, object, and currency of public justification, I can define public justification as follows: Public Justification Principle:  a moral rule is publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the rule.

In some cases, the “eligible set” of publicly justified moral rules has more than one member.31 Here an eligible set can be understood as including all proposals

29 I  borrow this language from unpublished work by Southwood and Kai Spiekerman, which Southwood mentions in Southwood 2016, p. 35 n. 51. 30 For an important discussion of internalizing moral rules, see Gaus 2011 pp. 202–​5. 31 Gaus 2011, pp. 321–​31; MPBW, pp. 110–​2.

36  Trust in a Polarized Age that persons lack sufficient reason to reject. Some eligible sets may have only one member, in which case that member is publicly justified. Other eligible sets may have more than one member, in which case we need some kind of social process to select one for public justification. But if the eligible set has more than one member, any member can be publicly justified, at least if selected by a proper process. I have spoken to the issue of procedures elsewhere.32 Most public reason views focus on the public justification of law and legal coercion rather than moral rules. I start with moral rules because they are constitutive of our practice of social trust, whereas the legal system is only part of this broader moral system. Further, my approach to trust and public justification enables us to explain why and how to publicly justify the law. But now that we have explained how to publicly justify moral rules, we can develop an account of how to publicly justify laws. To see this, we should understand a law as a supremely authoritative social rule typically coercively enforced through legal sanction by publicly recognized law-​applying institutions like courts.33 Laws resemble moral rules. They are rules backed by empirical and normative expectations: we ordinarily expect others to comply with the law, and community members all think they ought to do so. Laws are authoritative in that they are regarded as obligatory and violators are typically thought guilty. Laws are not mere advice or counsel but provide weighty reasons for obedience. Law is also tied to the existence of institutions that can authoritatively settle disputes about the law; laws require courts (though not legislatures, importantly). Laws are also typically backed by coercive force, such that the sanctions for violating the law are usually much more severe than the sanctions for violating moral rules, as well as more limited and formalized in scope. Legal systems combine law with adjudicative institutions. Adjudicative institutions, such as courts, settle disputes about the law based on their own legal norms. These systems also claim supreme authority over other institutions to interpret and apply the law. We need law because a social order that contains only moral rules will be uncertain, ineffective, and static, and so we will need laws to help solve these problems.34 When the law is used to solve these problems with the moral order, law can gain practical authority over persons; they will have good moral reason to obey the law. In MPBW, I  argued that one has a moral obligation to obey 32 MPBW, pp. 128–​32. 33 Ibid. Here my account of a system of social norms is related to Pettit’s account of social norms, and their structural similarity in their authority to legal norms is buttressed. See Pettit 2019. 34 Though we should not be too keen on the effectiveness of law, as we might get more order with less law in some cases (Bohnet, Frey, and Huck 2001, pp. 141–​2). Bohnet, Frey, and Huck also discuss the possibility of legal trust crowding out social trust, but as long as there are relations of trust, it is not essential that the norms be nonlegal in character. The sort of trust I’m after need not always be social trust but can be legal or political as well.

Must Politics Be War Here and Now?  37 a law if and only if she rationally believes that that legal system improves her mere moral order, or an incomplete legal system, and that obeying the laws generated by those legal systems improves upon her capacity to fulfill her moral obligation(s).35 So, laws are publicly justified when they help us to follow moral rules that are justified for us or amplify the existing legal order in a similar fashion.36 On this basis, we can formulate an account of legal trust: Legal Trust:  a public exhibits legal trust to the extent that its participant members generally believe that (i) legal officials are necessary or helpful for achieving public goals, (ii) legal officials are generally willing and able to do their part to achieve those goals, knowingly or unknowingly, by playing their prescribed legal role, and (iii) moral reasons normally motivate the legal officials to play that role.

The significance of legal trust vis-​à-​vis social trust is that legal trust is trust in particular legal officials to perform certain functions for moral reasons. The goals of these officials’ activities are typically public in that they are regarded as important by the community as a whole, like criminal justice. Our primary acts of legal trust are in trusting people to perform public functions that achieve these public goals. We sometimes trust legal officials to abide by moral and legal rules even if they are not consciously aiming at the public goal, so long as their obeying the rules helps to achieve the goal. We can now introduce a principle governing the public justification of the law: Legal Justification Principle: a legal rule is publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the law because each member rationally recognizes that compliance with the law improves upon her capacity to comply with one or more publicly justified moral rules and other legal rules.37 35 In MPBW, I understood improvement in terms of “moral efficiency,” which obtains when the balance of one’s justificatory reasons favors compliance, and not just reasons bearing on how an individual achieves her personal goals. See MPBW, pp. 148–​9. 36 In MPBW, I  argued, following Barrett and Gaus 2020, that often laws cannot change social norms because the norms are so engrained that the law will not be followed. But when social norms are not seen as justified, laws can perhaps “serve to chip away” at their support. For more on this, see Muldoon 2018, p. 144. 37 Once a number of laws are in place, the public justification of further laws may refer to prior legal arrangements and, in some cases, must refer to them. Jacob Barrett raises the possibility of someone approving a coercive law that would never help her comply with other rules but will help others. But I take it that if the law is not coercing her, then the legal justification principle does not apply to her in that case.

38  Trust in a Polarized Age We also need rules that can authoritatively change legal rules:  constitutional rules. I justify constitutional rules in three stages. The first stage identifies a scheme of basic rights that citizens adopt as constraints on public coercive power; these rights are then used to reject constitutional rules that would normally violate those rights. The second stage ranks the remaining unrejected constitutional rules by their minimization of legal error, that is, by how well they avoid both the imposition of unjustified (defeated) law and the failure to impose publicly justified law.38 The third stage involves showing that these constitutional rules can be stable for the right reasons. Stability for the right reasons, however, must be disaggregated into at least two other concepts. First, we should seek constitutional rules that help ensure conditionally cooperative agents that they are all prepared to cooperate with one another;39 this is what makes a constitutional rule durable. We should also adopt constitutional rules that discourage defection by more egoistic agents who engage in corruption and rent-​seeking; this is what makes a constitutional rule immune. Constitutional rules must therefore be stable in the following sense: Constitutional Stability: a constitutional rule is stable if and only if it is immune (in the specified sense) to abuse by political officials and durable (again, in the specified sense) among citizens subject to the rule (in virtue of the laws produced by the rule being durable and immune among citizens subject to the rule).

For a constitutional rule to be stable, that is, durable and immune for the right reasons, the laws that they produce must generally be stable in the same way. If the constitutional rule does not produce generally stable laws, that makes the constitutional rule not only less attractive but less stable, since its outputs are not themselves stable. We can now turn to the idea of a primary right as a kind of constitutional rule. A primary right is a moral claim-​right40 that anyone with a rational plan of life would want for herself in order to realize her conception of the good and justice and that she is willing to extend to others on reciprocal terms. Primary rights must satisfy several conditions to be publicly justified. They must be equal and

38 The term “defeater reason” on my view refers to a reason that successfully undercuts or rebuts a proposed justification for some moral, legal, or constitutional rule. One has a defeater when a law is not justified for her because of the defeater reason in question. 39 In MPBW, I defined conditionally cooperative agents as those prepared to cooperate if they believe others are likely to do likewise. For further discussion, see MPBW, chap. 6. 40 A person has a claim-​right to x if she has no duty not to x or have x, and others have an obligation to allow the person to x or have x.

Must Politics Be War Here and Now?  39 coherent: all persons have the same primary rights, and publicly recognized legal and political systems must resolve conflicts between primary rights when they arise. Primary rights can be negative rights (understood as rights against interference) or positive rights (understood as rights to assistance from others). They are rights to enjoy resources, which the philosophical reader can understand in terms of Rawlsian primary goods or along the lines of the capabilities approach to the distribuendum of distributive justice (the thing to be distributed), or other, related accounts.41 But however the resource is specified, persons will be entitled to access or pursue wealth and income, as well as capital goods. This is because primary rights are claims to make effective rights-​protected choices, and wealth, income, and capital are required to make those choices. Importantly, both negative and positive rights can secure these resources. When primary rights are publicly justified and entrenched in constitutional law, those rights are authoritative for real members of that society. This means that each person has moral reason to respect those rights when exercised by others. When members publicly respect those rights, they show that they are trustworthy with respect to the relevant regime of rights. They thus give one another evidence that justifies trust—​that is, evidence that provides sound epistemic reason to trust. So if its members generally do what they all have reason (grounded in each person’s own evaluative standards and her available evidence) to do, a society whose constitutional, legal, and moral orders are structured by the recognition and exercise of publicly justified primary rights can create and sustain high levels of social and political trust. In that case, politics will not be war. Moral rules are publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the rules. Legal rules are publicly justified when those subject to them see them as improving on the order of moral or legal rules. Constitutional rules are publicly justified when they tend to protect publicly justified basic rights, and to generate laws that are, on balance, publicly justified themselves, and that, together generate certain kinds of political stability. A non-​warlike politics flows from publicly justified constitutional rules, including constitutional rules that enumerate and protect basic rights. Since liberal order alone can be publicly justified, it alone sustains social, legal, and political trust for the right reasons. Thus, a fully liberal politics is not war.

41 I discuss these accounts in MPBW, pp. 163–​5. A primary good can be understood as any good that is essential for people to develop and exercise their capacity to form and pursue a conception of the good, whereas capabilities are person’s real freedoms to realize certain types of essential abilities.

40  Trust in a Polarized Age

6.  The Derivation of Liberal Rights To summarize the previous section, we count a moral or legal rule as publicly justified when each person has sufficient moral reason of her own to endorse and internalize that rule. The aim of public justification is to show that people from a range of diverse perspectives can endorse the rule as regulative of their behavior on the basis of their own personal commitments and values. But this does not tell us what rules and norms they can converge upon. I contend that they should be able to converge on a scheme of basic, primary rights. To show that liberal rights are publicly justified, I appeal to a “veil of ignorance” argument, though one quite different from John Rawls’s famous version.42 A veil-​of-​ignorance argument, in brief, tries to identify one or more schemes of rights that are mutually acceptable to parties to a social contract by denying contractors knowledge of factors that would bias them into endorsing rights and liberties that are hegemonic or sectarian to other contractors. As I argued in MPBW, my veil is only a heuristic to locate rights schemes that can be justified to multiple perspectives.43 Further, my veil is quite thin. It allows people to recognize a huge range of special facts about themselves and their environment, including facts about the failings of real people, like their tendency not to comply with the law. But the veil prevents persons from knowing whether they are the political hegemon or whether political hegemony is within their grasp. This is because they are restrained by a broader veil that denies them knowledge of their degree of social power or social status, as well as the social power or status of others. My veil of ignorance is just a power-​and-​status  veil. Depriving people of knowledge of their status and power does not require that they detach from their personal commitments or set their values aside; the veil is simply an abstraction that we can use to identify the rights claims we can fairly make of others. These rights should therefore not depend on our status and social power, even if they are shaped by our faith and values. Under these conditions, parties will adopt liberal rights as a risk-​avoidance strategy.44 They will recognize that it is generally better to have the liberty to 42 According to Rawls, the basic structure of society should be the one that parties agreeing to a social contract would choose if they had to choose from behind a “veil of ignorance” that prevented them from knowing their place in that structure, their personal characteristics, their life goals, etc. (Rawls 1971, pp. 136–​42). 43 Here my story somewhat mirrors the thin veil of ignorance discussed in Moehler 2019, p. 141. Unlike Moehler, however, I do not have a premorality contract; I begin the contract with a preexisting system of social trust. 44 Here my work is informed by North, Wallis, and Weingast 2013. North, Wallis, and Weingast argue that liberal democratic institutions (which they call “open access” states) emerged because political elites in previous, “natural” states found it more beneficial to credibly commit to restrictions on their power than to risk all-​out conflict. Even power elites in authoritarian societies may be willing to give up some power to protect against major costs like execution if they lose a conflict or just face back luck.

Must Politics Be War Here and Now?  41 pursue one’s own form of life, or the form of life of one’s group, than to risk being controlled, coerced, and dominated by those with rival and incompatible values. This will likely lead to the joint endorsement of a scheme of expansive, equal liberties. Even illiberal parties will choose liberal rights schemes because, while many of them dislike liberal rights, they will nonetheless tend to prefer liberal rights to a scheme of rights that risks making another sectarian group into the political hegemon. Muslims and atheists are liable to prefer a liberal regime of religious freedom to a Christian theocracy. Christians and atheists will prefer a liberal regime of religious freedom to a Muslim theocracy. And Christians and Muslims will prefer a liberal regime of religious freedom to an aggressively secular regime. My argument is not simply that liberal rights are everyone’s second choice. Parties are not merely settling for a rights regime because it is the best they can hope for. Rather, the rights chosen are those that persons can all comply with from moral conviction because primary rights protect each person’s ability to pursue her own projects and principles by herself and in concert with others. Thus, while these arrangements will give almost no one everything she wants, morally speaking, the regime will be sufficiently moral that it can claim the allegiance of its subjects, even if it is not morally ideal. This is true, I think, even for nonliberals, since truly liberal regimes secure liberty and integrity for each person and group equally. In this way, a system of moral, legal, and constitutional rules that protect a wide range of equal primary rights to resources for each person can be publicly justified to all. This means that diverse persons can comply with these rights freely, rationally, and in concert with one another. In virtue of an “overlapping consensus” or convergence on these rights, the recognition and exercise of liberal rights serve as public touchstones of trustworthiness for diverse persons.45 Each person can comply with and exercise these rights for moral reasons, displaying her trustworthiness with respect to the relevant rights practices. Others can, then, observe this compliant behavior and rationally increase their degree of trust in diverse others. In this way, people with diverse perspectives can rationally trust one another to abide by the rules of the regime. The resulting social and political trust ends political war and creates a moral peace. Ultimately, the bundles of rights chosen by parties behind the power-​ and-​status veil will have to be assessed by each group given their own diverse commitments. But since the veil already allows persons to appeal to their own diverse reasons, the rights choice when persons know their social power and status should be stable. Idealization should prevent unveiled persons from adopting

45

Rawls 2005, p. 133.

42  Trust in a Polarized Age hegemonic regimes based on biased and self-​interested reasoning. On top of this, for all of us, part of the case for adopting and abiding by a rights scheme is that the costs of conflict, instability, and mistrust are steep; trying to impose a regime that would not be acceptable to other groups would entail these costs, which a regime everyone would accept behind the veil avoids. However, the ultimate story about how each group can internalize the veiled choice of rights must be left to each group.46 Liberal politics is not war, therefore, because we can describe a feasible state of affairs47 where partially idealized versions of ourselves trust one another based on the trustworthy behavior created by a regime of liberal rights. In that condition, liberal politics is not war because liberal institutions sustain trust for the right reasons and thus end political war. And since that state of affairs is feasible, liberal politics need not be war for us in our current social condition. To put it one more way, liberal politics is not war because liberal rights are in a unique position to sustain social trust for the right reasons, ending political war. They are in this position because they can be publicly justified, or justified to each person in terms of her own moral reasons, commitments, and values. Recognizing that liberal rights are compatible with each person’s moral reasoning and agency, we can each see reasons of our own to respect those rights, and we can see that others have their own reasons to do so, even if their reasons are different from ours. As a result, liberal rights serve as touchstones of trustworthiness; by respecting those rights, we act from conviction, signaling to those we disagree with that they can depend on us. Liberal politics, therefore, should create and sustain trust for the right reasons.

7.  The Relevance of Real Trust But to fully demonstrate that liberal rights create trust for the right reasons, I must appeal to the empirical literatures on trust. Appealing to these literatures can deepen my analysis and strengthen my arguments in six different ways.

46 This is how I reply to the concerns along these lines found in Barrett 2020a. My strategy is similar to those pursued by Rawls 2005, pp. 386–​7, and Gaus 2011, pp. 359–​368. I discuss this objection further in appendix A. Two more points from Gaus: first, Gaus also argues that real people’s moral assessments tend not to be affected much by knowledge of their social position. If Gaus is right, that helps answer worries about bias. Second, another of Gaus’s arguments is that members of the public will recognize the great social costs of violating veiled reasoning; in this book, I extend Gaus’s argument by quantifying the noticeable social costs of unjustified institutions in terms of lost trust. Note that Gaus does not adopt my particular veil of ignorance. 47 Here a close feasible world is a possible world, an abstract description of an alternative state of affairs, and it is close and feasible because the state of affairs is much like our own and can be reached through ordinary human effort.

Must Politics Be War Here and Now?  43 Creating trust in the first place. Part of creating trust for the right reasons is creating trust. And the empirical literatures help to determine which institutional rules and rights practices cause trust. We need to appeal to public justification to determine whether these rules create trust for the right reasons, but if publicly justified institutions don’t even produce trust, then they will not create trust for the right reasons. Aligning the causes and justifications of trust. Imagine that we establish the institutions that furnish the rational basis for trust but, for unanticipated reasons, these institutions systematically undermine social and political trust in the real world. In brief, suppose the right reasons to trust turn out to conflict with the real causes of trust. If the institutions that justify trust reduce real trust, that is cause for concern. Real trust has positive consequences. Social arrangements that justify trust while undermining it, therefore, have bad consequences. We should therefore hope that the causes and justifications of trust align. Congruence with our good. Social contract theories attempt to show both that the contract establishes what we morally ought to do and that complying with the terms of the contract is somehow good for us. If publicly justified institutions undermine trust, then that is reason to think those institutions aren’t good for us.48 Similarly, if liberal rights practices undermine trust, those practices harm people, undermining their good. Proper idealization. If the causes and justifications of trust diverge, theorists should consider whether they have engaged in bad idealizations. If a theorist idealizes persons in order to show that certain institutional rules provide persons with reasons for action, and real persons complain that they can see no good or sufficiently weighty reason to comply with those rules, then perhaps the theorist has erred. Idealizing real persons risks getting each person’s commitments wrong, given how difficult theorists may find identifying those commitments. So, theorists should pay close attention to what persons say their reasons are. And they should pay that attention even if what persons say their reasons are is not definitive of the reasons they actually have; again, theorists can easily misidealize. Illuminate new sources of trust. By consulting the empirical literature, we will uncover hard-​to-​anticipate ways that liberal institutions create trust. For example, a lot of political trust is created simply by routine, noncorrupt interactions 48 Rawls spent much of A Theory of Justice trying to explain how the right and the good “congrue,” that is, why it is good for us to do the right thing, as that is specified by his social contract. If it turns out that Rawls’s proposal, which he calls “justice as fairness,” systematically undermines each person’s good, that is a problem for Rawls. That said, Rawls is concerned with an extreme idealization, a well-​ ordered society, and is only interested in congruence in that situation. So, it is unclear how worried Rawls would be if justice as fairness undermined personal good in the real world (Rawls 1971, p. 350). I, however, am trying to show that liberal rights practices can bring about political peace in the real world.

44  Trust in a Polarized Age with civil services, and political distrust often arises simply from bad encounters with the people who staff these offices. Insofar as liberal institutions regularize noncorrupt bureaucratic practices, they will create political trust. We also see that some apparently good ways of creating trust may be illusory. For instance, one might antecedently think that reducing economic inequality will increase social trust, but redistribution is probably not a good way to create trust. Instead, trust facilitates redistribution and higher levels of economic equality. I explore these claims in future chapters, especially c­ hapter 2. Motivating feasibility. In ­chapter  2, and in portions of each subsequent chapter, I argue that the recognition, protection, and exercise of various liberal rights creates real social and political trust. That is an excellent way to substantiate the feasibility of trust for the right reasons and political peace. If liberal rights practices create actual trust, then trust for the right reasons is feasible if the rights practices are publicly justified; for the practices produce real trust in the real world.49 Of course, some countries cannot feasibly reach the trust levels in other countries, but the fact that some countries have reached high levels of trust helps to strengthen the belief that others can too. Further, by showing how liberal rights practices create real trust, we can illustrate a feasible path toward high-​ trust institutions. We not only show that a high-​trust equilibrium, once reached, can be sustained, we help to show that we can reach a high-​trust equilibrium from our present state. We can now see six reasons to explore the empirical literatures: we can thereby (i) figure out how to create real trust, (ii) align the causes and justifications of trust, (iii) demonstrate congruence between publicly justified institutions and our personal good, (iv) verify that we are using proper idealizations, (v) illuminate new avenues for the development of trust and trustworthiness, and (vi) demonstrate the feasibility of political peace.

8.  Trust through Rights Practices Behind a thin veil of ignorance that deprived people solely of information about their relative power and status, individuals would tend to endorse strong protections of their basic positive and negative liberties and insist on resources sufficient to make the exercise of those liberties worthwhile. This list of rights would be liberal in the specific sense that it would accord extensive, equal rights to everyone, regardless of race, gender, class, faith, or even ideology. Liberal 49 Since liberal rights practices are feasible, given that we have such practices and that they are plausibly sustained at least in large part through morally responsible actions, the products of those practices are feasible in the same way.

Must Politics Be War Here and Now?  45 rights prohibit the state or private parties from using force to support a particular sectarian group, requiring a kind of neutrality between the active doctrines and belief systems in that social order.50 In MPBW, I argued that parties behind the veil would adopt at least four classes of rights, and associated rights practices, that serve as touchstones of trustworthiness. In this book, I also argue that they create or at least sustain trust in the real world. These classes are agency rights, associational rights, jurisdictional rights, and procedural rights. I focus on specific associational, jurisdictional, and procedural rights in subsequent chapters, but they all presuppose agency rights.

8.1. Agency Rights Agency rights protect the formation of coherent agent psychologies and the minimal capacity of persons to extend their projects, plans, and values into the external world. Since agency is required to pursue any conception of the good or of justice, everyone would want some substantive, weighty rights of agency for themselves and others. Given how important agency is, people unaware of their social power and status would sacrifice their potential authority to control the agency of others in order to secure authoritative protection for themselves. So, the rights of agency they would choose would secure the most essential resources and protect the most essential liberties to a great degree; moreover, those agency rights would have strong priority over other rights. In other words, these rights would be used to specify and limit other rights in cases of conflict. Agency rights can be negative or positive. Negative rights of agency include freedom of thought and protections from various types of mental or physical harm. They also include the right to life and bodily integrity, freedom of speech, and the formation of intimate relations with others, such as marriage or other contractual arrangements. Persons also have a primary right to own personal property and to freedom of occupation; the state may not seize personal property or dictate the terms of one’s work. I argued in the previous book, and I will argue again in ­chapter 4, that agency rights include a right to own private capital. We can justify this right based on more easily justified rights to freedom of occupation and personal property. If persons have a right to freedom of occupation,

50 Thus, I do not say that the state can take no position on whether anything is good in any sense; instead, the state must not promote goods whose goodness are denied by reasonable members of that society. States must therefore be what I call contextually neutral, neutral between the active doctrines and beliefs systems in that order.

46  Trust in a Polarized Age this should include a right to go into business for oneself by using one’s personal property as capital. Positive rights of agency are rights to meaningful access to the resources required for an agent to freely develop and exercise her agency. Persons need resources to meet uncontroversial needs, such as needs for food, healthcare, housing, clothing, and education. So, parties behind the veil would select primary rights to possess and consume these goods. Notice, though, that these are not necessarily rights to state provision of these goods; instead, they are rights to have social institutions organized in such a way that these goods are available and accessible in adequate supply. In general, societies should protect welfare rights in ways that do not curtail the negative rights of agency, so if institutions can provide vital goods and services through noncoercive mechanisms, then we should prefer using those methods. Another complication is that many members of the public would reasonably insist that social assistance be sensitive to the deserts or merits of recipients, which would lead the selected rights to be somewhat circumscribed. Agency rights circumscribed in these ways should bring even nonliberals to the table, so to speak. Nonliberals would select agency rights because they would not be willing to risk being a morally and legally unprotected minority, subject to a powerful, alien political hegemon.

8.2. Associational Rights From individual rights, we branch out to the rights of associations, such as families, churches, service organizations, and unions. Associations involve interactions among persons who already have agency rights, which is why associational rights are less essential for persons. But they are still essential for each person to fully enjoy rights of agency. Given how essential associational life is to most people’s conceptions of the good and of justice, associational rights should be publicly justified. And as with agency rights, associational rights should be publicly justified even for nonliberals, since in most cases nonliberals are communitarians who deeply value associational and community life. But behind the veil, since they would not know their level of power and influence, they would be prepared to ascribe moral authority to the associational rights of competing groups in order to secure moral protections for themselves.

8.3. Jurisdictional Rights Jurisdictions are spheres of social control where an individual or group has the sole authority to authoritatively determine what occurs within that social space.

Must Politics Be War Here and Now?  47 Jurisdictional rights establish this authority. Jurisdictional rights are not agency rights because they involve protections that are not strictly speaking required for one to exercise her agency; they have a unique basis because evaluative pluralism prevents people from making publicly justified collective decisions on a variety of matters. So, decision-​making must be decentralized through jurisdictions, allowing each person to make her own decision in her own partition of social space. Private property rights are a central jurisdictional right that can only be fully respected in a capitalist society, though property rights can be limited to provide social insurance and some economic regulations. We will see, nonetheless, that, while the right to private property is controversial, only certain aspects of it are subject to defeat by social-​democratic and socialist members of the public, namely private property in large amounts of capital. More modest forms of private property, say, in one’s home, the property required for freedom of association, and the capital necessary to run a small business, would be chosen behind the veil even by left-​wing members of the public.

8.4. Procedural Rights The procedural rights of citizens are rights to be governed by certain legal and political procedures. Legal procedural rights include the right to a fair trial, the right to a public defender, the right against self-​incrimination, and so on. Political procedural rights include the right to vote, the right to contribute to campaigns, the right to run for office, and so on. While some aspects of these rights are controversial, procedural rights are deeply entrenched in every free society, and with good reason. Anyone with a conception of the good and a conception of justice would want to be treated fairly by a legal system that could destroy her life and even kill her. Reflective and unbiased parties behind the veil would want these rights sufficiently badly that they will extend the same right to others on reciprocal terms. Political rights can be publicly justified on similar grounds: Reasonable citizens will want the right to influence the political process and to be able to have a genuine effect on policy inasmuch as this is possible.51 In the vast majority of 51 The term “reasonable” raises scruples among critics of public reason views, as I  discuss in MPBW, pp. 94–​7. Here I understand persons as reasonable just when they are prepared to abide by reciprocal terms of social cooperation. They do not demand that others follow moral rules that they would not want to follow. This is a thinner notion of reasonableness than one finds in the public reason literature and is, I think, more attractive as a result. It’s also important to stress that we do not need to go out of our way to exclude unreasonable people from social life. As Gerald Gaus argues, under diverse conditions where most individuals are willing to follow a suboptimal moral rule if enough others follow it (but they disagree on how many it takes to be “enough”), most of the

48  Trust in a Polarized Age cases, individual citizens cannot affect the political process, because citizens are too numerous for a single vote to matter. But, collectively, the right to vote has a critical role in preventing governments from engaging in disastrous policies. Democracies have many bad policies, but they are typically superior to policies in nondemocracies, such as authoritarian regimes like China.52 And parties behind the veil would limit the right to vote in order to prevent democratic majorities from violating primary rights, so majoritarian dictatorship cannot be publicly justified. Finally, parties will expect a certain degree of competence and impartiality in the administration of public policy, where citizen inputs are treated equally and citizens are treated equally and efficiently by political and other governmental officials. I will speak to the structure of a democratic constitution and democratic elections in ­chapters 7 and 8 respectively.

8.5. Constitutional Rights All publicly justified primary rights merit both legal and constitutional protection. They merit legal protection in order to stop violations by other citizens. Primary rights also deserve constitutional protection because members of the public need protection from the imposition of bad law by government. If primary rights are protected as constitutional rights, it will be harder for governments to violate these rights by altering or corrupting legal systems. Constitutional rules even provide remedies for rights violations, like judicial review, and can even imbue rights with a kind of sacred quality that lead people to see rights protection as fundamental to their national character, providing further protection still. We must also determine how to stabilize these constitutional rights. To achieve stability, the rights must both give citizens good reason not to violate them and prevent political officials from corrupting and undermining them. Constitutionally protected rights establish both goods. While subsequent chapters focus in part on showing that many liberal rights are publicly justified, I also show that the recognition and exercise of these rights creates trust in the real world. In this way, liberal rights practices generate real trust for the right reasons, ending political war.

population will end up following reciprocal rules, and unreasonable people will exclude themselves. See Gaus 2018. I thank Molly McGrath and Alex Motchoulski for discussion on this point. 52 Eduardo Martinez objects that moderately idealized agents may mistakenly, but sensibly, believe that properly structured authoritarian regimes like China will tend to have better policies than the average democracy. I address this point in ­chapter 8.

2

Social and Political Trust Concepts, Causes, and Consequences

This chapter, which is based on the extensive empirical literatures on trust in the social sciences, focuses on the creation and maintenance of social and political trust in the real world. The chapter’s overall conclusion is twofold. First, social trust and political trust are both critical social achievements for sustaining a diverse social order, but social trust is more important than political trust. Second, liberal-​democratic market institutions play a modest role in sustaining social trust and a large role in sustaining political trust. We can conclude, then, that liberal-​democratic market institutions are part of a positive causal-​feedback loop that sustains trusting social orders with diverse persons who disagree. That is how we get trust for the right reasons. I begin by outlining my accounts of social and political trust (section 1). I explain how social and political trust are measured (section 2), addressing some challenges to the usefulness of the empirical literature along the way. I examine the causes and consequences of social trust (sections 3–​4), as well as the causes and consequences of political trust (sections 5–​6). I  then answer a challenge about how people learn to trust one another (section 7). I conclude by linking the causes and consequences of social and political trust to liberal rights practices (section 8).

1.  Social and Political Trust Social trust is trust that is mass and mutual: it is placed by almost all community members in all or almost all community members. Because large societies typically lack shared goals of any real substance, I define social trust as requiring that people believe that the help of others is generally necessary or important for small groups and individuals to be able to achieve their own diverse ends. Further, people signal their ability and willingness to act in certain ways by following moral rules. Critically, social trust is not focused on any particular moral rule, but relates to a broad range of moral rules. Social trust, therefore, carries a generic expectation that members of society will comply with moral rules generally, especially moral rules essential to social order, like prohibitions on harm

50  Trust in a Polarized Age and lying. I do not affirm the common view that social trust must involve an affective state of security that others will act appropriately. Rather, social trust centrally requires the widespread belief that others will abide by a large set of “ordinary ethical rules,” that is, moral rules.1 When we socially trust, we believe others will choose “ethically justifiable behavior—​that is, morally correct decisions and actions based upon ethical principles of analysis.”2 To trust her society, a person must believe that other members of it will act according to shared moral rules that reliably allow her to pursue her projects and plans.3 Social trust per se arrives if most members of the public have these expectations: Social Trust:  a public exhibits social trust to the extent that its participant members generally believe that participants are necessary or helpful for achieving one another’s goals and that (most or all) members are generally willing and able to do their part to achieve those goals, knowingly or unknowingly, by following moral rules, where moral reasons normally motivate compliance.4

Societies create social trust by means of mass compliance with moral rules.5 If each person observes compliance with moral rules and thereby comes to think that people will comply with those rules, then those observations establish the empirical expectations required for social trust. Normative expectations must also be in play because social trust requires a belief that people ordinarily have adequate moral motivation to follow moral rules, which they can ignore only as a moral fault. And such faults are often taken to justify punishment, further indicating the presence of moral expectations.6 My view also provides reason to expect social trust to be highly stable over time. Remember that on my account, social trust is grounded in persons generally believing that others can be trusted. This belief is sufficiently broad that we

1 Messick and Kramer 2001, p. 91. 2 Hosner 1995, p. 399. 3 The person with the belief need not be aware that she holds the belief, however. 4 Notice here that I do not give sufficient conditions for social trust. I do this to avoid taking a stance on whether trust essentially involves affective/​emotional states, as it complicates the analysis. However, I think further exploration of the affective dimension of trust will help to explain certain social phenomena, like compounding distrust. For an examination of how the affective dimension of trust affects how trust functions, see Jones 2019, p. 2. Also, as noted in Chapter 1, readers of MPBW will also notice that I have changed my definition of social trust so as to no longer require that trusters see trustees as having moral reasons sufficient to motivate compliance all on their own. Instead, they must only believe that trustees are normally motivated to comply for moral reasons, allowing other kinds of reasons, like self-​interested reasons, to play a larger motivating role. 5 At least in combination with certain more-​ingrained traits, like personality, as we will see later. 6 Social trust may require punishing those who fail to punish; see Boyd and Richerson 2005, pp. 193–​203. In MPBW, I contrasted my account of social trust with that found in Uslaner 2002, p. 18.

Social and Political Trust  51 shouldn’t expect it to change if persons observe some particular moral rule being violated, such that social trust would fall if a common traffic norm is flouted. In such cases, we might partition our trust by context: the flouting of one traffic norm would not normally challenge one’s belief that most people follow most moral rules. I predict a drop in social trust, therefore, only when people observe widespread defection from a number of central moral rules. As we explore the empirical measures of social trust, which are often taken to show that trust is not responsive to observation, we will see that these measures do not conflict with my account of the observational basis of social trust. Many researchers are keen to distinguish between trust in society and trust in government, and they find that citizens distinguish between the two.7 But there is disagreement about their causal connection. Sonja Zmerli and Ken Newton find “robust and statistically significant correlations between generalized trust, on the one hand, and confidence in political institutions and satisfaction with democracy, on the other,” in twenty-​three European countries and the United States.8 Bo Rothstein and Dietland Stolle argue that citizens make even finer distinctions, including between legal institutions such as the police and judges and political institutions such as democracy and elected officials.9 Rothstein and Stolle understand these as order institutions and political institutions respectively. They document “a rather strong relationship between aggregate levels of confidence in order institutions and generalized [social] trust.”10 People “make strong connections between the impartiality of [order] institutions and generalized trust at the micro and macro levels.” Social trust is weakened when people experience “widespread corruption, inefficient institutions, unreliable police, and arbitrariness and bias of courts.”11 However, Rothstein and Stolle find that “there is no relationship between political institutions with elected office and generalized trust at the aggregate level.”12 In brief, generalized trust is connected to trust in the police and the courts, if not trust in politicians. In this way, theorists work with distinctions between social trust and forms of institutional trust. Following Rothstein and Stolle, we can distinguish at least three types of trust: social, legal, and political. Social trust is by now familiar, and legal trust is trust in order institutions such as law enforcement and the courts, as we saw in ­chapter 1. Political trust is distinguished from legal and social trust because it is placed in institutions of government that, in many societies, are directly affected by the political process. In particular, political trust is placed in government

7

Hardin 2004, p. 151, stresses this distinction and doubts a connection between the two. Zmerli and Newton 2008, p. 706. 9 Rothstein and Stolle 2008. 10 Ibid., p. 450; emphasis mine. 11 Ibid. 12 Ibid., p. 450; emphasis mine. 8

52  Trust in a Polarized Age officials: elected officials, political parties and their leadership, and members of the civil service. Social trust, recall, is trust placed by most people in all or most of the other members of society. I will characterize political trust as a subset of social trust because political trust is trust in government officials or some subset of those officials, who are members of society generally and typically not one’s friends or family. Political trust, as a type of social trust, involves relying on government officials to follow a particular kind of moral rule that I will call an institutional rule. Let’s understand institutional rules as norms that are (a) backed by empirical and normative expectations, (b) constitute social institutions, and specify (c) the aims of those institutions or (d) the procedures by which they achieve those aims. Institutional rules differ from other moral rules because they are tied to the public purposes of social institutions and are justified based on how they facilitate those purposes; moral rules as such need not constitute any institution or specify any overall social end. Institutional rules, therefore, are social norms that facilitate some institutional purpose or value. Examples include the norm that the Environmental Protection Agency preserves endangered species, that the Federal Reserve manages inflation and unemployment, and that the IRS collects taxes. The norms both constitute these institutions and specify their goals. Auxiliary institutional rules restrict how institutions can pursue those aims, such as rules that specify how these institutions may make policy. Institutional rules are backed by morally normative expectations: people think that members of the institution morally ought to follow the relevant norms, and we think that others think members of the institution should follow the norm. This means that a failure to follow institutional rules will ordinarily be seen as a source of justified public resentment, blame, and punishment. Some theorists may be inclined to de-​moralize institutional rules and argue that our expectations of compliance are not moral but prudential. This is a mistake because prudential normativity cannot render our resentment and blame appropriate. We do see political institutions as moral agents or as comprised of moral agents.13 Since institutional rules are directly tied to the publicly recognized purposes of the institutions comprised by them, trust in these institutions will also involve empirical and normative expectations that the institution ordinarily succeeds in achieving certain outcomes.14 Political parties are supposed to produce particular public policies, and civil servants are supposed to help their agencies or 13 Contra Paul Faulkner 2018, p. 631, I think our normal understanding of democratic citizenship involves holding political leaders morally responsible for outcomes (which Faulkner acknowledges) by means of the reactive attitudes (which Faulkner denies, in claiming that “affective trust” cannot be directed toward government appropriately). We hold institutions responsible in a similar fashion. 14 This also means that the outcome measure is correlated with the competence condition for interpersonal trust.

Social and Political Trust  53 bureaucratic bodies complete their objectives in an honest manner that does not exceed their authority. When they fail, citizens resent or are indignant with them. Normative expectations apply to both the political decision-​making process and the outcomes achieved. The empirical research shows that citizens expect government officials to be fair and avoid corruption in their decision-​making procedures, but also to achieve the substantive goals of economic growth and economic security for all. They are dissatisfied with officials who fail in either respect, and I think it is clear that they resent those officials as well. We can now define political trust: Political Trust: a public exhibits political trust to the extent that its participant members generally believe that government officials are necessary or helpful for achieving widely shared political goals15 and that (most or all) government officials are generally willing and able to do their part to achieve those goals, knowingly or unknowingly, by following institutional rules, where moral reasons normally motivate officials to comply with those rules.

We have political trust when we trust that government officials will observe institutional rules and produce certain outcomes. Political trust can in this way be understood as trust in government. In democracies, political trust can be understood as trust in democratically chosen officials and those they appoint. So measures of trust in democracy can be understood along those lines. Social trust is placed in most members of one’s society; we trust one another. But political trust can vary in its objects between different groups of officials, such as legislative bodies, agencies, parties, and so on: we trust Congress, the EPA, or the Democratic Party. In some cases, political trust may focus on a single person, such as the president of the United States, whereas in others it might be 15 Ryan Muldoon asks whether some people might distrust an institution because they think the institution itself is illegitimate, such that a person could distrust an institution even though the institution performs in line with institutional rules, if not because it does so. American conservatives might distrust the EPA because it effectively contains carbon emissions, and American progressives might distrust Immigration and Customs Enforcement because it effectively prevents immigrants from coming to the United States. But in both cases, resentment and indignation are present because the institution is seen as pursuing a sectarian value, or perhaps pursuing a value in a morally prohibited manner (say by violating a broader moral rule against excessive government control: this is how conservatives see the EPA and how progressives see ICE). This is why political trust is based on wanting and expecting institutions to achieve widely shared goals: it will be harder to trust an agency whose goals are not widely shared but are in many cases controversial. In this case, then, an institution’s failure to pursue a widely shared goal will be grounds for distrusting it among those who reject the goal in question. Here I admit a complication if government pursues values that are shared by the vast majority of the population but that are strongly rejected by a small minority: in that case advancing widely shared goals will result in indignation and resentment in the minority, and the resulting distrust, even if a society as a whole maintains a high level of trust in the institution. I thank Eduardo Martinez for raising this objection.

54  Trust in a Polarized Age placed in, say, NASA. I contend that political trust is always placed in those who are publicly seen as morally responsible agents, though this does allow for situations where people have low political trust in the leaders of an institution but high trust in the institution more broadly: one might trust the FBI but distrust its director. The empirical data focus on institutions and large groups, which is compatible with my broader definition of political trust. Once again, political trust must be understood as a kind of moralized trust.16 Political trust is not merely the empirical expectation that political institutions are reliable but the normative expectation that they are responsive to moral considerations, such as shared values and the importance of commonly recognized procedural rules.17 This is important in trying to explain when and why political trust falls. As we shall see, political trust falls when people believe some governmental institution has treated them unfairly or unjustly, which is arguably associated with how political institutions generate the negative reactive attitudes. Some of the benefits of political trust can be captured by a nonmoralized political reliance. After all, nonmoral agents like robots and psychopaths can produce good outcomes. But if citizens know that they are governed by robots and psychopaths, this will likely create fear, confusion, and alienation and undermine trust. In contrast, trust relaxes us in a way that reliance does not, since it establishes a kind of social relationship with the trusted person. This makes it easier for persons to become dependent on one another.

2.  The Measurement of Social and Political Trust The trust literature in political science and economics is decades old and there is a sizable metaliterature that assesses the viability of the trust measures. There are two main approaches to measuring trust found in the literature: direct and indirect. Direct measures of trust are based on people reporting their trust levels on surveys and questionnaires. Indirect measures infer subject trust expectations by observing individual decision-​making, behavior, and reactions, usually in structured experiments.18 However, direct and indirect measures of trust don’t always line up, and they seem to have conflicting implications. To draw clear lessons from the trust literature, we will have to determine the relative value of these measures. I will argue

16 In this way, Marc Hetherington’s definition of political trust “as the degree to which people perceive that government is producing outcomes consistent with their expectations” should include both empirical and normative expectations (Hetherington 2005, p. 9). 17 Or that they are rightfully held accountable by people for violating institutional rules. 18 Bauer and Freitag 2017, p. 3.

Social and Political Trust  55 that we can responsibly appeal to the empirical trust literature while taking these worries about measurement seriously. Direct measures draw from self-​report surveys that have been around since the early 1940s. They began with what I  will call the standard trust question. Here’s how the General Social Survey (GSS) phrases it:  “Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?”19 In more recent surveys, such as the World Values Survey, the following two questions are also often included: “Would you say that most of the time people try to be helpful or that they are mostly looking out for themselves?” And: “Do you think most people would try to take advantage of you if they got a chance or would they try to be fair?” Today, researchers use modified versions of these early questions.20 In many cases, responses are used to form an overall social trust index, though in cases where the three questions yield different responses, the standard trust question is privileged over the other questions.21 Some new surveys include the “wallet” question, “If you lost a wallet or purse that contained $200, how likely is it to be returned with the money in it if it was found by . . . ,” and then specify different groups.22 A very recent international wallet-​return test shows that trustworthiness, represented by rates of wallet return, correlate with “economically favorable geographic conditions, inclusive political institutions, national education, and cultural values that emphasize moral norms extending beyond one’s in-​group.”23 Direct measures of institutional trust (which includes political trust) arose during the 1960s and include questions about whether persons trust their government, certain political parties, or political officials.24 Surveys now often include questions with a structure like that found in the European Social Survey: “Please tell me on a score of 0–​10 how much you personally trust each of the institutions I read out. 0 means you do not trust an institution at all, and 10 means you have complete trust.”25 One of these institutions is “the government,” which is also invoked with survey items like the following: “People have different ideas about the government in Washington. These ideas don’t refer to 19 Jennings and Stoker 2004, p. 350. 20 Bauer and Freitag 2017, p. 8. 21 Indeed, these questions do come apart somewhat: roughly around 25 percent of respondents respond differently to the standard trust question and the fairness question, and around 30 percent respond differently to the helpfulness question. I thank Christian Bjørnskov for discussion. 22 Bauer and Freitag 2017, p. 7. One might worry that the wallet return test only measures observable trust, and not social trust, but given that I have defined social trust in part as being grounded in the observation of social norm compliance, it is fair to assume that observable trust is a proxy for social trust. 23 Cohn et al. 2019, p. 4. 24 Norris 2017. 25 European Social Survey.

56  Trust in a Polarized Age Democrats or Republicans in particular, but just to the government in general. We want to see how you feel about these ideas. For example, . . .”26 Today the study of political trust focuses on the World Values Survey and the American National Election Survey. Often these questionnaires are limited to a handful of questions with limited responses, but more recent surveys ask a larger number of questions with up to an eleven-​point scale, yielding richer data and in some cases challenging or strengthening older empirical results.27 This means that studies of political trust may deploy somewhat different conceptions of political trust, but it nonetheless appears that political trust research approximates a single underlying construct.28 Many readers, and philosophers in particular, will worry that these questions are too vague for hundreds of studies and dozens of books to justifiably rely upon. Perhaps people have different views about who count as “most people” other than, say, their countrymen. However, it turns out that generalized trust is a stable measure across a wide variety of contexts, both across cultures29 and within different radii of groups trusted.30 And many researchers argue that these measures of generalized trust really are getting at some stable, society-​wide attitude31 that other persons can be relied upon to engage in basic ethical behavior. The same goes for political trust. In the 1960s, lab work on game-​theoretic behavior began and laid the foundation for indirect measures of trust. The classic trust game is similar to the dictator and ultimatum games. In such games, one player has the authority to divide a prize between two parties and offer the division to the second party, and then the second party has the right to accept or reject the division. If the second party rejects the offer, neither party gets anything, but if he accepts, the parties receive the prize as the first party divided it. The trust game is a game of this type, though in many cases it involves trusting multiple people to act cooperatively. In canonical lab experiment, subjects are randomly, anonymously paired with one another and given different jobs. Players are given an endowment, and the first mover is asked whether she would be willing to send some part of her endowment to the second mover. The experimenter then triples whatever was sent to the second mover. The second mover is asked how much she wishes to send back to the first mover. Once the second mover completes her task, the players are paid and the experiment ends. Neither

26 Ibid. 27 Marien 2017. 28 Ibid., p. 98. 29 Tim Reeskens and Marc Hooghe find that “generalized trust . . . seems to refer to the same latent structure across Europe.” See Reeskens and Hooghe 2008, p. 530. 30 Delhey, Newton, and Welzel 2011, p. 786. 31 Nannestad 2008, p. 418.

Social and Political Trust  57 player knows the other, and they are paid in private.32 In brief, the game is a two-​ person sequential prisoner’s dilemma. The game is called the “trust game” because it appears to have elements of both trust and trustworthiness. The first mover exhibits trust because she might be motivated by a trusting attitude to believe that the second mover will return some of her endowment. The second mover exhibits trustworthiness because she doesn’t have to return any part of her gains. The trust game is thought to measure trust because the second mover has no self-​interested motive to return part of her endowment to the first mover, and the first mover can infer this and then send nothing. Empirical research finds that trust in this sense is common in the laboratory.33 The difficulty with the trust game is that motives other than trust might be at work, such as concerns with efficiency, reciprocity, and altruism.34 But many think that the behavioral measures of trust are, for all their flaws, superior to the survey measures. As economists often stress, talk is cheap, while action is costly, so behavioral measures are more likely to reveal the truth. A significant worry about much of the empirical literature derives from laboratory work on trustworthy behavior. Trust game experiments manifest a “lack of close correlation between behavior and questionnaire responses” such that those who say they are highly trusting are not in fact more likely to cooperate in microlevel trust games.35 Worrisomely, “The three questions most often used in survey research to measure generalized trust were not predictive of the likelihood that subjects will trust each other even in a repeated setting,” suggesting that the subjects may be “responding in a glib manner to this survey instrument.” In a recent survey of the literature on trust games and self-​report measures, Rick Wilson notes extensive research that shows that the two measures are uncorrelated.36 This line of research raises a serious concern: it could be the case that people say that they trust others, but act as though they do not. Some trust researchers have tried to answer this skepticism by arguing that in fact, the behavior and survey measures are congruent when it matters.37 For instance, trust researchers have found that “specific questions about past experiences of being trusted or extending trust in the past” were tied to trusting behavior and that the survey questions “were positively correlated with trustworthy, if not trusting, behavior.” Thus, since trust is arguably produced by the perception of trustworthy behavior in others, then the discovery that

32

Wilson 2017b, p. 3.

33 Ibid. 34

Ibid., p. 4. Also see Cox 2004. Ahn et al. 2003, p. 345. 36 Wilson 2018. 37 Sapienza, Toldra-​Simats, and Zingales 2013. 35

58  Trust in a Polarized Age people who tend to trust a lot behave in a trustworthy fashion in experimental settings strengthens the case for congruence. These individuals act in a trustworthy fashion because they believe, based on their observations, that others are trustworthy. Moreover, recent work has found a correlation between self-​report measures of trust and trusting behavior in the trust game when high financial payoffs are at stake; the behavioral and survey measures largely converge when the payoffs in behavioral settings are high. Paola Sapienza and colleagues found that the more payoffs in the trust game increase, the more the survey measures and the trust game behaviors correlate.38 And Wilson found that “when the appropriate controls are put into place (at least among students in the lab), it appears that the first mover’s behavior and the General Social Survey item [the main trust question] are correlated.”39 So it looks like the trust game data need not be taken to undermine the survey data. There is another reason to remain confident in direct measures of trust. Laboratory settings may elicit different motives than those generally at work “in the wild.” Indirect measures may generate misleading results in artificial games, by abstracting from the social context that real interactions involve, leading participants to conceive of their choices inside experimental scenarios in a way that does not reliably reflect their level of trust in strangers in their societies. So even if it is true that the trust game data and the survey data are inconsistent, it does not mean we should automatically favor the trust game data. It is reasonable, then, to appeal to the direct measures of social trust, as long as we keep the limitations of the data in mind.40 My conception of political trust diverges somewhat more from conceptions of political trust found in the literature because the empirical literature focuses on levels of satisfaction with government/​democracy and seldom makes the moral element of political trust explicit. However, when one is asked about whether one is satisfied with one’s government, it is natural to interpret this question as inviting a moral evaluation, particularly as to whether government obeys proper procedures and gets the promised results. If people think institutions generate good outcomes because, for example, officials have concern for citizens, or are generally ethical and possess good character, then we are liable to be more trusting. Consequently, in cases where satisfaction is measured, I presume a close 38 Ibid. 39 Wilson 2018, p. 286. 40 A third reason, which I discuss in MPBW, is that my account of social trust maps onto Marc Cohen’s survey of conceptions of social trust in the literature. See Cohen 2015, p. 465. And finally, cooperation in trust games varies over time, whereas survey responses do not, and it is more plausible that social trust is fairly stable than that it is highly variable.

Social and Political Trust  59 connection between trust and satisfaction; and indeed, researchers commonly treat the two concepts interchangeably. So high political trusters are more satisfied with government than low trusters; and those high in satisfaction should also be high trust. I do admit an important limitation, however, since the language of satisfaction may only pick up on respondents’ evaluation of the outputs of an institution and not any evaluation of the character or motives of members of the institution. Some citizens may simply look at institutions as mechanisms and think that moral evaluation is inapt. Others may think that, say, civil servants only look out for themselves most of the time but (perhaps because of incentives built into the institutions) still end up getting the right results. So people may be satisfied with their government even if they do not find the people who run it especially trustworthy. It would be odd to politically trust people one thinks are especially socially untrustworthy, but satisfaction might be neutral between positive evaluations of trustworthiness and no evaluation or a middling evaluation.41 All I can say is that surveys based on satisfaction measures are indeed ambiguous on this point, and so we should be cautious about drawing inferences from surveys that use the language of satisfaction to conclusions about whether people are politically trusting in the sense that I have outlined here. For the rest of this book, then, I will draw primarily from direct measures of social trust. The great advantage of self-​report measures is that they are cross-​ cultural and allow us to compare trust among different cultures and different types of institution. The trust game data are far less varied and rich. But, to be fair, the tensions between the survey and trust game data should make us cautious about drawing definite conclusions about which institutional structures generate trust, even when the direct measures establish a clear correlation between social trust and certain rights practices and their associated institutional structures. With political trust, we have no trust game data, so we must rely on the direct measures. Here I can only say that the survey measures do not make this distinction when “satisfaction” language is used.

3.  The Sustaining Causes of Social Trust: What We Do and Don’t Know Now let us examine what we know about the causes of social trust. It is vital to understand these causes because of how important social trust is for realizing



41

I am grateful to Scott Simmons for this objection.

60  Trust in a Polarized Age a number of goods. For instance, economists, sociologists, psychologists, and many others have argued that social trust is critical for maintaining well-​ functioning political and economic institutions. If members of a society do not trust one another, then they have little reason to take the risks required to create, build, and sustain good institutional structures. Thus social trust makes many great goods possible. However, one of the most remarkable features of the literature on social trust is how little we know about what causes it to form and endure. If trust is in fact a linchpin of a healthy democracy and a thriving economy, then it is a matter of some urgency to develop a fuller and more rigorous understanding of the underlying conditions that sustain it.

3.1.  Diversity Some of the conventional wisdom about the sustaining conditions of social trust is significantly exaggerated. Despite widespread belief to the contrary, there is, for instance, very little correlation between a society’s degree of ethnic diversity and the level of social trust that it enjoys. According to a recent study by Peter Dinesen and Kim Sønderskov, “Meta-​analyses report substantial variation in the relationship between ethnic diversity and social trust.”42 It does appear that ethnic diversity can bear negatively on trust, but this is usually due to “lack of contact and a segregation effect,” not diversity per se.43 Studies that control for the level of interaction across ethnic lines “indicate that the estimated effect of diversity becomes less negative/​more positive under contact.”44 And so, overall, Dinesen and Sønderskov’s survey of the research concludes that “exercising great caution, we do believe it is fair to say that the main finding from the literature is a negative—​albeit not always significant—​relationship between ethnic diversity and social trust,” though there’s “not really a consensus.”45 However, it is clear that when ethnic diversity is highly segregated, social trust does fall.46 Conflict can erupt between different ethnic groups, especially when there’s a conflict over limited resources, but if contact between ethnically distinct groups is properly structured—​and not structured around a resource or power conflict—​it can increase trust.47



42

Dinesen and Sønderskov 2018, p. 179.

43 Ibid. 44

Ibid., p. 183. Ibid., pp. 196–​7. 46 Enos 2017. 47 Putnam 2007, pp. 141–​2. 45

Social and Political Trust  61

3.2. Economic Inequality One might also think that economic equality causes social trust. It is reasonable to suppose that people who sense large gaps in the distribution of economic benefits within their society may come to believe that others who are much wealthier or poorer are too different to be trustworthy, or that those at the top got on top through unfair or corrupt processes.48 There is indeed a strong empirical correlation between a society’s level of economic equality and its level of social trust, but the direction of causation is hard to determine.49 Many trust researchers seem to think that economic inequality leads to more social distrust.50 However, the mechanism is unclear. It may be that perceived economic inequality is taken to be a sign of unfairness, corruption, and social institutions that are not organized to work to the benefit of all. On the other hand, economic inequality may simply generate observable social differences, leading people to withhold trust in direct response to their observations. In other words, economic inequality may make a preexisting sense of social distance more salient; it may not be the primary source of that distance. There is a strong case to be made that, in fact, most of the robust positive correlation between social trust and economic equality is explained by the fact that societies with a high level of social trust are more supportive of redistributive policies, which reduce economic inequality.51 When we trust others, we are less likely to be suspicious that those who receive economic transfers will misuse them or have secured those benefits unfairly. As a result, as Bergh and Bjørnskov argue, socially trusting societies are more likely to see redistribution as fair or just rather than as wasteful or a reward for sloth.

3.3.  Corruption and the Rule of Law There is one sustaining institutional cause of social trust about which we can be relatively confident. The evidence clearly shows a close connection between higher levels of social trust, lower levels of corruption in the legal system, and other indicators of reliable adherence to the rule of law.52 Corruption can be understood in a few different ways, occurring at a minimum when a political official abuses the power she has been entrusted with or betraying the public’s trust 48 Uslaner 2018. 49 Bergh and Bjørnskov 2014. 50 Uslaner 2002 is the most prominent of these researchers. 51 Bergh and Bjørnskov 2014 provide a rich and detailed argument to this effect, using multiple methods. 52 You 2018.

62  Trust in a Polarized Age in the official’s integrity or fairness.53 Corruption can also include nepotism and favoritism, as well as bribery and embezzlement. And in some cases, people consider abuses of private power as corrupt, such as using money for personal gain.54 If we focus on public corruption, the data are clear that a functional legal system resistant to public corruption leads people to trust others in their society. It is not entirely clear why this is the case. It is probably partially due to the fact that people’s legal property rights are better protected from strangers in noncorrupt legal systems.55 Functional legal systems punish theft effectively, for instance, so citizens do not have to worry that their neighbors will steal from them, which means they have less cause for suspicion. It is also possible that people use the behavior of judges and the police as proxies for the trustworthiness and goodness of the people as a whole, as judges and police might be seen as examples of the character of society generally. Alessandra Cassar, Giovanna d’Adda, and Pauline Grosjean have found that exposure to better formal enforcement, especially impartial adjudication of torts, increases informal norms of cooperation.56 Impartial treatment leads people to more cooperative behavior in laboratory trust games.57 Legal enforcement really can increase trusting behavior outside of formal institutions if it is impartial and regular.58 Recent empirical work on trust and fairness by Jong-​sung You backs up these claims. You argues that “the fairness of political and legal institutions affects people’s incentives for trust and trustworthiness” and that “individuals’ perceptions of the fairness of their society directly affects their trust in other people.”59 While many social trust researchers think that economic inequality directly reduces social trust and trustworthiness, You claims that it does so indirectly, by leading people to think that their political and economic institutions are unfair. You argues that “individuals’ perceptions of fairness are significantly correlated with social trust.”

53 Ibid, p. 477. 54 Ibid., p. 478. Indeed, there are corruption indices that focus on measuring favoritism and the influence of large interests in government, which the indices treat as essentially synonymous with corruption. See ibid, p. 480. 55 Here I do not mean protection from taxation but, rather, protection from theft, fraud, etc. by private parties. Though see Bjørnskov 2019 for some skepticism about whether legal property rights cause social trust. 56 Cassar, d’Adda, and Grosjean 2014. I am grateful to John Thrasher for bringing this paper to my attention. 57 Ibid. 58 This does imply that social trust can be crowded out by legal trust, perhaps because people start to be motivated by fear of legal sanction rather than moral grounds. See ibid., p. 859. But as I noted in ­chapter 1, this is not a problem for my view. We need trust for the right reasons, and that can often be through legal norms. 59 You 2012, p. 702.

Social and Political Trust  63 This suggests that further inquiry into the moral and political psychology of social trust may pay dividends. I would argue that public respect for rights is a primary driver of social trust and helps explain the connection between trust and perceptions of the integrity of legal and political institutions. A widespread sense of the security of rights promotes social trust by leading people to feel as though they can generally trust others, on the grounds that each person is treated fairly and equally in legal and political institutions. A general ethos of respect for equal rights seems likely to sustain trust through both cultural and political channels. First, visible public respect for rights will be taken to reflect widespread social commitment to the idea that each person ought to be treated fairly, which creates an expectation that others will at least feel they should conform to norms of fair dealing. Second, the social pressure to conform that accompanies the recognition of social norms helps to assure people that others feel the same pressure. This increases each person’s confidence that others are trustworthy. Third, as a political matter, rights will not be given vigorous and reliable legal protection without a general ethos of respect for rights. So even if we doubt the moral motives of others, salient public commitment to the protection of legal and constitutional rights helps people believe that political and legal systems will detect and punish untrustworthy behavior and that people will tend to avoid such behavior as a result.

3.4.  Communism There is an unambiguous result that communist governments destroy social trust. Countries that joined the Warsaw Pact following World War II are much less trusting than their counterparts.60 The change is manifest in Germany, which was divided following World War II, but can be seen in the Czech Republic. Focusing on the German case, Sarah Necker and Andrea Voskort find that people who grew up under the communist regime in East Germany are much less trusting than Germans who grew up in the German Federal Republic.61 This is arguably due to the fact that communist states use secret police and have unfair trials based on anonymous testimony. In the 1980s, 4 percent of East German adults worked for the secret service known as the Stasi. If one out of every twenty-​five people you meet can secretly ruin your life, that just might sow distrust in your fellow man.



60 61

Bjørnskov 2019 reviews the data. Necker and Voskort 2014.

64  Trust in a Polarized Age

3.5.  Monarchy One striking feature of the social trust literature is that “monarchy . . . appears to be one of the most robust determinants of social trust in cross-​country analysis,” and Bjørnskov has tried to explain this by arguing that “monarchs may provide nonideological focal points for societal expectations that are common to most citizens.”62 Monarchs, therefore, are a touchstone for salient observations of social unity because they have usually been nonpartisan. Notably, the monarchs in the data have almost no political power.

3.6.  Personality and Culture There are also cultural and psychological factors that determine one’s capacity to be socially trusting. The first nonrational factor is personality type:  some researchers claim that “personality mediates the relationship between biology and interpersonal trust.”63 At least one study suggests, for instance, that “interpersonal trust [is] positively related to openness, extraversion, agreeableness, and emotional stability, and negatively related to conscientiousness.”64 Social trust researchers disagree about the extent to which social trust is determined by “cultural persistence” or “experiential adaptation.” The debate concerns the “question of whether social trust is . . . a deeply held disposition socialized early in life that remains relatively immune to subsequent experiences or . . . an impressionable outlook shaped continuously throughout life by an individual’s experience.”65 Thus, one could argue that individuals’ personal social trust is sustained by cultural factors like their upbringing, family, and broader cultural experiences that are locked in at an early age, the culturalist approach; or by their observations and experience, throughout life, of trustworthiness in institutional settings: the institutionalist approach. There is evidence for both models. The culturalist approach is buttressed by the observation that social trust levels in particular countries are remarkably stable over time and through significant social and political changes, such as changes in public policy, changes in partisan control of government, and even changes in national wealth and form of government. So it looks like social trust may not be affected by a change in perceived trustworthiness in the population, and so social trust will have nonrational determinants, which I will address further later. 62 Bjørnskov 2019. 63 Cawvey et al. 2018, p. 125. 64 Ibid., p. 130. This may imply, importantly, that conservatives, who tend to have lower openness and higher conscientiousness, are less trusting. 65 Dinesen and Sønderskov 2018b, p. 206.

Social and Political Trust  65 But differences in institutions do seem to affect the trust levels of persons from the same culture. One way that trust researchers try to resolve the dispute is by studying immigrants’ trust. If the cultural-​persistence hypothesis is true, we should expect personal social trust among immigrants to remain largely unaffected by the social trust found in the societies to which they migrate; but if the experiential adaptation hypothesis is true, we should expect immigrants’ personal social trust to gradually change to match social trust levels in the places to which they have migrated.66 John Helliwell and colleagues estimate that social trust levels among migrants are probably about two-​thirds determined by the social norms in their adopted societies and only one-​third by their experience in their country of origin.67 Dinesen and Sønderskov have studied trust among emigrants from Sweden over long time periods. They found that, after forty-​five years, Swedish emigrants have the same trust levels as natives of their new home countries.68 So the preponderance of the evidence is consistent with the experiential hypothesis. And this accords with my story about the psychological basis of social trust. Social trust is responsive to the behavior of others and, presumably, to the trustworthiness of others that we infer through our experience of their behavior.69 Interestingly, social trust levels tend to stick for immigrants who immigrate when they are over 30 years of age. The change in immigrant trust, according to Bergh and Richard Öhrvall, is in fact “entirely driven by expatriates who came to the new country when they were at most 30 years of age,” but for those who are older, “trust is remarkably [sic] independent of both time spent in the new country and institutional quality.”70 On this basis, they argue that “high trust is a sticky personality trait.” One worry for my approach is that I base my case that we can restore social and political trust on the influence of trustworthy behavior on those who observe it in their society. But if high trust is a sticky personality trait, that suggests that trust isn’t much affected by observation but is, rather, culturally or even biologically native. In response, I admit that social trust is affected by putatively nonrational factors like personality, early childhood and cultural experience, and contact with other persons with different ethnic backgrounds. But even 66 Social trust is composed of personal social trust attitudes, as I explained in MPBW, p. 36. 67 Helliwell, Wang, and Xu 2014. Interestingly, cultural persistence appears to be stronger in the United States than in other countries, though Dineson and Sønderskov think this observation is due to the greater amount of effort in measuring cultural persistence in the United States. See their p. 226. 68 This point is probably not undermined by the possibility of a self-​selection effect, since “those who have just migrated are very similar in their level of trust independent of their destination.” See ibid., p. 24. 69 One intriguing possibility, proposed to me by Eduardo Martinez, is that people may update in response to experience generally, but do so slowly, slowing down further over time as their experiences accumulate. That would explain the observations I’m discussing. 70 Bergh and Öhrvall 2017, p. 21.

66  Trust in a Polarized Age cultural persistence can be driven by observations of trustworthiness in one’s home culture and then extrapolation from home-​culture experiences to the trustworthiness of persons from other cultures. The reason that ethnic diversity effects decrease once persons regularly interact with one another may be that interaction increases trust by leading to the observation of trustworthiness among persons in other ethnic groups. Moreover, experiential factors still play a major role in trust levels. If we removed a high-​trust person from a high-​trust country and placed her in a low-​trust institution, she would surely reduce her trust levels after a large number of encounters with socially untrustworthy persons.71 This intuition is backed up by the literature on Swedish emigrants previously cited.

4.  Consequences of Social Trust The social value of social trust is a function of its contribution to the formation and maintenance of effective economic and political institutions.

4.1. Social Capital Many researchers believe that social trust is critical for creating social capital, which Edward Glaeser, David Laibson, and Bruce Sacerdote define as “a person’s social characteristics—​including his social skills, charisma, and the size of his Rolodex—​which enable him to reap market and non-​market returns from interactions with others.”72 Social capital is built in part by being trustworthy in the eyes of others, and being trustworthy requires meeting public expectations of norm compliance. When we trust others over an extended period of time, and are trusted in turn, we are able to establish social networks we can rely upon in taking the risks needed to build social institutions. Social capital only accumulates within a system of social trust, as people can more easily come to know one another under high-​trust conditions. Social capital is valuable for any number of reasons. 71 There are also important differences in trust levels between different ethnic groups, where ethnic minorities tend to be less socially trusting than average, but more politically trusting than average, with the exception of black Americans, who are low trust all around: “Ethnoracial minorities have lower social trust than majority groups; but by and large, with the exception of black Americans, this pattern is reversed for political trust.” It turns out that it is very difficult to separate cause and effect in the case of ethnic-​minority trust, and attempting to do so requires touching on points of controversy that I worry would distract more from this discussion than add to it, so I have not examined ethnic effects on social and political trust here. I leave that to future work. See Wilkes and Wu 2018, p. 231. 72 Glaeser, Laibson, and Sacerdote 2002, p.  48. For further discussion and several related definitions, see Dasgupta and Serageldin 2000.

Social and Political Trust  67

4.2. Economic Growth But perhaps the most well-​documented benefit of social capital is its correlation with economic efficiency and growth. Paul Whiteley has argued that “social capital . . . plays an important role in explaining the efficiency of political institutions, and in the economic performance of contemporary societies.”73 He focuses on a study of the relationship between social capital and economic growth in thirty-​four countries between 1970 and 1992, and he argues that the impact of social capital on growth “is at least as strong as that of human capital,” which is positively correlated with growth. Social capital has a similar impact on the ability of poorer nations to adopt technological innovations introduced by richer countries and to “catch up” with rich countries in terms of their level of development. Social capital is thought to reduce transaction costs in markets and reduce the burdens of enforcing agreements. It also limits fraud and theft. Recent work by Fabio Sabatini has helped to quantitatively substantiate political scientist Robert Putnam’s famous comparison of northern Italian and southern Italian institutions, with social capital in the former appearing both more plentiful and more effective than in the latter.74 Sabatini finds that “strong ties,” such as familial relations, do not promote economic development. However, “weak ties,” which act as conduits for the diffusion of knowledge and trust among strangers, do promote development and growth.75 Vidmantas Jankauskas and Janina Šeputienė have found that social capital—​ understood as a form of social trust involving the maintenance of wide social networks—​is positively correlated with economic performance in twenty-​three European countries.76 Reino Hjerppe, in a survey of the relationship between social capital and economic growth, argues that social/​generalized trust positively correlates with many measures of economic performance.77 Robert Hall and Charles Jones find that in 130 countries, differences in “social infrastructure” lead to considerable social differences in the accumulation of capital, economic productivity, and even educational attainment, which impacts income across countries.78 And Rothstein and Stolle argue that a survey of the social trust literature finds that social capital is said to produce “well-​performing democratic institutions, personal happiness, optimism and tolerance, economic growth, and democratic stability.”79 73 Whiteley 2002, p. 443, equates social trust and social capital, but his point still stands if these ideas are distinguished in the way I have proposed. 74 Putnam 1994. 75 Sabatini 2007, pp. 19–​20. 76 They find no correlation, however, between the degree of civic involvement in these societies and economic performance. See Jankauskas and Seputiene 2007. 77 Hjerppe 1998. The term “generalized trust” derives from Fukuyama 1995, p. 29. 78 Hall and Jones 1999, p. 84. 79 Rothstein and Stolle 2008, p. 441.

68  Trust in a Polarized Age One especially remarkable finding is how these benefits of social trust build up over time. Yann Algan and Pierre Cahuc find a “sizeable causal impact of inherited trust on worldwide growth during the twentieth century.”80 “GDP per capita in 2000 would have been increased by 546 percent in Africa . . . if the level of inherited trust had been the same as inherited trust from Sweden.”81 Societies that can achieve high trust and keep it will dramatically improve their economic position. Trust means growth. One might think that economic growth promotes social trust, since richer people are less likely to be suspicious of one another.82 However, Bjørnskov convincingly argues that the strong empirical connection between social trust and economic growth has mostly the reverse causal explanation.83 A high level of social trust reduces the cost of commercial transactions and thus facilitates economic growth through free exchange. Scholars disagree about how social trust leads to economic growth. Some studies argue that trust allows physical capital and education to accumulate, whereas other studies provide support for the claim that trust increases overall productivity; very recent work suggests that trust “mainly affects the development of productivity and only weakly affects the rate of factor accumulation, if at all.”84 And the effect is large: a one-​standard-​deviation rise in social trust increases productivity about 20 percent of a standard deviation. This literature should lead us to conclude that social trust has great social value since it facilitates the effective functioning of many vital social institutions.

4.3.  Legal Trust and Corruption As noted earlier, researchers find a strong correlation between social trust and legal trust, which is explained largely by the fact that a trustworthy and noncorrupt legal system helps people trust one another. However, social trust can also lower corruption, and this is probably because legal officials are seen as exemplary members of the public, such that trust in most people strongly transfers to trust in legal officials and the legal system.85 Another reason that social trust can lower corruption and increase legal trust is that, if you trust that others will avoid corruption, you are less likely to act corruptly86—​especially if 80 Algan and Cahuc 2010, p. 2060. 81 Ibid., p. 2074. 82 Friedman 2005 offers a related argument that growth produces more social cooperation. 83 Bjørnskov 2018. 84 For a literature review, as well as his argument that social trust increases growth by increasing factor accumulation, see Bjørnskov 2020. 85 Bergh, Bjørnskov, and Vallier 2021. 86 You 2018, p. 476. In this case, people are often motivated by a desire for reciprocity and fairness.

Social and Political Trust  69 you detect a social norm prohibiting corrupt action, such that you expect to be sanctioned or punished if you engage in corruption. You has speculated that, because social trust lowers corruption and corruption lowers social trust, societies can establish high-​trust, low-​corruption equilibria and low-​trust, high-​corruption equilibria.87 And indeed, many countries appear to have stable trust and corruption levels, and these stable levels can vary considerably internationally.

4.4. Economic Equality As indicated in the previous section, many researchers think that social trust causes economic equality, not just the other way around. The main mechanism is that more trusting populations are friendlier to policies that compress income inequality.

4.5.  Education There is a close tie between social trust and the growth of schooling (though not the absolute level), at least since 1960. Some have argued that educational attainment makes people more trusting, but more recent work strongly suggests that social trust causes schooling improvements.88 This is primarily because when social trust is low, employers are less likely to use education level as a signal of employee productivity, since they must attend more to cues that indicate the general trustworthiness of employees. But in high-​trust countries, educational attainment becomes a more important factor. This leads to greater demand for educated employees, which leads to more investment in schools.

4.6.  Political Moderation and Consensus As we saw in the introduction, high social trust tends to limit partisan divergence. Higher-​trust people are more likely to have moderate opinions and are better able to establish consensus to make important policy changes.89 Social trust leads people to focus more on the public good90 and discourages politicians



87

Ibid., p. 479. Bjørnskov 2009. 89 Pitlik and Rode 2019, p. 4; Berning and Ziller 2017, p. 198. 90 Leibrecht and Pitlik 2015, p. 249. 88

70  Trust in a Polarized Age from vetoing policy changes.91 High social trust leads people to become more patient and so let reforms work themselves out before complaining.92 Falling trust also makes it harder for people to adjust to demographic changes.93 There are other potential causal channels from trust to divergence as well. For instance, falling social trust often leads to greater hostility to income redistribution, which may increase economic inequality, which Nolan McCarty has argued may aggravate divergence.94

4.7.  Love and Friendship While there are fewer data on this, social trust is also arguably required in order to form thicker moral relationships, especially relations of romantic love and friendship. Societies with high social trust strengthen our capacity to sustain love and friendship. They allow us to establish relations of love and friendship with people outside our in-​group—​and even preserve those relations with people who leave it.95 In contrast to political trust, it appears that social trust is an almost unalloyed good. Of course, being socially trusting is not good if one is surrounded by untrustworthy people, but people in such circumstances generally do not learn to be socially trusting. Societies that can sustain high levels of social trust enjoy enormous benefits with few costs.

5.  The Causes of Political Trust Political trust, as noted, can be understood in a broad sense, as trust in government or in democracy, or in a particular sense, for example as trust in the civil service, the legislature, or a specific elected official.96 As with social trust, political trust is determined by survey work, especially the World Values Survey and the American National Election Survey.97 Importantly, the diversity of survey instruments might suggest that work on political trust deploys several different conceptions of political trust. Yet it appears that political trust research is approximating a single underlying construct.98

91

Ibid., pp. 264–​5. Ibid, p. 7. 93 Heinemann and Tanz 2008, p. 182. 94 McCarty 2019, pp. 79–​80. 95 MPBW, pp. 54–​8. 96 Norris 2017. 97 Marien 2017. 98 Ibid., p. 98. 92

Social and Political Trust  71 Unlike social trust, political trust is not an unalloyed good. Democracies depend on a certain degree of political distrust of public officials and political parties.99 However, if we distinguish between political trust in government and democracy and political trust in parties and elected officials, we can offer a finer-​ grained analysis according to which broad trust in democracy and government is good, but a certain degree of distrust of particular officials and parties is also desirable. We want people to be general political trusters and specific political distrusters. Political trust varies more than social trust and seems responsive to citizens’ evaluations of the success of institutions in pursuing their publicly recognized aims. Political trust is therefore probably better explained in institutionalist rather than culturalist terms, which distinguishes it somewhat from social trust. The prime cross-​time institutional causes of political trust are economic performance and quality of governance.100 When people think that their individual economic performance is strong, they become more politically trusting, and when people observe low-​quality governance, their political trust falls. I now turn to examine the causes of political trust.

5.1.  Personal Economic Improvement One central, well-​documented cause of political trust is economic performance.101 When people feel that their own economic performance is strong, they exhibit more political trust. I should stress that people increase political trust in response to their perceived increases in personal economic welfare in the recent past, say over the past several years. Political trust also often decreases in response to major economic crises and the fear of loss of income. However, true economic welfare, and true economic improvement, may not increase one’s political trust if one’s perceived economic welfare departs from one’s actual economic welfare. Political trust also appears to have some noninstitutional causes, such as social trust, though this may be indirect. For example, if social trust increases economic growth, and economic growth improves personal economic welfare, that will tend to increase political trust.102

99 Warren 2018. 100 Zmerli and van der Meer 2017, p. 7. These factors tell us whether political trust will rise or fall within a single country, which does not necessarily imply that these factors explain differences in political trust levels between countries. 101 Ibid. 102 Zmerli and Newton 2017, p. 105.

72  Trust in a Polarized Age

5.2.  Corruption and Inequality A second central, well-​documented institutional cause of political trust is the observed level of corruption. Corruption’s reduction of political trust is much like its reduction of social trust. Cassar, d’Adda, and Grosjean find that high “public sector corruption . . . is consistently found to have a strong, negative impact on political trust.”103 The corruption that damages political trust is primarily “grand corruption” widely observed by the public; small acts of corruption (“petty” corruption) don’t decrease political trust much, with the notable exception of bribes and gifts to the police or the courts.104 Corruption is tied to economic inequality because some people see inequality as evidence of corruption. Uslaner argues that “inequality leads people to believe that leaders listen far more to the rich than to others in society” and that, generally, perceived unfairness reduces trust in government.105 This effect can be hard to correct because rich citizens can navigate corrupt environments more effectively than the poor, which can help to cement or even increase economic inequality.

5.3. Procedural Fairness Corruption is not the only feature of governance that leads people to adjust their political trust levels. What researchers call “procedural fairness” is also important, but it is a compound concept that includes both (i) citizen involvement and (ii) the predictable, impartial treatment of citizens. Citizens tend to trust systems more when they have been involved in making decisions in that system, or at least believe that they have.106 Margaret Grimes argues that “perceived ability to exert influence in a decision-​making process affects views of a group authority and its decisions because it engenders a feeling of control and being able to secure a more favorable decision outcome.”107 Thus, procedural fairness includes citizen involvement. Citizens also expect authorities to treat group members respectfully, impartially, and predictably. Another central factor in procedural fairness is the “predictability of behavior” and “neutrality and impartiality in the exercise of authority.”108 Citizens will trust an institution less if they perceive that its operation

103

Van der Meer and Hakhverdian 2017, p. 84. Uslaner 2017, p. 308. 105 Ibid., p. 302. 106 Grimes 2017, pp. 259–​62. 107 Ibid., p. 257. 108 Ibid. 104

Social and Political Trust  73 is “arbitrary” and “conspiratorial,” or that its officials exhibit “self-​serving behavior.”109 Citizens expect the publicly recognized rules of operation to be consistently applied to everyone and in a way that is regular and predictable.110 Corruption, then, is not the only way in which the behavior of political officials impacts political trust. Political trust rises when people believe that governmental officials implement policy in ways that are neutral, impartial, prompt, honest, and predictable. Political trust in the post office should increase, therefore, if people feel that postal officials treat all customers’ mail with the same high degree of care and competence. Political trust in the IRS should increase if its officials process tax returns quickly, effectively, kindly, and in a way that shows no favoritism between citizens based on who they are.111

5.4.  Quality of Governance Trust researchers often relate political trust to an even more complex measure they call quality of governance. The World Bank Governance Indicators measure quality of government each year in each country by averaging four indicators: (i) government effectiveness, (ii) control of corruption, (iii) rule of law, and (iv) voice. Effectiveness includes “perceptions of the quality of public services, the quality of the civil services, and the degree of [their] independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies.” The control-​of-​corruption indicator measures “perceptions of the extent to which public power is exercised for private gain, including both petty and grand forms of corruption, as well as ‘capture’ of the state by elites and private interests.” The rule-​of-​law measure is meant to record “perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence.” Voice picks out and documents people’s perceptions about “the extent to which a country’s citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media.”112

109 Ibid., p. 264. 110 Ibid. 111 Here, as with other measures of political trust, there is a partisanship bias, where corruption in one’s preferred political party is less readily acknowledged than corruption in another party. See Anderson and Tverdova 2003. 112 Van Ham et al. 2017, p. 163.

74  Trust in a Polarized Age It is now common to find a strong correlation between quality of governance and political support, and when quality of governance is high, it may explain nearly all of the cross-​national variation in political trust.113 Economic performance indeed matters, but it may be that “the effect of economic performance is contingent on the quality of governance,” such that economic outcomes matter the most when “quality of governance is lowest.”114 What seems to matter the most is impartiality of policy implementation by the national bureaucracy, and policy impartiality may even crowd out the explanatory reach of corruption measures, though I  would think that these two factors are pretty closely related.115 I cannot overstate the importance of quality of governance for political trust. As one researcher puts it: The quality of government, measured as impartiality of the executive and state bureaucracy, is the crucial factor that explains cross-​national differences in political support. It does not only explain differences between full democracies and semi-​democracies, or differences between established and young democracies, but also variation between and within the established democracies of Western Europe. These findings suggest that we should move beyond the more conventional explanations of democratic legitimacy in the literature that emphasize government by the people . . . and government for the people. . . . Rather, to a large extent public support for democracy and its institutions seems to hinge on the liberal aspect of governments that are bound to impartiality by their own laws.116

This is why, as I show in c­ hapter 8, elections matter less for social and political trust than one might think. Liberal impartiality is more critical for political trust than elections, as I argue in c­ hapter 7.

5.5. Social Insurance Effective social insurance also seems to increase political trust, as “the more positively an individual estimates the state of social protection in the country, the higher the satisfaction with democracy.”117 The reasons for this

113

Ibid., p. 199. Ibid., pp. 157, 161, 199. Ibid., p. 199. 116 Ibid., pp. 154–​5. 117 Kumlin and Haugsgjerd 2017, p. 289. 114 115

Social and Political Trust  75 are complex but probably have to do with decreasing economic anxiety and reducing inequality. Political trust often decreases in response to major economic crises and the fear of declining income. Importantly, there is evidence that means-​tested social insurance targeted at the poor can reduce political trust, while universal welfare supports that go to the middle class can increase social trust.118

5.6.  Authoritarianism versus Democracy One important finding in the political trust literature is that authoritarian regimes can exhibit high levels of political trust, as long as they manage corruption and economic performance effectively. Some countries that have replaced authoritarian regimes with democratic ones have even seen a decline in social trust, though this appears to reflect the experience of making that transition.119 People’s high hopes for democracy are often disappointed. Political trust is also affected by a winner’s bias, such that political trust is higher after an election win for one’s party and lower in the case of a loss, especially in winner-​take-​all voting systems. This can be trust-​reducing vis-​à-​vis authoritarian regimes. Another trust-​reducing effect of democracy is that citizens will now blame vast numbers of their fellow voters, rather than a few authoritarian leaders, for bad outcomes. Democracies will encourage people to be angry at other large sectors of the population. As a result, democracy can foster distrust of elements of the democratic system. And yet democracy can help build political trust. In general, observed compliance with political norms by officials, especially in the civil service, will tend to increase political trust of some varieties, as will observed compliance with norms calling for citizen involvement.120 In particular, when citizens feel that government officials treat them fairly, they are more likely to be politically trusting, and democracy is often seen as uniquely fair.121 Evaluations of democracy are especially important in developed, democratic societies, where economic growth is slower than in developing, authoritarian nations, where evaluations of economic performance are most salient. I will explore these and other factors in c­ hapter 8.

118 Ibid. Also see Kumlin and Rothstein 2005. 119 Letki 2018. 120 Van Deth 2017, p. 224. Though observed compliance with political norms can also increase political trust in nondemocracies. 121 Grimes 2017, p. 256.

76  Trust in a Polarized Age

5.7.  Media There is limited evidence that mass media can affect political trust; the effects are more modest than one might expect and may not be negative, depending on the media source.122 Before we can assess the effect of media on political trust, however, we should examine the data on trust in mass media themselves. Here are some of the main facts about trust in media.123 First, media distrust ranges from significant to high in most developed countries, and trust in media is often declining, especially in the United States (falling from 72 percent in 1972 to 40 percent in 2012). There are differences in trust between types of media—​say, newspapers and cable news—​and there is substantial variation within media types, such as among different cable news channels. Public-​service news sources tend to be more trusted. Somewhat surprisingly, trust in media and journalists is similar to that exhibited toward many other major institutions and occupations. The reasons for distrust are various, but a main source of distrust of media is that they are seen as influenced by powerful people and organizations. It looks like media exposure tends to reduce political trust in government, perhaps because people think that the media are “sensational, superficial, inaccurate” and concentrate on “bad news.”124 Importantly, mass media do not make already-​distrustful people less trusting but, rather, can hurt trust among the trusting. This trust-​reducing effect seems tied largely to trust in newspapers, however, and not trust in television news. Some researchers have postulated a general “media malaise” effect in reducing political trust. Media sources concentrate on bad news, a tendency compounded by negative election campaigning, uncivil behavior by politicians, attack journalism, the prioritization of stories with entertainment value, and even the emphasis television places on colorful, moving images. However, some find evidence of potential for a virtuous circle “in which attention to election campaign communications and individual feelings of political trust are mutually reinforcing.”125 The trust effects depend on the sort of media one consumes. Those who watch a lot of entertainment television or simply a lot of television have lower political and social trust, but watching a lot of TV news is more strongly associated with high political trust.126



122

Newton 2017. Ibid., pp. 354–​8. Ibid., p. 353. 125 Ibid., p. 363. 126 Though Fox News viewers seem to be an exception. See Benkler, Faris, and Roberts 2018. 123

124

Social and Political Trust  77 It is unclear what the effect of media distrust is on political trust. If people do not trust the media much, then they probably are more likely to “fall back on their partisan predispositions when evaluating political events and opinions.”127 Unfortunately, we know less about the connection between media trust and political trust in authoritarian regimes, though political trust will partly depend on the extent to which citizens in authoritarian countries trust mass media, given that they are universally controlled in those societies. Though some find a positive association between social media use and trust,128 others find no association between use of social media and social trust.129 Unfortunately, all things considered, the evidence for a causal connection between trust and social media is not yet strong enough for me to say much about it.

5.8.  Education Education improves citizens’ ability to track the behavior of governments. When government performance is seen as strong, education can engender political trust, but if performance is seen as weak, the opposite may occur.130 This is probably because education makes citizens more discerning consumers of mass media about political matters, and so they detect when their institutions are performing well or poorly and why. Education may also create trust by reducing the perceived risk of interacting with other citizens, which could affect social trust.131

5.9. Association Participation Participation in civic associations seems to increase political trust, though it is not entirely clear why.132 Associations may improve the quality of democracy, perhaps by helping with responsibilities the state cannot address. They may increase the representativeness of democracy because associations can influence politicians more effectively than individuals can.133 And people who live in communities with an active civic life may tend to be more trustworthy, although these



127

Ibid., p. 366. Warren, Sulaiman, and Jaafar 2014. 129 Pavić and Šundalić 2016. 130 Mayne and Hakhverdian 2017, p. 181. 131 Knack and Zak 2002. Trust may well also improve access to education. Bjørnskov 2009. 132 I develop an account of civic associations in c ­ hapter 3. 133 Liu and Stolle 2017, pp. 338–​9. 128

78  Trust in a Polarized Age differences are typically thought to be selection effects: high-​trust persons are more likely to join associations. I will review the evidence further in c­ hapter 3.

5.10.  Immigration Lauren McLaren finds that “perceived size of immigrant groups” has negative effects on political trust but actual sizes do not.134 Perhaps these effects arise because ethnic diversity decreases political trust and immigrant groups are perceived to increase ethnic diversity. However, these trust-​reducing effects can be addressed when immigrants and natives create “new, cross-​cutting forms of social solidarity and more encompassing identities.”135

5.11. Partisan Divergence As I argued in the introduction, there is good evidence that partisan divergence reduces political trust.136 This is probably because divergence often leads to more gridlock in presidential and bicameral democracies like the United States, which reduces quality of governance. Since, again, quality of governance is the main determinant of political trust, divergence should decrease political trust.137 One might also argue, along similar lines, that if gridlock also reduces economic growth and economic security, political trust will fall further still. And if divergence leads to more economic inequality, as Uslaner138 and McCarty139 argue, then since inequality seems to reduce political trust, that is yet another source of distrust. Falling political trust also sometimes leads to populist voting,140 and populist politicians often erode democratic norms, undermining political trust further.141 Lower political trust may then engender the erosion of central constitutional norms because when a public has low political trust, it sees its political system as broken and corrupt.142 134 McLaren 2017, p. 320. 135 Putnam 2007, p. 137. Putnam’s now-​classic piece finds that immigration and ethnic diversity tend to reduce social solidarity and social capital. 136 Introduction, see Introduction, section 1; see also Uslaner 2015, p. 363. 137 Uslaner 2015, p. 363. 138 Ibid. 139 McCarty 2019, pp. 78–​81. 140 Geurkink et al. 2020. 141 Norris and Inglehart 2019, pp. 65–​6. 142 McCarty 2019, pp. 154–​5. For a similar argument, see Rosenfeld 2018, pp. 282–​3. On the other hand, polarized/​divergent populations may be more inclined to monitor government for corruption and punish it when it occurs, so long as political competition is high and incumbent governors can run for office; for evidence, see Melki and Pickering 2020.

Social and Political Trust  79

5.12. Social Trust Some researchers find that social trust increases political trust, though often through indirect factors. For instance, the causal connection may be indirect if, say, social trust increases economic growth, which improves personal economic performance, which in turn increases political trust.143 Jeffrey Mondak argues that trustful people are more politically trusting.144 But the big reason that social trust may cause political trust is via the effect of social trust on corruption. Social trust may cause political trust to rise in democratic societies because fairness norms are salient in democracies, so unfairness should lead to reduced trust in those societies; second, public corruption in democracies tends to be widespread rather than concentrated at the top, as it is in authoritarian countries, where most people cannot draw from the public trough.145 Corruption might be harder to control in democracies because there are more political actors, which creates collective-​action problems.146 These effects are not present in authoritarian countries; indeed, the association between social trust and political trust in authoritarian countries is “nil.”147

6.  Consequences of Political Trust Interestingly, we know more about the causes of political trust than of social trust, but we know more about the consequences of social trust than of political trust. Researchers disagree about why political trust is declining across North America and in many Western European nations.148 Some argue that political trust is declining due to rising inequality and various widely observed events and governmental failures, such as Watergate and other corruption scandals. Others argue that as people grow richer and more educated, they become more discerning observers of political events and come to have higher expectations of democracy. “At the same time that people have become less trustful of government,” Russell Dalton has written, “other opinion surveys show continued and widespread attachment to democracy and its ideals, which may even have strengthened in recent decades.”149 143 Zmerli and Newton 2017, p. 105. 144 Mondak, Hayes, and Canache 2017, p. 146. See also c­ hapter 2, section 3.6 on personality and culture in the context of social trust. 145 You 2018, p. 480. 146 Ibid, p. 479. 147 Ibid., p. 487. 148 Or, if one follows van Ham’s interpretation, political trust is falling only in the United States, not Canada or Western Europe. 149 Dalton 2017, pp. 381–​2. For further attempts to explain falling political trust in terms of rising expectations rather than falling commitment to democracy, see the critical-​citizens approach in

80  Trust in a Polarized Age If Dalton is correct, then declining trust in government may “not represent alienation from the democratic process” and low political trust may not lead to a crisis of legitimacy for democratic governments generally. In fact, falling political trust could be a good thing, if it means that citizens are more attentive and expect more of their officials. According to Dalton, scandals and mass media, and even some economic disruptions, probably do not reduce political trust much or erode social capital. Instead, political trust is falling cross-​nationally regardless of these events, and the data do not show that social capital is decreasing cross-​nationally.150 Ronald Inglehart has argued that “public priorities are broadening to include new postmaterial values that stress autonomy, self-​actualization, and a more assertive political style” and that these values often lead “individuals to question authority,” reducing trust.151 These “engaged citizens” are “conceptually equivalent to postmaterialists,” and while they trust politicians and political institutions less, they strongly support democratic values, including equality and the protection of minority rights. Dalton argues, plausibly, that political trust falls when there is a “precipitating factor that makes people begin to question government in a new way,” such as a scandal or a recession. However, when the conditions improve, and a new government is elected, political trust “does not return to its previously high level.” Postmaterial values have already raised citizen expectations. Consequently, government failures make citizens less likely to have political trust in the future. This account of falling political trust is interesting because it tells us whether we should expect a decline in political trust to cause problems. If declining faith in political institutions is the result of increasingly postmaterialist citizens, who are confident in democracy and insist that it do better, then a drop in political trust should not create social instability. But if political trust has fallen due to factors that also undermine democratic values, then we should be more worried. I do admit that there is evidence that people in younger cohorts may not have lower political trust simply because of their dissatisfaction with their particular government. Roberto Foa and Yscha Mounk argue that “across a wide sample of countries in North America and Western Europe, citizens of mature democracies have become markedly less satisfied with their form of government and surprisingly open to nondemocratic alternatives.”152

Norris 2011, pp. 4–​5, and the allegiance-​assertion theory of citizenship in Dalton and Welzel 2014, pp. 283–​8. For earlier work on a related hypothesis, see Pharr and Putnam 2000, pp. 23–​6.

150

Dalton 2017, p. 386. Inglehart 1990. Also see Dalton 2017, p. 387. 152 Foa and Mounk 2016. 151

Social and Political Trust  81 We need reasonably high political trust for a number of reasons. Marc Hetherington argues that political trust is important to give leaders the confidence they need to enact programs that improve lives, because less-​informed citizens may not understand how the programs work and, if those citizens are also low in trust, they will not trust leaders to enact the policies anyway.153 Rothstein claims that countries with higher political trust have higher-​quality governments, and that such governments spend more on social policies and improve other outcomes.154 Low political trust may, therefore, make it harder for governments to give engaged citizens what they want in terms of better policy. Accordingly, despite the likelihood that political trust has fallen due to the emergence of more engaged, demanding, and critical citizens, the resulting decrease in political trust has made government less functional and responsive, which is bad. Democratic idealism, by leading us to expect government to do better, may have made it perform worse. Hetherington also argues that political trust is important when citizens must sacrifice something, especially when richer citizens face a risk of large losses through internal redistribution or generosity to foreign nations. He argues that “when political trust is relatively low, Americans are significantly less generous to other nations, in addition to poor people at home, than when they trust their government more.”155 This also seems true for racial policy: when political trust is low, people do not trust the government to improve racial equality.156 On the other hand, as noted, democracies depend on a certain degree of political distrust in public officials and political parties.157 So how much political trust is optimal? To answer this question, let’s distinguish between trust in democracy and trust in specific parties and elected officials. One can plausibly argue that political trust in democracy and government is good, but trusting particular officials and parties is often problematic. It may be best if people trust democratic institutions but remain somewhat skeptical of prominent and influential officials.158 However, again, if citizens are too skeptical and mistrustful, this may prevent political officials from enacting good policies that may be politically unpopular. And yet, if one thinks that many government policies are bad, then one may find low political trust advantageous.159

153 Hetherington 2005. 154 Rothstein 2011. Political trust can cause politicians to be trustworthy, engendering further trust in turn. 155 Hetherington 2005, p. 98. 156 Ibid., p. 105. 157 Warren 2018. 158 We may also expect most political officials to be trustworthy so that they can keep less trustworthy politicians in check. 159 Hetherington argues, for instance, that President Bill Clinton’s healthcare reform was derailed by Republicans’ associating the reform with their low political trust in government and in Clinton (2005, p. 127). It is also the case that “more trust increases support for military intervention,” and if

82  Trust in a Polarized Age

7.  Learning to Trust* In this section, I want to address a powerful challenge to my attempt to harmonize the empirical trust literatures with a philosophically respectable account of trust. I have claimed that trust is justified when people reasonably think that others comply with moral rules for moral reasons, not merely reasons of self-​ interest (or malice, or some other nonmoral consideration). But the empirical trust literature does not address whether people base their trust on whether they think trustees act from moral reasons rather than, or in addition to, reasons of self-​interest. For empirical researchers, that is all to the good, because it is difficult to sort out empirically whether persons act from moral reasons and whether trusters think that other people act from moral reasons. But that means I need something to say about how we learn to trust others, if indeed trusting others involves believing that they act from moral reasons in part or on the whole. The trouble is that moral rule compliance does not usually signal why people comply, since people’s reasons for action are frequently obscure. Given this, how do we learn that others are trustworthy from their rule compliance? How do we form social trust when we usually do not know the reasons that motivate people? My answer has two parts. First, any plausible account of learning to trust must postulate that people have some native disposition to trust others and to believe that others are basically morally decent. So, when we observe norm compliance, we take this to confirm our view that others are fundamentally moral and decent. However, if we are convinced that most people can’t be trusted, or worse, that most people are sociopaths, then no amount of observation is going to motivate trust. In fact, in this state of distrust, even direct communication with others about their reasons won’t increase trust because low trusters won’t believe that others are telling them the truth. This is why I take a modest level of trust for granted, and then argue that this level of trust can be increased and then stabilized at high levels. We cannot create high trust from zero trust. Nonobservational factors, such as upbringing, personality, and culture,160 as well as heuristics that people use to judge how trustworthy others are, such as character assessments, can also lead people to trust others. And this presumption can be strengthened by various kinds of trust-​building exercises.161 Once people have reached a threshold of trust in others, then observing rule compliance in certain cases can rationally increase trust in other people, since we

one thinks many military interventions are problematic, then, again, one will see a benefit to lower political trust. See Hetherington 2005, p. 143.

160 161

Cawvey et al. 2018. Kramer 2018, pp. 105–​10.

Social and Political Trust  83 can sometimes discern the reasons why persons act. In particular, we can learn about others’ motives in cases where self-​interest and moral commitment conflict, since self-​interested reasons are the standard reasons that lead people to ignore their moral reasons and so to demonstrate untrustworthiness.162 The key to learning to trust from observation is knowing something about the norms that embody a society’s shared moral beliefs (or at least the beliefs shared by people in the group of which the person one is observing is a member, or maybe even her own idiosyncratic beliefs). Since complying with social norms is usually at least somewhat costly, then rule violations can be attributed to self-​interest rather than from moral motivation. We can therefore increase our estimate of the trustworthiness of others as follows. John rationally increases his trust in Reba when (i) John observes Reba complying with a moral rule, (ii) John knows that the costs of following the moral rule are significant and recognized as such by most members of the public, including Reba, and (iii) John knows that people like Reba tend to recognize moral reason to observe the rule. For example, suppose that John has lost his wallet, which contains quite a bit of money. A week later John receives the wallet in the mail from Reba, with all the money accounted for. Reba’s self-​interest favored taking the money and not returning the wallet, as there would be no way to detect her self-​interested action. John therefore knows that Reba’s action in returning the wallet went against her self-​interest. He also has reason to think that people in his society generally think returning someone else’s wallet is the right thing to do. Conversely, John can learn to trust Reba less if he discovers that she found the wallet and chose not to return it.163 There is another, less common, kind of case where John can learn to trust Reba. John might come to trust Reba more if he observes her violating a moral rule where Reba’s self-​interested reasons favor complying with the moral rule but her moral reasons are known to require violating the rule. For instance, suppose Reba is (and John knows she is) a member of a pacifist religion, but there is a military draft and the social norm is to comply with it. He then observes Reba openly refusing conscription and being arrested. This may increase his trust in her. Conversely, John might rationally reduce his trust in Reba if she follows a moral rule in a case where her self-​interest favors it but her moral reasons are known to oppose it. For example, suppose John knows that Reba is a member of

162 In one way, since all social-​norm compliance seems to come at some cost to self-​interest by definition, observing social-​norm compliance may nearly always provide at least some confirmation that the person one is observing acts for moral reasons and not just self-​interested ones, unless one thinks the person is complying with the social norm simply because she fears sanction that she regards as illegitimate because she has not internalized the norm. 163 I choose the wallet return case for a reason: it is a standard way that recent trust research measures trust beliefs (Bauer and Freitag 2018, pp. 26–​7).

84  Trust in a Polarized Age a religion that forbids her from complying with a common social practice, such as saying the Pledge of Allegiance, but he sees that she says it anyway. He may thereafter trust her less. Learning to socially trust can function like learning to personally trust. Consider trust in the legal system. Most societies presume that police and judges have adequate moral reason not to take bribes. But bribes are cases where self-​ interest and moral reasons come apart. Cops may benefit financially from taking a bribe, but they act wrongly if they accept a bribe. So, when people observe cops taking bribes, that is evidence of untrustworthiness because people know cops are following self-​interest rather than acting on moral reasons. Conversely, when cops refuse to take bribes, then observers know that cops are typically acting on moral reasons rather than self-​interest, such that a refusal to take a bribe is evidence of trustworthiness. This is probably one main reason corruption in the legal system has a clear effect on social trust, because it is widely presumed that members of the legal system have sufficient moral reason to avoid corruption, even when it would economically benefit them. So cases of legal-​system corruption provide proper evidence of the trustworthiness of others. Similar cases abound in other institutions, such as when commercial institutions refuse to engage in profit-​generating fraud, or when motorists abide by the rules of the road even if it slows them down in reaching their destination. Thus, so long as there are practices that are widely and rightly presumed to be morally binding for all or nearly all people, and there are temptations to violate the rules that constitute the practice, then observing compliance with the rules should increase trust, and observing failures to comply with the rules should decrease trust. As that helps us understand, trust can be increased by observing actions and attitudes from others that tend to disconfirm our negative stereotypes about them.164 Our negative stereotypes may lead us to think that in cases of conflict between moral reasons and reasons of self-​interest, members of a mistrusted out-​ group will tend to follow reasons of self-​interest. But if we observe an out-​group complying with a moral rule in cases where doing so undermines its members’ self-​interest, that can provide reason to trust. Further, people can also learn about the moral motives of others through testimony, such as when a high-​status member of one’s group sanctions interaction with an out-​group.165 If a high-​status person in one’s reference network indicates that members of an out-​group have good character or are trustworthy, then one may infer that they—​and others more broadly—​typically do the right thing for the right reasons.

164 165

Pettigrew 2000. Kramer 2018.

Social and Political Trust  85 Thus, people can learn to trust others when they have ways of determining whether others comply with moral rules from moral motives rather than self-​interested  ones.

8.  Implications for Sustaining a Diverse Social Order I will now link the survey of the empirical literature to the problem of the book. For it is now clear that some liberal rights practices help to sustain social and/​or political trust in a large, diverse social order. Sustaining a diverse social order requires effective economic and judicial institutions. First, a functional and fair legal system, including both courts and police, is needed for people to trust one another and to trust government. Perceived corruption in political systems, and especially in legal systems, undermines both social and political trust quickly and sharply. Second, people are more politically trusting when they believe that their economic performance has improved and, sometimes, when they believe that the economy is doing well generally. Societies that can overcome the challenges of diversity are ones that know how to combat corruption and rent-​seeking; maintain fair, nonextractive economic institutions; and possess the policy flexibility and sensitivity to promote broad-​based economic growth. Developing countries like China experience torrid rates of catch-​up growth and, arguably, enjoy high levels of political trust as a consequence.166 But in developed nations, where growth is slower, economic performance matters less for political trust. Richer and better-​educated societies have more discerning and demanding citizens, who are politically distrustful in part because they are such strong supporters of democracy. Thus, we should expect political trust to fall worldwide as economic fortunes rise. Rising wealth and material security will lead the people of up-​and-​coming middle-​income countries to pay more attention to the noneconomic dimensions of policy and tune in to questions of morality, group identity, and individual rights. While the Chinese appear to have enormous trust in their authoritarian national government, this does not suggest that mature liberal democracies might do better by restricting democracy or limiting the electoral process. Instead, ensuring that institutions are widely responsive to the interests and beliefs of critical citizens is essential for maintaining political trust. We can expect that as China develops and, as it approaches the level of wealth found in developed nations, its economic growth slows, Chinese

166 Assuming that the China data are sound, which some trust researchers doubt. See Uslaner 2002, p. 226.

86  Trust in a Polarized Age citizens will come to have more multifaceted political evaluations and so will express lower political trust. While elections are associated with lower political trust, this is not because elections undermine democracy but because people are more focused on the warts of governmental policy and procedure. And if elections are essential for good policy, in terms of both quality of governance and economic performance, then electoral institutions are essential for overcoming excessive and dysfunctional levels of political trust. We have also seen that high participation in many kinds of associations may promote political trust by putting citizens into positive contact with one another, so protecting robust freedom of association is yet one more way in which the institutions of free societies can promote social and political trust across difference. We shall also see some evidence that associations help social trust in ­chapter 3. There is also reason to believe that policies aimed at reducing economic and ethnic segregation will promote social and political trust. The challenges of diversity can be overcome, at least in part, through encouraging diverse people to interact with one another regularly in all aspects of life. So there is an important imperative of integration to increase social and political trust among diverse persons in order to help achieve the goods trust brings.167 If we want to sustain social cooperation in diverse societies, and to reap the benefits, we need policies and institutions that maintain high social trust and high political trust in democratic government generally, while allowing for a modestly high level of political trust in particular politicians and political parties. We can do this by protecting liberal rights, such as freedom of association and freedom from discrimination and segregation. We can do this by protecting democracy, which reduces corruption and enhances economic and institutional functioning broadly. We can do this by protecting the rule of law, which reduces corruption and improves economic performance. And we must also make use of markets and social insurance in order to ensure that the economy performs well for everyone and protects most people from being exposed to dangerous, trust-​destroying economic and social risks. This is probably why the liberal-​ democratic, capitalist welfare states, such as the Nordic countries, have the highest levels of social and political trust. There is a positive feedback loop between these institutional structures and social and political trust generally. Their high trust is not merely the result of high ethnic homogeneity. Notice how the data bear on the distrust-​divergence hypothesis. It is prima facie plausible that both social and political trust decrease partisan divergence,



167

Anderson 2010.

Social and Political Trust  87 and that reducing partisan divergence increases social and political trust in turn. But we also have good empirical evidence that falling social trust produces divergence, which reduces political trust in turn. Thus, we have partially substantiated the distrust-​divergence hypothesis. And if that hypothesis is true, and if I have accurately represented the findings of the trust literature, then there is good reason to think that the distrust-​divergence feedback loop can be interrupted: at least some liberal rights practices have the consequence of creating and/​or sustaining social and political trust. I will now deploy four two-​pronged arguments that particular liberal rights practices create and sustain trust for the right reasons. This requires diving more deeply into the empirical literature and matching that literature with the public justification-​based arguments for the associated liberal rights. I  turn first to freedom of association.

3

Civil Society and Freedom of Association In this chapter, I first show how the exercise of the right to freedom of association creates trust in the real world. I then argue that a right to freedom of association is publicly justified and, therefore, creates trust in the right way. Combining these arguments shows that freedom of association creates trust for the right reasons, helping to vindicate the thesis of the book.1 I shall here understand freedom of association as a series of claim-​rights to form associations with others voluntarily and to govern the association according to the will of its members. Associational rights imply an immunity from interference by other citizens or government. Only association members control the nature and direction of the group. Most importantly, associations direct the activity of members in a certain domain; they have a claim on voluntary members to act in certain ways or face removal from the organization. These rights make it possible for persons to become robust association members, to join the institution and engage in the practices that give the association its purpose. In some cases, robust association membership implies a role in governing the institution. In this way, freedom of association as a rights practice occurs when citizens use freedom of association to form, maintain, and govern associations and act in concert with the values, principles, and demands of the associations in which they choose to participate. Freedom of association as a rights practice is measured in terms of association membership, and in some cases by participation in association events like weekly worship services in churches. In the literature, association membership is an empirical proxy for association participation, so the empirical measures do not always show that a member engages in the rights practice. But they measure something we can use as a proxy for engaging in the rights practice. Freedom of association creates real trust by putting people in contact with other members and with nonmembers in many cases. It creates trust for the right reasons because the recognition, protection, and exercise of the right of association serve as public evidence of the trustworthiness of association members and governments that recognize and respect and protect the rights of associations



1

Some of the content of this chapter derives from Vallier 2020a.

Civil Society and Freedom of Association  89 members. Abrogation of these rights and interference with associational practices can be public evidence of untrustworthiness. I begin by reviewing the trust literature on freedom of association, focusing on civic associations, broadly understood as nonprofit institutions with a moral aim rather than an economic one (section 1). I next develop an account of the kinds of moral rules that constitute associations (section 2). The next three sections (sections 3–​5) outline the ideas of a moral association and distinguish between the two main types of moral association, civic and commercial. I then explain how to publicly justify the primary right of freedom of association (section 6) and why the state does not have a priority over moral associations that would give it a blanket claim to interfere with them in constructing the social world (section 7). The next four sections address some common objections to robust freedom of association. I argue that freedom of association need not lead to private tyranny and balkanization (section 8), and I address feminist worries about freedom of association (section 9). I discuss the complex relationship between property and the right of association (section 10). Finally, I answer an objection from recent work by Jacob T. Levy, which holds that social contract theories cannot take on the insights of the pluralist liberal tradition, which celebrates freedom of association (section 11). I then conclude (section 12).

1.  Associations and Real Trust In MPBW, I argued that societies that protect freedom of association furnish the rational basis for trust and contribute to a non-​warlike politics by undermining our natural, robust propensity to distrust. But what are the consequences of freedom of association in the real world? The hope is that freedom of association makes it possible for societies to have a robust, flourishing civic life. When people join and participate in associations, trust may result from their increased interaction with members and nonmembers. The literature may bear this out with respect to civic associations. Unfortunately, we lack data on the relationship between membership in commercial associations and social and political trust, so I cannot speak to that here. But I think the causal channels from civic association participation to social and political trust should also work for commercial organizations. Civic associations are widely thought to contribute to social trust and political trust. Robert Putnam has famously argued that “an effective norm of generalized reciprocity is bolstered by dense networks of social exchange” like civic associations.2 The back and forth of favors and service to one another helps to “foster

2

Putnam 2001, p. 136.

90  Trust in a Polarized Age sturdy norms of generalized reciprocity and encourage the emergence of social trust.”3 Mark Warren argues that associational life can produce trust and that the forms of trust produced by reciprocity within associations can produce the sort of trust that helps sustain political institutions.4 Nancy Rosenblum argues that “the chief and constant contribution of associations to moral development is cultivating the disposition to cooperate,” even though these contributions are not guaranteed.5 Pamela Paxton has found that “membership in connected associations”—​ones that do not isolate themselves from society but permit membership in other groups—​increases trust, such that “at the national level, having more connected voluntary associations increases trust.”6 A survey of over thirty countries found that “at the individual level, membership in any voluntary association is a strong predictor of generalized trust.”7 Uslaner sharply disagrees.8 He argues that associational experience can increase in-​group identification and “particularized” trust, both of which can undermine social trust generally. But Uslaner is in the minority. Paxton thinks the work Uslaner cites is not worrisome once we distinguish between types of organizations.9 If organizations are ethnic in nature, then they can undermine social trust, but the voluntary associations we’re concerned with correlate with social trust when they’re connected to other voluntary associations and not self-​isolating.10 The trust-​association connection draws on what psychologists call intergroup contact theory, following famous work by Gordon Allport.11 It has sometimes been found that contact between members of different groups, including ethnic groups, helps to reduce prejudice and increase trust.12 Contact can also increase prejudice in many contexts, but in a huge literature review, Thomas Pettigrew and Linda Tropp have found that prejudice can be reduced when some or all of these conditions hold: (1) groups share equal power and status in the contact situation;

3 Putnam 1995, p. 67 4 Warren 2001, p. 74. 5 Rosenblum 1998, p. 59. For caution about how much these skills generalize, see p. 47. 6 Paxton 2007, p. 47. 7 Ibid., p. 65. See also p. 342. 8 Uslaner 2002, pp. 145–​6, 162. 9 Paxton 2007. 10 Jennings and Stoker 2004 also argue that generalized trust is tied to civic engagement via associations, but not for associations “based on exclusive identities.” Jennings and Stoker, however, raise the possibility that trust is a cause of civic engagement, not the reverse. 11 Allport 1954. 12 For a review of the contact literature in economics, see Bertrand and Duflo 2017. Importantly, Bertrand and Duflo find lots of evidence that contact does not reduce prejudice and distrust, though there do seem to be some contexts where the contact hypothesis holds.

Civil Society and Freedom of Association  91 (2) groups share interdependence with respect to common goals; (3) the potential for the emergence or development of cross-​group friendships exists; (4) the positive interactions run counter to and undermine prior negative group stereotypes; (5) group authorities or leaders sanction the intergroup contact.13 Pettigrew and Tropp claim that results “from the meta-​analysis conclusively show that intergroup contact can promote reductions in intergroup prejudice.”14 While recent work from Elizabeth Paluck, Seth Green, and Donald Green qualifies Pettigrew and Tropp’s review, especially in cases of reducing ethnic or racial prejudice, they nonetheless agree that contact usually reduces prejudice.15 Recent research also indicates that intergroup contact in these cases can promote trust, and that interpersonal trust can form quickly if these conditions occur.16 There are two ways intergroup contact through association participation may generate trust. First, associations can foster social trust by creating and sustaining social situations in which group members learn to engage in reciprocal interaction with one another. Participating in associations helps persons to learn to trust one another through observing trustworthy behavior within the group. Moreover, associations can sanction untrustworthy members, which provides further incentive to be trustworthy. Routine and repeat events also help to create more trustworthy members, especially by rehabilitating or reinstating those who have been sanctioned, and communication of reputation plays a role as well. Associations also create predictable behavior, which helps form trusting attitudes. Second, associations can create trust between group members and nonmembers. This may lead individuals not only to transfer their trust in members to nonmembers but to develop trust in nonmembers directly. Associations create opportunities for their members to interact with nonmembers. Admittedly, not all associations promote contact with other persons, and some encourage persons to form particularized trust alone, even at the expense of social trust. But association membership in toto seems to increase trust because it increases contact with the public. At least “bridging associations,” those that expose individuals to others unlike themselves, ethnically, religiously, and so on, should increase trust, even if “bonding associations,” those that



13

Pettigrew 2000; Pettigrew and Tropp 2006. Pettigrew and Tropp 2006, p. 751. 15 Paluck, Green, and Green 2018. 16 Kramer 2017, p. 16. 14

92  Trust in a Polarized Age expose individuals only to other group members, may sometimes reduce social and political trust. Associations that have formal or informal relationships with other associations, such as the relationship between a church and a soup kitchen, may be especially good trust promoters. Association membership promotes real trust when people with cross-​cutting identities interact with one another. For example, a political organization might organize persons with different religious beliefs, but who are members of the same economic class, thereby helping its members see people of different faiths as allies and equals. A religious congregation might consist of people with similar religious beliefs but different socioeconomic backgrounds, creating a context for members to build meaningful relationships across class lines and appreciate the shared humanity of rich and poor. In this way, multiple memberships create more trust than participation in a single organization: Even if all the members of one of your associations have a particular identity in common with you, the members of another of your associations may be diverse in that respect and homogeneous in another. Importantly, it looks like all of these causal channels by which associations create trust meet one or more of Pettigrew and Tropp’s conditions for Allport’s contact hypothesis to hold. These causal channels not only create real trust but provide their own right reasons to trust. This is because these causal channels do not subvert or conflict with the factors that justify trust but, rather, reinforce them. If trust is justified by trustworthiness and associations increase trustworthiness, then they justify more trust.17 And if trust is justified by experience of others’ trustworthiness and associations increase exposure to others’ trustworthy behavior, then for that reason too, associations justify more trust. However, while it is common for political scientists to affirm a causal connection between association membership and social trust, there are concerns that those who choose to join voluntary associations “may already be more trusting than those who do not.”18 Some data suggest that “high trusters join associations more frequently than low trusters.”19 And one paper argues that “the civic engagement hypothesis has not fared well empirically as the association between civic engagement and social trust at the individual level seems to be primarily driven by self-​selection” and that there is more evidence of the influence of 17 Trustworthy behavior also justifies trusting attitudes in others in ways compatible with respect for persons when that behavior involves compliance with publicly justified norms, as well as only holding others accountable for violations of publicly justified norms. For discussion, see MPBW, p. 48. 18 Paxton and Ressler 2018, p. 160. 19 Newton, Stolle, and Zmerli 2017, p. 43.

Civil Society and Freedom of Association  93 culture and major institutions on levels of trust.20 Further, when education, sex, age, and income are taken into account, “The correlations between trust and voluntary activity are generally reduced close to statistical insignificance.” Instead, long experience in “education, work, and family” increases trust.”21 But in a recent survey article on social trust and associations, Paxton and Robert Ressler note that “studies tend to find an effect between association membership and trust” and that while “trust does not influence association memberships,” it turns out that “association memberships do influence trust.”22 Yet even they caution that “empirical evidence on the relationship between associations and trust produces somewhat mixed results,” as some models “remove the effect completely.”23 Again, there are also concerns that some voluntary associations “are typically homogeneous social groups that usually engage with other groups like themselves,” which will only generate particularized trust rather than social trust.24 The evidence that association membership helps build political trust is similar.25 Some have hoped that associations would increase political trust by making democracy function better, by helping societies produce services that states cannot, and even by helping citizens articulate and press for positive political change, a process that could teach association members about self-​respect and democratic organizing. But “several studies have largely failed to find significant effects of associational activity on political trust.”26 So, there is some evidence that associations have a positive effect on social trust and political trust.27 However, it is rather more mixed than I’d expected. I do think the psychological story I’ve told about contact increasing trust is plausible and seems to explain trust in other social contexts, such as the market or interacting with government officials, which can explain the cases where the relationship between trust and association participation and membership is positive. But looking at the studies as a whole, the effect appears to be rather small, or masked by other factors. We can conclude, then, that protecting a robust right of freedom of association should help create at least some trust in the real world, but not to the degree that one might hope. 20 Bergh, Jordahl, and Öhrvall 2019, 4. 21 Newton, Stolle, and Zmerli 2017, p. 43. 22 Paxton and Ressler 2018, p. 161. 23 Ibid., p. 162. 24 Uslaner 2018, pp. 3–​14. 25 Liu and Stolle 2017. 26 Ibid., p. 341. 27 My view does appear to imply that freedom of association is weaker for bonding associations than for bridging ones. I think that a broad right of freedom of association is publicly justified and that states will have difficulty separating bonding and bridging effects in practice, so the trust literature supports my general conclusion but may better support a more nuanced view. I thank Eduardo Martinez for this point.

94  Trust in a Polarized Age I’m also not aware of any evidence that restricting association increases social or political trust, which would be a real cause for concern, since a publicly justified right of association, when exercised, would then reduce social and/​or political trust. Fortunately, again, the data suggest that association either has no effect or a positive effect on social and political trust,28 which means that respecting a publicly justified right of freedom of association can sustain real trust for the right reasons and interrupt the distrust-​divergence process.29

2.  Moral Rules and Standing* I now want to draw our attention to a special kind of moral rule that governs subpopulations. Rules of this kind specify that nonmembers of those subpopulations lack the standing to interfere with, or even criticize, certain actions of those subpopulations. These rules are the elements of associational life and play a central role in publicly justifying a primary right of association and its scope. I draw attention to this kind of rule because some theorists of social morality think that violations of moral rules are everyone’s business. Everyone has “standing to insist on performance and standing to hold the violator responsible for what she has done.”30 As Kurt Baier argues: “Whether a person conforms to the mores and laws of the group is not entirely his own business.”31 Yet there are many moral rules that not everyone has standing to enforce. In the Catholic Church, for instance, only some people have standing to insist that a member confess her sins or obey the directives of the church hierarchy.32 If an atheist insists that her Catholic friend confess, the atheist lacks standing to do so even if her friend acknowledges that she should go to confession. In this case, the Catholic friend is liable to think that the atheist, by making such a demand without being a Catholic, has violated a moral rule of minding her own business. To recognize the fact of restricted standing, we do not have to deny that all moral rules are subject to the scrutiny of public justification. There may be a 28 Wilson 2018, p. 283, notes data suggesting that association membership correlates with giving more in the trust game. Association members tend to be higher trusters. 29 It is possible that states could locate trust-​building associations and subsidize them, while suppressing or taxing trust-​undermining associations. You might get more trust with a more limited right of association. But I am not arguing that we can create trust for the right reasons with practices that maximize trust, just that the institutions that justify trust promote or do not undermine trust. 30 Gaus 2011, p. 224. 31 Baier 1958, pp. xviii–​xix. Here a group can be understood to range from local groups like families or associations to much broader social groups like political communities. 32 I use the Catholic Church as an example not because my aim is to defend freedom of religion in general, but because the Catholic Church serves as a good illustration of a civic organization, and defending the autonomy of civic organizations, both religious and secular, is one of my aims.

Civil Society and Freedom of Association  95 publicly justified moral rule that deprives nonmembers of standing to influence or direct the decisions and practices of a group association. Consequently, we feel appropriate resentment and indignation at the nonmember who sticks her nose in our group’s business. In contrast, within the group, members of the organization are governed by specific moral rules that function much the way generic moral rules function for members of a large public, establishing moral relations. To put it another way, all moral rules must be either publicly justified to the public or subpublicly justified to a subpublic, such as members of an association. So, some moral rules have more limited publics than others.33 Moral rules can be relative to all sorts of groups, from generic groups like races and genders to self-​consciously organized organizations like the Church of Jesus Christ of Latter-​Day Saints or the AFL-​CIO. In this chapter, I am focused specifically on associations, groups that have an expressly recognized set of goals and that are structured by rules for admitting, removing, and governing members. Once we grasp the nature of what I shall call moral associations, we can see why those who care about public justification must respect the integrity of those associations by prohibiting the law from interfering in their internal affairs, save in cases where unjustified associational rules can be clearly and predictably improved by legal and political institutions. But we must also recognize that in some cases, the rules that morally prohibit interference may be defeated by individuals who are denied membership. For instance, if many associations exclude oppressed racial minorities from membership, then the members of the racial minority may have good reason reject the rule excluding them from interfering with those organizations, in which case they have standing to criticize their exclusion and perhaps to seek legal or political remedy.

3. Moral Associations What I will call a moral association is characterized by a local unity of moral rules, and sometimes legal rules,34 with two features. First, as Lon Fuller has said of associations, a moral association’s moral rules are organized to promote a common end, a commitment shared by its members.35 Moral associations also include a legal principle, which “refers to the situation where an association is held 33 In Vallier 2018b, I identify a kind of moral rule, a jurisdictional rule, that specifies who has standing to interfere in the cases of the violation of other moral rules. They explain when people are immune to external judgment. 34 I use the term “legal rules” interchangeably with “law.” Both terms sometimes refer to the private laws of particular associations, which are what I am referring to here. 35 In this I follow Fuller’s first principle of associations (1969, p. 6).

96  Trust in a Polarized Age together and enabled to function by formal rules of duty and entitlement” and these rules are moral in character. Many moral associations have express, formal rules of operation; others are simpler, resembling families, whose moral rules are often tacit and unarticulated. But all moral associations have mechanisms for their own preservation; and these mechanisms typically involve the articulation and enforcement of internal rules. Fuller is far from alone in his understanding of associations. The great British pluralist G. D. H. Cole defines associations as any group of persons pursuing a common purpose or aggregation of purposes by a course of cooperative action extending beyond a single act, and, for this purpose, agreeing together upon certain methods and procedures and laying down, in however rudimentary a form, rules for common action.36

So, like Fuller, Cole finds in associations two principles of organization—​ shared commitment and a legal principle. I will understand moral associations similarly. Fuller contends that while all associations have both shared commitment and a legal principle, the two properties “stand in a relation of polarity—​they fight and reinforce each other at the same time.”37 Every association has legal principles that facilitate its shared commitments, but in some cases its legal procedures may frustrate its shared commitment. Similarly, the shared commitments of an association might undermine the proper functioning of its internal procedures, just as it might sustain those procedures. Fuller argues that as associations grow larger, the legal principle tends to dominate, and as associations shrink, shared commitment tends to dominate. Controlling the moral rules of an association is usually the exclusive right of group members or some subset of group members. If a moral association violates moral rules to which all members of the public are subject, then nonmembers will have standing to criticize association members, but otherwise not. The business of controlling the moral rules of an association is again given exclusively to its members. Associations thereby form jurisdictions that make their own decisions. Decision-​making in a society with many associations is thus decentralized. Associations have nonmoral rules, such as those governing etiquette, decorum, aesthetics, and the like. But an association’s moral rules mark it as an association, since moral rules specify the group’s normative aims, its membership



36 37

Cole 1920, p. 37. Fuller 1969, p. 8.

Civil Society and Freedom of Association  97 conditions, and decision procedures that, if violated, generate the reactive attitudes of resentment and indignation in its members.

4. Civic Associations One important type of moral association is the civic association, which goes beyond familial bonds or the bonds of friendship. Civic associations take many forms, including religious organizations, sports teams, universities and colleges, charitable organizations and service clubs, hospitals, media organizations, and neighborhood associations. “Any kind of formalized, non-​governmental, human interaction”38 not directed at making profits is a civic association. In civic associations, the legal principle is fairly well developed, especially in contrast to families. Civic associations typically have explicit shared commitments or aims. Christian churches aim to produce and improve Christian behavior, to confer the forgiveness of sins, and to offer a path to eternal life. A university usually aspires to educate students for professions and, in some cases, for citizenship; it often has other goals as well. Importantly, membership in civic associations is typically voluntary; even associations that assign membership by birth typically allow adult members to exit the group. This does not mean that members can always choose the rules of the association, though they sometimes can. Civic associations are distinguished from commercial organizations because they are typically nonprofits. While civic associations seek economic resources and provide economic benefits to their members, they seldom understand themselves primarily as profit seekers. In fact, civic associations often strongly prohibit all profit-​seeking activity in their name or by their officials. Similarly, simony laws, the prohibition on the sale of ecclesiastical offices and privileges, are recognized and enforced by basically all religious institutions. Standing to criticize the inner workings of these organizations is typically restricted to members. For example, not everyone has a stake in the internal behavior of the Catholic Church, so non-​Catholics typically lack standing to criticize internal church decisions. We do not generally think it is the business of society to determine whether the Catholic Church should ordain female priests or allow divorce in more circumstances than it does now.39 Civic associations are now thought of as institutions that constitute “civil society” or an arena “within which voluntary associative relations are dominant.”40

38

Tamir 1998, p. 216. Though some disagree. See Cordelli 2017 and Shorten 2018. 40 Warren 2001, p. 57. 39

98  Trust in a Polarized Age A  literature in political theory focuses on characterizing the relationship between civil society and the state, and civil society has been variously described as “apart from” the state, “against” the state, “in support of ” the state, and “beyond” the state.41 The idea of civil society arose gradually, starting in the seventeenth century, as a contrast to both the nation-​state and the growing commercial order.42 For most of human history, social institutions could not be divided into these neat categories because the functions of government were spread across these institutions. Churches had coercively enforced legal systems; the medieval Catholic Church was no mere religious organization. The fact that civil society is a relatively new social domain separate from the state might be taken to imply that civic associations cannot exist without a formal public legal system or legislative body and could not exist based on social norms alone. But this is not so. People still form churches, schools, hospitals, and professional organizations in the absence of public legal systems and legislators. These institutions are less stable and effective in such environments, but they form under many conditions, and so are fundamental to social order in a way that modern nation-​states are not.43 Critically, many civic associations did exist prior to the formation of modern nation-​states, for example in the medieval period in Europe. Social and religious organizations have existed for thousands of years. Given that civic associations can exist in the absence of many forms of law and politics, and have long ordered social life, they play a foundational role in establishing a stable social order. If civic associations can adequately perform various important social functions in ways that can be justified to their members, then on my account of public justification, the state should not interfere unless the organization imposes some harm or restriction on nonmembers based on an unjustified/​defeated rule.44 More precisely, if civic associations exercise their key social functions in ways that can be publicly justified to their members, that is, there is an adequate (sub)public justification to their members, then the state is not publicly justified in interfering. This means that associations can rightly constrain the reach of state power insofar as their constituent moral rules are subpublicly justified to their members. It is for this reason that the state has at best a limited role in determining the curriculum of universities, but does have a bigger role in regulating how universities handle cases of sexual assault. In the former, the state has no particular 41 Chambers and Kopstein 2006, p. 364. 42 Levy 2015, p. 19. 43 MPBW, p. 132. 44 Unless imposing these restrictions is the only way or far and away the best way for the state to ensure that those functions are performed for everyone.

Civil Society and Freedom of Association  99 expertise over the university, and the harms of a poor curriculum are modest. But in handling cases of sexual assault, universities are not as well equipped as states and have stronger incentives to resolve cases in ways that protect their interests. Further, the costs of getting incorrect judgments are enormous, for victims or the falsely accused. So, the latitude afforded to civic associations is by no means absolute. When associations cannot perform a critical social function to an adequate degree, a public legal system can permissibly intervene in some respects if it clearly can execute that function.45 In this way, we can establish the familiar liberal claim that civil society forms a bulwark against state power. Civil society is comprised of civic associations, and those associations play a role in structuring an order of publicly justified moral rules, that is, an order of public reason.46

5. Commercial Organizations We now turn to for-​profit institutions like the firm, and associations that aim to ensure a certain distribution of profit, like unions and professional bodies. Their common goal is typically to improve the economic and political resources of their members, both bosses and employees. Seana Shiffrin identifies two main features that distinguish commercial associations from other kinds of association: commercial organizations (i) are centrally concerned with “access to material resources and mechanisms of power,” and (ii) “have a fairly focused singular purpose whose pursuit is largely guided by [the] aim of profitable operation.”47 James Nelson argues that within what he calls “organizations” people “do not integrate collective goals and values with the core aspects of their identities, and they perform their tasks largely motivated by explicit or implicit threats of coercion or offers of inducements”;48 commercial associations resemble Nelson’s organizations, while civic associations resemble Nelson’s description of “collective communities” which are “constructed out of a cluster of identification relationships” where “individual members view their affiliation with the collective as a central aspect of their identities.” So commercial organizations are also 45 It is interesting to reflect on the extent to which the constituent moral rules of churches and universities are publicly justified to most of their members. Presumably liberal Catholics and conservative college students may doubt that their choice to join these institutions is sufficient to legitimize the rules applied to them. There is also a related question about whether all members or quasi members of an institution have equal standing to object to the rules of their organization: church parishioners and university students are usually seen as having rightly less influence over how an organization is directed than church leaders and university faculty and administrators. I leave these questions to the reader. And I thank Christie Hartley and Eduardo Martinez for raising related points. 46 Gaus 2011. 47 Shiffrin 2005, p. 877. 48 Nelson 2013, pp. 1581–​2. Also see Nelson 2015.

100  Trust in a Polarized Age distinguished by a lack of collective community among their members. This is not always true: a small-​business owner might deeply identify with a cause her business promotes, and members of a union could value their membership as a way of living out their political ideology. But the paradigmatic commercial organizations are importantly different from civic associations in these respects. Nelson argues that businesses are fairly unique in using threats and material incentives to achieve their ends, which makes their activities subject to more rigorous scrutiny.49 Since bosses typically lead and control commercial organizations, and they recognize that workers depend upon jobs to survive, bosses can exercise more control over their members than leaders of civic associations. Second, civic associations can also have a “conscience” in ways that commercial organizations frequently do not.50 As Nelson says, “For-​profit businesses are not generally bound up with deep ties of identity and attachment,” whereas nonprofit organizations tend to be “more hospitable to individual identification,” such that associational claims should be stronger in the latter case than the former.51 For these reasons, a publicly justified polity, one whose constituent moral and legal rules are justified for all, provides fewer protections for commercial organizations than civic associations. Civic associations are more essential in both creating and maintaining personal identities, and so the reasons persons have to insist on their integrity are greater. Further, the potential for threats and control is usually greater in commercial associations. There are exceptions. Closely held firms and many small businesses have practices and aims that make them more like civic associations.52 These businesses, therefore, are entitled to more protection because their operation is more tightly tied to the conscience and life projects of their owners and operators. The same is true of many trade unions. So, the mere fact that a group pursues profit is not enough to undermine its protections against state predation and control. Protection is especially weak for publicly traded firms, since there is no real case that such organizations have anything like a conscience, even if members exhibit some sense of social responsibility. This is because the leadership, operation, and ownership of the firm are open to the public for a price. Ownership can easily change hands; different people may own the firm tomorrow than owned it today.53

49 Nelson 2013, p. 1581. 50 Ibid., p. 1583. 51 Nelson 2015, p. 511. It is not clear how true this remains, given that modern corporate motivational programs aim to change this. I thank Mark Laufgraben for this point. 52 Like John Tomasi’s example of Amy’s-​Pup-​in-​the-​Tub. Tomasi 2012, p. 66. 53 In ­chapter 6, I will review the ways in which states can limit the power of publicly traded firms to protect workers.

Civil Society and Freedom of Association  101 Unions and professional organizations are harder cases.54 They are not exactly for-​profit organizations, even though they attempt to increase the income of their members and engage in quality control. We are typically less friendly to interventions into these organizations, which I suspect is because these groups are seen as engaging in primarily defensive activities or have nonfinancial motives in addition to financial ones. Unions are seen as protecting their members from the greater economic power of employers, whereas professional organizations like the American Medical Association are seen as motivated by concern for the people the profession they represent serves rather than their members’ income.55 This is not always true, however, and to the extent it is false, these institutions may be liable to legal interference. Focusing on unions, we can publicly justify protections for workers to unionize in whatever industry they like, in whatever way they like, so long as the contractual relations established between the union and its members’ employer do not harm third parties. It is possible that some firms will prod workers into signing contracts that relinquish their right to unionize. But state enforcement of these contracts might not be publicly justified, especially if the contracts are signed under pressure. That said, unions have no right to use state power or their own social power to destroy competitor associations. If a group of doctors wants to start an alternative to the American Medical Association, their right to freedom of association protects them from legal retaliation by the state on behalf of the AMA. And unions have no right to coercive protections from competition by firms within the same nation. Here I ignore political organizations—​groups dedicated to changing the behavior of government and the use of state coercion. But there is much to be said on their behalf.

6.  The Public Justification of Freedom of Association as a Primary Right* I will now explain why freedom of association is a publicly justified primary right.56 Recall that a primary right is one that all would want regardless of their rational plan of life or conception of justice and that persons are willing to extend

54 For some plausible arguments in favor of liberal protections for trade unions, see White 1998. 55 This is probably not a plausible way of understanding the AMA, historically speaking. See Blevins 1995. 56 Here I will not rely on my claims about what, in MPBW, I called “the legal state of nature” and its role in public justification as outlined in ­chapter 4 of MPBW. In general, while I think the role of the legal state of nature is important in the full case for freedom of association, it is not required to vindicate my claim that freedom of association creates trust for the right reasons.

102  Trust in a Polarized Age to others on reciprocal terms. Freedom of association is obviously crucial to pursuing one’s conception of the good, perhaps unless that conception is wholly atomistic. Also, though perhaps less obviously, freedom of association allows a person to pursue her conception of justice by organizing with others to engage in social and political activism or to realize relations of justice within one’s associations.57 It is also important that associations are derived from the actions of persons who already have publicly justified rights of agency, since associations are frequently ways in which people develop, protect, and expand their agency. Legal restrictions on these institutions will require powerful justifications. While the law is required to protect freedom of association, render it determinate, and resolve disputes about its extent, laws that interfere with these associations will often be illegitimate. To show that freedom of association is publicly justified, we must argue that moderately idealized members of the public behind a thin veil of ignorance would select an authoritative right of freedom of association as one of the principles that are to govern themselves and others on equal terms and that secure a range of resources like basic rights and liberties, income, wealth, capital, and the like. Given this, and my arguments in MPBW, we can justify a primary right of freedom of association as follows.58 First, moderately idealized agents know all the relevant factors about themselves and others save their relative social power and status. This thin veil of ignorance puts the contractors into a somewhat risk-​averse mood, making them keener on seeking protections for their rights than on seeking hegemony over others, given that they prefer to have protections if they turn out to be in the minority. Given the centrality of freedom of association for agency, and for following most conceptions of justice and the good, moderately idealized members of the public would choose to extend a primary right of freedom of association to others in order to secure the right for themselves, and so to protect their agency rights and the missions of the organizations they wish to join. Because everyone must have the same rights, the only eligible rights schemes are those that protect extensive rights of association for all those that protect less extensive rights of association. To restrict the associational rights of others, members of the public would have to be prepared to accept social and political interference in her own associations under similar circumstances. Thus, since each person would prefer extensive associational rights for herself, and would be more concerned to

57 In this way, freedom of association helps create a polycentric order where different conceptions of justice can govern different moral associations, and this is a plausible institutional consequence of the recognition of justice pluralism. For a case for polycentrism based on justice pluralism within a moral, traditional Rawlsian framework, see Kogelmann 2017. 58 MPBW, pp. 206–​7.

Civil Society and Freedom of Association  103 protect her own interests than to limit the liberties and resources of others, she would choose equal extensive rights of association. Behind the thin veil of ignorance, nonliberals, especially communitarians, would choose extensive rights of association. In fact, they would be especially committed to these rights, given that nonliberals tend to distinguish themselves from liberals by arguing that human flourishing and happiness require a robust communal life. Because the veil of ignorance would prevent communitarian members of the public from choosing rights schemes that give them privileged associational freedom, they would rank extensive associational freedom over limited freedom, lest their own highly valued associations be interfered with, degraded, or destroyed by a state with too much power. For communitarian nonliberals behind the veil would not know whether they and their values have enough social power and status to protect their associations if a rights principle that protects everyone’s is not chosen.59 It is critical for my purposes that a right to freedom of association restricts state power, in part because my account of public justification recognizes that most of social order comes from the moral order. Thus, many proposed government restrictions on associations are unnecessary in order for associations to produce positive outcomes. Indeed, in many cases moral associations may provide certain services, like poverty relief, more effectively than the state, in which case the less coercive, more pluralistic associations will be a superior mode of service delivery than the state.60 The state lacks any priority over the liberties of associations. If this argument is correct, activities that involve the codification, protection, appropriate reform, and expansion of primary associational rights should serve as touchstones of trustworthiness. Since all can see good moral reason to recognize and respect rights of association, activities that recognize and respect rights of association, as well as the exercise of those rights, should serve as evidence for the trustworthiness of citizens to multiple, diverse evaluative perspectives. So the practice of rights protection is a sign of trustworthiness. However, we will see that the exercise of this right can undermine trust in some cases, for example in building isolated associations that encourage people to distrust outsiders. We may, therefore, have stronger reason to protect associations that do not isolate their members and that do encourage them to interact with and trust people they otherwise might not get to know, on the grounds that such associations create more trust than isolated associations. However, while some associations may

59 There is an important question about whether nonliberals will continue to accept an equal right of freedom of association if the veil is lifted and they learn that they are the social and political hegemon. I address this question in appendix A. 60 I defend this claim further in Chapter 5.

104  Trust in a Polarized Age not create real trust, and indeed may undermine it, freedom of association is publicly justified as a general principle. That means it applies to all associations, both civic and commercial. Thus, if we were to suppress or regulate trust-​undermining associations in order to ensure that associations in general promoted trust, we would violate the strictures of public justification, and so undermine the creation of real trust for the right reasons.61 I now explore the relative priority of associations and the democratic state.

7.  The Priority of Democracy? An important feature of my account of freedom of association is that associations are part of the moral order; they can often sustain themselves without much help from the state, and they even often perform functions that the state performs. Given the centrality of the moral order in my theory, I agree with Gaus that law is not the preeminent response to moral disagreement.62 Legal and political procedures are not the only way in which we can solve problems, since we can appeal to the moral order to create, sustain, and alter moral rules. Law has various advantages, but if the moral rules comprising a society are publicly justified, then legal and political procedures should be limited, if for no other reason than that coercion might be used to impose unjustified rules. If we already live under justified moral and private legal rules, “We should be most reluctant to modify [them] through the political process.”63 Thus, if our moral rules are publicly justified and adequately perform their social functions, coercion is hard to publicly justify. Since there are multiple social orders that can protect and sustain associations, the state does not have priority to control or direct the nature or course of the operation of moral associations. If these institutions can perform key social functions without help or without much help from the political order, then the state should respect their autonomy. And if they can better perform functions than the state, then the state should defer to them because they are generally less coercive. If civic associations can better organize charitable activities for impoverished members of society, the state should not redistribute wealth in ways that would crowd out their efforts. If private firms succeed in generally producing adequate levels of food, the government should generally avoid interfering in ways that would undermine that capacity. And since moral associations, such as universities and churches, often have formal legal systems, if these

61

I thank Brian Kogelmann for raising this objection. Gaus 2011, p. 456. I discuss this in MPBW, ­chapter 4. 63 Gaus 2011, p. 460. 62

Civil Society and Freedom of Association  105 associations can apply and maintain private legal rules with adequate efficiency and justice, then the state’s legal system should generally avoid displacing those associational legal systems.64 My view somewhat resembles Paul Hirst’s associationalism, which holds that “the organization of social affairs should as far as possible be transferred from the state to voluntary and democratically self-​governing associations.”65 In contrast to Hirst, however, I do not insist that the associations be democratic, just subpublicly justified to their members. I  also resist Hirst’s aim of “recruiting associations as partners in governance,” because, as Victor Muñiz-​Fraticelli has argued, we can easily miss the “distinct and perhaps oppositional nature” of associations.66 The strong rights of moral associations must be stressed against philosophers who defend more extensive states on the ground that nation-​states have priority over other institutions, in that nation-​states have authority to alter those institutions, and not the other way around. For instance, Jack Knight and James Johnson have defended the priority of democracy over the market as essential for performing “crucial second-​order tasks involved in the ongoing process of selecting, implementing, and maintaining effective institutional arrangements,” such as institutions that coordinate ongoing interactions.67 Market order cannot have priority over democracy, in their view, because only democracies have the right amount of “reflexivity” in determining “which institutions arrangements to rely on across different domains.”68 In particular, “Markets do not facilitate the types of communication necessary for the tasks of institutional assessment and monitoring.”69 Only democracy, then, can set the boundaries and forms of institutions. Cécile Laborde has recently offered a similar argument that democratic states are sovereign over civil associations when “citizens reasonably disagree about where jurisdictional boundaries are drawn” between civil associations and the rest of social life.70 Both Laborde and Knight and Johnson thereby embrace the priority of the democratic state. These theorists make an important mistake that is often hard to see. Indeed, moral associations will disagree about their jurisdictional boundaries, and we will sometimes want democratic institutions to settle those disputes. And in 64 The priority of the association’s legal system can apply to disputes between members that might otherwise be handled by the state, at least when the matter is essentially internal and/​or the parties agree to use the association’s legal system. 65 Hirst 1997, p. 32. But I suspect that in many cases, the only decision rules that can be justified to association members are democratic in nature. 66 Muñiz-​Fraticelli 2014, p. 92. 67 Knight and Johnson 2011, p. 19. 68 Ibid., p. 23. 69 Ibid., p. 170. 70 Laborde 2017, p. 163. Also see p. 168.

106  Trust in a Polarized Age some cases, democratic intervention will be superior to allowing civil associations to decide these questions themselves. But this does not mean that the democratic nation-​state has authority over moral associations. Rather, democratic states are only licensed to restrict associational freedom, both in civil society and, in some cases, in the market, when the rules of associations are not publicly justified to their members, or the jurisdictional rules restricting standing for others to criticize the organization for exclusion are defeated or reasonably rejected.71 All Knight, Johnson, and Laborde establish is that we sometimes need democracy to intervene in associational life, but if my approach to public justification is correct, this does not establish the general priority of the democratic state. One reason that democratic intervention will be limited is that we often underestimate the resourcefulness of persons outside state law and even outside of any legal process at all. In many cases, we resolve disputes in the workplace or in church through deliberation and discussion or social punishment. We should not leap to embrace state power unless we are sure that state intervention will improve matters rather than make them worse.72 This is a common refrain in both Elinor Ostrom’s political economy of common-​pool resources and James C. Scott’s important work on social organization under anarchy.73 David Beito’s work showing how local fraternal organizations provided for private charity and health insurance in the absence of an extensive welfare state is also helpful.74 I should also note that states cannot always or even usually intervene to change social norms when people have internalized those norms as morally appropriate, and so often the only feasible institutional means of settling the boundaries of associations will have to be left to civil society and other parts of the moral order.75

71 I do not treat the family as an institution where state intervention is off limits. States may interfere in family life when the social norms that constitute family life are not subpublicly justified to its members and we can expect state intervention to shift families to more justified norms, for example by making laws that punish domestic violence. But we cannot push this insight too far in the direction of the state because, while familial social norms must be publicly justified, their need for recognition is not identical with a need for legal evaluation. Martha Nussbaum confuses social and legal recognition in her review of Rawls on the family, stating that while “individuals may call themselves ‘a family’ if they wish,” “they only get to be one, in the sense that is socially recognized, if they satisfy legal tests.” Indeed, societies must evaluate moral rules governing the family, but the law need not, and legislative review may be less necessary still. See Nussbaum 2002. 72 Nonstate actors might also engage in a kind of moral and civic creativity that can improve democratic institutions. 73 Ostrom 1990; Scott 2009. 74 Beito 2000. 75 Barrett and Gaus 2018.

Civil Society and Freedom of Association  107

8.  Private Tyranny and Balkanization Now I will address further objections to my defense of freedom of association.76 Private Tyranny: without more limits on freedom of association, associations will unjustifiably limit the liberties of their members.

We can elaborate the private tyranny objection as holding that associations should be structured much like liberal democratic government in order to avoid the inequality and oppression found within hierarchical organizations. John Stuart Mill, despite writing over a century and a half ago, provides an excellent characterization of this concern on the grounds that state tyranny can in some respects be much less worrisome than local associations. Mill: “Obedience to a distant monarch is liberty itself compared with the dominion of the lord of the neighboring castle.”77 As Levy notes, Mill was especially concerned about “the enforcement of . . . soul-​enslaving, individuality-​stunting norms through the oppressive combination of public opinion and local personalized power.”78 Today, the tendency to worry about private tyranny manifests itself as an insistence that we democratize associations to prevent private tyranny. Rosenblum calls this analogy between governmental and private tyranny the “congruence approach” to civil society.79 Chandran Kukathas articulates the private tyranny objection as the claim that “freedom of association, underpinned by freedom of exit, does not make for a free society because, in itself, it says nothing about the cost of exit.”80 Kukathas has argued, in response, that freedom of exit from oppressive associations is sufficient to counteract their tyrannical effects.81 One might answer that a citizen is not free to exit unless she can exit with relatively limited costs. In reply, Kukathas insists that high costs do not undermine the fact that persons freely consent to the organization; people are still free even if they can barely afford to exit the relevant, potentially oppressive, association.82 Kukathas’s response is inadequate

76 People will disagree about the proper boundaries between associations and with the state; some may argue on this basis that we need a sovereign state to resolve those boundary disputes. My response, given in Vallier 2020a, is that we can appeal to a number of different dispute resolution mechanisms, only some of which use state power. 77 Mill 1963–​, vol. 19, p. 416. 78 Levy 2015, p. 218. I do not think that Mill ignored the value of local associations, just that he clearly articulated the “rationalist” worry about private tyranny. Levy defends this position later in the chapter. 79 Rosenblum 1998, p. 46. 80 Kukathas 2007, p. 107. 81 Ibid., pp. 93–​103. 82 Ibid., p. 109.

108  Trust in a Polarized Age because we can easily imagine an unfree society of free associations, given that strong, potentially corrupt hierarchies can direct each person’s life. Fortunately, my account of public reason has more resources to respond to the private tyranny objection. Moral associations only have authority over their members when the moral rules imposed upon members are subpublicly justified. Thus, some uses of associational authority are illicit because they are based on rules that are not justified to their members. Consequently, there is moral ground for the member to disobey the oppressive rule, and if powerful members of the association punish those who justifiably disobey, there may be a role for the state to interfere with the association to protect the member. If a religious authority abuses a parishioner’s child and tries to discipline the parishioner for speaking out, the religious authority has violated the rights of the parishioner, not to mention the child. If the parishioner is powerless to seek justice for her child, then the state must interfere to protect the basic rights of parishioner and child. My account of public justification nonetheless faces hard cases. Civic associations and commercial organizations are often organized as strict hierarchies, and liberals have often argued that these hierarchies harm subordinates.83 Perhaps my approach has a similar problem. However, if the information demonstrating that these hierarchies are harmful is available to appropriately idealized members of the public, then the information can count toward a decisive objection to those social arrangements. Further, people voluntarily join hierarchical organizations all the time. And in many cases, voluntary submission is a good proxy for public justification: a choice provides powerful evidence that the agreed-​upon arrangement is publicly justified for the person that makes the agreement. Voluntary agreement can only justify so much, however. Government has no business enforcing contracts, such as slave contracts, that secure extremely harmful relationships. It plainly should not support or protect associations that attempt to control persons who have freely exited those associations. And government should be concerned that some putatively voluntary agreements are not voluntary due to unequal bargaining power between the parties to the association. In this way, I acknowledge that private businesses can operate as harmful, private tyrannies that dominate their employees, and my account of public justification requires that, in the absence of a moral solution, law and policy be used to eliminate or constrain those forms of domination and control. I discuss these issues at greater length in c­ hapter 6.84

83 For a detailed account of the costs and benefits of hierarchical firm structure, see Singer 2019. 84 And I discuss the way in which the law can enter into the moral order, of which civic associations are a part, at more length in MPBW, ­chapter 4.

Civil Society and Freedom of Association  109 I now turn to the Balkanization objection, which holds that extensive rights of freedom of association will be used to undermine social and political institutions: Balkanization: without more limits on freedom of association, associations will undermine social trust and political stability by producing inwardly focused citizens.

Rosenblum advances this objection when she argues that “the critical dilemma for liberal democracy in the United States today is not exclusion from restricted membership groups but isolation.”85 If we give associations too many rights and romanticize their role in preserving a free society against the state, then “freedom of association threatens to balkanize public life.”86 There are legitimate worries about isolated associations, especially groups that advocate hatred and bigotry.87 And indeed some institutions isolate themselves and their members, thereby undermining social trust. Pamela Paxton says they do: “Having more isolated associations decreases trust.”88 The question the public reason liberal must answer is whether the moral order is threatened by isolated associations. In many cases, the groups are sufficiently isolated that they will have little effect on the moral order, in which case their rights to association should be left alone. But in cases where associations garner enough power and influence to undermine the effective functioning of political and economic institutions, we must appeal to law and legislation.89 We should first attempt to restrain the influence and power of isolated associations through ostracism and criticism. But if those mechanisms are unsuccessful, then we may appeal to law and legislation to restrain these organizations, if only through the least coercive restraints available. I cannot say how often legal intervention will be justified, but it is fair to say that, given that legal options are on the table, we need not worry about excessive Balkanization created by isolated associations. When isolated associations undermine publicly recognized and justified moral rules that apply to all, there are plenty of social mechanisms that can appropriately prevent them from doing so, though we must be careful to recognize that government intervention comes with its own costs, since government too can abuse its power.

85 Rosenblum 1998, p. 102. 86 Ibid., p. 46. 87 Chambers and Kopstein 2001, p. 839. 88 Paxton 2007, p. 47. 89 Importantly, associations may have big effects on local communities but limited effects at the federal level, in which case we might think disputes about the boundaries of associational freedom should be determined at more local levels. This gets into the question of publicly justified federalism, which I sadly was unable to address in this book.

110  Trust in a Polarized Age I acknowledge that associations, including bridging associations, very often may pose a threat to democratic order. To push a particularly disturbing example, one study finds that membership in the Nazi Party in Weimar Germany was greatly facilitated by moral associations, as association members joined the Nazi Party in proportionately higher numbers than citizens who were not active association members did.90 However, it also appears that the association effect operated primarily in German states with low political stability, which suggests that associations can avoid these negative impacts when there is stable, significant state capacity.91

9.  Feminism and the Discrimination Objection I turn now to concerns that my account of freedom of association allows for too much private discrimination, including against women. I formulate the objection as follows: Discrimination:  without more limits on freedom of association, associations will discriminate against marginalized nonmembers.

The discrimination objection holds that freedom of association unjustifiably allows groups to exclude members on morally unacceptable bases, such as race, gender, religion, and the like, or to decline to provide services to groups with particular characteristics. And doesn’t my defense of freedom of association commit me to endorsing a right to discriminate? Before I answer, note that discrimination frequently involves a preparedness either to use coercion or to insist that the police use coercion on the discriminator’s behalf. Even where the law does not itself discriminate, the use of legal coercion to uphold private discrimination is not always publicly justified. If a social norm disadvantages a particular group and the practice of commercial organizations reinforces and is reinforced by that norm, then those organizations are contributing to a moral rule that cannot be publicly justified to the oppressed group. If Reba lives in a racist society, then if she wishes to prevent hiring blacks to work in her barbershop, her right of freedom of association does not protect her. But if Reba wanted to avoid hiring hipsters to work in 90 Satyanath, Voigtländer, and Voth 2017. 91 Here, however, there is a serious worry that societies with strong state capacity will risk incentivizing associations to become special interest groups and seek rents at the expense of society. I thank Christian Bjørnskov for raising this point. I will address rent-​seeking restraints in ­chapter 7, but I will have to wait to explore the connection between freedom of association and democratic constitutionalism for a future project.

Civil Society and Freedom of Association  111 her barbershop, then considering the lack of bigotry against hipsters, Reba has the right to discriminate. A business might also be entitled to discriminate, even against marginalized groups, if its services are nonessential and the marginalized groups have lots of alternative options. Many feminists may still think that I  have provided too few resources for women to defend themselves against discrimination and control by men in a variety of social settings, including within associations where they are members. For this reason, I would like to address Clare Chambers’s challenge to political-​ liberal views on the ground that they make a fetish of choice and that feminist values suffer if what people freely choose is treated as a “normative transformer,” making relations of domination seem morally permissible.92 Chambers stresses throughout her work that social norms can be oppressive and so a feminist state should use the law to end those norms, even if women subject to those norms freely accept them93: Formal freedom of exit is insufficient to excuse a culture or religious group’s imposition of unequal norms. Just as the state properly intervenes in discriminatory employment practices, so too it ought to intervene in discriminatory cultural or religious norms, even where those norms are not enshrined in state law, and even where members are “free” to leave groups.94

Unless the state interferes in this way, she argues, “The autonomy and fair equality of opportunity that liberals prize cannot be realized.”95 The reason for this is that preferences are often “socially formed”: the social norms create the preferences, and choices based on those preferences are therefore not autonomous.96 Chambers envisions an “equality tribunal” that would review cases of social unfreedom and domination in social norms and try to make sense of when intervention is appropriate. I reply, in agreement with recent work by Lori Watson and Christie Hartley, that public reason liberals can accommodate many of Chambers’s complaints.97 Public reason will address complaints about a loss of autonomy or the reduction of equality of opportunity and determine which laws and policies are publicly justified. I think this is even true on my heterodox convergence approach to public justification elaborated in ­chapter 1 since, as I have noted, there is a place for law to interfere even with voluntarily chosen social norms if some persons

92 93

94 95 96 97

Chambers 2008, p. 21. Ibid., pp. 78–​9. Ibid., p. 140. Ibid., p. 157. Ibid., p. 171. Watson and Hartley 2018, chap. 8.

112  Trust in a Polarized Age have sufficient reason to reject the authority of these norms.98 Many women may comply simply from social pressure, not because they accept and have internalized the norm based on their own reasoning. I grant my view does have to accept the authority of social norms that are publicly justified and internalized by female members of organizations, however, but I am not convinced many norms that we would otherwise consider oppressive can be justified to moderately idealized members of the public. We can now anticipate several further objections. First, Chambers thus far lacks an account of social norms that shows how social norms, and changes in social norms, can liberate women when conjoined with a requirement of public justification.99 When social norms are publicly justified, they treat all with respect and sustain trust among reflective persons. Thus, religious social norms can permissibly assign leadership positions to men alone if they are subpublicly justified for the women in subordinate positions; in fact, because these norms are subpublicly justified to them, they may even count as an extension of the freedom of these women of faith. However, if the norms are not so justified, and so illicitly restrict the freedom of women, the law may intervene. That said, I think Chambers is too quick to embrace legal solutions to problems raised by sexist moral orders. Legal systems can intervene in the wrong ways and can threaten feminist values if sexist politicians control the state. Insofar as Chambers wants a legal body, like an equality tribunal, to interfere with freedom of association, she puts feminist associations at special risk. Without the freedom afforded to universities, for instance, academic institutions could not have served as incubators for feminist ideas. Yes, the equality tribunal might successfully establish equality within, say, religious organizations, but hierarchical religious organizations would not trust such a body to treat them fairly and would feel politically emboldened to capture the governmental organizations that interfere with them to serve their own sectarian ends, at least in part because the tribunal’s power cannot be justified to them.100 Chambers’s work thus pays too little attention to the possibility of government failure in promoting feminist aims. Second, legal systems often simply cannot stop oppressive norms that are deeply entrenched.101 So, in some cases, legal interference can backfire. 98 Central statements of the convergence approach include Gaus 2011 and Vallier 2014a. 99 Here, as in MPBW, pp. 31–​5, I follow the account of social norms in Bicchieri 2006, though moral rules, as I understand them, are social norms whose normative expectations have a moral character, which she resists. 100 In my view, this is a pretty good account of much Christian involvement in right-​wing politics in the United States. People disagree about which issue led to the creation of the Christian Right (those on the left tend to stress racist and sexist motives, whereas those on the right tend to stress state intervention in schooling and the rise in abortion and divorce), but most agree that the Christian Right formed in response to new legal policy that changed how their churches functioned. 101 Barrett and Gaus 2018.

Civil Society and Freedom of Association  113 Readers may object to my argument as follows. Modern democratic states have tended to protect women from oppression and discrimination in the private sphere, so democratic citizens should want states to protect women from private organizations. We can’t simply throw up our hands after claiming that states sometimes make mistakes. We must instead look at the record of democratic states in protecting women’s rights over the last fifty years. Most feminists are bound to say that things have improved on balance. But notice that, on my view, if it is obvious that the democratic state has tended to protect women from oppression and discrimination, this will be clear to moderately idealized members of the public. In that case, my model of public justification will permit, if not require, that democratic states act to defend women’s interests. As long as members of the public at the right level of idealization can see that the state is better for women, they will have good reason to permit the state to use policy to limit the rights of associations. The nice feature of what I have elsewhere called pluralist contractarianism102 as opposed to more familiar pluralistic views is that the ideal of public justification is central, and the ideal of public justification gives both the theorist and the activist more resources to legitimately pursue protections for women. The discrimination objection therefore does not undermine my defense of freedom of association.103

10.  The Conventionalist Objection Associations require the capacity to own and operate property in order to function, so they require strong property rights protections against the state and others who would interfere with those ownership needs. But it is common in political philosophy to object to prepolitical property rights, even for associations. This is what we might call, following Ben Bryan, the conventionalist challenge, which holds property rights are not prepolitical constraints on the state because they are state creations.104 I formulate the objection as follows: Conventionalism: there are no prepolitical property rights because private property rights have a strong conventional component; they are necessarily the creation of political institutions. Consequently, property claims cannot provide a prepolitical restraint on the state, since they are not prepolitical. 102 Vallier 2020a. Here I  use “contractarianism” not in the narrow sense of a social contract grounded in instrumental rationality, à la David Gauthier, but rather as a catch-​all term for social contract views generally. 103 Some readers will object that it is unclear who decides whether certain conditions for associational justification have been met. I address this objection in the penultimate section of Vallier 2020a. 104 Bryan 2016.

114  Trust in a Polarized Age The conventionalist challenge is usually deployed against the familiar libertarian position that there are natural rights to private property. The most prominent defenders of this objection—​Thomas Nagel, Liam Murphy, Cass Sunstein, Stephen Holmes, Thomas Scanlon, and Philip Pettit—​have used the argument in this way.105 For instance, since the legal system can only be maintained by taxes, such that without taxes there could be no protection or even definition of these property rights, the state cannot possibly be prohibited to tax by property rights, since taxation is required for property rights to meet their most basic preconditions. Obviously, I am not defending natural rights libertarianism in this book. But I nonetheless expect people to reject my view about the rights of commercial organizations on the ground that these commercial institutions are the creation of a government-​run legal system. That is, I expect objectors to say that commercial organizations cannot use their property rights as a bulwark against state power because the organizations can only exist as the creations of a government-​run legal system. There are two simple, serious problems with this objection. First, if it is successful, it undermines all kinds of rights claims against the state, including many of the liberal rights endorsed by liberal egalitarians. For instance, how can we have a right to free speech against the government if the government (as the objection implies) is required to define and protect that right? Or how can we have a right to bodily protection against government if the government is required to define and protect that right? Few liberal egalitarians will bite this bullet. So why are private property rights different? The best reply, in my view, is that property rights (including the property rights of associations) are somewhat different from other liberal rights because political practices that protect highly articulated, extensive property rights in external objects require more deference from others than less controversial liberal rights. That is, property rights place relatively more restrictions on the actions of others than other rights.106 Critically, associations need external property. Families need homes to live in, and churches require land to build houses of worship. But liberal egalitarians don’t argue that the state can willy-​nilly deprive families and churches of their property, because families and churches depend on property rules. And yet the objectors refuse to extend the same protections to commercial organizations. Is this an inconsistency? Not entirely. Commercial organizations do seem to have 105 This is the general thrust of criticisms found in Murphy and Nagel 2002; Sunstein and Holmes 1999; Scanlon 2011; and Pettit 2013. 106 For a defense of this point, see Lomasky 1987, pp. 113–​4. Dan Moller has argued that much property in modern economies often takes the form of property more like the body, such as human capital. See Moller 2019.

Civil Society and Freedom of Association  115 more coercive power and control over their members, and the reasons to protect their autonomy from the state are weaker. However, the disanalogy is not so great that commercial organizations lack good prepolitical property claims. The second problem with the conventionalist challenge is that it depends on ignoring the critical distinctions between moral rules, legal rules, and constitutional rules. It is certainly true that we can only keep property rights in existence by means of socially constructed rules. But taxpayer-​funded, legislative rules are not required—​property rules are often stable moral and legal rules that are in equilibrium due to factors other than the actions of nation-​ states.107 So the conventionalist challenge only has force against property rights rules that would not be in equilibrium save for the actions of a taxpayer-​ funded nation-​state. The conventionalist challenge therefore has some force against large, publicly traded firms, but much less for small commercial and civic associations. But we must not push this point too far, as Bernard Harcourt arguably does in The Illusion of Free Markets.108 Harcourt argues that the idea of a truly free market is a “fiction” because markets require “massive government intervention” to create the property relations from which real markets spring, intervention that is often associated with severe legal penalties.109 Markets and property rights can be structured in “myriad ways,” and all include so much manipulation that “there is no neutral position, there is no one side that favors liberty.”110 Harcourt is certainly correct that property and markets always require a large range of norms to function, and that, in this sense, there is no perfectly free market because norms typically constrain behavior. And he is also correct that some property systems consist in norms enforced by severe penalties. However, Harcourt does not argue that this is a feature of most market property orders, nor that the punitive legal regimes that accompany many markets are necessary for the market to function. Perhaps these property rules would have survived with weaker penalties or based on a more limited range of legal and moral norms. So, while Harcourt is correct on occasion, it is unclear how often he is correct, and he does not do much to show that the problem he cites is a pervasive feature of property rights or the market order. Normless market orders are a fiction, norm-​structured market orders are not.

107 Judicial institutions, for instance, might be funded through voluntary donations and/​or user fees. 108 Harcourt 2011. 109 Ibid., pp. 181, 187. 110 Ibid., p. 242.

116  Trust in a Polarized Age

11.  Pluralist Public Reason Liberals can be distinguished from one another in a number of ways, including by what they regard as the greatest threats to liberty.111 Some liberals think that nonpolitical institutions, like firms and families, are the chief threats to freedom; in contrast, democratic government can liberate persons from these private tyrannies. Other liberals claim that government is the chief threat to liberty; in contrast, nongovernmental institutions like civil associations and families can limit the threat of government overreach. Jacob T. Levy has recently dubbed these two strands of liberalism as “rationalist” and “pluralist” respectively. As he puts it: On one side of this divide lies a liberalism I will call “pluralist”: skeptical of the central state and friendly toward local, customary, voluntary, or intermediate bodies, communities, and associations. On the other we see a liberalism I will call “rationalist”: committed to intellectual progress, universalism, and equality before a unified law, opposed to arbitrary and irrational distinctions and inequalities, and determined to disrupt local tyrannies in religious and ethnic groups, closed associations, families, plantations, and the feudal countryside, and so on.112

It is hard to formulate a liberal political theory that is both rationalist and pluralist. Levy thinks it can’t be done. But the public reason view I developed in this chapter helps answer his objection. My account of association and public justification blurs the contrast between Levy’s rationalist and pluralist liberalisms, yielding a pluralist public reason that has some of the attractive features of both rationalist and pluralist positions. Having placed associations at the heart of the moral order, I can now draw some lessons about how to reconceive the public reason project in light of the centrality of associational life and how to resolve this tension in the liberal tradition. Rationalist liberalism tends to be suspicious of associations because they are potential centers of private tyranny.113 According to Levy, rationalist liberals understand associations as mere appendages to the contrast between the individual and government. Even Locke’s conception of society “functions in his argument to choose a government, and, if needed, to reclaim the authority to do so. It is unitary and political. It is not pluralistic and extra-​political, as we today think of civil society as being.”114 For the social contract liberal, associations are annoying 111 “Pluralist contractarianism” refers to a range of social contract theories that includes public reason views. 112 Levy 2015, p. 2. 113 Levy excludes commercial organizations from his understanding of associations. 114 Levy 2015, p. 19.

Civil Society and Freedom of Association  117 barriers to answering important philosophical questions about the relationship between the individual and government. Pluralist liberalism, on the other hand, sees associations as sources of freedom, for on the pluralist view, associations function as “intermediate” institutions that usefully come between the individual and the state. Levy thinks that pluralist liberals have three general insights: (a) Within a larger, more diverse society, multiple social orders can emerge and survive, using local knowledge to evolve local norms that are locally functional.115 (b) Law can emerge from sources other than government, whether as the internal norms of such groups or as the norms that regulate relations among them. (c) Such orders are normatively attractive: perhaps they are absolutely attractive, because they provide the context for substantive life plans thicker than the formal rules of justice; perhaps they are attractive relative to the [moral] or legal orders enacted by deliberate state planning. We may be tempted, as Levy is, to think that the social contract tradition, including public reason views, cannot take on these insights comfortably. But my version of public reason liberalism can accommodate these insights within what might otherwise seem to be a rationalist framework. I acknowledge that social orders can emerge and survive pluralistically by making use of local knowledge and social evolution. In this way, I draw on the evolutionary contractarianism of F.  A. Hayek, Loren Lomasky, Brian Skyrms, and Gaus.116 I  also allow that law can emerge not merely from the state and its legislature but from common-​ law institutions, law grounded in the decisions of judges outside of the formal nation-​state. And third, I have argued that these orders are normatively attractive from a public reason perspective. Muñiz-​Fraticelli emphasizes that pluralism acknowledges that groups have authority over their members that “imposes an external . . . limit to the authority of the state.” I take on this insight as well.117 Levy claims that “there is no systematic way to combine all of the virtues and none of the vices of the two [rationalist and pluralist] mindsets, and no secure middle way that would allow us to know for sure which are virtues and which are vices.”118 And again, “A fully liberal theory of freedom cannot do without the insights of either rationalism or pluralism, and yet these are probably impossible 115 Ibid., p. 39. Here local knowledge is the knowledge the particular persons have about how to act in their local environment. 116 Hayek 1978, pp. 15, 24, 28–​9. See Sugden 1993 for an analysis of Hayek as an evolutionary contractarian. See Lomasky 1987, pp. 79–​83; Skyrms 1996; and Gaus 2011, pp. 409–​46. 117 Muñiz-​Fraticelli 2014, p. 178. 118 Levy 2015, p. 3.

118  Trust in a Polarized Age to fully reconcile.”119 I do not claim that we can combine all the insights of rationalist and pluralism liberalisms in one theory, nor do I claim that we can always know when the pluralist or the rationalist is correct in criticizing overreach by states or by associations. But we can go a long way toward integrating rationalist and pluralist insights into public reason liberalism. And pluralist public reason can save public reason views broadly from a powerful pluralist challenge to rationalistic social contract theory.

12.  Associations Create Trust for the Right Reasons Freedom of association helps to promote social and political trust. The evidence is not decisive, but since there is no evidence that associational life undermines social or political trust and there is a plausible prima facie case for how it might help, we can conclude that a robust right of freedom of association, when upheld and exercised, creates real trust. And since freedom of association is a publicly justified primary right, freedom of association creates real trust for the right reasons. I should stress the limits of my arguments, however. First, the trust literature largely strengthens the case for the freedom of bridging associations and may weaken protections for isolated associations. Second, all associations are subject to state intervention in cases where their constituent moral rules cannot be publicly justified to their members. This chapter is therefore not meant as a robust defense of pluralist liberalism to the exclusion of rationalist liberal insights. I now turn to more controversial questions about the status of economic liberties and the appropriate role of government in economic matters. These are challenging issues because commercial property rights are more controversial than other liberal rights, so we must grapple with a wider range of reasonable views on the matter.



119

Ibid., p. 253.

4

The Market Economy The market economy consists in the economic relations that generally arise from extensive private property rights, including rights in capital. In this chapter, I will argue that the market economy can create trust for the right reasons. Markets and property rights promote social and political trust in the real world, and they arise from a publicly justified primary right of private property. In this chapter, I first explain how I understand markets and private property rights (section 1). I then review the trust literature and argue that the market economy sustains trust in the real world both by encouraging diverse persons to interact peacefully and by creating economic growth (section 2). I will explore the bases for private property rights found in primary rights of agency and association, arguing that private property rights are central jurisdictional rights (section 3). I  then develop the connection between markets and economic growth (section 4). I review some of the complications that arise in appealing to judgments of desert in justifying the right to private property (section 5). Finally, I explain why strong forms of capitalism, where private property rights are fully extensive, are not justified for many members of the public (section 6). Chapter 5 qualifies capitalism with the welfare state, and c­ hapter 6 shows that economic regimes to the left of welfare state capitalism cannot be publicly justified. I defer assessing criticisms of capitalism to c­ hapters 5 and 6.

1.  Markets and Private Property: Some Definitions I understand a private property right as a bundle of particular rights to do particular things with some object. When an agent, A, has full private property rights over x, A has (1) the right to use x as she wishes so long as this is not harmful to others or their property; (2) the right to exclude others from using x; (3) the right to manage (A may give permission to any others A wishes to use x, and A may determine how they may use x);

120  Trust in a Polarized Age





(4) the right to compensation (if someone damages or uses x without A’s consent, A has a right to compensation for the loss of x’s value from that person); (5) the rights to destroy, waste, and modify (A may destroy x, waste it, or change it); (6) the right to income (A has a right to the financial benefits of forgoing the use of x and letting someone else use it); (7) immunity from expropriation (x [or any part of x] may not be made the property of another or the government without A’s consent, with a few exceptions); (8) liability to execution (x may be taken away from A by authorized persons for payment of a debt); (9) absence of term (A’s rights over x are of indefinite duration); and (10) rights to rent and sale, sometimes known as transfer rights (A may temporarily or permanently transfer all or some of his rights over x to anyone A chooses).1

Thus, a single private property right can be understood as the possession of one particular right over an item. A justified property right is one where a person has legitimate or morally binding claim to a particular right. She can permissibly and authoritatively demand that others not interfere with her exercise of the right. Full private property rights over an item include the possession of all the particular property rights to that item. And having full private property rights over one thing implies that the rights holder can acquire all particular rights over a wide range of other goods. These property rights are valid claim-​rights. A capitalist, or fully free market, regime, gives everyone the authority to acquire all of these rights over property through the acquisition of original resources or through voluntary exchange or gift. In this way, capitalist regimes are grounded in freedom of contract. Capitalism also involves few, if any, restrictions on which items can be owned or exchanged in this way. Sex, drugs, organs, and guns can be freely bought and sold.2 People are free to charge incredibly high prices for their goods and services or pay laborers an extremely low wage, provided only that they can find others who choose to take the deals they offer. Capitalism protects vast inequalities of wealth, so long as they form through processes that comply with property rights. And it generally prohibits the redistribution of wealth to help the least advantaged. We can therefore define pure capitalism as an economic system that allows persons to acquire the maximal 1 Honore 1961. 2 For libertarian arguments for legalizing markets in an enormous range of goods and services, see Brennan and Jaworski 2016.

The Market Economy  121 set of particular rights (listed previously) over maximal quantities of a maximal range of items.3 A purely capitalist economic system is thus a union of a vast array of laws that protect against personal and governmental interference with the acquisition, exchange, and use of the particular rights. I will understand markets, broadly, as institutions where persons may contract with one another to exchange particular rights over goods and services. Thus, there is a sense in which markets are institutions of free exchange and freedom of contract since buyers and sellers are not typically coerced into trading particular rights. I do not want to define markets such that it is a conceptual confusion to claim that markets can be oppressive and unfree. People may feel coerced into buying or selling particular rights in the free market; they may feel forced to buy or sell them at certain prices. Markets are simply institutions where particular rights are ordinarily voluntarily exchanged. But we can understand a free market as one where the generation and exchange of these particular rights proceeds without interference from government or private parties. And so, contra Harcourt, the idea of free markets is not an illusion.4 It makes perfect sense to acknowledge that we need norms that constrain behavior so that people can hold and exchange particular property rights while still holding that a market is free when interference with the holding or exchange of these particular rights is minimized. I will understand a market economy as one where a wide range of particular rights over capital, or productive property, may be exchanged. This means that the defining institution of the market economy is the stock exchange, an institution where capital is bought and sold in a wide variety of ways. But I will not define markets as involving full property rights, nor is my defense of the market a defense of pure capitalism.

2.  Markets and Trust Market economies arise from the protection of primary rights of private property. And since such property rights are (as we will see in the section after this) publicly justified to a diverse public, the outcomes of those processes should, within certain limits, be regarded as legitimate, authoritative, and so a rational basis for social, legal, and political trust. On top of that, markets produce prosperity in line with publicly justified principles, such as a principle of sustainable improvements, where a publicly justified polity permits everyone to partake in 3 Gaus 2009. It is standard for capitalism to bar a few potential objects of ownership, such as persons. 4 Harcourt 2011, p. 242.

122  Trust in a Polarized Age social and economic growth in ways that can be sustained over time.5 Let’s look at the evidence. Markets can generate social trust through a number of different channels, primarily putting diverse persons into contact with one another and creating opportunities for the production and observation of trustworthy actions. By allowing people to interact for mutual benefit, markets promote trust through social contact with diverse persons. People involved in market exchange realize that they can trust those who are different from them. This generalization is sustained by evidence gathered by Joseph Henrich et al. that “extensive markets transactions may accustom individuals to the idea that strangers can be trusted.”6 This is consistent with their findings across fifteen small-​scale societies that trustworthy behavior, understood in terms of ultimatum game offers, is strongly correlated with degree to which a society uses markets.7 These findings suggest that the temptation markets create to be untrustworthy may apply at the individual level but does not cause a detectable macrolevel decrease in trust. Markets may also impel people to be trustworthy by giving them an additional reason, namely financial gain, to abide by moral and legal rules, and the resulting perception of trustworthy behavior will increase social trust levels.8 On the other hand, insofar as markets promote economic inequality, they may reduce social trust, so care with the data is necessary. The first causal channel from markets to trust is from well-​defined, protected legal property rights. We can understand legal property rights in terms of the protection of personal property from theft or seizure. Protections for legal property rights are not understood as a function of tax levels but, rather, as a matter of security from theft, damages, fraud, and so on, and the ability to exchange and invest in the market. Nordic countries have high tax rates, but they also have strong protections for legal property rights in the sense described. Blaine Robbins argues that “in countries with fair and effective and efficient property rights, an increase in their effectiveness leads to an increase in generalized trust (especially in social democratic and Nordic countries).”9 Robbins has also found that “generalized trust is much greater when legal property rights and civic participation are acting in concert than when acting alone.”10 Jong-​sung You finds “a significant effect of legal property rights institutions on social trust.”11 Niclas Berggren and Henrik Jordahl offer a similar argument, namely that economic freedom, 5 MPBW, pp. 168-​70. 6 Henrich et al. 2005, p. 811. 7 Henrich et al. 2005. 8 Anomaly 2017, p. 106. This works so long as the financial reasons do not crowd out moral motivation, though Henrich et al. (2005) have argued that there is little or no crowding-​out effect. 9 Robbins 2012, p. 11. 10 Robbins 2011, p. 329. 11 You 2018, p. 486.

The Market Economy  123 which includes the rule of law and high-​quality property rights, increases social trust by reducing the objective risk in commercial activity.12 And Cassar has argued that “the quality of legal institutions has a positive causal effect on trust and trustworthiness” because good legal institutions reduce “the frequency with which subjects face opportunistic agents when trading,” and this reduction increases detectable trustworthiness, further increasing trust.13 Robbins offers a compelling causal story that fits the analysis of this book: that political and institutional incentives “foster expectations and beliefs about the reliability of anonymous others, and that effective legal structures and property rights institutions create an environment where common knowledge about the trustworthiness of strangers can grow.”14 When the legal system enforces the legal norms of property ownership and social norms back them up, individuals can observe trustworthy behavior in others when they comply with these norms and, thus, build social trust over time. With adequate legal enforcement, the punishments for deviation are clear, and when enforcement is ubiquitous, everyone knows that deviators will be punished and that people generally have an incentive to engage in trustworthy behavior. If legal property rights are publicly justified, as I will argue, then the norms of property ownership should tend to be internalized and punishments accepted as legitimate rather than seen as brute uses of force. Thus, the motive to avoid being punished is not merely self-​interested but moral in character. We might comply from fear of guilt, for instance. In this way, market order, and the private property rights that give rise to it, should increase social trust levels. A second source of social trust is that market orders with well-​defined property rights are less likely to be corrupt because people can seek financial gains through market activity rather than, say, the abuse of public office. They can make investments, save, and work without fear of government or private actors interfering with their activities in ways they cannot foresee or predict. This does not mean that high taxes will reduce trust: taxes can be predictable, they are not often seen as corrupt, and people can plan around them. It is unpredictable interference with property rights that reduces trust.15 When a society avoids the appearance of corruption, its members do not observe as much untrustworthy behavior. Social trust sustains a market economy in a number of ways. Most importantly, social trust reduces transactions costs, as we saw in ­chapter 2. High-​trust orders tend to be less corrupt and have better property rights protections, and 12 Berggren and Jordahl 2006. 13 Cassar, d’Adda, and Grosjean 2014, p. 859. 14 Robbins 2012, pp. 11–​2. 15 But corruption is probably best controlled through careful constitutional design, so I will focus on the connection between corruption and social trust in c­ hapter 7.

124  Trust in a Polarized Age much of the causal arrow runs from social trust to lower corruption and stronger property rights. Thus, insofar as markets create social trust and social trust, in turn, helps markets to function better, the market economy should initiate a positive feedback loop with trust. We do not have direct tests of this claim, but we do have evidence for the causal links in the loop. We know that social trust facilitates well-​functioning markets, and we have some reason to think that well-​ functioning markets, undergirded as they are by legal property rights, increase political trust. Finally, Christian Bjørnskov has found that trust “affects the efficiency and structure of management in the private sector” because “the clients of the bureaucracy may be more rule compliant in high-​trust countries” and because “the bureaucrats themselves may be more rule compliant.”16 We should be careful to avoid endorsing some causal connections. Economic growth and social trust are well correlated. One might think that economic growth causes social trust, perhaps based on the plausible assumption that when people see the economy growing they are less likely to grow suspicious of others as potential causes of poor economic performance. Moreover, when people are wealthy and expect to growth wealthier with time, one might expect them to be more relaxed and trusting in general. But the social trust literature does not bear this out. Social trust causes economic growth, rather than the other way around. Bjørnskov argues that the literature indicates “quite clearly” and with “multiple methods” that “causality in the long run must flow from trust to economic growth.” If the causality flowed the other direction, he argues, “Trust levels should have grown rapidly over time in some countries and slowly in others and in proportion to speed of economic development,” but that is not what happened.17 It is remarkable how big the effects are. Bjørnskov finds that a change in social trust of one standard deviation “is equivalent to an increase in per capita GDP of approximately 1840 USD, worth about 11 years of economic development at the sample mean growth rate.”18 How exactly does social trust facilitate economic growth? First, trust probably increases total factor productivity, mediated by education and institutional quality. Higher-​trust societies are better educators and have higher-​quality legal and political institutions.19 Trust also increases the quality of formal institutions, though this probably works primarily through “political channels, as the effects were increasing in political competition and only appeared significant in countries with stable democratic institutions.”20 And since social trust increases economic growth, it helps to determine the quality of legal systems

16

Bjørnskov 2020. Bjørnskov 2018, p. 552. Bjørnskov 2010, p. 342. 19 Bjørnskov 2018. 20 Ibid., p. 551. 17 18

The Market Economy  125 because “economic development is a strongly significant determinant of legal quality,” though there is also a feedback effect from legal quality to development. A society’s openness to international trade is also significant.21 Social trust also facilitates growth through the direct reduction of transaction costs. This effect is strongest when “the quality of formal institutions is low,” but that beneficial effect practically disappears when formal institutions are strong.22 The converse relationship is also supported:  institutional quality is most important in increasing growth where social trust is low and where “malfeasance is therefore more likely to be widespread.” This suggests that to get economic growth, a society needs a high level of either social trust or political and legal trust.23 Importantly, this might only be a short-​term effect.24 However, while economic growth does not help social trust, it does appear to help political trust. A person’s political trust increases when she feels that her own economic performance is strong and improving with time. It is important that political trust does not correlate well with objective economic performance or with cross-​national economic performance.25 Instead, “Subjective evaluation of the national economy is consistently one of the strongest (if not the strongest) determinants of political trust across a wide range of countries, including not only established democracies” but nondemocratic regimes as well.26 Moreover, when societies develop economic expectations based, say, on public awareness of the presence of natural resources, they will become less politically trusting when the benefits they expect do not arrive.27 Even so, insofar as markets create economic growth—​and insofar as persons recognize when their economic performance has improved—​markets should create political trust. This is true not only for people who benefit from economic growth directly, say through higher wages or profits, but for recipients of social insurance, since social insurance is more sustainable in high-​growth economies than in low-​growth economies. To the extent that free markets are combined with welfare state institutions that help ensure that growth benefits everyone, then to that extent free markets help to generate political trust. Importantly, all of these channels for increasing real trust should provide the right reasons for social and political trust too. Markets and property rights often 21 Bjørnskov 2010, p. 338. 22 Bjørnskov 2018, pp. 551. 23 Bohnet, Frey, and Huck 2001. 24 Ibid. 25 Van der Meer and Hakhverdian 2017, p. 98, find that political trust is not correlated with macroeconomic variables, but they allow that economic expectations, which are often untethered from economic reality, have an effect on political trust; that is consistent with the literature finding a connection between political trust and economic performance. 26 Van der Meer 2018, p. 603; emphasis mine. 27 Listhaug and Jakobsen 2018, p. 570.

126  Trust in a Polarized Age provide people with incentives to engage in trustworthy behavior, which will justify trust as well as causing it. In this way, the factors that cause trust do not bypass or conflict with our capacity for detecting and acting upon the reasons that justify our trust. The causes of trust for the right reasons may support one another. I conclude then that the market economy creates and sustains social and political trust and can thereby play a central role in interrupting the distrust-​ divergence process.28

3.  The Bases for a Primary Right of Private Property* Now we should turn to the public justification of private property. Here I argue that persons have an authoritative claim-​right to demand that others refrain from interfering with their acquisition and transfer of many particular rights over a broad range of goods and services that can be used to pursue their diverse conceptions of the good and justice. The market economy is the mode of economic exchange and production that arises from respect for these claim-​rights. In this way, members of the public have a primary right to participate in a market economy, though not, as we shall see in ­chapter 6, in a completely free-​market economy. Here is why. First, private property rights are publicly justified on the grounds that they are necessary to give form to more basic rights, like rights of agency and association. (i) Persons need primary property rights in order to protect their rights of agency. Private property provides persons with the resources they need to satisfy their urgent wants and needs and to pursue their projects, plans, and principles. These elementary considerations will motivate diverse persons behind a thin veil of ignorance who wish to pursue their conceptions of the good and justice to embrace at least some basic rights to private property.29 (ii) Persons have rights to participate in moral associations, and these moral associations require private property rights in order to exercise their collective agency. People need homes, faith communities need houses of worship, universities need campuses, and so forth. And individuals and associations need a wide array of goods and services to achieve their goals in their homes and the 28 Though I should note a caveat. I do not think business cycles are an inherent feature of the market economy, but I recognize that many people do. If so, then part of assessing the trust-​based case for markets is to look at the effects of business cycles on social and political trust; as noted in ­chapter 2, economic downturns often lead to lower political trust, at least temporarily. 29 Save the very few who think all property rights stand in the way of poverty relief, but I address these socialists indirectly in the discussion of socialism in ­chapter 6.

The Market Economy  127 places where they gather to pursue shared projects and values. (iii) This is true not merely of nonprofit moral associations like families and churches but also of economic organizations like unions and small firms. So insofar as members of the public will select rights of association, they must also adopt some property rights in order to ensure that those associations can flourish. (iv) Finally, property rights help to determine which associations can come into existence in the first place, and so we can justify property rights partly on the grounds that they are conducive to the formation of enduring associations; thus property rules that, say, lower transactions costs in forming associations may be preferable to property rules that do not.30 A second consideration that favors the selection of private property rights is that individuals’ and associations’ separate holdings serve as jurisdictions in a diverse society of people who have trouble making collective decisions due to crippling dissensus. Gaus and Eric Mack have rightly stressed the role of private property in addressing the problems of evaluative pluralism. Due to diverse perspectives, people will seldom be able to collectively agree about how to make various decisions. But if we employ jurisdictional rights as devices of public reason, then we can partition social space to give each person the right to follow her own evaluative standards in her own jurisdiction, while guaranteeing the same right to everyone else. That is, decentralization will make some rights practices eligible in order to resolve conflicts due to evaluative pluralism. Property rights are probably the main way in which decision-​making is decentralized, as people can acquire and transfer property without imposing too many costs, and while likely producing many benefits, for others. A person can buy or sell a home or business, contribute to a church, join a charitable organization, invest for retirement, save for vacations, and so on; she does not need a public consensus on which home, business, faith community, charity, investment, or vacation she should choose. Property rights free us from struggling to agree on which projects to pursue. Thus, contracting parties will adopt property rights because they help address evaluative pluralism. Like other rights, private property rights are publicly justified prior to the justification of legal and constitutional rules; thus, while all rights claims require recognition, articulation, and reform by the legal and constitutional orders, they also place considerable constraints on the activities of those orders. So we must speak to the relationship between law and the state in establishing property rights. One way to do so is to consider how private property rights function



30

I am grateful to Christian Bjørnskov for this point.

128  Trust in a Polarized Age under conditions of low state capacity, where states are weak and cannot sustainably or credibly interfere with property practices, markets, or restrictions on markets. In c­ hapter 3, I argued against the conventionalist challenge, which purports to show that states have priority over private property because property rights cannot function, nor even exist, in the absence of state action. I argued that property norms predate the nation-​state, and even legal systems, extending back to hunter-​gatherer tribes. Property rights are indeed socially constructed in that they depend on norms, but these norms need not be legislated, nor even embodied in state law. This means that, while property rights are not presocial, they are often prepolitical. Knight and Johnson’s argument for the priority of democracy in creating markets rests on the conventionalist challenge.31 They assume that markets require a fairly extensive state. However, while many features of markets seem to have developed first and best in strong nation-​states because strong nation-​states removed feudal restrictions on market transactions, this does not imply that markets are created primarily by democratic legislation. So, the need to construct markets does not bolster the case for the priority of a democratic, legislative state but, rather, reveals the role of social and legal norms in creating markets. Sometimes legislation is needed, but not always. Thus, arguments of the form “Markets must be created; therefore, they cannot ground claims to limit state power” rely on deep ambiguities about what sorts of social rules are required to create and sustain markets. Like property, markets are not presocial, and so we need norms (and trust!) to have markets. But that does not mean that markets always necessitate legislation. Here I appeal to generalizations about historical relationships between property norms and nation-​states solely to resist the conventionalist challengers who insist on limiting property rights based on the historical claim that states are required to create property rights. The historical claim is, as far as I can tell, false. And so it cannot play a central role in determining the scope of property rights.32 On the other hand, I do not claim that a publicly justified polity defaults to pure capitalism, such that primary property rights embrace full capitalism until a presumption against interference is overcome. Pure capitalist property rights have never existed, suggesting that they do indeed depend on the activities of other institutions. In this way, states may have partial priority over fully extensive property rights, but not over property rights as such. Without the state, we can expect private property rights to endure, but we see that many exchanges are nigh-​universally blocked, and in many 31 Knight and Johnson 2011. 32 Indeed, it is not clear how historical facts could figure into the justification of property claims at all; but even if they do, false historical claims do not.

The Market Economy  129 cases property ownership and market activity are restricted, as was the case in medieval social orders. Thus, the property rights that precede the state may end up reducing economic efficiency through restrictions on price formation and fluctuation. They may tend to privilege certain social classes over others, reinforcing unjustified hierarchical norms. And they may fail to provide adequate aid to the poor and protections for workers. Many prepolitical property norms will be defeated for many members of the public as a result. This means that the legislative state may justifiably intervene in property claims in many cases, so as to open up markets, protect vulnerable workers, expand property rights to larger firms that could not otherwise exist, and create and enforce intellectual property rights. But the legislative state is nonetheless limited by strong property rights. All of these considerations strongly favor some form of capitalism because they all provide grounds for embracing property rights in productive resources, that is, capital. Some of what I have said should be acceptable to egalitarian and progressive readers, at least with respect to limited economic liberties like a right to personal property (such as a home) and a right to freedom of occupation. But I argued in MPBW, and reviewed in ­chapter 1, rights of agency, association, and jurisdiction imply a right to own productive property, since a person can use her other rights to acquire productive property.33 For instance, a citizen can use her freedom of occupation to convert her personal property in her home into productive property, as does a person who decides to run a massage parlor out of her house.34 It’s hard to argue that productive property is not included in the primary right of private property. This means that nearly every publicly justified polity will have at least a loosely capitalist economy. The moral and legal rules that articulate and protect property rights will establish a capitalist economy because they often have priority over property rights legislation. For such legislation to be publicly justified, members of the public must see it (and, as we shall see, the body that creates it) as correcting important defects in property rights. But they arguably will see some forms of legislation as improving property rights and property practices. Legislation allowing prices to fluctuate, allowing buying and selling of property titles, redistributing large property holdings based solely on inheritance patterns, and so on, can arguably be publicly justified to each person. Thus, this primary right to private property will be strong but qualified by the need for certain kinds of economic intervention.35

33 MPBW, pp. 210-​2. 34 Or setting up the business in an area zoned for such purposes. 35 And since we finely individuate moral and legal rules, even more forms of interference with property rights will be defeated.

130  Trust in a Polarized Age

4.  Economic Growth, the Great Enrichment, and Sustainable Improvements A further reason to adopt a primary right of private property, and to institutionalize fairly extensive property rights, is the capacity of the market economy to yield incredible, compounding economic growth and prosperity. This factor is typically ignored in discussions of public reason and private property, but I think it is of tremendous importance. It will matter to moderately idealized agents36 and real citizens, who adjust their trust levels based on their perceived degree of economic improvement, as I argued in ­chapter 2. So let us explore the reasons for market productivity and then see how they figure into a contractarian argument for capitalism. The typical reasons for the productivity of capitalism are three:  incentives, information, and innovation. In a market economy, the profit motive will incentivize people to produce goods and services that others want to buy at a reasonable price, to improve those products, and to offer a wider variety of products over time. The free entry of firms into a market will create competition that should yield even more productivity and efficiency. Markets have the unique ability to convey information about how to make productive decisions through the price system, as prices stand in for much more complicated informational states about what buyers are willing to buy and what sellers are willing to sell. As Hayek famously argued contra central-​planning socialism, prices are the cognitive components of a highly intelligent social system of economic production.37 The individual buyer and seller need to know little in order for prices to coordinate them to make economically rational and profitable decisions in many cases. This does not mean that markets impel people to always be fully economically rational in the sense of Homo economicus, but rather that the market price system is the best system we have for harnessing that information. Finally, markets are innovative, as the profit motive and access to new information give people a reason to experiment with new production processes in order to increase their well-​being and financial position. Sometimes people make relatively small innovations, say by slightly increasing the efficiency of a production process. But in other cases, entrepreneurs take big risks with new ideas that, while failing in most cases, sometimes succeed in transforming and improving the economy, as the invention of the personal computer and smartphone did in recent decades. The capitalist system of profit and loss promises great financial and social gains from successful entrepreneurship.

36 37

I advanced a similar argument in MPBW, pp. 168–​70. Hayek 1945.

The Market Economy  131 On top of that, capitalism’s productivity greatly increases wealth over time, as economic growth benefits from compounding growth rates. Even low growth rates produce enormous gains as each year’s growth builds on what preceded it. Capitalism has produced gigantic improvements in economic well-​being over the last two centuries, along with improving many other measures of well-​being, such as quality and quantity of life. It is worth focusing on the stunning empirical fact that economist Deirdre McCloskey calls the “Great Enrichment.”38 Few today would dispute that a competitive marketplace, where firms are free to experiment with new methods of production that are then subjected to the withering scrutiny of millions of consumers, is a kind of golden goose. And it is a golden goose that we can kill; command economies nearly killed it. When we back off of pure capitalism, then, we must be mindful not to strangle the productive process. Even small costs to the growth rate have dramatic effects over time because of compounding growth rates. Without growth, we will lose enormous social goods not merely for the rich but also for the middle classes and the least advantaged. But diverse members of the public will judge moral and legal rules by their predicted effects on robust, energetic economic activity. Given that nearly everyone has an interest in greater economic well-​being, the fact that some moral and legal rules reduce economic growth is an enormous mark against them. Obviously economic wealth is not the same as human well-​being, but it surely has something to do with it, since economic wealth contributes to all kinds of goods, including food, healthcare, shelter, clothing, leisure time, enjoyable work, and so on, and high economic wealth increases the quality and abundance.39 Moderately idealized members of the public will insist on moral and legal rules that work to their benefit not only at present but over time.40 Environmentalists of various stripes will object, not unreasonably, that our present economic growth

38 McCloskey 2016. Critically, McCloskey, along with many others, argues forcefully that while private property rights are necessary for the extraordinary increase in economic prosperity all over the world, they are not sufficient. McCloskey, for instance, argues that the spread of certain pro-​ commerce moral ideals are necessary to fully explain the Great Enrichment. Here my aim is merely to argue that private property rights are publicly justified because they play a necessary role in economic growth, but including the other necessary conditions will help to determine which particular property rights claims are legitimate. I do not think the additional conditions undermine property rights claims, however, since lists of these conditions seldom include significant restrictions on property rights. I thank Brian Kogelmann for raising this point. 39 This contradicts the Easterlin paradox, which holds that while happiness varies with income up to a certain point, it does not continue to increase as income increases. For a standard critique of the Easterlin paradox, see Stevenson and Wolfers 2008. 40 This argument is made more fully in MPBW, chap. 5. This raises a fascinating, potentially large set of new research questions concerning the public justification of adopting a certain social discount rate. I thank Eduardo Martinez for raising the issue, which I’d love to address here. I assume that 0 and 1 are not reasonable discount rates (0 means the future matters not at all in comparison to the present, and 1 means that all of the future matters exactly as much as the present).

132  Trust in a Polarized Age path may be unsustainable. But the principle of sustainable improvements only undermines moral and legal rules that prohibit sustainable improvements. If the objections of environmentalists survive moderate idealization and the standards of what I call policy epistemology in ­chapter 5, then growth can be restricted to ensure that the growth we do get is sustainable over the long run. Restrictions on capitalism that undermine growth, especially those that restrict improvement of the economic well-​being of the least advantaged, will be hard to justify. Growth-​ decreasing policies—​both long-​term restrictions and short-​term restrictions that stick around past their intended expiration date—​pose enormous risks to prosperity. Members of the public will insist that we not kill the golden goose of productivity of the market economy. A further reason to care about economic growth is based on evidence that growing societies are more liberal and tolerant. As Benjamin Friedman has argued, growing societies have a large pie of resources to divide, which limits conflict-​generating scarcity.41 Given our aim of establishing trust for the right reasons, if economic growth promotes social stability and fellow feeling, that is good reason to think that growth-​promoting rules can be publicly justified. However, Friedman’s argument is incomplete because economic growth appears to increase political trust, but not social trust, as we saw in c­ hapter 2. The desirability of growth will not only strengthen the public justification of private property rights, it provides sufficient reason to reject restrictions on property rights. If property rights restrictions hurt economic growth that is broad-​ based—​growth that benefits everyone—​then many members of the public will have sufficient reason to reject those restrictions. The reverse is also true: if property rights protections hurt broad-​based economic growth, then many members of the public will have sufficient reason to reject those protections. Members of the public will be aware of these general consequences of the market economy. Even Marx acknowledged that capitalism is a fantastically productive economic system, despite the injustice and misery it can cause. So even socialists should recognize that capitalism has enormous productive potential. Of course, this doesn’t mean that we should have pure capitalism or that pure capitalism will be publicly justified to anticapitalists, but given that moderately idealized agents can see the predictable, positive economic consequences of capitalist economic systems, we should be able to begin public justification from a basic framework of private property rights that include modest rights to

41 Friedman 2005. Along the same lines, Uslaner 2002 argues that social trust and peace are promoted by economic equality. So we might be in a difficult position if we need economic equality to generate social trust, but redistribution of wealth reduces the growth rate. I  address Uslaner’s concerns about economic inequality and trust in c­ hapter 6.

The Market Economy  133 productive property. Then we can determine what sorts of government intervention can be publicly justified to amplify basic private property rights.

5.  Private Property and Deservingness One of the primary rationales for private property rights is that they reward persons for their contributions to society. As Mill said long ago, the justification for a system of private property is “the guarantee to individuals of the fruits of their own labor and abstinence.”42 The familiar idea is that only a system of private property can guarantee to people that they are rewarded for their contributions to society and for their investments, which involve abstaining from buying consumption goods. These considerations, however valid, cannot undergird the public justification of the right to private property, because a person’s contribution to the social good can only be determined once we have a system of property in place; we cannot make full sense of the idea of what an individual has contributed apart from a system of property rights that a society has chosen, collectively or through an interdependent spontaneous order, to sustain.43 We cannot say that private property is justified because it gives people what they independently deserve, since deservingness is downstream from the establishment of the right. And even once a right to private property is established, market outcomes are seldom, if ever, entirely a function of the deservingness of a worker or business owner. Markets are complex systems, and so it is difficult if not impossible to determine each person’s contribution to valued outcomes beyond a quite rough estimate. This is a point stressed by Hayek; Hayek even worried that basing society on deservingness would be potentially repressive.44 However, that does not mean that deservingness judgments play no role in determining the contours of the right to private property. Once some property rules are in place, we can develop a rough sense of the value that persons contribute by looking at their marginal product. And we can, in selecting among property systems, favor those in which people get more of what they contribute. For instance, feudal property arrangements reward inherited wealth over hard labor in the production of agricultural goods, and one can use a principle of 42 Mill 1963–​, vol. 2, p. 208. 43 For an argument to this effect, see Aas 2019. Aas argues that the less advantaged in a system of economic cooperation can complain that their contribution to upholding the economic system entitles them to larger holdings than they are given in a capitalist economic system, since respect for equality means that each person’s contribution should be treated equally. But what we collectively uphold is a function of the moral, legal, and constitutional orders and the extent to which they are publicly justified, so even if others contribute to our projects’ success, no egalitarian commitment falls out of that claim, just a commitment to structuring society in accord with what is publicly justified. 44 Hayek 2011, chap. 6.

134  Trust in a Polarized Age deservingness in order to defeat those arrangements in favor of more open markets, where inherited wealth is a relatively smaller factor in determining income and holdings. We can say, for instance, that a serf is not rewarded according to what she deserves by a feudal property system, and that this counts against such a system and in favor of one that better rewards those who toil on the land. Further, once a justified system of private property is established, people can come to deserve their holdings by acting in accord with those rights. Thus, one’s holdings under these property rules can sometimes be said to be deserved, and that can provide a reason to oppose the use of coercion to redistribute that wealth. The trouble with using deservingness to help flesh out the content of private property rights is that reasonable people disagree about who is deserving. This is for two reasons. The first is Hayekian: it is simply really difficult to determine what people contribute. The second is broader, namely that people have different theories of value contribution. So I am not sure how frequently deservingness judgments will be relevant to shaping the right of private property, but if there is an overlapping consensus judgment, say that feudal property holdings fail to give lower-​class persons what they deserve, then desert judgments can help determine the content of our property rights.

6.  Backing Away from Libertarianism Libertarianism implies that states, or at least private protection agencies, should codify and protect maximally extensive property rights over the maximal set of goods and services.45 It defends pure capitalism. Here we can understand libertarianism not in terms of foundational commitments to, say, self-​ownership, but rather as the advocacy of free-​market institutions. As I noted in this chapter, section 1, this means legally protecting a vast array of potential exchanges, such as selling sex, drugs, guns, and assisted suicide, and it means allowing total freedom to charge whatever prices one likes for a good or service, say an incredibly low wage for work or incredibly high prices for medical care. Pure capitalism will in principle allow for vast inequalities of wealth, and it prohibits most forms of redistribution.46 It is important to note, contra some libertarians, that pure capitalism is a coercive system when defined in this way. Every one of the particular rights listed

45 Rothbard 1973 discusses how private protection agencies might function in an “anarcho-​ capitalist” social order. 46 But a principle of rectification might require large-​scale redistribution, something libertarians recognize. See Nozick 1974, chap 4.

The Market Economy  135 is meant to be coercively enforced against violators by the officers of the minimal state.47 For instance, today Bill Gates has billions of dollars that the government coercively protects from expropriation by private persons and from further expropriation by other parts of the government. In a fully capitalist society, protection services would be expected to protect Gates against anyone who would take any amount of his holdings without his consent, save for things like damages for torts.48 We can focus our inquiry by asking whether the specific legal rules requiring coercive protection of these particular property rights can be publicly justified in most cases.49 I think it is quite clear that the coercion required to enforce maximally extensive private property rights cannot be publicly justified, because many people will have defeaters for particular functional legal units that add up to a pure capitalist level of private property rights protection. To show this, let’s imagine a progressive egalitarian, Reba, who affirms the justice of redistributive taxation and worries that capitalism will hurt the poor and produce unjust economic inequalities. Based on her egalitarian commitments, Reba would prefer no law protecting all of Gates’s holdings to a law that did so.50 So, if Reba were to use the political process to redistribute wealth from Gates, Gates lacks the rightful authority to forcibly prevent Reba from doing so or to prevent political officials from doing so on her behalf. If Gates prevents the expropriation of his holdings in at least a handful of cases, it is Gates who uses coercion without a public justification. The libertarian is bound to reply that the state will coercively prevent Gates from protecting his holdings from expropriation, and that this coercion cannot be justified for Gates. But in this case, Gates might be the initiator of coercion. Some will argue that the state’s expropriation would only lead to coercion insofar as it needed to stop Gates from interfering with officials’ expropriation of his holdings. The libertarian could complicate matters by arguing that, without a background assumption of the legitimacy of certain property rights, there is no way to determine who initiates coercion in this case. If Gates is the rightful owner of his property, then the expropriators initiate coercion against his property, and so 47 Or the agents of private protection services in a market-​anarchist regime. 48 Now, critically, a libertarian society might not, in practice, allow anyone to become rich in the way that Gates became rich, since Gates’s wealth derives from government interventions like extensive intellectual property rights, which many libertarians oppose. Any long-​term attempt to amass wealth would attract competitors, who might bid down the large profit margins that the very rich use to become very rich. 49 This is a consequence of individuating proposals in a fine-​grained way. 50 Reba’s concern is with Gates’s high level of wealth, not his holding property per se, so she will not have defeater reasons for anyone’s property rights. Similarly, she will not necessarily have defeater reasons for large groups of people holding similar levels of wealth collectively, such as in the form of a joint stock corporation.

136  Trust in a Polarized Age the coercion must be justified to Gates. But if Gates is not the rightful owner, then his acts of putative defense are not defensive, but aggressive. In this case, I think members of the public will reason as follows. Given that we face a potentially violent conflict over holdings, we should embrace property rules that can forestall these conflicts. We could establish such rules by first appealing to rights that are already publicly justified for both sides, such as rights against bodily harms in cases of conflict over enforcing the law. If so, then acts of expropriation should follow one of the justified legal procedures that both sides deem appropriate to avoid violent conflict in cases where the rights conflict cannot be settled to each group’s satisfaction. So even if we cannot agree on who initiates coercion, we can publicly justify legal procedures that violate libertarian proposals. This is due both to the defeater reasons that many have for libertarian property rights and the commitment many libertarians will have to procedural rules and other rights protections that will peacefully resolve conflicts. In this way, libertarians will recognize others’ authority to violate libertarian property rights in some cases. The difficulty for this approach is that the decision procedure may have to select among options all of which are defeated for at least some sizable segment of the public. In cases where the socially eligible set is empty and we cannot agree on what it means to default to no coercion, then we will have to appeal to a decision procedure that has more limited authority, perhaps on the grounds that it has authority in other domains without empty eligible sets. So we may get stuck with a partial modus vivendi arrangement that provides a shaky ground for trust. But there will be sufficient moral agreement on ways of resolving a dispute about which defeated option to choose. We can also back off of libertarianism because many on the left reasonably (if mistakenly) believe that a pure capitalist system would create a mass of extremely poor people who would be subject to considerable coercion. Many agree with Marx’s remark that “capital further developed into a coercive relation, which compels the working class to do more work than the narrow round of its own life-​wants prescribes.”51 Marx argued, and many non-​Marxists still agree, that capitalism is coercive because the working classes must work in order to survive and that this subjects them to enormous coercion by their employers, as well as by a society that keeps them poor by depriving them of the social means to better their condition. Since sincere, informed people of goodwill can still continue to affirm Marx’s complaint, even if Marxist political coercion cannot be publicly justified, then rights allowing employers to coercively direct their workers, and



51

Marx and Engels 2011, p. 338.

The Market Economy  137 allowing society to deprive workers of sustenance, will be defeated.52 I address concerns about workplace coercion further in ­chapter 6. But, and this is a critical point, primary property rights, including property rights in productive property, should be eligible even for reasonable socialists. First, as I argued in MPBW, socialists are primarily concerned about the private ownership of large amounts of capital, not with the capital owned by small firms, unions, churches, and other civic institutions. So as long as we do not insist that productive property rights be maximally extensive, without a quantitative limit per individuals, which would allow vast capital inequalities, socialists may be unhappy with primary property rights, but they should lack defeater reasons for them. They will also be able to see the ways in which primary property rights create trust and respect between themselves and nonsocialist members of the public. Given the great value of social trust, and given that socialists have their own economic interests that will be improved by private property rights, they should be able to adopt these primary property rights as authoritative, if suboptimal for them. To reject private property in capital entirely would be to insist that the only laws justified to them are those at the top of their ranking of property arrangements. Moderately idealized socialists will also recognize that some kind of democratic process is the only fair way to establish socialism.53 Until then, they must respect property rights established by legitimate democratic procedures, as well as publicly justified constraints on legislative activity such as property rights that normatively precede those procedures. Another argument is that, while socialist citizens might hope to transcend private property in various ways, in the meanwhile, private property rights, even in capital, may have authority if for no other reason than that socialists can make use of private property rights to advance socialism, such as for creating and maintaining spaces for socialist thought and political organizing.54 A final consideration is that socialists will recognize that socialism is defeated for most members of the public in liberal democracies, and so they will have to make up their minds about the economic institutions under which they’re willing to be trustworthy with respect to property rules in order to achieve the benefits of sustaining social trust in the right way. 52 This is not to say that, for instance, Marxist economic thought will survive moderate idealization in a way that could considerably shape public policy, as it may not pass the standards of policy epistemology, which I discuss in Chapter 5. 53 It is possible for socialists to revise our conception of democracy such that democracy requires socialism conceptually, and it is possible that there is some model that shows that having a stable democracy is tightly correlated with having socialism. In either case, however, socialists would have to insist, as many have, that what is widely taken to be democracy under capitalism lacks authority entirely, and few would still maintain this. 54 I thank Eduardo Martinez for this point.

138  Trust in a Polarized Age Libertarians are subject to the same line of argument in making strong, if not fully extensive, property rights part of a society’s legal system. They will recognize that libertarian property rights are defeated, and they will see good moral reason to live in a high-​trust society with many nonlibertarians. They must decide which nonlibertarian economic institutions to authorize and obey. The hard question is whether there will be any overlap between economic institutions that both libertarians and socialists regard as suboptimal but not defeated. In this chapter, I have argued that strong personal property rights and modest rights of productive property can be publicly justified. But in c­ hapters 5 and 6, I will argue that we must add some welfare state institutions to a liberal political order. This will help to show that property rights are eligible even for socialists, since they can be “bought off ” with the welfare state.

7.  Markets Create Trust for the Right Reasons We have seen both that markets create social and political trust and that the private property rights that give rise to markets are publicly justified. For this reason, we are on good ground in maintaining that markets create trust for the right reasons. However, a fully free-​market economy undermines trust because it will sow distrust among noncapitalist members of the public. We can now turn to analyze the case for the welfare state. We will see that certain kinds of economic security, like universal welfare programs, can increase real social and political trust. We will also see that, because maximally extensive property rights cannot be publicly justified and welfare rights often must be protected through redistributive antipoverty policies, welfare state institutions can be publicly justified.

5

The Welfare State The market economy plays a central role in establishing trust for the right reasons. However, a completely free-​market economy will be rejected by nonlibertarian members of the public based on their sense that some regulations and redistributive policies are necessary to improve well-​being and deliver economic justice. And as we will see, welfare state measures meant to guarantee economic security also have trust-​generating properties in the real world. So certain aspects of the welfare state should promote trust for the right reasons. In this chapter, I first argue that welfare state institutions of some varieties help to sustain social and political trust in the real world. I begin by defining key elements of welfare states and sketching their basic justification (section 1). I then discuss the literature on how the welfare state creates real trust (section 2) and extend the discussion to the empirical relationship between the welfare state and corruption (section 3). I then argue that central features of the welfare state are publicly justified, forming part of the justification for social and political trust. I show that a principle of social insurance is publicly justified, and that it requires polities to have basic social programs (section 4). I then argue that states should sometimes regulate the market process and produce public goods but that this the case for state intervention is complicated by the standards of evidence required to judge the effects of these policies, which I call policy epistemology (sections 5–​6). In the same sections, I also explore how to individuate policies for public justification. Thereafter, I address concerns about rent-​seeking in welfare state bureaucracies (section 7). Finally, I show that two social welfare programs can probably be publicly justified: Head Start in the United States and Bolsa Família in Brazil (section 8). I address economic inequality in c­ hapter 6.

1.  The Welfare State and Its Basic Justification Welfare state capitalism combines private property rights in land, labor, and capital with an activist government that engages in economic regulation, countercyclical policy, and the provision of social insurance. While there is no one thing referred to by “the welfare state,” we can understand it in terms of these broad institutional aims. Rawls argued that just governments will have four kinds of

140  Trust in a Polarized Age bureaucracies:  allocation, stabilization, transfer, and distribution.1 Allocative bureaucracies ensure that markets stay competitive; stabilization bureaucracies reduce the extremes of business cycles and bring about full employment; transfer branches provide for social insurance; and distribution agencies achieve distributive justice beyond the provision of social insurance. Welfare states typically have versions of the first three kinds of bureaucracy. They regulate markets to keep them competitive, engage in countercyclical policy to stabilize the economy, and provide social insurance. It is not essential to the welfare state that it redistribute for the sake of an egalitarian conception of social justice. In terms of particular rights, welfare state capitalism qualifies full private property rights in the following ways: The welfare state limits the right of persons to exclude the government from using their property, including their assets in productive property. It restricts rights to modify and use property in order to internalize negative externalities. The welfare state limits rights to income in order to impose redistributive taxation and regulations. In brief, welfare states redistribute wealth and control certain kinds of economic practices, both of which are departures from pure capitalism and full private-​property rights. We can now see the public reason case for the welfare state in broad outline. The welfare state is justified because full private-​property rights are defeated, particular right by particular right. Thus, when persons claim that they authoritatively possess these particular rights, others have sufficient intelligible reason to reject those authority claims. The public justification of the welfare state thus depends on being able to publicly justify legal and constitutional rules that systematize these restrictions on purely capitalist property rights and that use these restrictions in order to redistribute wealth, control economic activity, and the like. While neither libertarians or egalitarians will regard the welfare state as morally optimal, in order to sustain social trust and social cooperation, they should both find some elements of markets and the welfare state as morally suboptimal and so authoritative for them. But let’s focus first on the empirical connections between the welfare state and social and political trust.

2.  Real Trust and the Welfare State My review of the empirical literature on trust and the welfare state provides limited evidence that welfare state institutions have a positive effect on social trust. Some policies, such as limiting economic inequalities believed to signal



1

Rawls 1999, pp. 245–​8.

The Welfare State  141 unfairness, can help promote trust. Further, better-​educated citizens can make better risk assessments about whether others are trustworthy. But the data are limited. On the other hand, there is good evidence that the welfare state helps create political trust. Nonetheless, the causal connection remains unclear. But the welfare state has significant effects on the main determinants of political trust, namely (i) economic performance and (ii) public corruption, and on the basis of those effects, I argue that the welfare state plays a critical role in sustaining political trust. First, insofar as welfare state institutions help to spread the gains from economic growth across the population, welfare state programs should increase the perceived economic well-​being of the large majority of society to give as many people as possible a sense that their economic well-​being is increasing with time. Effective social insurance seems to increase political trust by increasing one’s economic security: “The more positively an individual estimates the state of social protection in the country, the higher the satisfaction with democracy.”2 This is consistent with the finding that political trust falls during recessions, when economic security is reduced. Economic security, then, should support political trust because citizens judge that political institutions are performing their expected public functions of increasing public prosperity.3 It is true that some of the rich who pay higher taxes may resent redistributive programs, but we have little reason to think that paying higher income taxes is well correlated with lower social or political trust, especially if the taxation is predictable and applied uniformly and the revenues are not seen as wasted. Welfare state programs, if they slow economic growth, can reduce the public’s political trust, but programs that both spread economic growth and plausibly increase it in various ways will be a central practice in publicly justified social orders. Second, effectively run welfare state programs should reduce the public perception of corruption in government. Progressive and centrist members of the public closely tie inequality and corruption on the premise that the social choice to protect property rights that allow for large economic inequalities is a moral failing. Furthermore, a great deal of political corruption may be generated by the fact that in highly economically unequal societies, the rich have relatively more

2 Kumlin and Haugsgjerd 2017, p. 289. 3 There is a complication here, however, because many citizens cannot follow changes in policy, much less have any sense of whether those changes in policy are responsible for their economic condition. Citizens are mostly rationally ignorant of such factors, given how hard they are to track. My assumption here is that citizens simply compare their level of economic well-​being, security, and projected increases between the recent past and the present, and then decide to trust on that basis. This judgment will often be inaccurate, but with a trustworthy media and reliance on policy experts, the tracking problem might be resolvable. I thank Christian Bjørnskov for raising this challenge.

142  Trust in a Polarized Age political power than the poor.4 Some will trust less because they see those political inequalities as intrinsically unfair and so as in themselves signs of corruption. Welfare state programs might reduce political trust, however, especially if they are seen as taking money from trustworthy members of the public to give to less trustworthy members of the public or are administered in ways that seem unfair or corrupt. This suggests that the trust-​promoting effects of welfare state programs will depend on the form those programs take. As we noted in ­chapter 2, the effect of welfare state programs on trust depends somewhat on whether the programs are targeted or universal. First, many members of the public are bound to see some welfare state programs as unfair if they are means-​ tested for certain kinds of poverty. For example, Aid to Families with Dependent Children (AFDC) arguably had trust-​reducing effects for those who believed they were providing aid without receiving aid themselves. Food stamps in the United States raise these concerns. However, as Staffan Kumlin has argued, the policy aims of these programs can be addressed through universal welfare state programs, that is, programs whose benefits goes to most members of society based on simple, largely non-​means-​tested criteria. These programs are not seen as directed at helping marginalized groups that are regarded as untrustworthy. Universal welfare programs are therefore correlated with higher social trust than welfare programs that target the poor; under universal programs, the needy are not singled out for stigma and “there are fewer incentives and opportunities to cheat,” reducing the opportunities and sources of real distrust.5 Universal welfare programs help to avoid these appearances because they’re easier to administer and so have more limited prospects for trust reduction due to fewer opportunities for corruption and bad encounters with the public. This is arguably why welfare state institutions in Nordic countries, where welfare state programs tend to be universal, do not seem to undermine political trust in those countries, and may well improve it.6 Compare American experiences with Social Security and food stamps.7 Social Security is universal, and is seen as benefitting everyone, not just 4 I am skeptical, as we shall see in c­ hapter 6, that increasing redistribution is likely to reduce the political influence of the rich, since I think their influence and riches are due to common factors like having other markers of high social status, but many members of the public see things differently, and public perceptions influence trust. 5 Kumlin, Stadelmann-​Steffen, and Haugsgjerd 2018, p.  399. There is recent counterevidence, however. See Bergh, Jordahl, and Öhrvall 2019, p. 1. 6 We do not know whether it matters if universal welfare supports take certain forms. For instance, are cash distributions superior to in-​kind distributions, and are private assets superior to public ones? 7 One important factor not assessed in the empirical literature is the connection between social and political trust, social insurance, and the public rationale given for social insurance. Nancy Fraser has distinguished between three types of entitlements in the welfare state:  those based on need, programs based on desert, and programs based on citizenship in general. Head Start is based on need, Social Security is often thought to be based on desert (people work and pay into it, and then receive benefits when they retire), whereas public education may be seen as required by citizenship.

The Welfare State  143 the poor and marginalized. Food stamps are seen often as benefiting only the poor and marginalized, and many people see that form of redistribution as unfair because they think some of the poor are undeserving. Social Security is also easier to administer well, and so people seldom interact with Social Security officials in comparison with food stamp programs, which require means testing. Moreover, the benefits of Social Security are more private, as people receive their benefits through checks mailed to them or directly deposited, whereas food stamps have public markers as sources of potential resentment, such as the grocery checkout line.8 It is almost unheard of to complain that some receive excessive Social Security payments, but complaints that food stamps are excessive are manifold.9 And it is rare to hear anyone complain about their interactions with the Social Security bureaucracy, but complaints about need-​b ased assistance are much more common. We can see, then, that at least in the United States, two of our best-​ known forms of social insurance may have greatly divergent effects on political trust (and perhaps social trust as well). That means we should probably want to err on the side of universal welfare programs to help the poor and marginalized. Two brief, important caveats. First, universal welfare programs can also be needlessly expensive because they transfer wealth in a circular fashion, taking from the middle class, cycling those transfers through a bureaucracy, and then returning those benefits to the original contributors. That inefficiency may reduce growth and thereby lower political trust. Second, while high-​trust countries are more likely to yield trust-​increasing and trust-​maintaining encounters with public bureaucracies, lower-​trust countries are more likely to generate mistrust when citizens interact with officials. That will undermine political trust further, undermining the capacity of states to run successful social programs and thus yielding further distrust.

But it is plausible that the trust effects of different welfare state policies will depend on what people think justifies them. For instance, since food stamps are based in need, then if people who appear not to be in need receive them, that may reduce trust in government for violating the justification for the policy. That is, the public rationale for the policy may set up the normative expectations behind the policy, which is essential for making trust judgments. However, we lack data on whether public rationales mediate trust judgments in these cases, so I set aside this line of reasoning, with the hope of returning to it in future work. For the original distinction, see Fraser 1994. 8 Though the validity of these complaints is reduced by the fact that they are frequently racialized. For evidence of this association, see Kinder and Kam 2010, pp. 182–​99. 9 This might be because Social Security payments are seen as deserved, whereas food stamps are seen as going to people who are not truly in need. Perhaps the trust-​conducive effects of Social Security would fall if there were, say, a visible public scandal where undeserving persons were receiving Social Security.

144  Trust in a Polarized Age

3.  Corruption and the Welfare State Any attempt to establish social and political trust must address the state’s vulnerability to corruption, behavior where officials use public office for private gain.10 Since public bureaucracies are responsible for implementing the public policy process I endorse in this chapter, we must look at the role that corrupt bureaucracies play in undermining trust. The effect is considerable, such that an essential feature of a trust-​creating welfare state is that its processes are seen as procedurally fair and noncorrupt. Both in his own work and in a survey of the trust literature, You has found strong correlations between “the levels of control of corruption and the levels of social trust at the country level” in democracies.11 Similarly, political trust correlates strongly with corruption in general.12 These correlations remain even after “controlling for regulation of business, fairness of legal system, gross national product (GDP) per capita, ethnic fractionalization, and Freedom House score of democracy.”13 It is not hard to fit the results of You’s findings into our causal story. Corrupt behavior among political officials is plainly a strong marker of untrustworthiness. So, the observation of corruption will reduce trust in those officials, and may even lead them to be less trustworthy in light of observing bad behavior by other officials. In many cases, ordinary citizens have the opportunity to engage in corrupt behavior, and observation of corrupt behavior in one’s fellow citizens will lower social trust levels. It is also plausible that people take the behavior of legal officials as a proxy for how trustworthy people are in general, such that corruption of legal officials may lead individuals to reduce their trust in people in general. Thus, institutions that manage corruption, such as the rule of law, democracy, strong judicial institutions, and the like, will increase social trust, and since these institutions are likely to be publicly justified, then these paradigmatic liberal institutions help to sustain social trust between diverse persons. Differences in the fairness of the political process are able to explain cross-​ national differences in political trust “measured as satisfaction with democracy and confidence in national political institutions.”14 Indeed, “degree of public level corruption” matters a lot, and “The more widespread corrupt practices are, the less citizens trust national political institutions” and the less pleased they

10 Bjørnskov 2011. For an earlier discussion of various measures of corruption and some difficulties with them, see Treisman 2000, p. 399. 11 You 2018, p.  487. Here r  =  .74. You finds that the correlation entirely evaporates in nondemocracies, r = .04. 12 Ibid., p. 484. 13 Ibid., p. 481. 14 Van der Meer and Hakhverdian 2017, p. 98.

The Welfare State  145 are with democracy. In general, observed compliance with political norms by officials, especially in the civil service, will tend to increase political trust of some varieties, as does compliance with shared norms of citizenship.15 In particular, when citizens feel that government officials treat them fairly, they are more likely to be politically trusting.16 Importantly, small acts of corruption, “petty” corruption, don’t decrease political trust much; “grand corruption” that is widely observed by the public does. The main exception is that observing gift payments to police or the courts reduces political trust.17 These effects are especially strong in developed, democratic societies, where economic growth is slower than in developing, authoritarian nations, where evaluations of economic performance are most salient. Political trust is also affected by a winner’s bias, where individuals’ political trust is higher after their favored candidate or party wins an election, especially elections with winner-​ take-​all voting. This effect is arguably most pronounced in the United States. Corruption can also affect political trust by affecting key economic indicators. First, corruption can reduce political trust by reducing the rate of economic growth, since one causal factor in determining political trust levels is perceived economic improvement. This effect could be enormous. Christian Bjørnskov has found that a standard deviation shift has a “corresponding income variation of 12440 USD,” which corresponds to “38 years of economic development” after some controls.18 Second, economic inequality appears to decrease political trust, though largely when the inequality is seen as corrupt. Uslaner argues that “inequality leads people to believe that leaders listen far more to the rich than to others in society” and that, generally, perceived unfairness reduces trust in government.19 This effect can be hard to correct because the rich can navigate corrupt environments more effectively than the poor, which can help to cement or even increase economic inequality. I will discuss inequality and trust further in the next chapter, but insofar as economic inequality is taken as a sign of public corruption, we have further evidence for a connection between corruption and political trust. So, we want a welfare state that is operated in a procedurally fair and non-​ corrupt fashion. Welfare states can vary greatly in their structure from one another, but regardless, we must insist that the welfare state be operated in a procedurally fair and noncorrupt way. And that probably means that we should respect the welfare rights of citizens through forms of social programs that are universal, routinized, and predictable and that leave little room for corruption

15

Van Deth 2017, p. 224. Grimes 2017, p. 256. Ibid., p. 308. 18 Bjørnskov 2010, p. 342. 19 Uslaner 2017, p. 302. 16 17

146  Trust in a Polarized Age and incompetence in their administration. That is, social programs should resemble simple transfer programs that have relatively simple rules and that can benefit everyone. Social Security fits these requirements well; Bolsa Família does too, as we will see. There is a real worry that governments that conduct large-​scale redistribution will invariably be subject to corruption and incompetence, since the power to tax and spend will be co-​opted by special interests, leading to redistribution that cannot be publicly justified. This is surely true sometimes, in particular when it comes to social benefits for small or medium-​sized groups like professional bodies, unions, and the like. But universal welfare programs can sidestep at least some of these worries insofar as we can design a liberal constitutional order that can only engage in redistribution in a universal fashion. One way to do this would be to discourage government from engaging in targeted social programs and instead require that it reallocate through a simple, predictable negative income tax. Again, the more general the structure of the program, the more easily it will establish political stability; in my terms, it will be essential for ensuring that constitutional norms possess the quality of immunity, that is, that they are insulated from manipulation by special interest groups who act primarily from self-​interest. Concerns about corruption are therefore going to ground public choice-​style restraints in bureaucratic and administrative law, the law that governs administrative bodies. This is because regulatory bodies are frequently captured by special interest groups. Insofar as regulations slow economic growth and are subject to corruption through rent-​seeking, the agencies that make them cannot be publicly justified. One source of egregious rent-​seeking is in US agricultural subsidies from the Department of Agriculture. On the other hand, the Social Security Administration’s provision of old-​age pensions could be made constitutionally protected, perhaps with some generic limits on its ability to provide targeted benefits to the disabled to prevent fraud. The closer we can approach a social insurance Rechtsstaat, that is, a regime governed by the rule of law, the more essential the welfare state will be for generating social and political trust in the real world.20 What we want is to avoid large bureaucratic bodies with the power to tinker with political and economic life, as these agencies will be targeted for capture and corruption.21 We should instead entrench certain universal welfare programs in the constitutional order. Doing so will also create greater stability.

20 Hayek 2011, pp. 287–​307. 21 Though we will have to pay close attention to social science on this question; while public choice models predict lots of capture and corruption, administrative bodies might also be sources of policy expertise, and institutional design may allow us to insulate these bodies from capture.

The Welfare State  147 One way to do this would be to replace most social welfare programs, with the possible exception of catastrophic health insurance, with a negative income tax along the lines proposed by Milton Friedman.22 This would reduce bureaucratic waste and negative experiences with government agencies and agents, and it would be universal in structure. The difficulty is that the negative income tax would be seen as redistributing wealth in an unconditional fashion that many would reasonably find unfair. Conservatives and libertarians are entirely reasonable in insisting that people only receive welfare benefits if they are unable to work. But tracking whether someone is able to work is bureaucratically complex and would undermine the streamlining effects of a negative income tax. Thus, the negative income tax may be trust-​increasing in some respects but trust-​reducing in others. The genius of Bolsa Família is to provide largely unconditional cash transfers to the poor by combining assistance with relatively low barriers to qualification in the form of vaccination requirements and school attendance. So, if qualifying for aid under a negative income tax could be somewhat conditional in ways that are hard to fake, that would speak in favor of the tax. Such a policy could be publicly justified, and would help to create social and political trust in the real world. I would like to end the section by identifying a feedback loop between trust and the welfare state. Welfare state institutions of certain kinds can increase political trust and perhaps, to some extent, social trust. But both forms of trust help to preserve the welfare state, perhaps in as many as four ways. High levels of trust prevent free riding on public goods.23 Social trust helps make public bureaucracies efficient because members of high-​trust societies tend to be better civil servants. Social trust also helps to reduce tax cheating, which makes welfare states easier to fund and reduces opportunities for overuse and corruption of tax authorities. High trust can also lead to better initial agreements on welfare policies and their implementation, so high trust can avoid costly misconduct investigations and renegotiations. And when welfare states have these features, as we have seen, they should increase the political trust of people who interact with the welfare state, so we have a positive feedback loop. We can therefore expect that some social programs can help interrupt the distrust-​divergence process.

4.  The Principle of Social Insurance We have seen that central welfare state institutions promote trust. Now I must show that they do so for the right reasons. So I begin by establishing the public

22 23

Friedman 2002. Bjørnskov 2019, p. 639.

148  Trust in a Polarized Age justification of a principle of social insurance that can be used to support a number of different kinds of social programs.

4.1.  The Public Justification of the Principle of Social Insurance* As I have argued, progressive and egalitarian defeater reasons do not undermine the public justification of private property. Private property is already a primary right and, on many issues, we individuate policy proposals finely, so any proposed defeat of private property rights must often proceed piecemeal, particular right by particular right. Further, progressive and egalitarian defeaters do not license state intervention to redistribute wealth in general. The redistribution of wealth, as a recurrent social process, is coercive in many respects, so it must itself be publicly justified as an ongoing, official government procedure. But many conservative and libertarian members of the public will have their own defeater reasons for some, but not all, redistributive policies, which may be sufficient to defeat the case for a permanent redistributive governmental power. In particular, more radical forms of redistribution, say those justified by luck egalitarian or other egalitarian conceptions of justice, will not be publicly justified for many members of the public. If a policy can only be justified on luck egalitarian grounds, then since even some thoughtful egalitarians reject luck egalitarianism (perhaps because luck egalitarianism is thought to require leveling down), the policy cannot be justified.24 In this way, luck egalitarian reasons function a bit like religious reasons in public justification: they can defeat coercion, but they cannot justify coercion on their own.25 Other reasons for redistribution are not so sectarian, however, such as the widespread belief that government provides some essential services that cannot be adequately provided by the market process and civil society. Some goods, like roads, defense, and police services, are nigh universally regarded as good, and taxation for those purposes is widely regarded as publicly justified. The only people who disagree are anarchists and radical libertarians.26 And their numbers are small enough that they can be otherwise accommodated.27 24 Nozick 1974, p. 229; Raz 1986, pp. 227, 235. For a classic egalitarian critique of luck egalitarianism, see Anderson 1999. 25 As I argue in Vallier 2014a, pp. 196–​216. 26 I think that the radical libertarian is unreasonable if she insists that the only acceptable scheme of social institutions is the one she advocates, but she is reasonable if she insists on the right to secede or partially separate from her state in conjunction with other libertarians. I discuss these matters in appendix B. 27 In particular, I would argue that libertarians deserve a kind of global exemption from various coercive state policies on the condition that they give up some government benefits, much as the

The Welfare State  149 I now argue that basically all members of the public, save for a very small number of libertarians, will adopt what I  shall call a principle of social insurance. I understand social insurance as a type of public policy where government provides income and resources to people who, for one reason or another, lack access to goods and services vital for them to remain normally functioning persons and citizens. Varieties of social insurance include tax-​financed health insurance, unemployment insurance, disability insurance, old-​age pensions, and food stamps. These policies are typically justified based on the widely recognized badness of not having access to food, healthcare, shelter, and other resources necessary to meet one’s basic needs, as well as the badness of fear about one’s ability to secure those goods. The point of social insurance is to insure persons against the loss of these goods in order to both directly promote their welfare and reduce the fear associated with the loss of such goods. Many members of the public will also recognize that social insurance is required to alleviate injustice against the poor and needy. Because the poor and needy have welfare rights—​rights to goods and services required for physical, mental, and social functioning—​social insurance may be needed to protect those rights. Thus, the principle of social insurance states that members of the public will be generally prepared to endorse policies meant to secure certain goods for all persons, especially the socially and economically marginalized, such that a general test for any group of public policies is the degree to which they put the well-​ being of the least advantaged at risk.

4.2.  Objections to the Principle of Social Insurance Critics of the principle of social insurance typically offer four objections to state provision, all of which I think can be answered. Answering the objections helps to explain why the principle of social insurance is a basis for evaluating the public justification of laws and policies, since the main reasons to reject it are only reasons to qualify it rather than to reject it entirely. The four objections claim that social insurance (i) goes to the undeserving too often, (ii) generates economic inefficiency, (iii) crowds out private charity, and (iv) is excessively coercive. (i) Many complain that recipients of social insurance are often undeserving, either because their need is their fault or because they are not in need in the first

Amish have a broad exemption from public schooling after a certain grade level and from Social Security.

150  Trust in a Polarized Age place.28 Importantly, this is not an objection to the principle of social insurance itself, but rather implies that the principle is hard to institutionalize because cheaters are hard to detect. Even so, worries about abuse are insufficient to justify ending social insurance programs in general. The government must simply be pushed to be more scrupulous and efficient, even if that is difficult.29 Some complain, along similar lines, that social insurance incentivizes sloth, which is considered an immoral use of taxpayer funds, but the same reply is appropriate here—​the government should try to take steps to avoid incentivizing sloth.30 Those who genuinely need social insurance have morally urgent needs that can be easily met, such that taking away their social insurance cannot be justified based on the potential sloth of other recipients of government funds. We cannot deprive the genuinely needy to avoid rewarding the slothful, unless the slothful overwhelmingly outnumber the deserving poor. (ii) Some claim that social insurance is economically inefficient. Social insurance reduces the cost of making bad economic choices, such as not keeping one’s job or not saving enough money for retirement. If we reward poor decisions, we can expect more poor decisions in the future. 31 Similarly, the relevant forms of social insurance are funded by taxing the rich, who work and produce significant economic value. Taxation often discourages the rich from being economically productive. If we discourage the productive and reward the unproductive, we can expect a slowdown in economic growth, which, when growth rates are compounded, can reduce economic prosperity over time. This objection is more serious than complaints about deservingness for several reasons. First, it does not appeal to controversial moral ideas like the undeservingness of the poor, but rather makes a straightforward, verifiable empirical claim about a country’s economic well-​being, which everyone has reason to care about. Second, the model of human behavior at work is relatively simple and easy for members of the public to grasp: people respond to incentives in the long run such that if you pay people for something, you get more of it (in this case, poverty) and that if you reduce the pay people get for something, you get less of it (in this case, economic production). 28 Goodin and Schmidtz 1998, pp. 10–​2, reviews the question of blame and its importance. 29 In my empirical view, the difficulty is partly a function of social trust. More socially trusting orders are better providers of noncorrupt social services. However, there are other difficulties in pressing government to be more effective, such as universal programs leading to undeserving recipients because at least some recipients will be seen as undeserving, or that the program is badly targeted to the undeserving rather than the deserving, or that some people are legally entitled to obtain the relevant benefits without officials knowing this. 30 For a review of the empirical literature on cash transfers and the behavior of recipients, see Banerjee et al. 2017. 31 Goodin and Schmidtz 1998, pp. 14–​20.

The Welfare State  151 Concerns about efficiency can justify limits on social insurance. At some level of taxation, and some amount of social insurance, economic growth will slow, making many worse off in the long run, including even the recipients of social insurance. And the income and wealth available to them through social programs will be smaller than it would have been following a period of greater growth. Such a large amount of taxation and redistribution is defeated due to the principle of sustainable improvements. However, reasonable people disagree about which level of taxation and redistribution is ruled out, so the legal and constitutional orders should settle upon the appropriate balance within the bounds set by other primary rights. (iii) Critics of social insurance sometimes argue that government provision of social services and the regulation of service provision crowds out nongovernmental groups from providing those goods and services.32 When government gives money to the poor directly, some nongovernmental charities are crowded out because the government is doing their job for them. Similarly, private employers are crowded out of providing jobs to the unemployed if the government provides a public jobs guarantee.33 Defenders of social insurance will counter that government provision is both more effective and just and that crowding out private institutions is not a problem so long as the needy are cared for. But recall that institutions are not automatically under the authority of government in such issues: if institutions within the moral order can satisfy the demands that the poor be cared for to an adequate degree, then the coercion involved in government provision probably cannot be publicly justified. This implies a presumption on behalf of civil society in providing for the needs of the poor and protecting economic justice. The presumption can only be overcome by empirical data that strongly support the greater effectiveness34 of government-​provided services.35 That said, members of the public who justifiably believe that government services will be less effective than private organizations cannot defeat government-​ provided services based on these concerns alone. These individuals presumably think that some government provision is better than none and that the coercion

32 This is part of the argument in Beito 2000, which documents the robust fraternal order system that once alleviated poverty but that was gradually replaced by the welfare state. 33 Abrams and Schmidtz 1984 and Bolton and Elena 1998 find significant crowding-​out effects. For counterevidence, see Boberg-​Fazlić and Sharp 2013. 34 Here we can understand greater effectiveness in terms of the efficient improvement standard for intervening in the legal state of nature, which I discussed in MPBW, chap. 4. But to roughly review it, the policy will need to be an improvement on civil society, either because civil society does not provide the relevant good or because government intervention can considerably improve on what provision there is. 35 In my view, there are works that meet this standard of evidence. See Kenworthy 2013.

152  Trust in a Polarized Age involved in government intervention, while unfortunate, is morally suboptimal rather than defeated.36 So even if members of the public reasonably wonder whether government provision is a necessary supplement to civil society, some social insurance will be justifiable for even many conservative and libertarian members of the public, so government provision can still be publicly justified. It is reasonable to worry that crowding out is significant and that many forms of government-​provided social insurance are inefficient, particularly in large, diverse social orders like the United States with a great deal of bureaucratic waste,37 but this should not bar redistribution in general. (iv) Finally, many will complain that the taxation required to provide social insurance is excessively coercive. I take it as obvious that taxation is typically, if not always, coercive. In cases where people want to build their lives around having access to a high level of income and wealth, government expropriation can prove quite coercive indeed.38 Further, higher tax rates will tend to be more coercive because they will reduce the options some have to spend their own money. Progressives and egalitarians are likely to respond that a world in which people tolerate high taxes on the rich would increase the choice options of sufficiently many people that the tax rate could be justified. But plenty of sincere, informed people of good will not only reject this claim, but think it is obviously false. They will have sufficient reason to reject the progressive economic forecast, such that some rich persons who reasonably reject the claim will experience the extraction of their income as coercive and alienating, and so these taxes will probably not publicly justified. Progressives and egalitarians will likely offer two replies. First, they might argue that, while imposing taxes coerces the rich, the state “interacts no less coercively with [citizens] when it uses the threat of legal punishment to protect private property.”39 As we will see later, however, the claim that protecting property is as coercive as redistributive activities is implausible. For this implies that all economic regimes are equally coercive, as though Denmark’s market economy is just as coercive as Soviet Russia’s planned economy was. Second, progressives and egalitarians might argue that the revenue collected could be used to provide more options for recipients of the revenue, so redistribution may reduce coercion overall. But public reason is not in the business of 36 Again, there will be some holdouts, like radical libertarians, but these groups are likely small in number and may be able to be accommodated in another way. See appendix B. 37 See Tanner 1996 for a discussion of AFDC, a good candidate for an ineffective and counterproductive program. 38 Gaus 2011, p. 522. 39 Bou-​Habib 2015, p. 658.

The Welfare State  153 maximizing free choice; it insists rather on respecting persons. Thus, it insists that coercion be publicly justified, even if using unjustified coercion would minimize coercion on net. My claim that taxation is coercive does not commit me to a natural rights approach to property rights or to preconventional property rights. Instead, I’m simply combining a basic primary right to private property and my model of public justification to defend my conclusion. The state is not necessary in order for society to have property rules that will protect income and wealth from expropriation, at least up to a certain level of economic holdings. Many members of the public think that social insurance can be adequately provided without high levels of taxation and that high levels of taxation are inefficient and excessively limit their range of choices. Consequently, they have sufficient reason to reject extensive social insurance, but not the principle of social insurance itself. Once again, the principle of social insurance, properly hedged, will serve as a basis for evaluating the justifiability of particular laws and policies. I conclude that a publicly justified polity provides various forms of social insurance for its citizens, and that doing so will be essential in creating trust for the right reasons. However, the extent of social insurance will be qualified by concerns about deservingness, inefficiency, crowding out, and coercion.

5.  Justifying Regulations and the Provision of Public Goods* Any polity that combines robust capitalist property rights with social insurance is a kind of welfare state. But welfare states also include a regulatory apparatus and bodies that provide for public goods. So, we must now determine the extent to which a regulatory apparatus and the provision of public goods can be publicly justified.40 One hurdle in determining the appropriate limits of the regulatory state is that people will reasonably disagree about whether some regulations are increases or reductions in coercion. Libertarian members of the public will have strong objections to a variety of regulations, given that they view such regulations as coercive impositions. But progressives and egalitarians will reject coercive enforcement of extensive private-​property claims, so they may support regulations that are protections against coercion, such as regulations protecting workers from control by their bosses. Fortunately, most regulations do not raise these complications: they’re plainly coercive, and some think the coercion is justified.



40

Here I expand upon Gaus 2011, pp. 509–​44.

154  Trust in a Polarized Age So, in this section, I focus on regulations that most can agree are increases in coercion.41 The simplest public justification for a coercive regulation is that the regulation prevents a harm, given that moderately idealized members of the public recognize the wrong in harming others.42 (I understand a harm as a setback to another person’s interests or to the interests of an association or the public as a whole.)43 For this reason, some regulations required to protect people from workplace harms, like health and safety regulations, should be publicly justifiable. Importantly, we must have reason to think that the regulation prevents the relevant harm without imposing a greater harm, but there are regulations that plausibly meet this requirement.44 There is an important barrier to justifying regulations to prevent harms, however. In many cases, the nonstate legal system can constrain or limit harms that regulations are meant to prevent. They do so through the tort law, the law of damages. The tortfeasor, the one who commits a tort, is frequently forced to compensate the person whom she has harmed. These private legal rules can perform the absolutely vital function of resolving property disputes. No legislative apparatus is required for the existence and functioning of at least a fairly simple and primitive tort-​law system.45 Since people with a financial interest at stake will often recognize, at least in vague terms, that they are liable to civil suit for harming others, this threat may prove enough to discourage them from creating the harm or damage in the first place. In this case, then, state regulation is not needed. And in many cases, torts are a superior method of addressing damages, as torts can be more responsive to particular cases than general regulations imposed by a central bureaucracy. Of course, in other cases a regulatory solution may be superior, especially when it must be enforced by the state against a large economic power that has the ability to control or distort the tort-​law process. It is true that these large economic powers can distort the regulatory process as well. But if it can be demonstrated that the regulation is more likely to protect people from the relevant harms or damages, despite the threat of regulatory capture (where the regulated group comes to control the regulatory body), then state regulations can be 41 I set aside cases where the coercive regulation will indirectly bring about the reduction of coercion in some other part of the economic system, such as carbon taxes that reduce the rise of sea levels that would impinge upon the land of many property holders. These indirect effects can be built into the case for the public justification of the regulation. 42 I discuss harm and public justification briefly in MPBW, pp. 202–​3. 43 Ibid. 44 For an argument that purports to justify carbon mitigation to conservatives, see https://​ niskanencenter.org/​blog/​new-​study-​the-​conservative-​case-​for-​a-​carbon-​tax/​. 45 For a classic discussion of the origins of the common law prior to the existence of powerful states, see Hayek 1973.

The Welfare State  155 publicly justified. Though as we shall see, it is a further question whether a regulatory body can be publicly justified, since it may output both justified and unjustified regulations. Public goods, recall, are goods whose consumption is nonrival and nonexcludable: the fact that one person can consume the good does not prevent others from doing so, and one person’s consumption does not undermine another person’s ability to consume the good. Public-​good provision is easier to publicly justify than regulation because public goods are not necessarily coercive: the only coercion typically required is the coercive financing of the good. Moreover, I already acknowledge the productive function of the state as an essential part of the constitutional order.46 Nonetheless, public-​goods provision faces a barrier of its own, since public goods can be provided by civil society and the market economy, at least in part. Moreover, government might be able to construct markets that can provide public goods instead of providing them directly. For instance, education up to a certain level is often thought to be a public good, but government can provide the good by providing taxpayer-​funded vouchers for use in private schools.47

6.  Policy Individuation and Policy Epistemology Publicly justifying regulations and public goods provision faces another hurdle. For we can only publicly justify coercion if members of the public have some way to convince one another that the policy in question will have certain effects and that the benefits of the policy will exceed the costs associated with lost opportunities for choice. I think this is arguably an implicit part of what Rawls called the “guidelines of inquiry,” and I will call this part policy epistemology.48 A policy epistemology includes a set of standards that specify the level of evidence required to show that a public policy will have the results the person arguing for or against that policy claims it will have. In particular, policy epistemology specifies the level of evidence that should convince any (moderately idealized) member of the public to assign relatively high probabilities to predicted effects of the policy. For instance, policy epistemology might establish standards of evidence that would rationally motivate some members of the public to assign a probability of .8 that a proposed policy will produce a 0.2 percent increase in the growth rate, but that

46 MPBW, p. 185. 47 There are other important hurdles as well. For a remarkably helpful outline of the many conditions required to justify government provision of public goods, see Anomaly 2015. 48 Rawls 2005, pp. 223–​4.

156  Trust in a Polarized Age would give other members reason to adopt a .55 probability that the proposed policy will produce a 0.2 percent increase in the growth rate.49

6.1.  Policy Individuation* A second component of policy epistemology is that it sets how members of the public individuate, or distinguish between, different laws and policies as subjects of justification. We can ask, for instance, whether two legislative provisions should be subject to public justification alone or in conjunction with one another. When, in other words, should members of the public treat multiple legislative, legal, or governmental actions as one issue, and when as separate issues? This is the question of individuation in public reason, which I have addressed in MPBW. There I argued that we should individuate finely, that is, at the level of laws and policies. What ultimately must be justified are particular laws and policies; principles and rights are adopted by members of the public in order to evaluate those particular laws and policies. Since MPBW was published, however, I have had the opportunity to develop a more detailed criterion of individuation with Colin Manning. In a coauthored essay, we raised problems for the criteria of individuation in the public reason literature and then offered our own criterion as a solution.50 Readers who want a review of the literature and an opportunity to assess our arguments against other criteria are referred to our article. Here I will merely reproduce the criterion that we developed. Manning and I  think members of the public would individuate policies in terms of what we call functional independence, which we define as follows: Functional Independence:  law x is functionally independent from law y for (moderately idealized) member of the public P if and only if (1) P rationally believes that x will achieve at least one of her highly ranked goals [g1 . . . gn] with or without y based on (2) her model51 of the effects of x and y.

For a law to be functionally independent from another law, a member of the public must determine whether the law can satisfy at least one of her goals, or more, without y. That is, she must determine that the law can, by itself, realize one of her political or moral aims. She does this by appealing to an implicit or explicit 49 Note that policy epistemology recognizes that the same evidence will not convince everyone that some policy will have predicted results, at least not to the same degree. This is in part because people have different prior probability assignments. 50 Manning and Vallier 2021. 51 Or the model held by one she regards as a reliable testifier.

The Welfare State  157 model of how two or more laws affect the world, and, using that model, whether she thinks the two laws have interactive effects or not. It is true that the laws will interact in many cases, at least obliquely, but as long as her model of the effects of the laws predicts that a law can satisfy her goals independently from another law, then that is sufficient reason to individuate them.52 The foregoing allows us to formulate a general criterion of individuation: General Functional Independence: law x is functionally independent from law y if and only if (1) each member of the public believes that x will achieve at least one of her highly ranked goals [g1 . . . gn] with or without y based on (2) her model of the effects of x and y.

This, then, is how we should individuate issues for the moderately idealized public as a whole. We acknowledge in the essay that it will often be difficult, if not impossible, to determine how to individuate issues based on these considerations, especially because members will have different models of effects. But in cases where the need to individuate is unclear, the principle of general functional independence should be sufficiently action-​guiding to be of use; for at least we can ask persons whether they see the law in question as achieving their goals and what they think the effects of the law will be. Manning and I then argue that public reason liberals should individuate down to functionally independent legal or policy units, that is, the smallest functionally independent units. We individuate finely on the ground that members of the public will often have refined, specific opinions about certain laws and policies and that allowing them to assess these laws and policies one unit at a time enables an overall greater satisfaction of their evaluative standards than insisting that they assess a range of laws and policies as a whole. By analogy, it is often better to allow legislators to vote on particular bills than to require that they vote for a raft of legislation. But that is only an analogy. A unit may be smaller than a bill. It may even be smaller than a single legal rule, since some members of the public may favor distinctions between cases that some rules would treat as the same. For example, just because capitalist laws might cover landholdings of all sizes under the same rules, with no special provision entitling people to establish large holdings, it does not follow that these large holdings are not covered by a different, yet still functionally independent policy unit than smaller holdings.

52 A policy might undermine another one of her goals, in which case we would have to take that further risk into account. But the advantage of speaking in terms of “highly ranked” goals is that, in virtue being highly ranked, they are likely to outrank the goals of avoiding certain kinds of losses.

158  Trust in a Polarized Age The criterion of general functional independence is part of policy epistemology. In this way, policy epistemology contains (i) a criterion specifying the level or grain of individuation of laws and policies and (ii) a criterion specifying the level of evidence required to convince diverse members of the public of the effects of those laws and policies.53 And in fact, the latter is in some ways a component of the former, since individuation itself requires a prediction about the effects of laws and policies, which requires an account of how evidence works in these cases. I cannot say much more here for reasons of space, but throughout the rest of the book, I will appeal to the language of functionally independent policy units. If a law or policy is a functionally independent policy unit, then we assess its public justification independently from other laws and policies.54

6.2. Policy Epistemology Policy epistemology is critical for determining which public policies can be publicly justified, especially in the case of coercive regulations. Part of showing that persons have reason to submit to coercion is demonstrating that the coercion in question is an improvement according to each person’s reflective perspective. This will require, at a minimum, some serious public method of counting the economic costs and benefits of various regulations and policies. Otherwise, only those who think that a proposed policy realizes moral values that override economic well-​being will have sufficient reason to endorse it as if it comes at an economic cost.55 While I lack the space to develop a satisfactory account of the standards of policy epistemology, the idea is sufficiently intuitive that it can help us to evaluate the standards of evidence that American politicians and bureaucrats often have available. However, I am worried that the models we use to explain and predict economic phenomena are more limited than is commonly understood, and so I am not sure the empirical case for most public policies is epistemically sound. I am also concerned about my inability to articulate all the different dimensions of a complete policy epistemology, so here I will focus entirely on how to predict whether a policy will have certain purely economic costs and benefits. In the end, 53 It will also include a specification of the role of political testimony in conveying information, which I cannot address here, but see van Wietmarschen 2019. 54 One benefit of the criterion of functional independence is that it can answer Andrew Lister’s concerns about finding the right level of individuation in Lister 2013, chap. 4. Manning and I have a more detailed reply in Manning and Vallier 2021. And I have already addressed Lister’s argument for decoupling public reason and classical liberal institutions in Vallier 2014b. 55 Perhaps not even then, since whether a policy realizes certain moral values will have an empirical dimension.

The Welfare State  159 that will not be the only consideration in whether a policy’s ratio of costs and benefits can be predicted, since not all costs and benefits are purely economic. And indeed, noneconomic costs and benefits may be of such significance that they outweigh the monetary implications, such as the significance of considerations of the justice of a policy. On top of this, even the basic monetary cost-​ benefit analysis is going to be plagued with difficulties of its own, as I will discuss further.56 One reason to think that policy epistemology will vindicate few regulations draws on an interview with Ronald Coase, a Nobel laureate in economics and one of the founders of the law-​and-​economics movement. He reported the following about the costs and benefits of regulations: When I was editor of The Journal of Law and Economics, we published a whole series of studies of regulation and its effects. Almost all the studies—​perhaps all the studies—​suggested that the results of regulation had been bad, that the prices were higher, that the product was worse adapted to the needs of consumers, than it otherwise would have been. I was not willing to accept the view that all regulation was bound to produce these results. Therefore, what was my explanation for the results we had? I argued that the most probable explanation was that the government now operates on such a massive scale that it had reached the stage of what economists call negative marginal returns. Anything additional it does, it messes up. But that doesn’t mean that if we reduce the size of government considerably, we wouldn’t find then that there were some activities it did well. Until we reduce the size of government, we won’t know what they are.57

In other words, Coase believes that some regulations must be beneficial, but in his attempt to summarize decades of research, he cannot recall a single instance where a regulation passed even the simplest cost-​benefit test. Perhaps Coase is biased, but it would take a remarkable level of bias to lead him to claim falsely that he cannot recall a single case of a regulation passing such tests. Whether there was any such case could be easily checked. And if there are none, that is remarkable. If Coase’s assessment of the evidence is correct, that will leave at least many conservative and libertarian members of the public deeply skeptical

56 One very simple difficulty is that all cost-​benefit analyses are partial equilibrium models, demonstrating an equilibrium between different causal factors in, say, a single sector of the economy, whereas predicting effects requires a general equilibrium model, looking at the causal interactions across the economy. This is related to the problem of complexity I grapple with subsequently. I thank Brian Kogelmann for this point. 57 https://​reason.com/​archives/​1997/​01/​01/​looking-​for-​results/​1.

160  Trust in a Polarized Age of these regulations, casting doubt on the claim that coercive regulations can be publicly justified. Let me offer another example. Consider organizations that try to determine the budgetary impact of various pieces of legislation, like the Congressional Budget Office (CBO). The CBO fully admits that “it is often difficult or impossible to determine, even in retrospect, the incremental impact on the budget of a particular piece of legislation.”58 Their admission is remarkable, so I reproduce the agency’s explanation in some detail: CBO regularly prepares cost estimates for legislation when bills are reported by committees of the House of Representatives or the Senate. In some cases, such legislation is changed before enactment. Although CBO often provides updated cost estimates (especially for direct spending provisions) prior to the enactment of legislation, proposals are sometimes amended after cost estimates are prepared. Moreover, in many cases the actual costs or savings resulting from enacting legislation cannot be identified; they may be a small part of a large budget account or revenue stream, and there may be no way to know for certain what would have happened if the legislation was not enacted. In fact, most of the cost estimates that CBO completes are for legislative proposals that are not enacted, so it is not possible to determine their accuracy.59

Anyone worried about the coerciveness of legislation ought to find this admission troubling. The CBO is not usually able to demonstrate the cost-​effectiveness of public policy even after the legislation passes. In many cases, it is, in the CBO’s own words, impossible to determine the impact of the legislation the agency assesses. There is no plausible policy epistemology where this admission does not weaken the public justification of many proposed public policy provisions. However, sometimes legislation appears subject to successful monitoring. The example CBO uses is Medicare Part D, which it originally estimated to cost $552 billion in 2003 but which has turned out to cost $358 billion. The original estimate assumed that the growth rates for drug spending would remain close to the long-​term trend from 2003 to 2013, but growth rates did not follow the trend. So, one of the CBO’s best, if self-​selected, cases is one where the agency plainly overestimated the cost. The CBO has recently touted its low forecasting error for general economic trends as equivalent to administration and private sector consensus, saying its largest errors are the same as the large errors of other forecasters. But its forecasts are only for general economic trends, and not for the costs and benefits of social

58 59

https://​www.cbo.gov/​publication/​44017. https://​www.cbo.gov/​publication/​44017.

The Welfare State  161 programs. Developing general forecasts is typically much easier than determining the specifics of the effects of policies. For determining overall economic numbers does not require specific knowledge of particular legal conditions but of overall aggregate measures. The CBO’s admission is bad news for defenders of an extensive administrative apparatus. Public reason liberalism, again, requires a policy epistemology where the costs and benefits of major legislation can be generally assessed as economically efficient in the rough and inadequate sense of the program’s monetary benefits exceeding its monetary costs. While the principle of social insurance can be publicly justified, many forms of social insurance will not meet the standards of policy epistemology, since groups like the CBO cannot demonstrate to members of the public that the legislation creating or revising some form of social insurance will be worthwhile. It is even harder to justify regulations and policies that go beyond the provision of social insurance. Policymaking must also face up to the limitations of expert prediction in general. Recent work on the ability of experts to predict social outcomes has shown that their abilities are quite limited. Philip Tetlock has found that less ideological and open-​minded experts, those with diverse and eclectic cognitive styles, are not much better than random predictors and fall behind basic statistical models, though these “foxes” are better predictors than the “hedgehogs” who stick to one theoretical approach.60 Expert political judgment isn’t hopeless, but the best empirical study of expert ability is remarkably humbling. This means that even the most honest attempts to predict the outcomes of public policy by consulting expert opinion are bound to meet with limited success. So, meeting even the most basic standards of policy epistemology is going to be difficult given the limitations of expert prediction when statistical models are inadequate or unavailable. To illustrate, I will use the rough standards of policy epistemology to examine President Obama’s 2009 stimulus package, which was a salient, relatively recent countercyclical policy. Determining whether the Obama administration’s 2009 stimulus package was effective requires counterfactual determinations that are extremely difficult to make. Paul Krugman influentially argues that our prolonged economic difficulties in the United States are due in part to Obama’s stimulus package being too small.61 But many others think the package was just right, too large, or wholly ineffective.62 While they do not represent anything near an academic consensus, more than two hundred academic economists reject the view that “more government spending is a way to improve economic 60 Tetlock 2006, pp. 73–​4. 61 Krugman 2012, p. 118. Also see Krugman’s explanation of his old-​Keynesian approach: http://​ krugman.blogs.nytimes.com/​2011/​10/​09/​is-​lmentary/​. 62 For a range of opinion at the time, see the roundup in the Wall Street Journal: http://​blogs.wsj. com/​economics/​2009/​01/​29/​economists-​debate-​diverse-​perspectives-​on-​stimulus/​.

162  Trust in a Polarized Age performance.” Policymakers, these economists say, should instead “focus on reforms that remove impediments to work, saving, investment and production. Lower tax rates and a reduction in the burden of government are the best ways of using fiscal policy to boost growth.”63 Reasonable dissensus among expert economists suggests that it is hard to determine whether the stimulus package was worthwhile, which in turn suggests that the evidence favoring the stimulus package does not meet the standards of policy epistemology. To be more specific, during the Great Recession, economists in constant dialogue with one another in the economics blogosphere could not agree on the size of the fiscal multiplier, which determines how much economic activity government spending produces. Some thought the multiplier at that time was under 1, such that each dollar of government spending would stimulate less than a dollar of private spending, whereas others thought the multiplier was greater than 1, such that for each dollar of government spending, more than a dollar of private spending would be stimulated. Still other economists thought the multiplier was near zero or even negative, given that government spending might crowd out private spending, that government was not competent to spend in ways that would be stimulative, or government spending might crowd out stimulative policy pursued by the Federal Reserve. And remember the importance of securing economic growth. If some regulatory and public-​goods programs undermine economic growth, that can serve as a defeater for those programs. Regulations generally restrict choice options and, thus, frequently reduce the ability of businesses to engage in economic experimentation. One comprehensive study of economic regulations imposed between 1977 and 2012 found that economic growth since 1980 has probably been reduced by about 0.8 percent per year. The authors claim that if regulation had been held to its 1980 level, “the economy would have been nearly 25 percent larger by 2012 (i.e., regulatory growth since 1980 cost GDP $4 trillion in 2012, or about $13,000 per capita).”64 Defenders of a more extensive administrative state will surely quickly and reasonably criticize the study, but it could easily be correct. And given that the costs of regulation may be enormous, if the authors are even merely likely to be correct, say at a probability of .51, then we have reason to worry about regulation as a severe economic harm.65 The principle of sustainable improvements will tell against bureaucracies that generate a glut of destructive economic regulations.

63 http://​object.cato.org/​sites/​cato.org/​files/​pubs/​pdf/​cato_​stimulus.pdf. 64 Coffey, McLaughlin, and Peretto 2016. 65 However, see the next section for worries about complexity creating barriers to assigning any probabilities to outcomes, or even to knowing all the members of the outcome set.

The Welfare State  163 Public-​goods provision faces a similar challenge, since the government may provide public goods less cost-​effectively than the private sector could or provide it in place of private provision of a more important private good or service. So public-​goods provision can sometimes undermine economic growth as well; however, since a number of public goods are arguably necessary for growth, the principle of sustainable improvements might actively require the provision of public goods. And since public goods are typically not coercive in nature, even though they’re often coercively financed, the justificatory bar is somewhat easier to meet.

7.  Insulating Bureaucracies from Rent-​Seeking The final potential barrier to justifying coercive regulations and public policy is the threat of rent-​seeking, where citizens, lobbyists, or officials use coercion to redistribute economic gains to themselves by abusing the coercive power of the state, such as through a tariff on imports pushed by a domestic manufacturer to raise its profits by making its competitors’ imported goods more expensive. This threat is a potential barrier because the bureaucracies that issue regulations and public policies may be so vulnerable to rent-​seeking that members of the public have sufficient reason to reject the existence of those bureaucracies, which would make it pragmatically difficult if not impossible to impose new regulations and public policies. Remember from c­ hapter 1 and MPBW that for a constitutional regime to be stable, it must be immune to abuse by officials; that is, it should discourage efforts by the self-​interested to use it corruptly to benefit themselves at the expense of others. One worry that we might have about especially active and extensive regulatory bodies, for instance, is that engaging in regulatory capture will be attractive given how much power such regulatory bodies possess. What’s worse, rent-​ seeking produces regulations that can accumulate over time, dragging down the overall efficiency of the economy and thus violating the principle of sustainable improvements, as Mancur Olson famously argued.66 Many of the same problems hold for governmental processes for providing public goods, such as the use of infrastructure spending as largesse for congressmen. Institutions vulnerable to rent-​seeking, then, are harder to publicly justify than systems that are less vulnerable. Add to this the data reviewed in c­ hapter 2 suggesting that publicly detectable acts of rent-​seeking can undermine social and political trust in virtue of being seen as corrupt, and we have a powerful argument against granting



66

Olson 1984.

164  Trust in a Polarized Age too much power to governmental bureaucracies.67 This means that even good regulations and public-​goods provision can be undermined on the ground that the regulatory body required to generate them, being insufficiently immune, is vulnerable to defeat as a set of constitutional rules. Now there is at least one consideration that cuts in the other direction, which is that weaker bureaucracies may be less immune than stronger ones. For instance, a small regulatory body may be more easily lobbied or controlled by outside forces and may lack a cohesive internal culture and modes of self-​monitoring that could prevent its officials from becoming corrupt. If so, then the body will output more unjustified coercion and inefficient spending than a larger, more powerful bureaucracy. It is indeed less tempting to capture a small, limited bureaucracy with relatively little power, but it may also be easier to do so. So, while it is common for libertarians to use concerns about rent-​seeking to argue for a smaller state, it is an empirical matter whether larger or smaller bureaucracies are more vulnerable to rent-​seeking and regulatory capture. Thus, concerns about rent-​seeking may tell in favor of stronger regulatory bodies rather than weaker ones. In sum, state regulations and public-​goods provision can be publicly justified only if the following three conditions are met: (i) the empirical case for the policy must meet the standards of policy epistemology, (ii) the policy must not reduce economic efficiency too much, and (iii) the policy must be enacted by a body that is adequately immune from rent-​seeking groups, inside and outside of the government. These conditions suggest that few regulations and public policies can be justified, but how few, or how many, is far from clear. Given these standards, all is not lost for the policy wonk. Yes, determining whether policies are economically worthwhile is often difficult, if not impossible. But policy epistemology cannot be so demanding that no public policy or regulation is ever publicly justified. Moreover, we do not want a policy epistemology that is so demanding that policymakers are never publicly justified in engaging in policy experiments to uncover just the sort of information that they need to justify the public policies they favor. In some cases, policies can be temporarily justified as attempts to discover what works and what doesn’t, though we will want to take care that the coercion involved in imposing and funding the policy can be publicly justified on that experimental basis.68 And there may be ways of improving our ability to assess and predict the outcomes of various policies, say, through microcosmic deliberation and the use of prediction markets, as I show in ­chapter 7. 67 Here I do not mean to equate rent-​seeking with corruption, just that some forms of rent-​seeking will be seen as corruption. I thank Brian Kogelmann for pressing me to make this point explicit. 68 But given the dangers of experimenting on a large public, members of the public might insist that states, cities, and localities experiment with proposed public policies first so that the experiments’ effects can be evaluated.

The Welfare State  165 Thus, I should state here that policy epistemology should recognize the need for policy experiments and so allow many projects to go into the red when it comes to their ex post costs and benefits. Total real costs are just too hard to estimate. For this reason, I think policy epistemology may yield a policy process that frequently yields policies that exceed their estimated costs. However, if we recognize that the policies we implement usually cost more than expected, then we should at least demand that projections fall well below what we are willing to spend. Over time, then, if it becomes clear that we underestimate costs, we should be more hesitant to engage in policy experiments, but if we appear to overestimate costs, we should be more permissive in policy experimentation. One way to go would be to try to publicly justify some degree of policy risk, a specified extent to which members of the public would be prepared to gamble in order to secure policies that are accurately predicted to have benefits, as well as policy experiments that provide insight valuable enough to improve policies or even the policy formation process in general. I think this approach has merit, since all members of the public would be prepared to take at least some risk to implement their own policies in the hope that they turn out well. But libertarian and conservative members of the public would be more risk-​averse when it comes to coercion-​increasing policies than progressive and egalitarian members of the public, so they would limit the size of the risks that policymakers can take.69 One challenge for the whole enterprise is the difficulty in assigning probabilities to policy outcomes. My hope here is that policymakers can appeal to prediction markets in order to make rough estimates of the likelihood of various outcomes. I discuss the merits of prediction markets further in ­chapter 7. An additional way to proceed would be to rely heavily on federalism by encouraging policy experimentation at the local and state levels, then allowing the federal government to review the results and respond accordingly. However, even prediction markets will have trouble generating accurate probabilities if policies are applied to parts of society that are subject to complexity, a social state that arises when different institutional features interact. Thus, a complex system is something more than a complicated system, a system with many parts: it is a system whose complications causally interact with one another in ways that are hard, if not impossible, to predict. One kind of complexity is what Jacob Barrett calls combinatorial complexity, which “arises when predicting what outcome will be realized by the interaction of multiple institutional features.”70 Combinatorial complexity implies that “the effect of any two 69 Though libertarians and conservatives might want to run similarly risky experiments in decreasing coercion, and progressives will be hesitant to run those experiments, so perhaps an overall favorable attitude toward policy experiments could be secured if each side agreed to go along with the others group’s desired experiments. 70 Barrett 2020b. Also see Gaus 2019.

166  Trust in a Polarized Age institutional features cannot be reduced to the sum of the effects of each feature by itself.” If social systems exhibit combinatorial complexity, then we will not only have trouble predicting their effects, we may not be able to assign probabilities to predicted outcomes; worse still, we may not even be able to identify all the possible outcomes. Thus, if policy covers a complex system, then even the best modes of prediction will fail, and the public justification of a risk profile will be of little or no use. So, in sum, we can probably publicly justify some substantial risk level for policy experimentation with policies that increase coercion, and we can garner estimates of the probability of policy outcomes through various social mechanisms, like prediction markets. Unless, of course, we face complexity, in which case we are out of luck. I am not in a position to determine which social systems exhibit complexity, however, so I am unable to determine the proper level of risk. Legislatures and other policymaking bodies are much freer to experiment with policies that reduce coercion, since they do not face the same justificatory hurdle. It is typically permissible to reduce coercion as long as we anticipate that it will not force institutions to initiate unjustified coercion at some point in the near future. Thus, coercion-​reducing policy experiments are far easier to justify. This does not entail that policy experiments should tilt to the political right. For instance, tax cuts do not always face an easier hurdle than tax increases. If tax cuts yield deficits, then they are likely to eventuate in future tax increases, in which case we have simply deferred coercion to a future time rather than reducing it on net. Similarly, policies redistributing wealth are not necessarily more coercive, since protections for property rights are coercive too. So, some policy experiments redistributing some wealth may count as coercion-​reducing. Even fairly demanding policy epistemology standards do not rule out all or nearly all experiments.71 Here, then, we can distinguish between different “tiers” of policy in accord with how easily those types of policies can satisfy the standards of policy epistemology. The most permissive tier includes policies that are coercion-​reducing, or that are not coercive, such as legalizing the use of a recreational drug. These policies do not require demanding standards because the point of the standards is to ensure that coercion is publicly justified. The least permissive tier includes policies that are coercive by nature, such as the nationalization of an entire industry. 71 One might also think that if an economy or society is in a dire position, it is easier to publicly justify policy experiments in order to find a way out of disaster. On the other hand, given how little we know, and that terrible policies could make things worse, we should perhaps remain cautious even in those conditions.

The Welfare State  167 There are also middle-​tier policies that are not necessarily coercive, but must be funded through coercive taxation, such as the provision of public schooling. While my standards of policy epistemology may seem to yield enormous skepticism about the effectiveness of public policymaking, I think some policies will pass the test. For instance, I think carbon taxes are sufficiently important as means to reduce the threat of climate change that the economic costs will be worth paying. The scientific evidence satisfies the standards of policy epistemology, the economic effects of simple taxes are fairly easy to predict, and while carbon taxes will be subject to some rent-​seeking, a tax policy could be designed that is transparent and simple enough that tampering and cheating will be detectable, which will hopefully discourage these activities in advance.72 The case for providing public goods is again easier, given the obvious need for public goods like defense spending, disease prevention, and transportation infrastructure.73

8.  Head Start and Bolsa Família: Publicly Justified Social Insurance To demonstrate how to apply policy epistemology, and to vindicate the public justification of the welfare state, I will attempt to show that at least two antipoverty policies can be publicly justified. This will help answer the concern that my standards for publicly justifying policy are too demanding. It will also help to show that key poverty-​relief functions of the welfare state can be publicly justified in their specifics. Here I will focus on only two policies, but my sense is that there are many other good candidate policies eligible for public justification.74 Consider, for instance, one of the most basic American antipoverty policies, Head Start. Head Start provides early-​childhood education and health, nutrition, and parent involvement services to low-​income children and their families.

72 Though part of my confidence comes from the fact that even libertarians should see carbon taxes as protecting property rights and, thus, as reducing coercion. For my argument, see https://​ niskanencenter.org/​blog/​natural-​rights-​libertarians-​support-​carbon-​mitigation-​answer-​may-​ surprise/​. 73 And as we will see in c­ hapter 7, section 5, governments may play a noncoercive role in collecting information about policies they have run or that have been run by others. This is one way in which it is easier to overcome evaluative pluralism about empirical matters, since data collection can at least potentially resolve disagreements, in contrast to evaluative pluralism about moral and religious matters. In this way, it will be harder to use intelligible reasons to defeat policies over which we have empirical disagreements than those subject to moral and religious disagreement; a good-​faith effort can be made to determine which policies have effects that can make them eligible for multiple members of the public, given reasonable agnosticism about empirical effects. 74 For instance, policies aimed at closing the “word gap” between poor and middle-​class or upper-​ class children entering primary school.

168  Trust in a Polarized Age The goals are to produce stable families and improve the physical and emotional well-​being of children. Head Start is arguably a functionally independent policy unit, as defined in section 6: it is causally independent of other programs in the eyes of members of the public. Most, if not almost all, members of the public will regard the policy as helping them achieve their highly ranked goal of alleviating childhood poverty; they can do so without appealing to other closely related policies; and moderately idealized members of the public will be able to evaluate the effects of such a simple wealth transfer based on relatively simple economic models. Head Start probably violates no primary right, including rights of private property, because the kinds of extensive private property rights that would prohibit redistribution to impoverished children will be defeated by basically everyone who isn’t a radical libertarian. Second, the redistribution might slightly decrease the incentives for wealth creation among the very wealthy, but the amount of taxation required to fund Head Start is small, and the gains from improving the welfare of children probably vastly outweigh the economic losses even just in terms of economic gains from having healthy and productive workers. Head Start is one way to provide people with social insurance, so it obviously comes under the principle of social insurance. We also do not have to worry much about rent-​seeking and corruption regarding the recipients of Head Start aid.75 There is an interesting question regarding whether Head Start can meet the standards of policy epistemology. Remember that if a policy reduces coercion, it does not have to satisfy the same high evidentiary standards as a policy that increases it, but since Head Start is funded with taxes and taxes are usually coercive, Head Start may have to meet high policy epistemology standards. Even so, I think Head Start can satisfy the standards of policy epistemology. In its multi-​ decade history, many studies on its effectiveness have been conducted.76 The large majority of studies find direct improvements for three-​and four-​year-​old children who would otherwise be in poverty, though there is controversy over whether the benefits of Head Start persist into late childhood and adolescence. But since most members of the public, if not basically all, will regard directly improving the welfare of young children as an important public policy goal even if the benefits do not persist into late childhood, we only need good evidence to show that Head Start benefits its recipients when the aid is received. And the evidence supports that hypothesis.

75 Though one does wonder about rent-​seeking from service providers. 76 For accessible literature reviews of the evidence, see https://​www.brookings.edu/​research/​the-​ long-​term-​impact-​of-​the-​head-​start-​program/​, https://​www.nber.org/​chapters/​c13489.

The Welfare State  169 One of the most impressive antipoverty policies is the Bolsa Família (BF) program in Brazil.77 Under BF, poor families with children are given money in the form of a direct transfer, but the transfer is conditional on keeping children in school and taking them in for regular health checkups. If families do not do this, they lose the benefit. The program directs funds to women, who are often heads of households due to absent fathers. The goal of BF is twofold: reducing current poverty and encouraging families to invest in building human capital in their family, with the hope of breaking the intergenerational transmission of poverty. BF reaches a huge portion of Brazilian society, including many people who have never received any other social transfer.78 Many studies have shown that families spend the money on food, school supplies, clothes, and so on, and not on drugs and alcohol. There is a great deal of evidence that suggests that BF has played a central role in Brazil’s dramatic reductions in poverty and economic inequality since the program was introduced in 2003. By 2013, BF is said to have cut Brazil’s extreme poverty rate in half, from 9.7 percent to 4.3 percent of the population, and Brazil’s sky-​ high Gini coefficient fell by 15 percent points. The chances of a 15-​year-​old girl attending school have increased by 21 percentage points. Moreover, the program is seen as having very low overhead costs, as it is a simple program requiring little administrative oversight. It has not reduced the incentive to work, and it may well increase national economic growth.79 BF’s results are so impressive that over forty countries have adopted similar programs. The poverty reductions are fairly well documented over the past fifteen years, as noted. However, there is a concern among Brazilian policymakers that could ground a defeater reason: BF may distort industrialization.80 Due to the receipt of more cash transfers, local economies are devoting themselves increasingly to the production of consumption goods for recipients of funds, rather than generating new concentrations of capital, which would lead to long-​term poverty reduction through higher wages, and the other benefits of economic growth. The prime method of permanently ending poverty is the gradual accumulation of capital, which enables broad-​based wage growth and an increase in available bundles of goods and services. Remember that one aim of BF is to reduce long-​ term poverty through human-​capital investments. So if BF undermines capital accumulation in Brazil’s most impoverished regions, then that is a strike

77 BF is not a form of social insurance, strictly speaking, since it is not triggered by the loss of some benefit or job. That is, it is not insurance against a loss. However, it is insurance against extreme poverty, so we can understand BF as a social insurance scheme all the same. 78 For a summary of these studies from the World Bank, see http://​www.worldbank.org/​en/​news/​ opinion/​2013/​11/​04/​bolsa-​familia-​Brazil-​quiet-​revolution. 79 Rougier, Combarnous, and Faure 2018, p. 200. 80 Ibid., pp. 212–​3.

170  Trust in a Polarized Age against it. BF does not appear to have this effect in areas that already had an industrial base, but it may well have a negative effect on areas that had no industrial base when BF was introduced. The goal of BF was to eliminate the need for BF in the future; if BF undermines capital accumulation, then it may become a permanent feature of the Brazilian poverty landscape. A recent overview of the literature on BF that draws attention to growth-​reducing effects concludes that BF “has largely achieved its main goal, i.e., the reduction of poverty, while significantly improving educational performance in the poorest regions and social groups,” which is likely to have significant positive effects on economic growth. That said, “It is increasingly acknowledged that, beyond their direct impact on poverty reduction, [conditional cash-​transfer programs] also drive changes in local economies” by creating consumption economies, and this can generate “weak prospects of a sustained and autonomous growth pattern.” There is evidence that BF spurs “light industrializations in the poorest localities,” but the same evidence also shows that a negative effect on industrialization has occurred. Critically, however, even the authors of this study conclude that BF has had transformative effects in poverty reduction and likely has at least some positive effects on economic growth through building human capital, even if it also has negative effects on growth in poor regions with a narrow industrial base. We can hope that the Brazilian government finds a way to address slowing industrialization. Of course, even the most touted and successful social programs have costs; no one would deny this. But the costs are not enough to supply some members of the public with defeater reasons. It seems that there is a very broad consensus that BF has played an enormous role in the reduction of poverty and increasing educational attainment in Brazil. Whether it will succeed in its long-​term goal of breaking cycles of poverty is, as of yet, unclear, but there is no evidence that the growth-​reducing effects outweigh the growth-​increasing effects, on balance. So it may well be that BF is not only reducing poverty but increasing economic growth at the same time. If so, that is a powerful reason to think that BF is publicly justified. Only radically libertarian members of the public who reject all cash transfers as unjust are likely to think that conditional cash-​transfer programs like BF are worse than no cash transfers, as evidenced by the fact that even many libertarian thinkers, like Milton Friedman and Hayek, supported the government provision of some form of social minimum.81 Thus, while BF may be seen as suboptimal for many members of the public, it is unlikely to be defeated. Policy epistemology suggests that most moderately idealized members of the public should conclude that BF has poverty-​reducing effects, which can lead one to conclude



81

Hayek 1979, p. 55; Friedman 2002, pp. 191–​4.

The Welfare State  171 that BF satisfies the principle of social insurance and the principle of sustainable improvements at the same time. Thus, members of the public are likely to have defeaters for property rights protections so strong that they would prohibit the modest amount of redistribution required to fund and implement BF.82 It appears then that Bolsa Família passes the standards of policy epistemology and does not reduce economic efficiency too much. Given that it is a simple cash-​ transfer program, there is little evidence that Bolsa Família is implemented by a bureaucratic apparatus that is insufficiently immune to rent-​seeking. And given that BF may reduce coercion in some respects, by reducing unrest among the poor, that may make publicly justifying it even easier. I suspect other policies can pass the relevant tests, especially reforms that not only provide the poor with wealth but ensure that they partake in economic growth. Michael Tanner, for instance, has recently argued that we can promote growth among the poor in a number of ways, such as reforming education to give more control and choice to parents and to break up the public school monopoly, reforming the criminal justice system, and curtailing the war on drugs.83 He supports bringing down the cost of housing to make it easier for the poor to bank, save, borrow, and invest. Land reform is crucial: we need to avoid various forms of rent-​seeking that increase the price of housing, such as exclusionary zoning and the abuse of historic-​preservation rules.84 I  think these policies can be publicly justified, but I  cannot show that here.

9.  A Modest Welfare State Creates Trust for the Right Reasons Some institutions characteristic of the welfare state can be publicly justified. These include simple antipoverty programs, certain essential regulations, and the provision of certain widely valued public goods. There are hurdles to justifying these programs, and there are greater hurdles to justifying further programs. But the basic institutions of the welfare state will likely be part of any publicly justified polity. And in virtue of this fact, welfare rights, in virtue of being publicly justified, create the right reasons for trust. The practice of protecting and exercising these rights justifies trusting attitudes among diverse persons.

82 Though perhaps men who object to the gender imbalance of the policy may have defeaters for it; this might be managed by allowing fathers to opt in to the program. 83 Tanner 2018, pp. 12–​3, 125. 84 Ibid., p. 174–​5.

172  Trust in a Polarized Age Our review of the data also allows us to conclude that certain welfare state institutions, especially well-​run, noncorrupt universal welfare supports, create real trust, both social and political. Welfare state institutions thus create trust for the right reasons. I will now argue that a publicly justified polity should not tilt further left.

6

Against Egalitarianism In this chapter I want to determine whether trust and public reason require economic institutions further left than those of the welfare state. In particular, I assess three left-​wing proposals: redistributive taxation to reduce large inequalities of wealth and income, primarily to preserve democratic institutions and increase trust, and two types of left-​wing economic regime that are receiving renewed attention: property-​owning democracy and liberal socialism. I will first argue that highly redistributive taxation meant to save democracy can be publicly justified in principle but that current empirical work suggesting that redistribution will help democracy is too inconclusive to satisfy the standards of policy epistemology (section 1). I then argue that the data showing a strong correlation between economic equality and social trust do not justify compressing income and wealth inequality (section 2). The next four sections (3–​6) argue that property-​ owning democracy, a left-​wing economic regime favored by many contemporary political philosophers, cannot be publicly justified, and the section after those (section 7) does likewise with liberal socialism. I address concerns about workplace coercion and firm structure next (section 8, though the issue is first raised in section 6). I then conclude by arguing that the economic regime most appropriate for a morally peaceful polity full of the diversity of views present in liberal democratic regimes is a market economy supplemented by a modest welfare state (section 9).

1.  Inequality and the Democratic Process Economic inequality has become an increasingly important focus in US politics and in left-​wing intellectual circles. For several decades, the Left has focused primarily on defending and extending the welfare state, largely on sufficiency grounds. The dispute with the political Right has been whether the government should provide all persons with a basic standard of living and whether persons have a right to that support. But with the supposedly increasing divergence between income quintiles in the United States, and with similar divergences in European nations,1 some have argued that welfare programs are not enough to 1 Piketty 2014 provides empirical documentation of these claims, though it has been the subject of considerable criticisms. See http://​www.ft.com/​cms/​s/​2/​e1f343ca-​e281-​11e3-​89fd00144feabdc0.

174  Trust in a Polarized Age realize economic justice—​that differences in income and wealth must also be compressed.2 The arguments for income and wealth compression can be categorized based on whether they regard economic equality as intrinsically or as instrumentally valuable. Income equality is said to be intrinsically valuable because it is required by principles of justice: that is, equal incomes and equal wealth are just unless inequalities are somehow chosen or found mutually acceptable. A number of conceptions of distributive justice are rooted in such a commitment, such as luck egalitarianism.3 As I see it, these views are excessively sectarian: the reasons generated by these doctrines will be neutralized by reasonable disagreement about justice. The luck egalitarian, for instance, advocates coercive redistribution based on a doctrine that many, if not most people, reasonably reject.4 In this way, the luck egalitarian who advocates radical redistributive policies is in the same position as an establishmentarian Roman Catholic who wants the Catholic Church to become the national church of the United States. Both have sincere normative commitments, but most members of the public will have defeaters for the functionally independent policies proposed by each group, such as the coercive establishment of Catholicism through congressional action or sharply redistributive taxation that can only be justified on a luck egalitarian basis. That said, public reason liberals must be concerned with claims that economic equality has instrumental value, particularly with respect to the effective functioning of democratic institutions. Many have argued that societies with greater income inequality have worse-​functioning democratic institutions and that greater inequality causes this greater level of dysfunction.5 I find these causal connections hard to assess. In particular, I am skeptical of views that superrich individuals and small groups are powerful enough to systematically distort the democratic process.6 Our best evidence, instead, is that the difference in the responsiveness of democratic institutions to economic classes is not a matter of greater responsiveness to the top 0.01 percent.7 Say what

html. Piketty’s response: http://​piketty.pse.ens.fr/​files/​capital21c/​en/​Piketty2014TechnicalAppendix ResponsetoFT.pdf. See Sutch 2017. 2 Piketty, for instance, advocates an annual, global tax on capital to constrain inequalities. See Piketty 2014, pp. 515–​39. Rawlsians, of course, have long maintained that the welfare state is insufficiently attentive to reducing economic inequalities. 3 For a discussion of luck egalitarianism, see Arneson 2013. 4 For a variety of objections to various forms of egalitarianism, see Frankfurt 2015. 5 Bartels 2010. 6 For instance, some have found no significant correlation between levels of wealth inequality across advanced democracies and how well their democratic institutions function. For discussion, see Gaus 2011, pp. 517–​21. 7 Gilens 2012, pp. 70–​96.

Against Egalitarianism  175 you will about Charles Koch, George Soros, Warren Buffet, Sheldon Adelson, Bill Gates, Mark Zuckerberg, Michael Bloomberg, and other superrich Americans who attempt to influence elections and policy, they cannot undermine democratic institutions on their own. They simply have too little power with respect to other political forces, and their work and contributions are plagued both by principal-​agent problems and by divergent political activities that often cancel one another out. The real worry, as Martin Gilens has argued, is that democratic institutions in the United States are solely responsive to the top 10 percent of income earners.8 A recent major study by Gilens and Benjamin Page finds that the top 10 percent of income earners and large corporate powers and interest groups have a huge effect on policy vis-​à-​vis everyone else.9 In fact, disturbingly, they find that when the top 10 percent want policy X and the bottom 90 percent want not-​X, the top 10 percent always get their way. Gilens acknowledges that the top 10 percent and bottom 90 percent often agree on policy, such that in many cases, the poor get what they want;10 but in cases of disagreement, the top 10 percent win out, which means that democratic processes under conditions of significant economic inequality are ignoring what the bottom 90 percent believe is best or appropriate. If Gilens is correct, public reason liberals have two reasons to worry. First, democratic institutions will not track what is publicly justified for members of the public. The production of legislation will more likely track what is publicly justified for the top 10 percent. Thus, if income inequality causes this lack of responsiveness, then inequality undermines public justification. Second, members of the public have primary rights to vote and influence the political process. If Gilens is correct, most Americans will have ineffective political rights, something that provides strong defeaters for the institutions that sustain these responsiveness-​undermining political rights. It is possible that the top 10 percent have a better sense for what is publicly justified for everyone than the public as a whole. In that case, these voters getting their way politically produces a higher number of publicly justified policies and a lower number of defeated policies than we would have if the government responded to all of us. But given the difficulties in determining what is publicly justified in a diverse world, we have reason to defer to the judgments that members of the public make concerning themselves, rather than consulting the judgments of elite members of society.11

8 Gilens 2012. 9 Gilens and Page 2014. 10 Gilens 2012, p. 78. 11 It is possible that some have more informed and accurate views than others about which policies are best for most people, but it is unclear whether that group overlaps with the wealthiest people.

176  Trust in a Polarized Age Gilens and Page’s work has come under scrutiny.12 And we need more than an interpretation of a single, albeit impressive, data set to determine whether Gilens and Page’s work meets the standards of policy epistemology. Otherwise, their work cannot serve as the prime evidential base for policy that compresses income and wealth inequality. But let’s assume for the moment that their interpretation is the only reasonable interpretation of the data and that the data are sufficiently strong evidence to justify the belief that, in the United States, over the last several decades, the preferences of the bulk of the population count for little or nothing in democratic outcomes, while the preferences of the rich and the moderately rich determine everything. What policies can be justified in response? The best argument from the left is that our high levels of income and wealth inequality are generated by coercive property rules. Given that these property rules produce inequalities that undermine primary political rights and lead constitutional rules to fail to track what is publicly justified for most people, the property rules are defeated. If government imposes policies aimed at compressing economic inequalities, the affluent have no legitimate complaint given that preserving a well-​functioning, rights-​protecting democracy is more important than having enough money to pursue extremely expensive life projects or to do philanthropic work. However, the presumption against coercion and in favor of solutions offered by voluntary, civic institutions means that we should look for coercion-​reducing policies that can reduce income inequality. I  think two coercion-​reducing policies can go a long way toward reducing income inequality in the United States: repealing or reforming many residential zoning laws and weakening intellectual property protections.13 One of the controversies raised by Thomas Piketty’s well-​known work on income inequality is that much of the inequality he documents can be explained by the differing values of the real estate held by the very rich and that owned by everyone else.14 If so, then reforming zoning laws to prevent them from creating artificial shortages of real estate should be an excellent way to reduce inequalities of wealth. Limiting zoning laws can also boost economic growth: a recent study finds that in 220 metro areas, zoning constraints on land use “lowered aggregate US growth by more than 50 percent from 1964 to 2009.”15 That’s staggering. 12 For a critique of Gilens’s focus on “responsiveness” as the appropriate standard of democratic equality, see Sabl 2015. For alternative interpretations of the Gilens-​Page data, see Enns 2015 and Branham, Soroka, and Wlezien 2017. Both purport to show that the middle class has far more influence than Gilens and Page recognize. 13 https://​obamawhitehouse.archives.gov/​sites/​default/​f iles/​p age/​f iles/​20151120_​b arriers_​ shared_​growth_​land_​use_​regulation_​and_​economic_​rents.pdf. For a detailed discussion of these two cases, see Lindsey and Teles 2017, pp. 109–​26 and pp. 64–​89 respectively. 14 Rognlie 2015. 15 Hsieh and Moretti 2015.

Against Egalitarianism  177 The same is true of intellectual property laws that artificially inflate the wealth of the creators of intellectual property, especially with respect to software copyrights. It has also been estimated that extremely strict and long-​term intellectual property rights have promoted the spectacular growth in the wealth of Silicon Valley billionaires and the many people who work at technology firms.16 Some libertarians oppose all intellectual property rights, not to mention the extended protections that information technology firms insist upon for themselves, so a libertarian-​egalitarian alliance might be productive, given that each group has its own strong, but distinct, reasons to reject the coercive imposition of strong intellectual property laws.17 The hard question is whether reforming zoning laws and intellectual property laws will make the democratic process more likely to protect primary political rights and reduce legislative error. I find this determination difficult to make, given all the counterfactual judgments involved, the limitations of our empirical tools, and the other difficulties in ranking constitutional rules. For this reason, I think we should first pursue coercion-​reducing methods of compressing income inequality rather than coercion-​increasing methods, and only pursue coercion-​increasing methods if we are confident that economic inequality causes democracy to function less well.18 We should also pursue the relatively modest coercive restrictions brought about through novel campaign finance reform before we attempt to substantially restructure the economy through taxes, regulations, and redistributing capital holdings.19 My skepticism about coercively compressing inequality is driven by the view that inequalities in wealth and income cannot by themselves cause democratic institutions to function poorly. Both inequality and poor democratic functioning could have the same cause, such as constitutional rules that allow excessive rent-​seeking, which could both increase inequality and make democracy function less effectively. If it is relatively easy for people to gain special government favors, we can expect the politically well connected to benefit from these favors by becoming richer than others.20 And if it is relatively easy for certain people to gain special government favors, then democratic institutions must be

16 See Joseph Stiglitz’s discussion at http://​opinionator.blogs.nytimes.com/​2013/​07/​14/​how-​ intellectual-​property-​reinforces-​inequality/​. 17 http://​f reenation.org/​a/​f31l1.html,https://​mises.org/​library/​ideas-​are-​f ree-​c ase-​against​intellectual-​property. 18 Unless there are sources of defeater reasons other than concerns about inequality. I think there probably are other sources of defeater reasons for zoning laws and intellectual property protections. 19 Gilens 2012 is long on diagnosis and short on remedies; he focuses almost exclusively on campaign-​finance reform and gerrymandering. But the campaign finance reform would have to be aimed at limiting the power of the top 10 percent, which differs from current campaign finance reforms aimed at the top 1 percent. See pp. 247–​52. 20 Lindsey and Teles 2017.

178  Trust in a Polarized Age more responsive to their wishes than to the wishes of everyone else.21 The concentrated benefits and dispersed costs of a rent-​seeking society22 could explain both the failure of democracy to track the will of the people and increases in inequality between the affluent and everyone else.23 And before we can justify the use of redistributive taxation, we would need to determine that we lack alternative, feasible, less coercive methods of trying to protect the integrity of the democratic process. There are other confounding factors with respect to the social traits of the politically influential classes. The affluent, for instance, may secure greater influence through better education about the political process. So, if we could increase the quality of civics education for most members of the public, we might thereby enable most people to play the political game more successfully than they can now. And we may be able to fix educational inequality without extensive redistribution to compress inequalities of wealth and income.24 It is also possible that the influence of the affluent is due less to how much more wealth and income they have by comparison to everyone else than to their higher social rank, which could persist even with lower levels of economic inequality. If influence is due to social status, then even if we coercively compress income inequality, we will not make the democratic process track the will of the people better. The democratic process will remain tied to persons with high social status. Let’s combine this point with Gordon Tullock’s observation that, given how high the stakes are in politics, it is peculiar that more isn’t spent on campaigns.25 Given how much influence the affluent have and how little they spend on purchasing politicians, this strongly suggests that they’re affecting political outcomes in some other way, probably by way of their high social status.26 Material inequalities do seem to translate into greater social power and status all on their own, however, so perhaps this explanation only goes so far.27 Inequality and democratic responsiveness might be related in other ways. Inequality might lead to divergence, divergence might lead to gridlock, and

21 We will see in ­chapter 8, however, that Achen and Bartels 2016 suggests that democracy may not be responsive to any coherent electoral coalition. 22 Tullock 1993. 23 Gilens 2012 and Gilens and Page 2014 find that interest groups have powerful effects on legislative outcomes, though they frequently lose when their policy preferences diverge from those of the top 10 percent of the income distribution. 24 For instance, educational inequality could be reduced by loosening zoning laws, reducing the cost not only of housing but of building new schools, and reducing the capacity of the affluent to zone their communities to ensure that their local public schools exclude children from lower-​income families. I thank Eduardo Martinez for this point. 25 Tullock 1972. For a much more detailed discussion, see Ansolabehere, Figueiredo, and Snyder 2003. 26 I thank Will Wilkinson for this important point. 27 Harkness 2014, 2017.

Against Egalitarianism  179 gridlock might prevent the enactment of policies that redistribute wealth, even if such policies are supported by most of the public.28 In this case, it would not be wealth causing a lack of responsiveness to the poor, but institutions that function poorly in divergent political environments. If we could reform electoral institutions, then perhaps divergence would not have this effect. When divergence is introduced as a variable, nonresponsiveness can be explained in other ways. The absence of a clear, decisive cause of economic inequality shows the importance of policy epistemology. The standards of evidence for determining causal connections between social variables might be relatively permissive with respect to what individuals are entitled to believe, but the standards of evidence required to publicly justify coercion are much higher. If you’re going to tax people making $1,000,000 per year at 80 percent rather than 40 percent, you’d better be able to show with an adequate degree of certainty that reducing their income increases the protection of primary political rights. Otherwise, the coercion required to redistribute cannot be publicly justified to the person coerced. One might reply that the redistribution is not coercive: the coercer is the rich person resisting redistribution, a scenario I discussed in c­ hapter 5. However, the discussion in ­chapter 5 concerned cases where redistribution funds a small number of vital social programs and the taxation does not diminish the range of choices available to rich persons. Redistributive taxation in favor of economic equality must be much larger, so much larger that it removes many valuable choices from rich persons. So it coerces the rich person, often greatly.29 Right now, in my estimation, coercively redistributing wealth in the name of greater economic equality cannot be publicly justified, so such redistributive coercion undermines the rational basis for social trust between diverse persons. It is true that many economically advantaged persons have been able to rig the political system in their favor, and these persons have no valid objection to having their wealth redistributed, nor can they reasonably object to changes in legal rules that would make it difficult for them to stay wealthy or increase their wealth. Yet blanket redistribution cannot be publicly justified until we have better data demonstrating cause and effect with respect to inequality and 28 McCarty 2019, pp. 150–​2. 29 Here it might be objected that the rich and poor face unequal threats of coercion, since the rich are coerced relatively little by a property system compared to the poor, and we may want to have a presumption against coercing the weak that trumps a presumption against coercing the rich. I prefer to make public justification sensitive to concerns about undue coercion, not as part of the presumption against coercion, but as part of the set of defeater reasons for coercive property rights; I think much of what I have to say in this chapter bears on how public reason deals with inequalities of various kinds. A second objection is that rich persons will still have lots of valuable options available to them after such taxation, but my point is that while they may have many valuable options after the taxation, they will still have lost other valuable options. I thank Scott Simmons for the latter point.

180  Trust in a Polarized Age undermining the democratic process. Gilens and Page have made an important empirical breakthrough in connecting economic inequality to the responsiveness of democratic institutions, but they have yet to satisfy the higher standards of evidence required to justify coercive redistribution to those who hold reasonable non-​left-​wing perspectives. I am much more confident in this skeptical conclusion when it comes to the work of Thomas Piketty, who famously attempts to explain increasing income and wealth inequality in industrialized nations by arguing that the natural rate of return to capital (r) exceeds the general rate of growth (g), as it did in the nineteenth century.30 That is to say, the amount of money that is generated by capital holdings is greater than the wealth generated by the general rate of growth, which means that large capital holders will grow richer faster than the rest of the population. That means inequality will not only expand over time, but do so quickly. Yet despite his large data sets, Piketty does not spend much time in his work to explain why r exceeds g. He also ignores serious problems, such as the diminishing marginal returns to holding capital. A standard claim in economics is that additional units of some good or service will produce less value for a person than the previous unit, such that the benefits of each additional unit of the good or service should gradually fall. The point is simple enough that it should apply to the profits made on capital. As the rich accumulate more capital, the profits yielded by additional units of capital should fall. This means that eventually the returns to an additional unit of capital should converge with the general rates of growth, such that r should eventually equal g. That is, the rates of return for r and g should equalize, stopping the slide into a class-​stratified society.31 But Piketty leaves us to wonder why the returns to capital don’t diminish; he just claims that they do not, empirically speaking. Here again we have a purported connection between relatively free economic conditions and dangerous levels of income and wealth inequality without a clear causal connection between the two. Without a causal connection demonstrable by means that meet the standards of policy epistemology, Piketty’s work cannot by itself publicly justify the coercive redistribution of wealth. There are also increasing worries about Piketty’s work. Some have criticized Piketty’s interpretation of the data on inequality in the nineteenth century, and Piketty has been suspected of overestimating how much income inequality has risen in the last several decades.32 30 Piketty 2014, pp. 25–​7. 31 For an accessible version of the criticism that Piketty’s model doesn’t address diminishing returns to capital holdings, see http://​larrysummers.com/​2014/​05/​14/​piketty-​book-​review-​the-​ inequality-​puzzle/​. I develop a charitable version of Piketty’s criticism of welfare state capitalism in Vallier 2019b. 32 Sutch 2017. Auten and Splinter 2017 argue that Piketty overestimates the increase in inequality in the United States by about 200 percent. Piketty argues that the share of national income captured by the top 1 percent increased from 9.1 percent in 1979 to 15.7 percent in 2014. Auten and Splinter

Against Egalitarianism  181 What’s more, the causal connection between inequality and trust is obscured by the fact that Piketty may have misidentified who the richest people are: he hypothesizes that passive rentiers are capturing most rents, meaning that the richest people are not engaged in much personal effort to produce wealth. But other recent work suggests that most top earners are active managers and owners of “mid-​market firms in relatively skill-​intensive and unconcentrated industries.”33 Redistribution is much less likely to reduce economic production if most of the rich are passive rentiers than if, for instance, they are actively managing their own private firms. If the rich are making their money through work, redistribution that reduces their incentive to work may reduce growth and thereby drag down political trust. For these reasons, in our attempts to make democracy responsive to the poor, we are on good grounds in pursuing equality through coercion-​reducing policies. But we should only advocate coercive policies if less coercive policies fail—​and even then, only if we have strong empirical evidence to support our proposals. I want to draw attention to the fact that I’m contradicting a causal story that is widely embraced on the left. Rawls’s critique of welfare state capitalism depends on his surmise that even with welfare state institutions, capital will tend to concentrate in the hands of the few, such as the top 1 percent; Alan Thomas extends Rawls’s story, saying that “the best off will, acting in concert, use their leverage to exert monopoly control over capital and then start to introduce a range of ‘democratic’ bads.’ ”34 According to Thomas, “Over time, those inequalities will generate a drift to oligarchy as large concentrations of power exert undue influence over the competitive space of the political process.”35 Here Thomas follows John Roemer, who claims that “the problem in a capitalist economy is that there is a very small class of wealthy people who receive huge amounts of income as their share of firms’ profits, and it is generally in the interests of these people to have high levels of the profit-​increasing public bads.”36 Rawls, Thomas, and Roemer, along with many other critics of capitalism, believe that capital has a natural tendency to concentrate and that, as capital concentrates, the capital-​rich will obtain engage in artificial restrictions on the market that redistribute wealth upward to them. Importantly, Rawls and Thomas argue that mere income redistribution

find an increase from 8.4 percent in 1979 to 10.1 percent in 2014. For similar concerns about Piketty’s estimates, see Magness 2019.

33

Smith et al. 2019. Thomas 2017, p. 302. 35 Ibid., p. 109. 36 Roemer 1994, p. 57. 34

182  Trust in a Polarized Age can’t stop this process, as it is a central part of their criticism of welfare state capitalism. Something more radical is required. But the causal story that Thomas and Rawls take for granted is deeply contentious, and the best recent attempts to demonstrate the story’s accuracy—​the work of Piketty and that of Gilens and Page—​are themselves disputable even by economists without a promarket bias. This means that the narrative of concentrating capital cannot be taken for granted; in fact, the evidence is weak enough that the standards of policy epistemology are not met. Consequently, members of the public as a whole lack sufficient reason to permit coercive redistribution of income and capital in response. Because it is easier to justify reducing coercion than increasing it, this does not contradict my earlier statement that the evidence is strong enough to ground left-​wing objections to many coercive policies. Egalitarians have good grounds to advocate policies that reduce coercion, such as limits on real-​estate zoning and intellectual property. Similarly, egalitarian citizens even have sufficient reason to reject extensive property rights in capital. But this will not publicly justify a coercive governmental apparatus that regularly redistributes income and capital. First, the apparatus may lack immunity from rent-​seeking. Second, we have established that some primary property rights in capital are publicly justified, if suboptimal for egalitarian citizens, and these property rights will limit the legitimate redistribution of capital, though not prohibit it.

2.  Inequality and Trust One robust finding in the social trust literature is that societies with more economic equality have higher social trust. As Christian Bjørnskov puts it, “A conclusion reached by virtually all studies is that income inequality is among the most robust cross-​country determinants of trust.”37 Some trust researchers think that income inequality is the primary cause of low trust. Uslaner even ends his book with the following: “If you want to develop the cooperative spirit that will help overcome collective action problems, don’t get rich, get equal.”38 Since I am concerned with establishing social trust across diverse viewpoints, it appears that I may be committed to redistribution in the name of promoting trust: if we want to increase social trust, we must “get equal,” right? Not quite. Remember that establishing trust for the right reasons is a matter of sustaining trustworthiness by means of public justification. And trust can

37 38

Bjørnskov 2006, p. 5. Uslaner 2002, p. 255.

Against Egalitarianism  183 only be sustained in a diverse populace if moral, legal, and constitutional rules are publicly justified to each person. Therefore, if there is a publicly justified primary right to property that restricts redistribution, preserving trust for the right reasons means not violating that primary right (unless it is overridden by some other primary right). We are not in the business of simply doing whatever maximizes trust; our goal is to create and sustain trust in the right way. Since members of the public want to trust each other, social-​democratic members of the public may have good objections to—​defeater reasons for—​ strong property rights because they think that such property rights will undermine our capacity to trust one another. However, just because societies with high inequality have reduced trust does not imply that redistribution will increase trust. In fact, as far as I can tell, we have little evidence that a country which increases its level of redistribution will become more trusting. Social trust is fairly stable over time despite big changes in tax policy.39 And to publicly justify the relevant redistributive coercion, the evidence that redistribution promotes trust needs to be strong enough to satisfy the standards of policy epistemology. Further, we cannot publicly justify redistribution to increase trust unless we have a good guess as to why income inequality yields lower trust, and it is hard to guess well in this case. You has argued that income inequality lowers trust not because of “social distance” between the affluent and everyone else but because people see income inequality as a sign of unfairness.40 He also finds that political inequality undermines trust. You suggests that both income inequality and political inequality are seen as unfair, which leads persons to distrust those who have benefited from these kinds of inequality, as well as those who have implemented the policies that produced the inequalities. Insofar as citizens see inequality as unfair, then moving toward more economic equality should increase trust. But people disagree about whether economic inequality is in itself a sign of unfairness. An alternative interpretation is that people use economic inequality as a proxy for other forms of unfairness, such as unequal connection to political power and the ability to siphon funds from the public trough. If redistribution does not reduce the apparent influence of special interests in politics, then redistribution alone may not increase trust.41 For example, shifting to “inclusive” institutions, those which allow everyone to participate in beneficial cooperation and interactions, from “extractive”

39 And many other policy changes, as noted in c­ hapter 2. 40 You 2012, p. 702. 41 Alan Thomas agrees that collective trust can be eroded by the appearance that people are getting away with something and that redistribution of income may not be enough to undermine that appearance. He instead argues for property-​owning democracy, but we will see later that property-​ owning democracy is not a good solution. Thomas 2017, p. 200. I address Thomas’s arguments for property-​owning democracy in appendix E.

184  Trust in a Polarized Age institutions where the many are exploited by the few, may be what reduces trust.42 Extractive institutions will produce inequalities and reduce trust at the same time, explaining the correlation between inequality and trust without postulating inequality as a direct cause. What I would argue then, is that we need more evidence that redistributive taxation will promote trust. After all, we might be able to promote trust in other ways, such as by insulating the democratic process from the effects of economic inequality. If economic equality can be increased through the reduction of coercion, such as by liberalizing zoning laws and intellectual property laws, then we should begin our redistributive efforts with these policies first. The preceding discussion presumes that income equality causes an increase in social trust and not the other way around. But Bergh and Bjørnskov find “no evidence that welfare state redistribution increases trust, as net inequality does not affect trust.”43 Instead, they find that social trust reduces income inequality. One reason for this may be that social trust increases support for redistribution, which lowers inequality. When people trust each other, they are less likely to believe that redistribution will be abused or misused by recipients and are likely to feel more social solidarity with strangers. The main equality-​to-​trust causal story requires that people can track income and wealth inequality through observation and adjust their trust levels downward as a result, probably because the observation of inequality suggests either unfairness or a degree of social distance. But there are two problems with this story. First, social-​distance theories of social trust are only weakly supported by the data, if they are supported at all. If even ethnic diversity has either a small negative effect on trust or no effect at all, why would economic inequality have a substantial negative effect? It is far easier to distinguish between ethnic groups than economic classes in the United States, at least.44 Second, if perceptions of unfairness reduce trust, this may be due not to the perception of differences in income and wealth levels but to well-​publicized cases of public corruption. And as we saw in ­chapter  5, the observation of grand corruption, especially in the legal system, has a major negative effect on social and political trust. So it seems to me as likely that social trust leads to economic equality as that economic equality leads to social trust. And even if I’m wrong, the fact that there

42 For the importance of this distinction, see Acemoglu and Robinson 2013. 43 Bergh and Bjørnskov 2014, p. 192. 44 Indeed, people can notice economic inequality through other signals, like the presence of luxury goods that only a few people can afford, but there are similar cues for ethnic difference.

Against Egalitarianism  185 is good empirical work arguing strongly in favor of inequality being caused by distrust, and not the other way around, should sow doubt about the trust-​based case for redistribution. I think there is a more serious concern about inequality and trust. Perhaps sufficient economic inequality will make it infeasible to insulate the democratic process from extensive rent-​seeking for the superrich. When the superrich have enough economic power, limiting economic inequality becomes too difficult, if Gilens is correct. Gilens argues that democratic “responsiveness to the affluent” grew along with economic inequality for the past three decades.45 Now, as Gilens stresses, the mere fact that the affluent rule doesn’t mean that they always engage in rent-​seeking, or that poor never get the policies they want. But he does argue convincingly that when the affluent and everyone else differ on a policy question, the affluent get their way. And this implies that if the affluent pursue rent-​seeking policies, they will tend to succeed. Some of the best evidence for this is the pattern of increased real-​estate inequality; the affluent have enormous control over zoning at the local level. The difficulty with this common argument is that we lack data showing that increasing redistribution will allow a society to recapture democratic responsiveness after it is lost, and so it is not clear whether introducing new coercion will help to restore trust and stability.46 If responsiveness is degraded by high economic inequality, and it is hard to recapture resonsiveness once it is lost, this suggests a powerful argument for redistributing preemptively in order to maintain adequate responsiveness. If Gilens’s view meets the standards of policy epistemology, then, one can make at least a prima facie case that preemptive redistribution can be publicly justified as a means of preserving the democratic rights of lower-​income members of the public, limiting publicly unjustified rent-​seeking, and preventing the erosion of social trust. This will make coercion-​reducing redistribution easier to publicly justify, though it is less clear whether coercion-​increasing redistribution will become easier to publicly justify as well. I think this is the best argument for the redistribution of wealth beyond the provision of social services and public goods, and I am open to its providing a public justification. However, I do worry about whether preemptive redistribution, which some call predistribution, makes sense in the real world, because taxation is an ongoing political process without any recent beginning.47

45 Gilens 2012, p. 252. 46 In email correspondence, Gilens notes that we do not know whether responsiveness can be regained once it is lost. 47 For a version of this criticism, see Wilson 2017.

186  Trust in a Polarized Age

3.  Property-​Owning Democracy I will now turn to assess whether proposals further to the left, specifically property-​owning democracy and liberal socialism, can be publicly justified. Here I argue that neither regime type can be publicly justified because many moderately idealized persons will have sufficient reasons to reject both. Since these regimes have not existed, I can’t marshal trust data for or against them. Social trust levels in communist and post-​communist countries are considerably lower than in liberal-​democratic market orders, but advocates of property-​owning democracy and liberal socialists will probably insist that this is due to institutional factors that their preferred regime types will not share, such as a lack of democracy.48 However, given my arguments in c­ hapter 5, the facts that legal property rights will be more restricted in these orders, that large governments are given to increased rent-​seeking and public corruption, and that these regimes will experience slower economic growth than market-​based regimes, will generate defeater reasons and make these regimes less socially and politically trusting.49 Therefore, I think these regimes are unlikely to create trust for the right reasons. Before I  can assess these regime types, we must note that defenders of property-​owning democracy or liberal socialism usually argue that their preferred regime type is most appropriate for Rawls’s idea of a well-​ordered society, and so defending these regimes is an exercise in Rawlsian ideal theory. But this book sidesteps ideal theory by allowing many forms of noncompliance into the model of a morally peaceful polity, a society much like ours, but with high levels of social and political trust. Thus, to be fair to advocates of further-​left economic regimes, some of my concerns apply only in nonideal theory and should only worry property-​owning democrats and liberal socialists insofar as they wish to justify transitioning to their ideal regime from real-​world liberal-​democratic welfare state capitalism. But, as we shall see, other concerns apply to Rawlsian ideal theory as well. Property-​owning democracies (hereafter PODs) would combine the regulative and redistributive functions of the welfare state with the governmental aim of ensuring that wealth and capital are widely dispersed. Rawls, political philosophy’s most famous advocate of this approach, argued that property-​ owning democracy and liberal socialism were the two regime types that best realized his principles of justice, though he was vague about how their institutions 48 Bjørnskov 2019, p. 643. My guess is that black-​market participation, widespread corruption, and police-​state behavior, such as anonymous reports to secret police, keep social trust low, and those factors plus poor economic performance lead to lower political trust. 49 Note that Nordic countries are capitalist welfare-​states, and are far from property-​owning democracy and liberal socialism, so they will not count as counterexamples to my position. For evidence of Sweden on this point, see Bergh 2016.

Against Egalitarianism  187 are meant to realize his principles.50 To compensate for this deficiency, a number of political philosophers have recently tried to add institutional and policy content to the idea.51 Property-​owning democracy is usually contrasted with a Rawlsian form of welfare state capitalism. I shall understand a Rawlsian welfare state as combining private property rights in land, labor, and capital with an activist government that engages in countercyclical policy, extensive economic regulation, and the provision of social insurance.52 Property-​owning democracy adds the key governmental aim of broadening capital ownership “so that income returns from capital are broadly rather than narrowly distributed,” thereby producing “a more equitable pre-​tax distribution.”53 In other words, PODs engage in the “predistribution” of capital, setting pretax incomes at a more equal level from the economic start.54 In contrast to state socialists, who seek to abolish markets in productive capital entirely, however, Rawls, Martin O’Neill, and Thad Williamson stress that PODs have markets and a price system. But price signals are used primarily for “allocation” and not “distribution.”55 This means that the price signals are used to inform people what it makes sense to both produce and consume, but without leading to unequal ownership of income and assets. The extent of markets is also limited, as PODs have extensive public sectors to provide public goods that would otherwise be underprovided. Nonetheless, while their states will heavily regulate and control the market, PODs would still allow people to choose their careers and jobs, but not careers and jobs whose existence relies on large concentrations of capital. Further, competitive markets will help to decentralize economic power. O’Neill and Williamson propose five institutional features of a POD: (i) a right to equal public education, (ii) a right to minimum income and/​or the means for supporting oneself and one’s family at a minimal level of social acceptability, (iii) a public system of campaign financing and explicit limitations on corporate political activity, (iv) a right of individuals to share a society’s productive capital and/​or wealth, and (v) a collective right to sufficient productive capital to sustain viable democratic communities at the local level.56 Since (iv) and (v) are the 50 Again, we cannot point to real examples, because they do not exist. Nevertheless, for simplicity’s sake, I will usually refer to PODs as if they do or will exist. 51 For many of their articles, see the edited volume O’Neill and Williamson 2012b. Political-​ economic work on property-​owning democracy began with economist James Meade, whose work inspired Rawls. See Meade 1964. In this chapter, I focus on the Rawlsian interpretation of property-​ owning democracy. 52 In this way, a Rawlsian welfare state is more extensive than the welfare state I  defended in ­chapter 5. 53 O’Neill and Williamson 2012a. 54 Also see Thomas 2017. 55 Rawls 1971, p. 273. 56 Ibid. It is unclear what “local” refers to.

188  Trust in a Polarized Age focus of this section, I define PODs as recognizing individual capital rights and collective capital rights in the senses just described.57 To implement collective capital rights, that is, to help the economy better sustain local democratic bodies, both O’Neill and Williamson argue that the workplace should be largely democratic.58 Workers, in their view, should not only govern their own terms of employment but, possibly, be given part of the task of enforcing workplace regulations and, perhaps, nontradable coupons for fixed stock ownership. Or a government institution could hold the stock for them. We can see then that property-​owning democracy is democratic in two senses: it protects equal civil and political liberties via a representative democracy, and it protects worker control of the workplace. I argue that, in comparison to a market-​based order with social insurance, property-​owning democracy cannot be publicly justified because in normal liberal democracies many members of the public will have defeater reasons for the additional coercion, over and above the coercion required to establish a welfare state, that it takes to establish and run a POD.59 A POD must also avoid violating the principle of sustainable improvements by reducing growth, but it is likely to do so, as we will see. On top of this, policy epistemology requires meeting demanding standards of evidence to show that regulation and public policy can be publicly justified, and the more radical policies of a POD will face an even greater epistemic hurdle to being publicly justified. Finally, concerns about economic inequality, while valid, have limited power to justify even coercive redistribution of income, which is far less coercive than restructuring the workplace. Let’s explore these concerns.

4.  Property-​Owning Democracy Is Inefficient Property-​owning democracy cannot meet these hurdles due to concerns about its economic efficiency. PODs distribute capital holdings with the express purpose of preventing high concentrations of capital in the hands of private actors (though, oddly, not in the hands of government). But capital dispersion can undermine growth by destroying the effective accumulation of capital needed to improve the production of goods and services and raise wages. This is due to three kinds of problems:  incentive problems, information problems, and 57 See Williamson 2013 for a discussion of how to guarantee these rights through constitutional amendments. 58 O’Neill 2009, pp. 33–​42. 59 In Vallier 2015, I  argued that the case for property-​owning democracy fails on expressly Rawlsian terms on the grounds that PODs are inefficient and unjust compared to the welfare state, but I will not advance that Rawlsian argument here.

Against Egalitarianism  189 compliance problems. In a POD, citizens lack incentives to produce efficient outcomes, they lack the information to produce efficient outcomes, and they face great temptations to engage in destructive rent-​seeking.

4.1. Incentive Problems Everyone in a POD knows that capital accumulation has legal limits dictated by whichever POD bureaucracy tracks and modifies capital accumulation across an entire society. As a result, effective capital accumulators will only seek to accumulate capital up to the point where doing so exceeds the legal maximum. PODs effectively tell capital producers to stop producing capital that they would produce if they could own and operate it. What’s worse, the positive externality created by the accumulation of capital, namely the signaling generated by the presence of a few successful capital accumulators in a particular area of the economy, is distorted. Much like consumer prices, profits are relative informational signals: they tell entrepreneurs how to redirect their resources to increase profits. If these profit signals are significantly blunted by a POD bureaucracy’s distribution of capital, there will be less incentive to engage in the activities that create the relevant signals. To put it another way, capital accumulation signals unmet demand, meaning opportunities to obtain large profits in some markets.60 Over time more firms will enter those markets, lowering the rate of profit and the price of goods, along with increasing the supply of those goods. In this way, capital will flow into parts of the economy where it is most needed. POD administrators interfere with this process. Consequently, capital will not be allocated to its most efficient uses. A natural reply is that capital recipients will do something useful with the capital they are given. But the question is whether the recipients will, statistically, produce more efficiently than those from whom capital is confiscated. Markets tend to draw relatively more effective producers to sectors with high profit margins, and PODs block that process. PODs thus tend to take capital from those who are better producers and give it to those who are less effective. So it is hard to see how the new uses of capital will balance out these disincentive effects. What’s worse, PODs face transfer costs. So far, I  have assumed that redistributions are costless, but this is false. Moving capital around will inevitably waste part of the capital stock. Another set of disincentives applies to the recipients of state funds: the marginal benefits of working decrease for a great many people. To the extent that the 60 Save in cases where capital is acquired through rent-​seeking, or there are high barriers to entry. I thank Eric Schliesser for this point.

190  Trust in a Polarized Age state provides citizens with unconditional funds, say through savings, checks, and services, those citizens can avoid working to acquire what the government gives them. Thus, the productivity that would be generated by citizens who would otherwise have to work to earn these resources will be lost. A POD might provide a 25-​year-​old with all of the following: free public education through graduate school, a $100,000 savings account awarded at adulthood, public housing, a permanent stock holding in the company that offers him his first job, and a guaranteed, unconditional minimum income. Surely people will be less productive workers if they receive all these goodies for free and are taught that they are owed them as a matter of justice. This is not to say that any pre-​or redistribution from haves to have-​nots will generally create disincentive effects but rather that at this level of distribution, there will be significant disincentive effects. Welfare states suffer similar problems to the extent that they engage in redistribution, but they redistribute much less, so their disincentive problems are less severe. Three replies suggest themselves. First, O’Neill and Williamson could argue that the property-​owning democratic governments are sufficiently able, in ideal conditions, to spend the redistributed income in socially productive ways, counteracting the disincentives I  point to. But it seems obvious, at least to me, that offering people $100,000 for free is not that sort of social investment. A second reply concedes that redistribution has disincentive effects but preventing inequality in the first place, also known as predistribution, actually counters growth disincentives.61 The difficulty is that even O’Neill and Williamson admit that predistribution requires all kinds of redistribution, so it is not clear how predistribution can avoid disincentive effects.62 Finally, one could argue that if people did not have to fear financial ruin, they would be willing to engage in more creative risks, which would in turn have positive spillover effects for the productivity of the economy as a whole. But welfare states already provide safety nets, investment in public education, and bankruptcy and limited liability law. It is hard to believe that offering people, say, an additional $100,000 would increase economically beneficial risk-​taking sufficiently to counteract disincentive effects, given what welfare states already have in place. A third set of disincentive effects is even more obvious. PODs will block a wide variety of market exchanges. Thus, insofar as these exchanges would benefit both parties, creating consumer and producer surplus, PODs prevent citizens from generating these benefits.63 This is a general point about restrictions on market transactions, but it is especially acute for PODs because they block

61 For some evidence on this issue, see Alesina and Rodrik 1994. 62 Narveson 2013 raises some similar worries about how to distinguish predistribution and redistribution. 63 This point is familiar from introductory economics textbooks. See Mankiw 2011, chap. 8.

Against Egalitarianism  191 free exchanges en masse through enormous regulation of the economy. POD proponents could try to argue that the economic benefits of a POD will outweigh these costs. Perhaps they could point to effective public investment or even negative externalities produced by these market exchanges. But POD proponents have provided no argument for either position. Welfare states can invest in education and correct for negative externalities, and they can do so more “cleanly” by interfering less with market transactions. A fourth set of incentive problems arise from the creation of worker-​owned-​ and-​operated firms. The standard worries about worker-​operated firms are twofold:  underemployment and underinvestment relative to capitalist-​operated firms. In mainstream neoclassical economic theory, worker cooperatives have a disincentive to hire more employees when their capital stock expands because a new worker would divide current workers’ shares and therefore reduce their value. So, while in the long run it may be in the interest of workers to hire more people, in the short run they would have to divide their shares and so reap fewer benefits from them.64 A capitalist boss does not face this disincentive because adding an additional worker does not reduce his stock or income. To see the underinvestment problem, consider that worker cooperatives will have workers at different stages of life. Middle-​aged workers will have an incentive to invest in the company so as to expand the value of their holdings. But younger workers face a disincentive to invest because typically worker cooperatives in socialist or left-​ wing economic theory do not allow the sale or exchange of shares to nonworkers. If the cooperative did not impose these restraints, shares would quickly move into the hands of nonworkers and the cooperative would cease to exist even if the firm survived. So, if younger workers would like to be able to change jobs, they will prefer to receive income than to invest in their nontransferable shares. A similar story holds for older workers. Older workers nearing retirement will prefer a stream of income to investment, as their time horizon is far shorter.65 Capitalist bosses do not face these disincentives, as they typically have an interest in increasing the capital stock of their firm.66 There is empirical evidence that these effects do not manifest in many firms with partial worker ownership and operation.67 But these firms all allow for 64 For modeling and data to this effect, see the classic work of Vanek 1970. Vanek finds that employees tend to maximize their shares rather than profits. 65 For a model of underinvestment effects, see Ward 1958. For a review of classic models of this sort and empirical evidence, see Dow 2003. Note that standard proposals for workplace democracy prevent people from divesting their capital stock into income when they leave that workplace, so those who leave face a substantial economic loss. 66 Though as with workers, we may expect them to become more risk-​averse as they age, though less so because of extremely generous retirement packages. 67 See the introduction and articles in Kruse, Freeman, and Blasi 2011. The effects that do manifest can include fewer principal-​agent problems and less rent-​seeking. Singer 2019 canvasses a lot of recent evidence on this topic.

192  Trust in a Polarized Age types of capital transfer that workplace democrats often seek to prohibit. For instance, firms where workers own most of the capital stock allow workers to trade it under certain conditions. A defender of property-​owning democracy will likely reply as follows. All PODs aim to increase the number of capital owners relative to society’s capital stock. If capital redistribution is successful, the economy will have more capital producers, both because it will have more firms and because it will have more worker ownership of those firms. A broader distribution of capital has a great many economic advantages. First, it increases the diversity of market actors, which may generate thus-​ far-​ unacknowledged innovations. Further, more-​ diverse markets are less prone to disruption if a single firm goes out of business. In other words, there are a number of “dynamic effects” generated by PODs that I have failed to consider. For these replies to succeed, however, we need reason to think that these “dynamic effects” are sufficient to outweigh the more concrete economic costs I have cited relative to welfare state capitalism. Unfortunately for POD advocates, we have little reason to think that. For instance, if a wider dispersion of capital would be that much more productive, why haven’t markets already converged on such a distribution? One could point to regulations that favor larger concentrations of capital, but we may be able to remedy these concentrations by repealing the relevant regulations rather than engaging in redistribution. POD advocates would do better to argue that most capital markets under welfare state capitalism are market failures in the economist’s sense, in that they fail to realize a large number of feasible Pareto improvements. But this reply would require supplying an economic model where market failure occurs in welfare states but not in PODs, and I cannot detect any attempt among property-​owning democrats to provide such a model.

4.2. Information Problems PODs have an extensive bureaucratic apparatus that is charged with extraordinary tasks, such as making specific numerical judgments about appropriate price levels, wage levels, stock distributions, and the like.68 They must do so without the information available to market actors, namely that provided by a free and functioning market. While POD advocates like to claim that they advocate a free market to allow for the allocation of goods and services, what they describe is a state that constantly intervenes in the market. After all, what are all



68

They need not fix price and wage levels, but rather delineate an efficient and just range.

Against Egalitarianism  193 their enormous, powerful bureaucracies for if not to interfere with the market process? To see how intervention distorts and misrepresents market signals, let’s imagine a mostly capitalist society that only redistributes wealth via a negative income tax and is gradually transforming into a POD. In the beginning, markets will suffer from few distortions given that there is (a) just one major redirecting force (the negative income tax), and (b) the force is predictable, so firms can adjust their production schedules around it. As we add POD bureaucracies, multiple forces begin to interact, constantly readjusting and redirecting the market process. POD bureaucracies block exchanges that signal profit opportunities. They take over industries to increase the provision of public goods but lack the requisite profit-​and-​loss accounting faced by ordinary firms that rely almost entirely on sales rather than government funding. PODs block capital exchanges through regulations and monitor profit levels to stop excessive capital accumulation, preventing profit signals from reallocating capital as the economy develops. They prevent the transfer of capital stock from workers to capitalists to maintain workplace democracy and, in doing so, prevent others from determining whether such exchanges are in their economic best interest. Furthermore, the need for administrative law to maintain the right distribution of capital will require considerable regulatory flexibility, so consumers and producers will have an unstable economic environment. As a result, they will have a harder time predicting which purchases and sales make economic sense. This is why Hayek approved of a social minimum but strongly opposed a state that would regulate the economy through administration-​made law.69 Markets are information systems, and extensive regulatory states impoverish the information available, both in quality and in quantity. This generalization holds even if the government does not own or operate firms. Administrative tinkering dumbs down the informational coordinating power of the market process. These information-​based arguments illustrate a general concern about the Rawlsian attempt to separate the “allocative” and “distributive” functions of prices. Rawlsians want the benefits of efficient market allocations without producing unattractive distributive outcomes, such as excessive income inequality. However, income inequality itself is a market signal, as those who make money in a particular economic sector will grow relatively wealthier than others, indicating that there is money to be made in that sector. POD advocates believe in a government that regulates in just the right ways to ensure that productive resources go to the same places they would if the relevant regulations and controls



69

Compare Hayek 1979, p. 55 with Hayek 1973, pp. 134–​6.

194  Trust in a Polarized Age didn’t exist, or to related places. But how are the bureaucracies to know what those places are? What prices to set? What profits to limit?70 Of course, welfare states engage in many of these same regulations. If so, why think the informational problems are unique to PODs? The answer is that welfare states usually do not attempt to monitor and control the capital stock as deliberately and intricately as PODs. Accordingly, distortionary effects are considerably reduced. POD advocates will probably respond by trying to make their ideal more flexible. Maybe property-​owning democracy, for instance, does not require the massive bureaucracies that Rawls endorsed. Perhaps instead we could restrict the state to pursuing the least distortionary means to achieve the wide distribution and wide operation of productive capital. But what would these methods be? Williamson and O’Neill give us little help here, Rawls even less.71 To realize individual and collective capital rights as understood in property-​ owning democracy, then, a society needs a powerful government that will invariably be constrained by information problems, even in ideal theory. And while we might try to overcome incentive problems by demanding, through regulations, that people do what the system needs them to do, that’s a nonstarter here. Informational problems make it difficult for even properly motivated people to determine how to act. Strict compliance does not increase access to economic information.

4.3.  Rent-​Seeking Finally, in nonideal theory, we have excellent reason to worry that the great power of these bureaucratic organizations will attract rent-​seekers, who will gradually capture the organizations and run them for their own benefit. These “public choice” problems could cripple POD economies vis-​à-​vis capitalist welfare states and, insofar as such corruption is publicly observed, will hurt social and political trust. Administrators often have incentives to alter prices or distribute holdings in ways that benefit their friends and allies. Political leaders can order 70 Waiting for feedback from elections will take a long time, and is unlikely to yield the fine-​ grained information required. 71 In conversation, Alan Thomas has suggested to me that one route to property-​owning democracy is through the arguments of Keynes and Marvin Minsky that some concentrations of capital are productive and others unproductive. The goal of property-​owning democracy, he said, is to have government move capital from more productive to less productive uses, not to impose a capital ceiling. This is a unique feature of Thomas’s model, but for reasons cited in this paragraph, I’m highly skeptical that states could systematically distinguish between productive and unproductive uses of capital. Economists can’t even agree on what the distinction comes to and whether it applies in a market economy. And even if we had a clear understanding of the distinction, I have serious doubts that governments either could or would track it in their public policy.

Against Egalitarianism  195 administrators to engage in popular but destructive policies that sacrifice long-​ term gains for the short-​run benefit of winning elections. Individuals can lobby politicians to provide them with pork-​barrel benefits that aren’t worth the social cost.72 These familiar criticisms, I want to emphasize, are fair game. Rawls, O’Neill, and Williamson claim that even in ideal theory, capitalist societies allow firms to misbehave by co-​opting the political process. If powerful interests can control and corrupt government in ideal welfare states, they can in nonideal PODs.73 Given the extraordinary power that Rawls, Williamson, and O’Neill must assign bureaucracies, the incentives for regulatory capture are enormous.74 A state powerful enough to realize a POD’s individual and collective capital rights is worth staffing with friendly regulators, as the state could potentially destroy an individual’s or group’s economic livelihood through regulation and redistribution. For instance, suppose that Reba, a businesswoman, has accumulated just enough capital to come under the scrutiny of a POD bureaucracy. She is now subject to monitoring. Perhaps Reba’s workers have complained to administrators that they have too small a share of the business owner’s capital stock, or perhaps a competitor has tipped off a friendly regulator that Reba has been making new capital stock purchases. In both cases, she has a powerful incentive to resist regulation in any number of ways, perhaps by placing her friends on local regulatory boards. Alternatively, she could join an otherwise legitimate lobbying organization and push its activities in her direction, or she could use her nonmonetary influence and energy to elect politicians friendly to her views.75 A critic might claim that the new social motivations and norms that citizens acquire in a POD will make rent-​seeking behavior manageable. And we already see that social expectations and norms restrict government corruption in some advanced democracies. Nordic nations have relatively large governments but relatively little corruption. In a POD, given that people will have a number of new or strengthened moral motivations, these social norms might be easier to enforce via our ordinary moral practices. Accordingly, POD political cultures can restrain some amount of rent-​seeking. The problem with this “social norms” reply is that it can be used to restrain rent-​seeking behavior under other regime types. After all, Nordic states aren’t PODs, they’re welfare states with extensive markets, 72 For a review of rent-​seeking models of regulation and the political process, see sections 15.2 and 15.3 of Mueller 2003. 73 And if Chris Freiman (2017) is correct, they must allow the same institutional failures in PODs as they do in welfare states in ideal theory alone. 74 For a classic discussion of models of regulatory capture, see Stigler 1971. 75 Note that this point is not undermined by the claim that in a POD there will be far more businesses and business owners, such that social coordination between them to influence political life will be weakened. Regulatory capture can occur gradually through the accumulation of local rent-​seeking groups. This process is most famously described in Olson 1971.

196  Trust in a Polarized Age and those markets are almost as free—​and in some respects freer—​than markets in the United States.76 In some ways, then, appeals to social norms can undermine the justification for property-​owning democracy, since if social norms can be used to reduce corruption and rent-​seeking in welfare states, we have even less reason to move to a POD. So it is fair to say that rent-​seeking will be worse in PODs than in welfare states, given their greater concentration of state power. POD defenders must reply by arguing that social restraints against rent-​seeking will be more prevalent in PODs than in welfare states, but they have not made this argument yet.

5.  PODs Cannot Be Publicly Justified Now let’s bring the rest of my justificatory apparatus into the argument. First, we have a publicly justified principle of sustainable improvements. If my arguments in this section are successful, then PODs violate the principle of sustainable improvements vis-​à-​vis welfare state capitalism because PODs will have lower growth rates. Second, PODs require substantial coercive interference, far beyond redistributive taxation, and this creates vast opportunities for defeater reasons to undermine the legal rules required to institutionalize a POD. Moreover, given the incentives for corruption in a nonideal POD, such a state is likely to encourage corruption that reduces political trust, and probably social trust as well. Real-​world PODs are likely to be poorer, more coercive, and more corrupt than capitalist welfare-​states. These considerations serve as defeaters for property-​ owning democracy. But suppose that my claims about property-​owning democracy are all false. So long as many reasonable members of the public share my concerns, they will have defeaters for the additional coercion involved in imposing property-​ owning democracy on a diverse public. From their perspective, the policies that jointly realize a POD are individually defeated, primarily because these policy units will not meet the standards of policy epistemology for them, and they are likely to form reasonable beliefs that property-​owning democracy will reduce growth, such that property-​owning democracy will, in their minds, run afoul of the principle of sustainable improvements. It is also worth remembering that actual people have been surveyed on their conceptions of distributive justice, and, as Frohlich, Oppenheimer, and Eavey 76 https://​www.heritage.org/​index/​ranking. In most cases, the Nordic countries fall behind the United States due solely to their high tax rates and high government spending. On nearly all other metrics of market freedom, Nordic countries have freer markets. In the 2019 ranking, Denmark and the United States are within one tenth of a percentage point of one another on overall economic freedom, with Sweden and Finland close behind.

Against Egalitarianism  197 found thirty years ago, most people’s intuitions align with an “average utility with a floor” approach that Rawlsians reject.77 And an “average utility with a floor” conception of distributive justice has far fewer resources than Rawlsian justice as fairness to justify property-​owning democracy, since property-​owning democracy is likely to reduce average utility vis-​à-​vis welfare state capitalism and welfare state capitalism can provide an adequate floor. This suggests that many members of the public in the real world reject egalitarian conceptions of distributive justice such as those required to justify property-​owning democracy. Consequently, many of them will have defeaters for POD-​based coercion. We can conclude, then, that property-​owning democracy and public reason liberalism are incompatible given the present diversity of opinion in liberal democratic orders. Perhaps POD advocates can convince those of us who disagree.78 But until then, property-​owning democracy is not a viable governance structure for a free society. Thus, the property-​owning democrat is a kind of sectarian, much like a Catholic establishmentarian. The one may have the correct conception of justice (say, justice as fairness) or the other the correct conception of the good (traditional Roman Catholicism), but neither form of hegemony can be justifiably imposed on those who reasonably disagree.

6.  Workplace Coercion and Coercive Taxation Consider two replies to my argument that property-​owning democracy cannot be publicly justified. First, defenders of property-​owning democracy can argue, as I have, that unions are civic associations that have a primary right to organize to control their economic lives. If capitalists use coercion to prevent them from exercising their freedom of association, say by threatening to fire them, then the capitalist may be using publicly defeated coercive threats to get his way. In that case, the state might have a role in prohibiting the capitalist from engaging in these threats. On this basis, we can build various labor-​rights protections into the legal order of a publicly justified polity. So, defenders of property-​owning democracy have a point in complaining about workplace tyranny. Indeed, employers frequently coerce their employees, and the mere fact that an employee chose to take that job does not mean that she is not coerced, as Elizabeth Anderson has forcefully argued.79 If you go to work for a firm and your boss threatens to fire you unless you have sex with him, that’s coercion, and the state has reason to 77 Frohlich, Oppenheimer, and Eavey 1987. Lissowski, Tyszka, and Okrasa 1991 replicated this result. Also see Beck 1994. While Swope et al. 2008 find some evidence for Rawls’s conception, they find much less risk aversion than Rawls postulates. 78 Or rather, convince those who disagree at the right level of idealization. 79 Anderson 2017.

198  Trust in a Polarized Age interfere to prohibit such coercion if it cannot be stopped through social morality. Similarly, labor protections of many varieties can be publicly justified in the name of restricting employer-​initiated coercion. Anderson’s work on private government helps to make this point. You are, she says, subject to private government whenever (1) you are subordinate to authorities who can order you around and sanction you for not complying over some domain of your life, and (2) the authorities treat it as none of your business, across a wide range of cases, what orders it issues or why it sanctions you.80

Employers often exhibit intense control over their employees, and they can only do so because the state protects their property rights with coercive force when they engage in acts of private government. Returning to the case of sexual harassment, if a boss sexually harasses his female employees, and they refuse his advances and he fires them, that usually means they can no longer enter their workplace or collect an income from their employers because they will be forcibly prevented from doing so by the police. Again, employer coercion must be publicly justified. And it surely will not be. Importantly, Anderson does not advocate property-​owning democracy but instead offers more modest policy arrangements, such as increasing worker exit rights, and requiring companies to have some kind of rule-​of-​law arrangements that prevent arbitrary interference, establish a worker’s bill of rights, and give employees voice in the workplace.81 Depending on how these policies are fleshed out, and insofar as the rights are coercion-​reducing, they may be eligible for public justification even among people who think that extensive property rights in capital are best. The biggest worries are that these policy proposals might not be the least coercive policies for limiting workplace coercion and that they might undermine economic growth.82 So even if we recognize workplace coercion, the case for property-​owning democracy remains weak. Property-​owning democrats cannot merely show that members of the public have strong objections to workplace coercion—​ that they possess defeater reasons—​they must show that publicly justified economic institutions cannot effectively be established with any less coercion than 80 Ibid., p. 45. 81 Anderson 2017. 82 These points are explored by Tyler Cowen’s reply to Anderson, especially Anderson 2017, pp. 111–​3. Here one might reply to Cowen that, while economic growth is important, these policies might be best for affirming workers’ standing and equal freedom as citizens. But remember that I am capturing concerns about standing and equal freedom with my standard of public justification, and the principle of sustainable improvements is part of what publicly justifies certain arrangements to most members of the public. I thank Christie Hartley for raising this concern.

Against Egalitarianism  199 is needed to establish a POD. That is difficult. Labor laws in welfare states can go a long way toward protecting the primary right of workers to resist employer control. For example, welfare states can protect the associational right to unionize. Imagine a society in which all persons are free to unionize on what terms they like and in which few large firms have the right to coercively penalize them for doing so. Unionization rates could increase substantially, and the welfare of workers might be much better protected than it is now. Under such a regime, workers in a free economy could do quite well. Also remember that a publicly justified social order almost certainly provides social insurance, which may considerably reduce the costs of losing one’s job; this too will strengthen the position of workers relative to capital. Additional coercion, then, may be unnecessary.83 Now, economists widely agree that some union activities actively reduce the incentive of employers to hire new employees, since unions fight for higher wages, and generate higher consumer prices, because employers pass some of the cost of those wages on to consumers. As a result, the principle of sustainable improvements, when combined with standard economic concerns about the effects of unionized labor on wages and prices, may well meet the standards of policy epistemology and so justify some restrictions on collective bargaining in the workplace. The second, and best, reply to the argument that PODs are defeated is that PODs may not be more coercive than welfare states. Paul Bou-​Habib has provided several arguments for this claim. Bou-​Habib first argues that PODs’ higher tax rates are not necessarily more coercive than the lower tax rates required to have a pure capitalist state or a welfare state. He claims that “the extent to which citizens will comply with tax laws depends on, among other things, their sense of justice—​particularly whether they morally endorse the economic regime in question.”84 Bou-​Habib backs up this conjecture by claiming that, in cross-​cultural studies, the empirical literature on tax evasion is “decidedly mixed”; he cites empirical evidence that tax compliance may not fall as tax rates increase and that, across cultures, where tax rates are higher, so is tax compliance. However, several tax-​evasion studies conclude that tax-​evasion rates are determined by a variety of factors and that higher tax rates do correlate with more tax evasion. For instance, Gamze Oz Yalama and Erdal Gumus find that “increases in the tax rate and in tax burden increase

83 However, unions also engage in coercion and insist on coercive protections for themselves, including from workers in their section who do not wish to join the union. These forms of coercion are probably often defeated, which will limit the capacity of unions to resist employers. But unions would nonetheless be strengthened if most of the coercive restrictions they face were removed. 84 Bou-​Habib 2015, p. 656.

200  Trust in a Polarized Age tax evasion,” confirming results in five other studies.85 Bou-​Habib cites Grant Richardson’s work on the determinants of tax evasion, but Richardson focuses on factors other than high tax rates.86 The main finding that Bou-​Habib appeals to is that tax evasion is partly determined by “tax morale,” which Erzo Luttmer and Monica Signhal define as “an umbrella term capturing nonpecuniary motivations for tax compliance as well as factors that fall outside the standard, expected utility framework,” such as feelings of “guilt or shame for failure to comply.”87 They note that tax morale can be raised when government attempts to “increase[e]‌ voluntary compliance with tax laws” while “creating a social norm of compliance.”88 Attempts to raise tax morale include shaming tax evaders and instituting public campaigns to encourage tax compliance. Bou-​Habib argues on this basis that higher tax rates might not lead to more tax evasion because the taxed might have higher tax morale, perhaps because they might believe the higher tax rates are fairer, and this would lead them to comply with the higher rates. The problem with Bou-​Habib’s position is that evaluative pluralism suggests that moving to property-​owning democracy will not raise tax morale across a diverse population. Some may be sufficiently inspired by POD institutions that they will happily pay higher taxes, but people who do not endorse property-​owning democracy may nonetheless be strongly tempted to evade higher taxes; such individuals will experience the higher tax burden as coercive. Libertarians and conservatives will surely experience the higher tax burden as coercive. Bou-​Habib disputes the idea that higher tax rates are more coercive on other grounds as well. The only reason that higher tax rates appear to be more coercive, and so a greater cost on our options, he argues, is because “we fail to appreciate that the structure of social relations that make up a distribution of private property in our society is a revocable product of social choice.”89 Property rights are generally products of coercive enforcement, so the state interacts with citizens no less coercively when it protects private property than when it imposes high tax rates. The problem with this argument is that it is subject to a simple reductio ad absurdum. I have already granted that the enforcement of property rights is coercive, but different schemes of property rights are not equally coercive. If they were equally coercive, then the Soviet Union’s enforcement of collective property



85

Oz Yalama and Gumus 2013, p. 21. Richardson 2006. Luttmer and Singhal 2014, p. 150. 88 Ibid., p. 149. 89 Bou-​Habib 2015, p. 657. 86 87

Against Egalitarianism  201 rights would have been no less coercive than is the enforcement of property rights in present-​day Denmark. Surely, then, there is some way to distinguish between more and less coercive regimes while simultaneously recognizing that property rights are coercively enforced. One way to do so is to compare the options available to citizens under different regimes, and this is partly a function of a society’s level of economic prosperity, as greater riches are associated with more choices. If market orders are likely to be more efficient than property-​ owning democracies, as I  have argued, then people should generally have more options under welfare state capitalism than under property-​owning democracy. Increases in taxes that move us toward property-​owning democracy should therefore prove more coercive over the long run in virtue of reducing economic growth. There is also a more immediate coercive impact due to the fact that some property rights can be sustained without legislation. In the absence of state intervention, we can expect moral and some legal rules to form around the basic property holdings of citizens, as property norms form in a vast and diverse array of human social circumstances. This means that tax legislation will increase coercion because it disrupts the rules that would be in equilibrium in the absence of such laws. While the state uses coercion to protect property rights, legislation and legislative bodies are not always required to enforce them, which means that state interference with these rights is often an increase in coercion. Thus, while property rights involve social decisions, they do not always require legal decisions, and certainly not always legislative decisions. I am sometimes struck that people who use arguments like Bou-​Habib’s almost never distinguish between the different types of norms required to establish property rights; they frequently assume that these norms must be imposed from the top down by force. But this assumption is unsupported. Justified property rights are vague and do not rule out redistributive taxation in principle. However, moral and legal rules that specify property rights rule out some redistributive taxation. Consequently, we can expect redistributive taxation to become less optimal, and eventually less than eligible, as tax rates increase. Given that PODs require high levels of capital taxation, they require considerable coercion. The levels of taxation will violate property rights because they are too coercive. For this reason, PODs cannot “constitute the correct way in which to partition social space into individual spheres of authority.”90



90

Ibid., p. 611.

202  Trust in a Polarized Age

7. Liberal Socialism Liberal socialism combines the standard list of liberal liberties with government ownership and operation of most productive resources.91 It is typically seen as further to the left than property-​owning democracy, since it replaces the wide private dispersal of capital with a concentration of capital in the hands of the government. This regime cannot be publicly justified for many reasons. The first and most obvious reason is that we have strong reason to believe that such an order will dramatically reduce the bundle of resources available to each person, given the inefficiencies in a planned economy attributable to information problems and incentive problems. The efficiency-​related concerns that apply to property-​ owning democracies apply much more strongly to liberal socialism, given that it lacks a competitive pricing mechanism for capital.92A second concern is that socialism would limit the scope of the projects and life plans that people can pursue; some, such as followers of Ayn Rand, think that “entrepreneurship is itself a form of human flourishing. Start-​ups, innovation, risk-​taking, organizing groups to solve problems and implement new ideas—​all are basic to the projects, plans and ideals of many.”93 Liberal socialism, therefore, will stymie those whose personal ideals include the ownership of private capital, since under this type of regime capital is primarily owned and operated by the state. Liberal socialists will reply by arguing that participation in a socialist order is included in their personal ideals. But liberal socialism is extremely coercive, and there is, as we have seen repeatedly, a presumption against coercion that must be met by a public justification. A further problem can be drawn from Hayek’s The Road to Serfdom.94 I read Road not as a defense of laissez-​faire capitalism but rather as an argument that liberal socialism cannot be stable for the right reasons. Hayek’s argument is that the liberal rule of law and individual freedoms are incompatible with a socialist economic order and that a socialist economic order must invariably collapse into a totalitarian socialist order. The argument is remarkably proto-​Rawlsian, as it assumes that the “scales of value” of rational and moral persons “are inevitably different and often inconsistent with each other.”95 Given that everyone’s 91 Rawls 2001, p. 138. For an argument that Rawls was a “reticent” socialist, see Edmundson 2017. Unfortunately, Edmundson’s book lacks an adequate defense of liberal socialism’s relative economic viability vis-​à-​vis the welfare state and property-​owning democracy. It focuses primarily on arguing that justice as fairness supports socialism, but on my view, justice as fairness’s socialist implications will be defeated, at least as a result of justice pluralism. 92 As described in Hayek 1945, a famous piece surprisingly ignored by Rawls and other advocates of liberal socialism. 93 Gaus 2012, p. 100. 94 Hayek 2007. 95 Ibid., p. 102.

Against Egalitarianism  203 fate is bound up in the details of the central plan, there will be rampant disagreement about how to proceed as a collective, and this disagreement will be present among democratically elected officials. But given how much is at stake, Hayek foresaw that interminable squabbling would make democratic leaders look weak and ineffectual, particularly during economic downturns, when public anxiety increases. This would leave the liberal-​socialist regime vulnerable to takeover by some group or sect that promises to end the debate and impose a single plan based on the sect leader’s empathic connection with the people. Under liberal socialism, people are not allowed to go their own separate ways when it comes to decisions about economic production, a freedom that markets protect. This winner-​take-​all aspect will lead to ferocious fights about how to plan the economy, and the fights will undermine the stability of democratic regimes, leading to totalitarian government. An advantage of private property in the means of production is that it allows people to pursue different ends while coordinating the operation and production of capital; that is far less feasible in a socialist regime. Some have attacked Hayek for making a bad slippery-​slope argument, perhaps based on his purported claim that welfare states must lead to totalitarianism.96 Critically, Hayek expressly embraces the principle of social insurance by supporting “the assurance of a certain minimum income for everyone.”97 Without getting into the weeds of Hayek exegesis, I will simply note that I have elsewhere weighed in on the debate about Hayek’s concerns about the welfare state by distinguishing, as Hayek does implicitly, between the welfare state of law and the welfare state of administration.98 The welfare state of administration is not the same as a social insurance or safety-​net state. It is instead based on a robust conception of distributive justice that requires interference with a vast number of the internal workings of the market. Hayek is concerned only about coercive redistributive institutions that permit constant state tinkering. Hayek affirms the welfare state of law, where clear, public, general principles, rather than extensive administrative bodies, regulate social insurance and other state functions.99 The problem with liberal socialism, then, is that it is a welfare state of extreme administration. It does far more than engage in regulation and the equalization of incomes; it pursues distributive justice by owning a society’s major concentrations of capital and operating them according to a central plan. Such a

96 Farrant and McPhail 2010 both advance and discuss other versions of this criticism. They argue that Hayek thought the welfare state would lead to serfdom, in contrast to Hayek scholar Bruce Caldwell in Hayek 2007, pp. 29–​31. 97 Hayek 1979, p. 55. I discuss Hayek’s endorsement of a minimum income as a necessary condition of legitimacy while simultaneously rejecting social justice in http://​bleedingheartlibertarians. com/​2012/​05/​hayek-​enemy-​of-​social-​justice-​and-​friend-​of-​a-​universal-​basic-​income/​. 98 http://​bleedingheartlibertarians.com/​2012/​05/​hayek-​on-​serfdom-​and-​welfare-​states/​. 99 I discuss these details of Hayek’s views in Vallier 2020b.

204  Trust in a Polarized Age regime will prove unstable for the reasons outlined.100 We can see that the logic of evaluative pluralism applies to deciding what to produce. Deep disagreement will lead to the intense struggles that weaken democratic orders and their ability to resist totalitarianism. Understood in this way, Hayek was right. We have never seen a liberal-​ socialist regime, only liberal welfare states and illiberal-​socialist regimes. While many liberal welfare states moved in a socialist direction following World War II, they did not approach the liberal-​socialist ideal. Clement Attlee’s Labour government in Britain socialized only 20 percent of society’s productive resources.101 Furthermore, after a few decades, liberal-​socialist-​leaning parties retreated from their programs and started selling off state enterprises. Real-​world socialist regimes destroyed their liberal democratic institutions; liberal democratic regimes eventually backed away from socialist policies. Liberal-​socialist regimes can be expected to undermine social and political trust in many ways, if they can manage to exist at all. They assault legal property rights, they’re bound to invite enormous rent-​seeking and public corruption, as we found in all real command economies, and they’ll be inefficient, hurting economic performance. They will also create legislative gridlock, which we’ll see in the next chapter is unpopular and undermines political trust. And if Hayek is right that liberal-​socialist regimes are unstable and become totalitarian or authoritarian, then they’ll lead to social and political trust getting reduced by authoritarian rule.102 Importantly, Thomas rejects liberal socialism for similar reasons; in a liberal-​ socialist order, government control of capital “will be likely to prevent workers in democratically organized firms from exercising any real control over the operations of the enterprise.”103 Public authorities can engage in coercive control and domination just as private authorities can.104 I would like to end by stressing that, from empirical work, we know that communism is one of the few regimes that can consistently destroy not just political trust, but social trust. Bjørnskov puts it simply: “Communism destroys trust.”105 This may be because communist orders are nondemocratic, or because they 100 Note that this argument holds without the claim that a principle of social insurance leads to serfdom. 101 Thorpe 2001, p. 124. 102 This should not be undermined by the empirical finding that authoritarian regimes sometimes have higher political trust than democratic regimes, as few of these authoritarian regimes are socialist and those that are or have been have destroyed previously high levels of trust. 103 Thomas 2017, p. 183. Thomas also has a series of useful criticisms of arguments for liberal and democratic socialism, in particular that justice as fairness rules them out. See pp. 216–​71. 104 Schemmel 2015 argues that universal welfare programs can help answer the criticisms of the welfare state offered by Rawls, whereas the remaining concerns can be answered by democratic socialism. But Schemmel does not address the concerns about socialism raised here. 105 Bjørnskov 2019, p. 643.

Against Egalitarianism  205 usually have serious corruption in their legal systems, especially with the institution of secret police. But it is also plausible that communism destroys political trust through central control and the impoverishment of the public, which is present in liberal socialism.

8.  Singer on Restructuring the Firm In his recent, excellent book, The Form of the Firm, Abraham Singer argues for a compromise between the Chicago School emphasis on the need for efficient firms with hierarchical structures and the liberal-​egalitarian arguments that firms and corporations should take on participatory and democratic forms.106 Singer acknowledges that markets need to be efficient and that this often requires that firms have hierarchies, but he also thinks that markets are not so perfectly efficient that firms cannot reform inequality-​generating social norms that lack a moral justification. Singer thus searches for institutional wiggle room between hierarchy and efficiency that allows firms to take on participatory aspects. We can’t have pure workplace cooperatives for a number of reasons, especially because they face big “decision-​making costs that are a major obstacle for worker ownership” since workers more directly bear the costs of poor decisions and may avoid taking risks.107 And he rejects strict analogies between the roles of citizens and employees since firms also have shareholders that have claims of justice. He therefore rejects many of the arguments for more socialist forms of business organization. Instead, Singer argues that firms should cultivate “cooperative norms, and not merely the contractual allocation of governing rights, that allow firms to economize on market failures.”108 While we should be concerned with efficiency, “We ought generally to prefer structures of corporations where those who most closely relate to the corporation have a say in its governance.”109 Within my framework, this is all well taken. Socialists are bound to have defeater reasons for hierarchical firm structures due to what they regard as excess workplace coercion of workers by bosses, whereas classical liberals concerned with growth are bound to have defeaters for firm structures that are too participatory to satisfy the principle of sustainable improvements. So, my account of public justification has to engage in a careful balancing act in order to ensure that the market order is publicly justified for all, which is one reason to follow Singer’s nimble balancing of these two concerns.



106

Singer 2019. Ibid., p. 151. 108 Ibid., p. 118. 109 Ibid., p. 201. 107

206  Trust in a Polarized Age We must also remember that private property rights include freedom of contract, which is a real constraint on firm structure. Freedom of contract has a great impact on the structure of firms, since firms can be understood as systems of contractual agreements for how capital and labor are to interact within a firm and between firms. So freedom of contract can help to legitimize boss-​led firms insofar as freedom of contract, in practice, tends to produce more boss-​led firms than worker-​led firms. If boss-​led firms tend to arise from respect for freedom of contract, then that helps to publicly justify boss-​led firms. But let’s not push this point too far. While individual contractual arrangements may not impose coercion on others, the social dynamics that arise from freedom of contract may impose coercion on others and violate their rights, such as through negative externalities. And, more importantly, true freedom of contract will also allow for the formation of cooperatives in far more cases than are currently permitted. Unfortunately, Singer’s suggested reforms face real difficulties. Singer proposes a cooperative bank given a monopoly over cooperative financial services and charged with the responsibility of providing loans to cooperatives, in order to counter capital constraints; sectoral insurance and risk-​pooling arrangements among cooperatives to overcome the lack of risk diversification; and, most ambitiously, a cooperative stock market that would facilitate a market for membership and provide price information to discipline cooperative governance.110

Given that my social contract theory is a nonideal theory, we must face two worries. First, there are questions about whether a cooperative bank can solve the relevant information problems in deciding which firms to aid and which to ignore. Second, the bankers themselves might create incentives for rent-​seeking or engage in rent-​seeking themselves. The proposal for a cooperative stock market will likely suffer from similar defects, but on a grander scale. I should say that I do not mean to reject having a cooperative banking system exist alongside a traditional banking system, with the same kinds of protections for cooperative banks and traditional banks. I think the welfare state policies proposed in both the previous chapter and this one will pressure firms to coerce their workers less and will, in time, produce more democratic firms if that is what workers desire. Since people will have more options for employment, more wealth, and more economic security, they will have more freedom to exit domineering firms. So they will threaten to leave if their conditions are not improved. Second, an order of public reason will certainly have fewer restrictions on unionization than we have, which is a



110

Ibid., pp. 216–​7.

Against Egalitarianism  207 traditional route that workers use to gain a foothold in the structures of their firms. So, while I think Singer is right that we should try to balance the liberal democratic interests of workers with the need for market efficiency, I am not sure his proposed reforms can be publicly justified. Other policies, however, can help achieve some of his goals.

9.  Egalitarian Economic Institutions Do Not Create Trust for the Right Reasons A commitment to public justification does not imply political libertarianism. To the contrary, a principle of social insurance can be publicly justified. Further, we can justify regulations and additional public policies, like the provision of public goods, so long as they meet the standards set by policy epistemology. We can justify some redistribution in the name of preserving the integrity of the democratic process as well, though the present empirical evidence linking inequalities of wealth and democratic dysfunction is inconclusive. We can even justify coercion in order to encourage firms to take on more participatory dimensions, following Singer’s work. However, public reason rules out redistribution for the sake of egalitarian conceptions of distributive justice. Property-​owning democracy and liberal socialism are simply too inefficient and too violent. Public reason is therefore neither libertarian nor egalitarian, but rather tends to favor modest welfare state capitalism—​a market economy combined with a principle of social insurance, a strong right to unionize, relatively low taxes, demonstrably effective regulations and public policies, and strong protections for the integrity of the democratic process.111 Such a regime is one that all members of the public can see as justifiable to all by engaging in a piecemeal evaluation of its central component laws and policies. Even socialists and libertarians will prefer such a regime to warlike conflict because the main institutions of a modest welfare state capitalist regime should be publicly justified for them even if the relevant moral and legal rules are suboptimal from their perspectives. For despite their radicalism, libertarians and socialists typically care about more than establishing their sectarian ideology. They also care about being able to trust and be trusted by people who reject their views. And most of them will tolerate regimes that support trust given the great benefits of a trusting order. Of course, libertarians and socialists need not acquiesce in what they regard as a deeply unjust order. They are free to agitate for their preferred regime types. But at present, neither the imposition of a radically free-​market economy nor



111

Compare Gaus 2011, p. 526.

208  Trust in a Polarized Age an extensive state can be publicly justified to diverse members of the public. Libertarians and socialists must advocate for their views in the same way that Protestants and Catholics advocate for their views: by reason and persuasion, not by using force against those who reasonably disagree. This means that we can establish strong but limited primary rights to property and freedom from government control in our economic lives. Members of the public can easily see the ways in which property rights can promote their conception of the good and enable them to pursue their conceptions of justice. And given the benefits of private property rights, many will be willing to extend these rights to others in order to preserve them for themselves. However, many social-​ democratic members of the public will support restrictions on these liberties for themselves in order to impose the restrictions on others, and this will limit the scope of primary property rights. In particular, social democrats will have defeaters for strong property rights claims that prohibit the provision of adequate social insurance and the implementation of important regulations and public policies. Social democrats may even have defeater reasons for property rights claims that prohibit egalitarian redistribution, since many social democrats reasonably believe that we can only protect the integrity of the democratic process and create a more trusting environment with direct compression of income differences. However, other members of the public will reject the limitations on primary property rights required to implement invasive forms of government interference, such as the coercive interventions required to impose property-​ owning democracy or liberal socialism. Consequently, the protection of publicly justified primary rights to property means that we must have a market economy, though one checked by limited regulatory and redistributive mechanisms. The security of property rights is also grounded in the evidence that market economies tend to have high levels of social trust—​that interacting in markets can create generalized trust by means of transactions that “demonstrate that those who are different can be trusted” and by treating persons on the basis of productivity rather than race, gender, and so on. Moreover, market orders can generate trust by helping persons to reach out beyond their particular social group into life with strangers.112 Though, of course, if markets create the perception that certain groups unfairly benefit from market exchange, they will undermine trust, such that limitations on these processes may be able to be publicly justified. In sum, we can conclude that moving further left than the traditional welfare state is unlikely to create trust for the right reasons. First, many far-​left economic policies are sufficiently coercive and contestable that other citizens will have



112

Berggren and Nilsson 2013, p. 181.

Against Egalitarianism  209 defeater reasons for these policies. Thus, even if these policies create trust, they do not do so for the right reasons. Second, it is unclear whether redistributive taxation can restore trust even when it is publicly justified: it may not be able to interrupt distrust and divergence. And finally, property-​owning democracy and liberal socialism will probably reduce trust, aggravating the distrust-​divergence process. Thus far, I have not discussed the form that government must take and the justification for liberal governmental structures. I turn to this task in the next two chapters. Chapter 7 focuses on legislative power and procedure, while c­ hapter 8 focuses on elections and other activities at the foundation of the democratic process.

7

Democratic Constitutionalism In this chapter, I argue that legislative processes should conform to democratic constitutionalism. Democratic constitutionalism holds both that the legislative process should appeal to extensive citizen input (democratic), and that government officials should convert citizen input into policy via processes that are predictable, effective, and neutral between citizens (constitutionalism). I first argue that many elements of democratic constitutionalism create real trust. I  then argue that democratic constitutionalism is publicly justified. Regarding the public justification story, recall that I have argued that constitutional rules are justified in three parts: their protection of primary rights, the degree of legislative error they generate, and the levels of durability and immunity they produce. Here I will assume that all discussed constitutional rules have survived the first stage: they adequately protect primary rights. I will instead focus on increasing the quality and stability of legislation. Most public reason liberals justify constitutional orders in a different way: they first appeal to the democratic input of citizens, and then justify many other rights, laws, and policies in terms of whether a democratic citizenry or legislature would choose them under the right conditions. In contrast, I am outlining how to publicly justify the legislative process before I justify electoral rights and democratic elections. The main reason for my departure is that the best justifications for rights to participate in the political process are that their exercise protects more fundamental rights, like the rights of agency, association, and jurisdiction, and that they tend to increase legislative quality and stability. Democratic rights can be limited if they systematically violate primary rights, generate bad law, or destabilize the moral, legal, and political orders.1 For those skeptical of my approach, let me note that the contrast with mainstream democratic theory is not as sharp as it may first appear. This is because public reason liberals effectively treat liberal rights as restrictions on the democratic process, even if they insist that these early-​stage liberal rights are justified because they are an essential part of a functional democratic process. The deep

1 Also notice that my order of justification is mandated by arguments in MPBW for placing the primary rights of agency, association, and jurisdiction early in the order of public justification. See MPBW, chap. 5.

Democratic Constitutionalism  211 difference is that I allow a broader range of liberal rights, such as the right to private property in productive resources, to constrain the democratic process. I will also spend some time in this chapter explaining how to publicly justify political obligations and the state’s right to rule. The reason for this is that states can only sustain trust for the right reasons if they provide persons with good moral reasons to abide by their directives. Thus, we need a theory of political obligation and authority in order to account for the right reasons for people to trust a constitutional liberal-​democratic state to execute its publicly justified functions. This will include an account of the authority of legislation, as the authority of political officials, on my view, is derived from their tendency to impose publicly justified laws. Accordingly, I proceed in eight parts. I argue that democratic constitutionalism creates and sustains real social and political trust by limiting corruption, protecting legal property rights, facilitating broad-​based economic growth, and implementing procedures that are seen as fair and impartial (section 1). I then argue that democratic constitutionalism is publicly justified. I explain how to justify political obligations, that is, obligations to comply with constitutional rules and the directives of officials acting in accord with those rules (section 2). I outline and justify the protective function of the legislative process, the realm of legislation devoted to the protection of upstream primary rights (section 3). I shall call this protective policy. I then outline and justify the productive function of the legislative process, the realm of legislation devoted to creating public goods, regulating negative externalities, addressing the business cycle, and other matters that are typically not concerned with the violation of basic rights (section 4–​6). In those sections, I review constraints on what I call productive policy and identify institutional norms that satisfy these constraints, such as restrictions on rent-​seeking, policy minipublics, and prediction markets. I next answer the concern that nothing can be publicly justified under the conditions I outline (section 7). I then conclude (section 8).

1.  Trust and Governance I will now argue that four features of democratic constitutionalism create and sustain social and political trust in the real world: legal property rights, limitations on corruption, procedural fairness, and economic performance. I then argue that political trust in this contact increases both social and political trust. So, I defend six claims: (1) Protecting some primary rights creates social and political trust. (2) Restricting corruption increases social trust.

212  Trust in a Polarized Age (3) Procedural fairness and broad-​ based economic growth promote political trust. (4) Social trust contributes to good governance, which creates political trust. (5) Political trust creates social trust, creating positive feedback between social and political trust. (6) Political trust increases itself in some cases. The clearest and most obvious way in which democratic constitutionalism creates real trust is by recognizing and protecting the primary rights whose exercise creates social and political trust. Thus, insofar as a constitution includes rules recognizing and protecting freedom of association, the constitution will secure more of the trust created by freedom of association than if the association-​ protecting rules were not in effect. Similarly, insofar as a constitutional regime protects legal property rights, then to that extent it secures the real trust created by legal property rights; the same is true for welfare rights and all other rights that tend to produce trust. Protecting rights is a central way to create real trust. In ­chapter 2, we saw that limiting corruption is essential for sustaining social trust. Again, there is robust evidence that “both regional-​level quality of government and individual-​level corruption perceptions and experiences are strong determinants of social trust.”2 Constitutional rules limiting corruption create real social trust. They do so by discouraging observable acts of corruption, which are well-​documented trust reducers. Also recall from ­chapter 2 that the two main determinants of political trust are perceived economic performance and procedural fairness/​quality of governance.3 Procedural fairness, recall, requires citizen involvement and that citizens be treated predictably and impartially; quality of governance is similar but broader, including government effectiveness, control of corruption, rule of law, and voice. People normatively expect political institutions to make them all more prosperous over time, and so when people become more prosperous, they will express greater satisfaction with government because they think government officials are doing their jobs.4 The constitutional rules I  defend subsequently should help to create broad-​based economic growth by limiting rent-​seeking and increasing policy quality. As we shall see, legislative systems that frustrate rent-​seeking and improve the selection of empirically well-​grounded economic policies should raise political trust.

2 You 2018, p. 486. 3 See c­ hapter 2, section 5. 4 Here we can understand economic performance as broad-​based economic growth. That is, improvement in everyone’s economic performance and security. Economic performance thus satisfies the associated principles of sustainable improvements and social insurance.

Democratic Constitutionalism  213 Similarly, citizens of most societies seem to expect their political institutions to have high quality of governance and to be procedurally fair. As we saw in ­chapter  2, this means both that political institutions should be seen as including citizens in political decision-​making and that officials implementing policy should so implement it respectfully, impartially, and predictably.5 They should regularly meet the normative expectations associated with their particular branch of government, treating all applicants, files, and so on, in a neutral and timely fashion. We shall see that some legislative procedures should increase procedural fairness in both respects, such as the use of small deliberative bodies known as “minipublics”6 and prediction markets. First, policy minipublics should increase the public’s sense that they have a role in determining the direction of public policy, since “one of the strongest methods for restoring trust is enhancing citizens’ involvement in governmental policy making.”7 Second, policy minipublics and restrictions on rent-​seeking should avoid the appearance that policy implementation is irregular and corrupt. Restrictions on rent-​seeking and corruption, as well as requiring that imposed rules be general and abstract, rather than particularized and easily gamed, should increase perceptions of procedural fairness and thereby increase political trust. This is important because when “institutions and policy-​making adhere to high standards of quality and impartiality, political support is expected to be less sensitive to short-​term fluctuation in the economy,” which helps maintain political trust throughout often destabilizing business cycles.8 Trust researchers disagree about the relative importance of economic performance and procedural fairness for political trust, which complicates my argument. Perhaps only one of these two factors really matters for political trust. We can address this concern by looking at regional variations in the determinants of political trust.9 For instance, in Latin American societies with higher levels of corruption, economic performance explains most of existing political trust. That could be explained if Latin Americans have more lenient attitudes about corruption or if poorer nations place a higher priority on economic development.10 In Western and Southern Europe, Mariano Torcal detects the influence of both economic conditions and fair procedures on political trust.11 One way to explain 5 At least in developed countries with states that perform the basic state functions of peace and security. In African states, political trust is often affected by failures of basic state functions. See Hutchison and Johnson 2017. 6 I outline the structure of a kind of minipublic in this chapter, section 6. 7 Vigoda-​Gadot and Mizrahi 2014, p. 2. This means it will be important to frame minipublics as supplements to participation rather than as replacements, lest people feel as though they were being disenfranchised. 8 Van Ham et al. 2017, p. 14. 9 Van Ham et al. 2017. 10 Bargsted, Somma, and Castillo 2017, pp. 408–​9. 11 Torcal 2017, p. 420.

214  Trust in a Polarized Age this regional variation in the determinants of political trust is that, as economic well-​being increases, citizens demand more from their institutions than mere economic performance. In poorer countries, citizens prioritize economic performance because they cannot meet many of their basic economic needs. To them, procedural fairness and limiting corruption may seem like a luxury. But as societies develop, citizens place increasing importance on limiting corruption and establishing procedural fairness. Along these lines, Tom Van der Meer finds that “the more importance people attach to impartiality, rule of law, and professionalism, the stronger the effect of quality of government on political support.”12 Thus, countries with low governance quality, but good economic performance, can create high political trust.13 So both procedural fairness and economic performance matter, and procedural fairness matters comparatively more as countries grow wealthier. Thus, if the liberal constitutional order I  defend secures both procedural fairness and economic performance, then the causes of trust in both poor and rich countries are consistent with its producing trust for the right reasons.14 Social trust also helps to facilitate good governance, probably because “politicians in high-​trust democracies respond better to voter demands for good governance” and because “bureaucrats and other economic agents behav[e]‌honestly.”15 Social trust also helps legislators introduce beneficial policy innovations; a comparison of US states finds that “social trust levels are positively associated with the likelihood of legislatures introducing policy innovations” that allow high-​trust states to introduce “reforms strengthening economic freedom.”16 Christian Bjørnskov speculates that this effect may arise because “politicians whom voters perceive to be trustworthy are better at credibly signaling or communicating the objective necessity of innovations and thereby better at avoiding skepticism among the electorate.” Social trust probably promotes good governance by creating more trustworthy civil servants, whereas electoral mechanisms create trust “only indirectly through the quality of the legal system.”17 Since social trust creates good governance and good governance creates political trust, social trust creates political trust by proxy Political trust is sometimes thought to cause social trust. For instance, Zmerli and Newton find that institutions that enforce trustworthy behavior, such as 12 Though remember from van Ham that more quality means a weaker effect for economic growth. 13 Thomassen, Andeweg, and van Ham 2017, pp. 520–​2. 14 It appears that political trust is higher in local officials because the public may be “more confident in how their [local] politicians spend the taxpayers’ money” (Listhaug and Jakobsen 2018, p. 569). This speaks to the relationship between public reason and federalism, but unfortunately, we lack enough data on federalism and trust to say much more than this. 15 Bjørnskov 2010, p. 325. 16 Ibid., p. 329. 17 Ibid., p. 340.

Democratic Constitutionalism  215 the rule of law, democracy, and egalitarian social policies, “combine to create conditions that favour political trust and particular and general social trust.”18 Political trust may help to stabilize political institutions like “credible threats of sanction,” and transparent government, which allow for the formation of social capital and social trust.19 Van Deth notes that people may also take untrustworthiness among political officials to imply that the public is untrustworthy; believing that the public is untrustworthy may lead people to depart from social norms.20 Van Ham stresses that “there is evidence that a trustworthy government may not only generate political trust, but also facilitate the development of social trust,” though she worries that there might only be a “spurious correlation between social and political trust.”21 In all these ways, political trust could create social trust, which helps establish a positive feedback loop between social and political trust. Now we can say that the institutions that create political trust thereby may create some social trust. Political trust helps to increase policy quality, which can in turn create further political trust; thus, political trust increases itself. This can occur in three ways. First, political trust can help “elected representatives to place collective interests ahead of parochial concerns by affording legislators greater leeway to depart from constituency preferences.”22 This might improve the proportion of policies that are justified for the public as a whole. For example, having more policy leeway may allow political officials to base policies on the recommendations of policy minipublics and prediction markets that their constituents might dismiss. That means policy will hew closer to what is publicly justified for citizens because policy will be closer to what moderately idealized members of the public would endorse since minipublics should improve the informational quality and rationality of policy recommendations. Now, it is true that political trust can lead to either progressive or conservative initiatives, so while policymaking in high-​trust countries and states need not be biased in favor of one group of partisan policies or another, political trust might ease the passage of unjustified, partisan policies.23 But insofar as political trust encourages or allows political officials to heed expert opinion on policy issues, it may increase the number of high-​quality (and so often publicly justified) policies.24 18 Zmerli and Newton 2017, p. 121. Moreover, legal trust, trust in police and the legal system, may help generate trustworthy behavior in political officials and in citizens, creating both political trust and social trust at the same time. 19 Wilson and Eckel 2017, p. 125. 20 Van Deth 2017, p. 213. 21 Andeweg and Farell 2017, p. 84. 22 Randolph 2017, p. 197. 23 Ibid., p. 209. 24 We do have recent, limited evidence that trust affects policy choice, at least at the local level. See Cooper, Knotts, and Brennan 2008, p. 459.

216  Trust in a Polarized Age Second, political trust can decrease populist voting patterns and help increase public support for civil liberties in the face of perceived threats.25 Thus, insofar as members of the public have defeater reasons for populist policies, political trust can decrease the number of unjustified policies. And it is plausible to think that especially coercive populist policies, such as tight restrictions on trade and immigration, cannot be publicly justified—​though we should be careful here: as I noted in ­chapter 2, restrictions on immigration may increase political trust. Third, political trust helps people to accept policies that adversely affect them materially or that they disagree with ideologically.26 In this way, political trust is essential for keeping the losers of particular policy debates on board with publicly justified institutions. The importance of this effect for my argument is considerable, since political trust will be essential to help persons recognize that they have sufficient reason to go along with policies that are not ideologically optimal according to their evaluative standards. And since all policies create some losers, it is essential for political stability (and durability in particular) that people stay on board and remain trusting and trustworthy even when they lose.27 This is another reason that quality of governance is important. Quality governance creates the kind of political trust that allows losers to stay on board with their regime, in the sense of trusting it. As Van Ham notes, “Citizens maintain political support even when a policy decision can hurt them when the political system has ‘procedural fairness, compromise decisions, fading memories, and constitutional arrangements for vertical division of power.’ ”28 In this way, democratic constitutionalism helps to create real social and political trust, and creates conditions whereby social and political trust manifest various positive feedback loops, playing a central role in stopping the distrust-​divergence  cycle.

2.  A Theory of Political Obligation* My account of public justification has three layers: moral, legal, and political. The moral layer is focused on the public justification of moral rules; the legal layer is focused on the public justification of legal rules; and the political layer is focused on the public justification of constitutional rules understood as law-​creating and 25 Ibid. 26 Randolph 2017, p. 201. Also see van Deth 2017, p. 213. 27 One caveat of these findings is that strong institutions might crowd out trust through the use of coercive punishments, such that strong institutions may undermine the need for trust. But since strong institutions will only be stable and effective when there is a high degree of social trust in these institutions in the first place, this worry is not very damaging to my argument. Wilson and Eckel 2017, p. 137. 28 Van Ham et al. 2017, p. 14.

Democratic Constitutionalism  217 law-​altering institutions. The justification of the political layer depends upon that of the deeper legal layer. Specifically, the principle of constitutional justification builds on the principle of legal justification just as the principle of legal justification builds on the principle of public justification. To see this, recall both principles: Public Justification Principle:  a moral rule is publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the rule. Legal Justification Principle: a legal rule is publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the law because each member rationally recognizes that compliance with the law improves upon her capacity to comply with one or more publicly justified moral rules and other legal rules.

Functionally independent legal rules are justified insofar as each person has sufficient reason to internalize them because she recognizes that compliance with laws improves upon her capacity to comply with the moral rules and other legal rules that apply to them. The principle of political justification is defined likewise: Political Justification Principle: a constitutional rule is publicly justified only if each member of the public has sufficient intelligible reason to comply with and internalize the constitutional rule because each member rationally recognizes that compliance with the constitutional rule improves29 upon her capacity to comply with a set of laws (and/​or moral rules).30

To determine whether a constitutional rule improves upon one’s capacity to comply with laws and moral rules, we must revisit the analysis of legal authority in MPBW. A constitutional rule has authority when it passes all three stages of the constitutional choice process, which is why it constitutes an improvement on the legal order. To see how this works, imagine a political state of nature that runs parallel to the legal state of nature. In such an order, there are moral and legal rules, but no legislative decision procedures. There are no laws that govern the process of

29 Here I  implicitly understand improvement as morally efficient, as I  did for the Public Justification Principle and the Legal Justification Principle. For elaboration, see MPBW, pp. 148–​9. 30 Notice here that my view diverges from John Hasnas’s social peace view, since he claims that the social peace view “cannot give rise to a duty to obey the law of a state-​administered legal system” because legislative rules “are not necessarily peace-​promoting rules” (Hasnas 2013, p. 471). My view, in contrast, is that insofar as legislation improves upon our ability to comply with our legal obligations, the moral peace approach can ground the authority of legislation.

218  Trust in a Polarized Age laying down laws or adjudicating disputes about whether legislative decision procedures act within their authority.31 In this state, laws can secure trust by structuring moral rules, but they cannot reliably and systematically structure other laws, since there are no decision procedures for resolving conflicts between laws. In such a state, laws will exhibit problems similar to those moral rules that exist in the legal state of nature: they will be uncertain, static, and ineffective. They will be uncertain because there is no authority charged with interpreting the law. They will be static because there are no publicly recognized means of changing the law. And the laws will be ineffective because they can only be imposed by the legal system through judge-​made law and trials and not by democratic decision. There is no institution that can preemptively and deliberatively alter the law. Now, legal systems greatly improve the moral order, so the problems of the political state of nature are more limited than those in the legal state of nature, and they call for a more limited response. But the need for a legislative system is manifest. We can now develop an account of political obligation, which I understand as our obligation to one another to follow constitutional rules. It is not defined as a duty to comply with the directives of political officials; rather, if we are obligated to follow constitutional rules, then insofar as those rules empower officials, we are obligated to follow the directives of those officials when the directives are constitutional. My account of political obligation begins with rules for this reason: the authority of officials is entirely derivative from the authority of rules. Here’s a full definition of the conditions of political obligation as I understand them: A person has a political obligation to comply with a constitutional rule if and only if (i) she is epistemically entitled to believe that general compliance with the constitutional rule will effectively improve her legal order’s capacity to control moral and legal rules (or other constitutional rules) and that her compliance with the constitutional rule improves32 upon her presently weak ability to discharge her moral and legal obligation(s) (and her other constitutional obligations); (ii) she rationally observes that sufficiently many other agents are complying with the constitutional rule; and (iii) complying with the constitutional rule does not violate her other moral and legal obligations of equal or greater weight.33

31 Raz 1970, p. 147. 32 The improvement here is a morally efficient one. Again, see MPBW, pp. 148–​9. 33 Here I assume that the obligation only obtains when a presumably justified law is applied to the subject of justification. It might be that the law does not affect the citizen, in which case the obligation does not apply to her, but were it to affect her, she would have an obligation to obey it.

Democratic Constitutionalism  219 Let’s streamline this account. First, constitutional rules are social norms, so constitutional rules just are those that are generally obeyed. So we can drop the reference to obedience.34 We can also shorten the improvement conditions and contract the entitlement condition to rational belief. I can focus on the relationship of the constitutional rule to the legal order rather than the moral order (where the legal order can contain constitutional rules). And I will assume that the relevant moral and legal obligations are not overridden, undercut, or otherwise defeated by other obligations. Constitutional rules have authority when the following conditions are met: A person has a moral obligation to obey a constitutional rule if and only if she rationally believes that the constitutional rule significantly improves her legal order and that obeying the constitutional rule improves upon her capacity to fulfill her legal obligation(s).

The arguments that derive political obligation from legal obligations parallel the arguments in MPBW that derive legal obligations from moral obligations.35 Reba has an obligation to follow the law when it better enables her to comply with her moral obligations than she could in its absence; similarly, John has an obligation to follow a constitutional rule when it better enables him to comply with his legal and moral obligations than he could in the absence of the constitutional rule.36 Political decision procedures are admittedly imperfect and will often generate bad law. Further, constitutional rules might be unfair or immoral in themselves. In these cases, persons lack obligations to comply with and endorse those constitutional rules. However, constitutional rules can have authority even if they pass defeated laws. Reba will have no obligation to comply with laws that cannot be publicly justified for her; but if her constitutional decision procedures do a reasonably good job of providing publicly justified laws, she must not seek to overthrow or undermine her constitution. This is how a constitution can be legitimate and authoritative when it is not fully just, because we can be obligated not to undermine the constitution even if it is fallible. So constitutional rules can retain their authority even if they sometimes impose unjustified laws, so long as they meet the other conditions of public justification. To do this, constitutional rules must be limited in certain respects; I will discuss those limits later in the chapter. Now we ask which forms of democratic decision-​making can be publicly justified. 34 This is not an empirical claim, but a conceptual one. The true constitutional norms are social norms; constitutional rules written on paper that are not in effect are not social norms. 35 I argue for this claim in MPBW, chap. 6. 36 Here we can see that constitutional rules gain their authority from addressing conflicts between laws, whereas laws in the political state of nature can address conflicts between moral rules but have trouble addressing conflicts between legal rules. I thank Alex Motchoulski for raising this concern.

220  Trust in a Polarized Age

3. Protective Policy Protective policy choice concerns the articulation and protection of primary rights. To illustrate, let’s consider the familiar example of the First Amendment to the US Constitution: Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

The First Amendment articulates a series of rights, specifically a right against religious establishment, and rights of free exercise of religion, free speech, freedom of the press, assembly, and petition. It prohibits Congress from making laws that violate those rights in order to prevent government from violating these rights. The First Amendment can be reformed or repealed by legislative action, but only with very large supermajorities in Congress, plus supermajorities of state legislatures, voting in favor of those changes. So, First Amendment rights prohibit violations by public actors, as well as private actors should they even assume public office, and these rights are themselves protected by strong supermajority rules. First Amendment rights protections seem to pass all three stages of constitutional choice. They protect important primary rights. They have historically functioned to prohibit or lead to the repeal or judicial review of laws violating those rights. Citizens comply with and respect these rights and procedures in some of their public political activities. And special interest groups interfere with First Amendment political processes, like altering First Amendment protections, relatively less than other political processes, like budget  allocations. For this reason, First Amendment protections impose duties of obedience upon most Americans under the right conditions. They have authority. We can generalize from First Amendment liberties to generate a more abstract account of the protective function of the legislature. First, there seem to be good grounds to use supermajority rules to protect basic rights, since supermajority rules are biased against changes. The rules protect against false positives (rights-​ violating laws) at the price of tolerating more false negatives (having fewer rights-​protecting laws). But this is arguably the appropriate weighting of these two priorities regarding protective policy, since legislation may tend to violate rather than protect these rights.

Democratic Constitutionalism  221 A common complaint about supermajoritarian protections is that they assign unequal political power to minorities who reject proposed legislation vis-​ à-​vis majorities that favor it.37 I reject the complaint for the following reason. Constitutional rules are publicly justified in large part based on their legal effects, and so the prime public justification for a protective rule is that it will tend to stop violations of primary rights. This is how equality is realized—​through widespread public justification and equal rights. If supermajority rules are publicly justified to protect equal primary rights, then supermajority rules better express a commitment to equality than simple majority rules. We also cannot forget the stabilizing function of supermajority rules. Because they make it difficult to change rights-​protecting laws, this allows the public time to internalize those laws and the constitutional rules themselves, which helps ensure that the rules will be durable, and because the rules are publicly justified, durable for the right reasons. Further, gaming First Amendment political processes, such as free speech law, will be difficult, ensuring immunity, and since the rules are publicly justified, these processes will be immune for the right reasons. Such a system is more stable than one in which a simple majority can undermine fundamental rights at any time. However, supermajority rules can cause problems in promoting stability because they reduce policy responsiveness, given that laws are harder to change when preferences change. This can lead to policy entrenchment.38 Similarly, supermajoritarianism may make the law more conservative, biasing it toward citizens who tend to be skeptical of legislation, and this can undermine stability as well by making policy seem more partisan.39 And supermajority rules, in restricting legislation, raise the stakes for the policies that look like they can pass, since legislative changes become rare and so acquire importance. This can create conflict and acrimony.40 So we cannot vindicate the public justification of supermajority rules easily, even for protective policy.

4.  Constraints on Productive Policy Now we turn to the voting procedures that should govern productive legislation—​ laws that provide goods and services where primary rights are not typically at

37

For discussion, see Gaus 2011, pp. 487–​90 and Ganghof 2013. Gilbert 2017, p. 631. 39 I thank Micah Schwartzman for this point. 40 I thank Christian Bjørnskov for this point. 38

222  Trust in a Polarized Age stake.41 Paradigm cases of the state’s productive functions are the provision of infrastructure, countercyclical policy, and some aspects of social insurance. Based on my previous arguments, there are six constraints on policymaking. (1) Productive policy must not violate primary rights of agency, association, jurisdiction, and legal procedure. It also must not violate primary rights of property, which means it must not excessively regulate property and must protect property from negative externalities. Policy processes that reliably violate these rights are illegitimate. However, since productive policy typically concerns issues that do not bear on primary rights, at least not directly, this first constraint is relatively modest. (2) Productive policy must not undermine sustainable economic growth. This is not merely because persons have primary rights to income, wealth, and capital, but because the principle of sustainable improvements42 requires that the economy be shaped so as to allow economic growth and to ensure that the benefits of growth redound to all. (3) If social insurance is required to ensure that a sufficient amount of resources are produced, then productive policy must impose and sustain an appropriate amount of social insurance, as required by the principle of social insurance. (4) Productive policy must minimize corruption, including rent-​seeking, not only because the coercion involved in rent-​seeking is unjustified, but because public corruption is a major threat to social and political trust. Immunity for the right reasons is essential here. Officials can cooperate with one another because the number of officials is usually small, and this can draw self-​interested persons into the political process in order to lobby or become one of those officials. (5) If productive policies increase coercion, they must meet fairly high standards of policy epistemology. Our predictions of the effects of policy must be reliable enough that members of the public from a range of diverse perspectives can see the policy as satisfying their own goals, which is largely ensured by satisfying the first four constraints to some degree. (6) There must also be incentives for political officials to make policy decisions based on our best available scientific evidence and a consensus of experts when expert opinion is reasonably believed to be superior to the reasoning of the public as a whole.43 Toward this end, productive policy should try to structure deliberation so as to incentivize high-​quality debate and discussion. 41 I grant that the line between productive and protective policy is not sharp and must be settled upon by the constitutional process, but I think it fair to assume for the sake of argument that such a line can be drawn. 42 MPBW, pp. 168–​70. 43 Superiority may be determined through prediction markets and superforecasters, discussed later.

Democratic Constitutionalism  223

5.  Discouraging Rent-​Seeking I lack the space to show which constitutional rules meet  all six of these constraints. I  will instead focus on determining which institutional structures can protect against rent-​seeking and meet the standards of policy epistemology. It is commonly assumed that productive policy in real-​world democracies is often plagued by the influence of lobbyists and special interest groups, who constantly attempt to push legislation that favors their interests. This is to be expected for the familiar reasons presented by public choice economists.44 Officials are often self-​interested, so they can easily be tempted by campaign contributions and the promise of social and economic benefits once they leave office. Furthermore, politicians who are not focused on their self-​interest, and who wish to legislate based on other factors, can often be stopped and intimidated by politicians who are “bought and sold.” Rent-​seeking can corrupt the legislative process not merely through the purchase of legislative votes but also through legislative tactics like agenda control.45 While Gerry Mackie has made a convincing case that legislators do not use these tactics as often as one might expect, there are still serious attempts to co-​opt the legislative process for the benefit of the few at the expense of the many.46 Legislation for which many members of the public have defeater reasons can be passed nonetheless when relatively few people expect great benefits from it, because of the logic of concentrated benefits and dispersed costs: the few who each gain a lot from a bill have more incentive to invest time and energy in lobbying on the issue than do the many who each lose a little. It may appear that rent-​seeking tends to produce type-​1 errors (where the legislature does what it shouldn’t) rather than type-​2 errors (where it fails to do what it should). Yes, special interest groups work to defeat legislation, but, more often than not, they attempt to legislate benefits through coercion rather than increase their power through the prevention of legislation. The best counterexamples are fights to oppose regulation, such as the regulation of carbon emissions.47 But, on balance, lobbying and rent-​seeking increase the incidence of type-​1 errors

44 See the literature reviews on the topics of rent-​seeking and the inefficiencies of bureaucracy in Mueller 2003, pp. 333–​84. 45 Ibid., pp. 112–​4. 46 Mackie 2003, pp. 158–​72. 47 Though restricting carbon emissions can protect persons and property from the destructive effects of large carbon emissions.

224  Trust in a Polarized Age because fewer rent-​seeking benefits are garnered by stopping than by passing legislation. However, the best research suggests that, contrary to what one might expect, it is not actually the case that lobbying is obvious and successful, nor is it the case that the successful lobbying that exists tends to favor an increase in coercion rather than a decrease. First, it is very hard to know how lobbying leads to influence.48 Some studies find a large effect, others none.49 The biggest clear source of influence is the “provision of information and other policy-​related aid,” but it is not clear whether this function is morally problematic based on the constraints of public justification. On the other hand, if money buys the ability to provide information to government officials in ways that shut out poorer groups, then that is a big concern.50 The best research suggests that interest groups are successful when they have enough lobbying resources, they know a member of Congress allied with their policy preferences, and that congressperson is able to act successfully to help the group.51 So here the threat of special interest group lobbying is more the control of information than any direct purchase of power. There does appear to be successful lobbying in some cases, but the influence is subtle and conservative. Lobbyists don’t generally bribe so much as they encourage committee chairs to kill bills before they are voted on. Firms value influence with “legislative leaders endowed with formal institutional powers more  .  .  .  than [influence with] other leaders,” and they value influence with committee chairs most.52 When legislators become leaders of a committee or a party caucus, “They experience a significant boost in corporate campaign contributions.”53 Firms also value a kind of access to negative political power, which includes controlling the voting agenda and killing bills.54 It’s important to recognize that, in general, interest groups aren’t trying to change the law, but to “protect, not to upset, the status quo,” which can have a huge but harder-​to-​ notice effect.55 And these interest groups have very different priorities from the public, a pattern that has been documented over and over again.56 It looks like two reforms can address these patterns. First, campaign-​ contribution limits, insofar as they apply to committee chairs, do seem to reduce special interest influence on the democratic process. For instance, they keep committee chairs from getting more contributions when they advance to

48 49 50 51 52 53 54 55 56

Leech 2010, p. 550. Ibid., p. 534. Ibid., p. 551. Ibid., p. 15. Fouirnaies 2018, p. 177. Ibid., pp. 179, 189. Fouirnaies and Hall 2018, p. 132. Baumgartner 2010, p. 523. Ibid., p. 530.

Democratic Constitutionalism  225 that new, more powerful position. Term limits are also probably effective in this case.57 While they do not erode direct access to committee members, they “entirely remove” the value of indirect access. This is probably because “access to the committee assignment process requires a longer time horizon, necessitating a relationship built and maintained over many electoral cycles in order to bear fruit.”58 The public choice economists are right, however, that more heavily regulated industries have a stronger incentive to engage in rent-​seeking, one way or another. The industries most heavily regulated at the state level do the most to influence committee and party leaders at that level.59 This is also true at the federal level: “Firms that are highly regulated by the federal government are likely to spend the most on lobbying.”60 This suggests another, broader reform, which is to try to control political outcomes in ways that do not involve direct regulation and control. In the past, I have advocated constitutional structures to restrict rent-​seeking, such as supermajority rules and bicameralism.61 Both have the effect of blocking legislation and so avoid lots of type-​1 errors. However, if special interests primarily seek to block legislation, then type-​2 errors are the bigger problem, and we may want to avoid supermajority rules in order to reduce the advantages of those who seek to preserve the status quo. Policies like term limits and limiting contributions to committee chairs may reduce those advantages even in systems with supermajority rules. But even still, there are reasonable worries about excessive supermajority rules creating higher stakes around the legislation passed and undermining stability by allowing for more distance between what most people support and the state of policy. We can also better structure the power of the purse. At present, Congress passes general appropriation bills and directs the funds the government has (and often funds it does not have) to specific legislative purposes. These funds are often directed toward and by rent-​seekers. But a constitution could tie expenses to methods of payment, so as to reduce the problem of excessive taxation and inappropriate spending. For instance, we could follow James Buchanan’s recommendation to require that every bill contain provisions specifying how it is to be financed.62 This would require scoring each bill to determine its costs and then specifying the forms of revenue collection that would cover those costs. Tying 57 However, term limits might forbid the formation of subject matter expertise that could increase quality decision-​making, and so hurt political trust. I thank Mark Laufgraben for this point. 58 Fouirnaies and Hall 2018, p. 145. 59 Fouirnaies 2018, p. 177. 60 Leech 2011, p. 13. 61 For a discussion of the effects of bicameralism, see Buchanan and Tullock 1962, pp. 233–​48. 62 Buchanan 1999, p. 141. On the role of Buchanan’s proposal in public reason, see Gaus 2011, pp. 496–​7.

226  Trust in a Polarized Age funding and spending would help stop those who would appropriate funds from a general pot. It would also provide the public with estimates of the costs of proposals and make clear where revenues would come from. Based on my argument in c­ hapter 5, I am dubious as to the value of future cost estimates because I am not sure predictions are feasible, but perhaps Congress could consider the highest-​quality predictions available and, guided by those predictions, specify the maximum that could be spent on each program. Though admittedly these caps would be hard to enforce in practice. I anticipate two concerns about such a proposal: that less legislation will be passed and that the restrictions on legislation will be biased in favor of libertarian points of view.63

5.1.  Less Legislation Passed Such a constitutional requirement would dramatically reduce the amount of legislation passed. Many people would be far more hesitant to support spending bills if they also included tax increases. Moreover, policy epistemology will rule out excessively vague and sketchy cost-​benefit analysis as insufficient to justify coercion. In response, I  would argue that reducing the amount of legislation passed is often good given that many bills cannot be publicly justified to members of the public. And the bills that do pass will stand a fair chance of being publicly justified, which amplifies the legitimacy and stability of the constitutional order.64 Another benefit of restricting legislation is the reduction of rent-​seeking. This is, again, important for many reasons. For one, Mancur Olson has convincingly argued that rent-​seeking groups tend to accumulate over time, gradually slowing the process of economic growth by gumming up the works with inefficient spending and other activities.65 Since even small decreases in the growth rate significantly reduce economic well-​being over time, we must be vigilant against even small, short-​term, rent-​seeker-​generated friction in the productive policy process. This is especially true in light of the principle of sustainable improvements. Checking extensive rent-​seeking is also critical for rendering the 63 And perhaps a third: that such a requirement would make Keynesian macroeconomic policy impossible because deficit spending during recessions would be difficult, if not impossible. I’m not sure Keynesian macroeconomic policy survives policy epistemology, so I’m not sure having an anti-​ Keynesian bias is problematic. However, legislatures could pass economic stimulus bills that spend money immediately and raise taxes only later. 64 Though this kind of gridlock could reduce political trust by making it appear as if government can’t get anything “done”; the hope is that by not passing bad laws that undermine trust that the overall effect of more limits on legislation we get more political trust on balance than otherwise. 65 Olson 1984.

Democratic Constitutionalism  227 legislature relatively immune from invasion by more rent-​seeking politicians and lobbyists. Without protections, we can expect rent-​seeking behavior to gradually crowd out efficient cooperation as more rent-​seekers enter the system and more within the system, observing their colleagues’ behavior, become inclined to engage in rent-​seeking too. Agents within the system may consistently distort the rules of the game within which they operate. However, given the research I  have cited, we do not want to simply give legislators the blanket ability to destroy legislation, lest they do so in ways that benefit special interests. The idea instead is to properly package legislative bills so that their costs and benefits, or at least rough estimates of their costs and benefits, are available and hard to game. Placing these restrictions on legislation should also be compatible with weakening supermajority rules if they are causing problems. We could, for instance, have simple majority rule for bills that include both spending and the power to raise revenue to pay for it.

5.2. Libertarian Bias A second concern is that these restrictions on legislation will tend to yield libertarian outcomes and, therefore, will be biased in favor of libertarian points of view. For it looks as though we are simply making it difficult for legislators to make productive policy. Now, in one way, if public justification implies libertarian conclusions, this is not in itself problematic. We are inquiring into which institutional structures best sustain social and political trust for the right reasons. If libertarian institutions are uniquely suited to that end, then so be it.66 That said, I  don’t think there is a bias. Brink Lindsey and Steven Teles say that “beating back rent-​seeking . . . will sometimes require increasing the size of government.”67 In particular, they argue that we must increase the “analytical capacity” of government to collect and distribute information, and to impose policies that may “cost taxpayers more up front” but that will be less vulnerable to rent-​seeking than the institutions we currently have. Lindsey and Teles present four case studies of what they call “regressive regulation,” understood as “rent-​creating policies that undermine growth while exacerbating inequality.”68 On my view, these studies identify policies whose defeat in the process of public reason is overdetermined. These policies reduce growth, 66 This is distinct from the concern that libertarians are social dictators in matters of public justification: here libertarian outcomes aren’t due to libertarian reasons but due to the reasons of nonlibertarians canceling one another out. Gaus 2011, pp. 501–​6, addresses the libertarian-​dictator objection. 67 Lindsey and Teles 2017, p. 10; emphasis mine. 68 Ibid., p. 32.

228  Trust in a Polarized Age often coercively increase inequality, and create the public perception of corruption. They profoundly undermine the rational basis for social and political trust, and they are likely to reduce social and political trust among actual persons insofar as they pay attention to these practices. The four policies are (1) Giving financial institutions subsidies that incentivize excess risk in borrowing and lending; (2) Granting excessive monopoly privileges under copyright and patent law; (3) Protecting of incumbent service providers under occupational licensing; and (4) Regulating land use in ways that make housing artificially scarce. All four policies use coercion to increase rents in ways that increase inequality. This is plain for policies (2)–​(4) because coercion is used to create intellectual property, occupational licensing, and land-​use restrictions. But it is also true for (1) because subsidies for financial institutions are paid for by taxes and often produce new regulations during or after financial downturns. Lindsey and Teles make a compelling case that the economy will be better off without these policies. Their best argument, in my view, is against land-​use restrictions for housing. This occurs when upper-​class homeowners use city-​and county-​level land-​use restrictions to increase their home values, giving poorer people fewer places to live. Mortgage and housing policy “redistributed [wealth] upward” in clear ways.69 If Matthew Rognlie is correct, land-​use policy may account for as much as two-​thirds of the increase in inequalities of wealth and income over the last several decades.70 Land-​use restrictions also undermine economic growth by preventing the integration of much-​needed laborers into the economic networks in large cities, New York City, San Francisco, and San Jose in particular.71 Workers are unable to join highly productive metropolitan economies because they cannot afford the literal rent. This reduces productivity. Lindsey and Teles argue that land-​use restrictions can be beaten back by centralizing regulations in the hands of state and federal governments, depriving local governments of the power to engage in upward redistribution. For instance, states and the federal government might assign municipalities a “municipal zoning budget,” where a city government would receive a capped fund which they must spend by deciding “every year on a target for how much the overall housing stock should increase,” and until a city reached its target, it could not impose new land-​use restrictions. But more radically, states could simply take over

69

Ibid., p. 109. Rognlie 2015. 71 Lindsey and Teles 2017, p. 121. 70

Democratic Constitutionalism  229 land-​use policy, which would undermine the power of municipalities entirely and make rent-​seeking more onerous for the local rich.72 Now, in one way, this is not an increase in government power on net; rather, it reduces rent-​seeking by centralizing power in the hands of state and national government. But it helps to illustrate that policies that reduce rent-​seeking are not necessarily libertarian in character in that they do not involve restrictions on state power. In actual practice, then, productive policy will sometimes involve increases in government power, and at other times require reductions. The policies that Lindsey and Teles criticize are amply undermined by the defeater reasons of members of the public. Changing those policies will sometimes require increases in government power and other times decreases.

6.  Minipublics and Policy Epistemology Let’s explore how to promote deliberative quality in the productive policy process with an eye toward giving officials an incentive to act on the results of high-​quality deliberation. Current public deliberation in the real world is too unmoored from direct participation by citizens, and citizens have little incentive to deliberate well based on good information, such that real deliberation does not approximate the reasoning of moderately idealized agents. I believe policy minipublics can increase deliberative quality.73 A  policy minipublic could be composed of citizens selected by random sampling and paid a salary over the course of a month74 or up to half a year to engage in policy review. Members of the minipublic would learn about a particular policy proposal and its costs and benefits; they would then deliberate about, and finally vote on, whether to recommend the policy for ratification. They could be recalled to sign off on changes to their proposals made by political officials. Unique minipublics could be selected for each piece of major legislation, with a congressional committee deciding which legislative proposals are significant enough to call for a minipublic. Minipublics must be small in size to preserve face-​to-​face discussion. Face-​ to-​face discussion has great benefits that anonymous discussion lacks, as persons have an incentive to behave in a calm, reasonable, and moral manner. If a

72 Ibid., pp. 168–​9. There would likely be some costs here, as policy stability might be lower at the state level, and state governments may lack the local knowledge municipal leaders have. 73 Here I am influenced by Fishkin’s idea of a minipublic. 74 Service would be voluntary, so the salary should be high enough to enable economically disadvantaged persons to participate in the process. Ideally, a society would also establish a social norm where citizens are expected to take part in the process, not unlike jury duty, but without legal compulsion.

230  Trust in a Polarized Age group becomes too large, these dynamics are undermined and more unsavory dynamics can take over. To control minipublic costs, there would be only a limited number of minipublics at work at a time. Minipublics will review the policy recommendations of legislators and make their own recommendations public in almost all cases (save some national security issues). The recommendations must be public so that citizens can determine whether politicians are following the recommendations of deliberators and hold politicians responsible if they ignore what the deliberators recommend. The hope for minipublics is that their recommendations will come to be seen as a proxy for the informed consent of the populace. This is because the deliberators will not be career politicians but randomly selected citizens with limited terms and confidential identities. Their recommendations should be capable of acquiring a high degree of democratic legitimacy if they are seen as representing the public as a whole. If minipublics, as an institution, gain that legitimacy, then when a minipublic endorses a piece of legislation, that should increase the likelihood that it will be ratified and stabilized. James Fishkin’s empirical work on microcosmic deliberation gives us reason to hope that policy minipublics can increase the quality of public deliberation. Fishkin has found that citizens almost never have cyclical collective preferences, because deliberation tends to produce single-​peaked preferences; he has also found that polarization effects only crop up from time to time.75 So some worries about democratic failure appear to be addressed through careful construction of deliberative practices. However, Fishkin’s experiments have one considerable drawback: they tend to cover political issues of minor significance. In cases where a minipublic could make recommendations about important policies like healthcare policies or stimulus bills, we can expect a dramatic increase in attempts to manipulate the process.76 Legislators will try to control the way citizen-​deliberators are chosen (not unlike attorneys in the jury selection process), along with the way in which information is presented to them. Legislators and outside groups may try to influence the deliberative process, say by sowing division and discord by making racial, gender, class, religious, or cultural differences salient in discussion. They could even promise these deliberators benefits following the process, like private sector jobs based on their policy experience, if they can circumvent anonymity. These threads could delegitimize the minipublics. The protections for jurors found in the jury system could be used to resist these forms of tampering. The identities of the deliberators could be withheld from the 75 Fishkin 2009, pp. 143–​6. Here Fishkin means polarization, but not sorting. 76 Brennan 2016a, p. 66; Neblo 2015, pp. 181–​4. We may also find a more stubborn divergence of opinion among deliberators.

Democratic Constitutionalism  231 public and from the majority of legislative officials. For instance, legislators could select moderators, and the moderators alone could be allowed know the identity of citizens. This would help to counteract the temptation to pay off deliberators and thus reduce the incidence of corruption.77 Selecting deliberators at random would reduce bias as well. If we want to go further, we could also give representatives of different ideologies and groups equal time to speak with the deliberators. One potential benefit of randomly selecting deliberators from a diverse public is that we are likely to end up with a diverse minipublic that is harder to bias in a particular direction. As Hélène Landemore has argued, diverse groups can often contribute to good decision-​making better than a group of persons with high ability, so long as deliberation is appropriately structured.78 Critically, diversity can trump ability, so deliberators’ diversity might compensate for their weak deliberative skills. That said, elite deliberative groups would be small, so Landemore’s numbers-​trump-​ability theorem may not apply.79 Some will complain that, despite Fishkin’s findings, deliberation on matters of great significance will lead to polarization.80 If a group is diverse, opinion could diverge based on those differences. Further, polarization would be difficult to counteract once deliberators become aware of their ideologies and the ideological balance of their minipublic. However, encouraging them to discuss policy face to face may exert pressure not to polarize. To stop moderates from growing disillusioned and dropping out of the discussion, moderators could require a certain level of participation. This is consistent with recent empirical work showing that discussions with facilitators and explicit deliberative norms “reversed tendencies to group polarization,” but free discussion without facilitators or explicit deliberative norms did produce “the undesired polarization patterns” that Cass Sunstein and others report.81 So minipublic structure can address and ameliorate concerns about partisan divergence. Contact might also avoid marginalizing shy or insecure speakers. It is worth noting a recent finding on this matter, that “in fairly easily replicable circumstances, the expected forms of privilege have little effect on participation and no effect on influence.”82

77 Here I think the benefits of secret decision-​making would exceed the costs, following arguments in Kogelmann 2021. 78 Landemore 2013, pp. 102–​4. 79 Ibid., p. 104. 80 For evidence to this effect, see Sunstein 2011. Fishkin finds little polarization, but that is probably an artifact of the low stakes involved in the deliberative bodies he put together. For an argument that we should expect polarization, see Talisse 2017, p. 119. Here the term “polarization” appears to refer to changes in view, not sorting. 81 Bächtiger and Beste 2017, p. 111. 82 Siu 2017, p. 126.

232  Trust in a Polarized Age One might also worry that the information that deliberators appeal to could be limited or distorted by the political process. And even if politicians did not try to control the flow of information, media and partisan bias would still corrupt the information minipublic members are presented. Thus, the minipublic’s deliberation may not be based on good information, and it may fail to approximate the standards of policy epistemology. An attractive remedy for this problem is to require deliberators to consult prediction markets for information about which policy proposals are likely to be effective. Prediction markets are structured as follows:  participants in the market bet on the effects of various policies in the future by buying and selling predictions, and so, through supply and demand, affecting the odds offered in the market. Participants without good information will tend to have views that are randomly distributed across the relevant range of predictions, and so we can expect their views to cancel one another out. But people with good information will tend to tilt the market in favor of their position and against others. So, prediction markets serve as good information aggregators. They are surely imperfect, but they could dramatically improve the quality of information available to deliberators. As Robin Hanson argues, prediction markets are our most epistemically effective method of gathering information.83 Further, prediction markets have the advantage of anonymity of recommendation, which means that citizen-​deliberators can get information untainted by the celebrity and charisma of a popular cultural figure. Landemore has complained that prediction markets are not themselves a political decision-​making mechanism, but prediction markets can still be a useful supplement for citizen deliberation.84 Philip Tetlock has argued that a few “superforecasters” can beat prediction markets, but there is no reason that citizen-​deliberators cannot listen to superforecasters as well.85 Another concern is that legislators will ignore policy minipublics if the minipublics go against public opinion.86 Since politicians often genuinely care about their constituents’ untutored views, they may have a strong motive to ignore the minipublic, since it is constituents as a whole who vote. And since the minipublic has no binding political power, it would have no way to successfully fight back. In reply, we can only hope that citizens will recognize that, because the minipublic is randomly selected and knows more than they do, they can treat the minipublic as representing their considered views. Citizens often approach their elected representatives in this way, only sometimes blaming them for



83

Hanson 2013. Landemore 2013, pp. 173–​84. 85 Tetlock and Gardner 2016. Superforecasters could advise anonymously. 86 I thank David Estlund for encouraging me to address this concern. 84

Democratic Constitutionalism  233 voting differently than they would like on complex matters. The promise is that the minipublic will be seen as more like the public than the elected representatives and will inherit some moral authority in that way. If this is not convincing, we might be able to get minipublics widely recognized as serving Congress in an advisory role; that would create an expectation that Congress will listen to the minipublics without raising concerns about the minipublics having equal authority with Congress itself.87 We also defer to the judgments of juries in civil and criminal trials, and juries are akin to minipublics. This shows that citizens of many democracies are already prepared to allow a group of randomly selected, representative peers to make important decisions. It is true that, in certain high-​profile cases where the public pays close attention (rare enough), jurors are sometimes blamed and attacked for going against public opinion. But data on minipublics give us reason to think that minipublics can increase the quality of deliberation in the formation of public policy. Another way to increase the influence of minipublics is to require that they list the reasons and arguments they find most convincing, as Michael Neblo proposes.88 Listing reasons and arguments would provide evidence that the minipublic’s recommendations are reflective and careful, it would generate more information for the public to consider, and it might create trust by revealing to the public that the deliberators are good-​willed, transparent, and informed. It would also allow ordinary citizens who addressed those arguments in their own public comments to feel as though they were participants in the minipublic. It could help nonparticipants decide which policies and candidates to endorse. And it would be harder to “jam” these signals, because people could bypass elite interpretation of the minipublic’s recommendations by looking at the arguments for themselves. Politicians would also respond to the minipublics’ reasons and arguments, especially if the public expects them to. Politicians will also have a better sense for how and why some policies have legitimacy in the eyes of the public, since the minipublics would be based on random sampling.89 There is some empirical research that tries to assess the influence of minipublics on policy. John Dryzek argues that, while there are cases where minipublics have had “influence and impact, they are outnumbered by cases where a minipublic is established but turns out to have little or no effect on public decision-​making.”90 And one study has found that the influence of a minipublic



87

I thank Jenny Lambe for this point. Neblo 2015, p. 187. 89 Ibid., p. 188. 90 Dryzek 2010, p. 170. 88

234  Trust in a Polarized Age is inversely proportional to how deliberative it is.91 However, André Bächtiger and Simon Beste have found that the more voters knew about minipublics—​ in particular how they recruited people and how they were free from partisan instructions—​“the more likely they were to vote for the mini-​public’s policy recommendation in the later citizen referendum.”92 A final concern about the use of minipublics takes the form of a dilemma raised by Cristina Lafont. She argues that democratic theorists who appeal to minipublics have two goals:  to represent public opinion, and to filter uninformed and otherwise problematic public opinion. Minipublics have the dual priorities of being “mirrors” of real opinion and “filters” that remove problematic opinion.93 But these priorities conflict. To the extent that we appeal to unidealized or real deliberation, we’ll represent public opinion, but epistemic quality will be low. We get the mirror, but not the filter. Yet to the extent that we appeal to highly idealized and structured deliberation, we’ll improve epistemic quality, but lose representativeness. I believe that my emphasis on moderate idealization provides a way of interpreting the mirroring function of minipublics that reduces the tension. Minipublics are meant to increase epistemic quality so as to bring the decisions of real deliberators closer to what moderately idealized agents would endorse, such that highly structured deliberation can both mirror moderately idealized agents and filter out problematic opinion at the same time. Another reply is that even Lafont allows that minipublics have other positive effects, such as allowing for open contestation and empowering ordinary citizens. So we may have additional reasons to use minipublics.94 Deliberation has flaws, both due to discourse failure95 within the group and due to attempts to influence the group from the outside.96 But the incentives of the group, combined with Fishkin’s long experience with such small-​ group deliberation, give us reason to hope that consulting these deliberators could reduce the errors produced by the productive policy process. We can elevate deliberative quality and tie the passage of legislation to high-​quality minipublic deliberation. There is reason to hope that, through high-​quality deliberation, we can bring laws and policies closer into line with what is publicly justified.

91 Smith, Richards, and Gastil 2015. 92 Bächtiger and Beste 2017, p. 113. Examples include the British Columba Citizen Assembly and the Irish Citizen Convention. 93 Lafont 2017, p. 88. 94 Ibid., p. 95. 95 Pincione and Tesón 2006. 96 We might even find that various parties engage in “jamming” by exploiting uncertainty about deliberators’ preferences in order to manipulate hearers and receivers (Minozzi 2011, p. 310).

Democratic Constitutionalism  235

7.  An Anarchist Challenge* The last three sections explain how to publicly justify the productive policy process. Constitutional rules governing productive policies must not violate the following six conditions. The normal operation of the rules must (i) avoid routine violations of primary rights, (ii) avoid undermining an economy’s robust potential for economic growth, and (iii) provide adequate social insurance. The rules must (iv) avoid incentivizing excessive corruption and rent-​seeking, (v)  yield policies that tend to meet the standards of policy epistemology, and (vi) incentivize political officials to base policy on political epistemology. While it is impossible for any particular constitutional rule to fully meet all of these conditions, a legitimate and authoritative constitution that sustains trust for the right reasons must be comprised of rules that collectively do not violate these six conditions. The immediate worry about this list is that it implies anarchism, since no set of constitutional rules can meet these demanding standards.97 In response, I would first point out that a set of constitutional rules can be publicly justified if it is seen as the best available means of satisfying these constraints and as superior to not having constitutional rules. Since having at least some constitutional rules is likely to be superior to having none, at least some constitutional rules should be in the optimal eligible set of rules, that is, potentially publicly justified and undefeated. Second, once protective policies are in place, modest supermajoritarian voting rules, supplemented by policy minipublics and prediction markets, can satisfy conditions (i), (iv), (v), and (vi). Strong supermajority rules already protect against violations of certain primary rights. Further, modest supermajority rules applied to the productive function of the state help ensure that the policies that are passed are supported by large majorities; this will, in turn, discourage shifting coalitions that can lead to excessive corruption and rent-​seeking.98 If we supplement supermajority voting in the legislature with policy minipublics and prediction markets, we can help to ensure that legislation meets the standards of policy epistemology. It is harder to determine whether a legislature structured in this way will undermine economic growth and social insurance and thereby violate conditions (ii) and (iii). Democratic countries often have strong growth rates and provide fairly extensive social insurance, even if they usually fail to grow as quickly as feasible and leave some marginalized groups economically insecure. But the 97 For a version of this objection, see Enoch 2015. I  first addressed this objection in MPBW, pp. 114–​5. For an excellent, thorough response to Enoch and others, see Schultz-​Bergin 2020. 98 It is worth noting that supermajority rules may not create as much gridlock as one might think; for instance, the US Senate’s use of the filibuster may not create more barriers to legislation than there would be without it. See Ramey 2017.

236  Trust in a Polarized Age fact that the voting public cares about economic growth and economic security means that there is a democratic constituency for policies supporting broad-​ based economic growth and economic security, even if the public is often confused about which policies those are. This will help improve representatives’ ability to produce policies friendly to economic growth and social insurance.99 And while representatives will be tempted to pass policies that hurt growth and contract social insurance, supermajority rules, policy minipublics, and prediction markets should contain those temptations to some extent.

8.  Democratic Constitutionalism Creates Trust for the Right Reasons In this chapter, I have argued that democratic constitutionalism creates political trust for the right reasons and probably plays a role in creating social trust for the right reasons. Since democratic constitutionalism is a traditional component of a liberal regime, this is one more argument for the thesis of the book, that liberal institutions create trust for the right reasons among diverse persons, ending political war. We can also see that the causes of real trust themselves furnish the right reasons to trust. The determinants of real trust count as good reasons to trust others. If, for instance, when people observe “clean” political processes, they become more politically trusting, then that itself is one of the right reasons to trust those officials; we trust them because we see that they follow procedural norms. This means that many causes of real trust not only create trust, but create trust for the right reasons by themselves.

99 Insofar as the public is not anti-​reliable in judging which policies produce economic growth and security. If the public is anti-​reliable, and cannot be persuaded otherwise, then representatives may become less likely to produce the right policies.

8

Elections and Process Democracy This chapter attempts to show that electoral rights, rights to participate in the selection of political officials, create trust for the right reasons. I hope to show first that electoral rights and the resultant democratic practices help to sustain social and political trust in the real world. I will then argue that electoral rights are publicly justified. The arguments I give will also provide support for a new kind of democratic theory I call process democracy, a version of democratic theory that seeks to structure different stages of the democratic process in accord with different democratic values. I begin by looking at the relationship between electoral rights and social and political trust (section 1). The positive relationships found in the literature are weaker than one might hope, but they still illustrate how democracy can help to create social and political trust for the right reasons. I then develop public justifications for electoral rights, appealing to both instrumental and deontological considerations (section 2). I also argue that procedures that institutionalize electoral rights can make authoritative decisions that all or nearly all members of the public are obliged to follow. I next explore the different values that a publicly justified electoral process should embody, in particular widespread participation and equal voice (section 3). I explain the centrality of responsiveness to the reasons of citizens in justifying electoral rights (section 4). I then answer arguments against the responsiveness of democratic institutions advanced by Christopher Achen and Larry Bartels and return to an argument from Gilens to which they respond (section 5).1 The next section addresses a variety of concerns about the public justification of electoral rights, such as voter ignorance (section 6). I discuss the selection of political officials (section 7). I end by developing my account of democratic governance that I call process democracy, which combines elements of the three major classes of democratic political theory—​aggregative, deliberative, and agonistic (section 8).

1.  Elections and Social and Political Trust I begin by connecting electoral institutions, those required by electoral rights, with social and political trust—​trust in democracy, government, and

1

Achen and Bartels 2016.

238  Trust in a Polarized Age political officials in particular. I hope to show that rights to participate in elections, and the exercise of those rights, helps to increase political trust in the real world. Before we look at the data, we must recall the differences between trust in the democratic process and trust in particular political officials. The two can come apart, and, critically, distrust in particular politicians may be good if it leads citizens to better detect when officials deviate from their constitutionally assigned roles. For this reason, I will focus on trust in democracy and the democratic process, not trust in officials. The empirical literature also distinguishes between trust in democracy and trust in government per se. Since much of the empirical work is done on democratic countries, these two questions are sometimes answered with the same data. But in nondemocratic countries, trust in democracy is not applicable. Third, many of the measures of political trust concern satisfaction with democracy. This can differ from trust in democracy, but insofar as satisfaction means citizens think democracy is doing what they trust it to do, as I argued in ­chapter 2, we can use satisfaction measures as proxies for trust. That said, satisfaction with democracy is subject to more fluctuation than other measures of political trust. So, we cannot always assume that trust in democracy and satisfaction with democracy are the same trust judgment.2 I once hoped the empirical literature would demonstrate that there is a positive correlation between democracy and social trust and that democracy improves social trust. I was disappointed. To return to two studies already noted in ­chapter 2, Zmerli and Newton find “robust and statistically significant correlations between generalized trust, on the one hand, and confidence in political institutions and satisfaction with democracy, on the other.”3 The associations are significant in twenty-​three European countries and in the United States. However, Rothstein and Stolle find no relationship between social trust and political trust.4 Rothstein and Stolle’s study is more compelling because they distinguish between trust in political institutions and trust in legal institutions, and they find that trust in order institutions is connected to generalized trust but not to political trust. Further, Bjørnskov finds that “electoral institutions are not associated with social trust.”5 Thus, previous data showing a connection between democracy indices and social trust may



2

Hooghe 2018, pp. 627–​9. Zmerli and Newton 2008, p. 706. Rothstein and Stolle 2008, p. 450. 5 Bjørnskov 2010, p. 323. 3 4

Elections and Process Democracy  239 “prove to be spurious.”6 You indicates that “democracy increases trustworthiness of government and thereby promotes a trusting citizenry” and that “people tend to perceive the same outcome as more fair when they have participated in the process which produced it and when everyone has been given equal rights.”7 In this way, democracy might increase social trust. And yet You admits that the causal effect of democracy on trust is hard to determine due to conflicting evidence.8 Uslaner finds an effect of democracy on trust, but only after forty-​six years of continuous democracy, and he has argued that trust produces democracy and not the other way around.9 Margaret Levi and Laura Stoker have found a connection between political trust and social trust at local levels of government, where citizens can regularly interact with officials.10 They find that “whether citizens trust each other is strongly influenced by whether they have confidence in the government that they share,” so democracy may lead to social trust, but it does so by way of political trust.11 However, Levi and Stoker only connect democracy and social trust at the local level. What’s worse, the introduction of democracy may reduce social trust. New democracies seem to lose social trust as a result of democratization. This happens through two processes. First, social trust is often high in nondemocracies because people must rely on informal networks to acquire many goods and services:  authoritarian governments are generally less responsive to the needs of citizens. When countries democratize, new institutional trust may crowd out preexisting social trust. Second, many citizens of new democracies have high political trust because they are enthusiastic about democratic governance; but as reforms are introduced, expectations are often disappointed, and political trust declines. This means that many newly democratic countries have lower political trust, which can adversely impact social trust.12 Mark Warren has tried to distinguish between types of social trust in order to determine which trust relationships support democracy and which harm it. He 6 Ibid., p. 343. 7 You 2012, p. 407. 8 Ibid. 9 Uslaner 2002, p. 227. 10 Levi and Stoker 2000, p. 495. 11 Ibid., p. 494. Peggy Schyns and Christel Koop have found that general political trust in democratic countries affects interpersonal trust: “The more people distrust politicians and people in government, the less they trust other people in general, even when controlled for all other variables” (2010, p. 145). 12 Letki 2018 reviews these two dynamics. Ideally, we could remove newly democratic countries from the data set, but from what I can tell, doing so does not reveal a significant positive correlation. We would also have to take media effects into account, since transitions from authoritarian to democratic governances often are accompanied by a significant freeing of the press, who can then report on negative events freely.

240  Trust in a Polarized Age argues that if trust relationships “could be publicly justified to those they affect” then they should be compatible with democracy, since democracy is publicly justified.13 Democracies, in his view, may support social trust in three ways: (1) By reducing the risks of generalized trust through security and the rule of law; (2) By reducing particularistic dependencies through universal welfare supports, thus freeing individuals from clientelistic or other kinds of dependency relationships; and (3) By increasing the likelihood that interpersonal and socially mediated warrants can spread through institutions, such as public schools and universities, that cut across particularistic ties.14 I think there is some reason to think that the first connection holds, since it is plausible that democracy reduces corruption, thereby limiting one of the leading causes of distrust.15 I am somewhat skeptical of Warren’s second connection, as I do not think the best explanation of the correlation between social trust and universal welfare supports is that the former causes the latter. As we have seen, social trust leads to support for universal welfare programs, even if there is also a modest causal arrow running in the other direction.16 Warren’s third connection suggests that democracy improves the quality of other institutions and thereby affects social trust at a remove. This might be true as well, but it is a highly mediated effect. The data are more promising with respect to whether elections are correlated with political trust. Marc Hooghe argues that the “entire body of research makes clear that political trust and elections are strongly intertwined, with various potential directions of causality.”17 The main causal relationship seems to run from political trust to participation in electoral institutions: “Not only will political trust help to mobilize voters, it also partly determines voter choice, and subsequently it might be affected by the way an election has been conducted or by election results.”18 Van der Meer finds that the data on political trust do suggest that “elections tend to bolster political trust,” as political trust rises during election cycles.19 13 Warren 2018, p.  81. Warren seems to have a shared-​reasons account of public justification in mind. 14 Ibid., pp. 84–​5. 15 For discussion, especially in comparing the functioning of “natural” states versus “open-​access” states like liberal democracies, see North, Wallis, and Weingast 2013. 16 Bergh and Bjørnskov 2014. 17 Hooghe 2018, p. 617. 18 Ibid., p. 620. 19 Van der Meer 2017, p. 273.

Elections and Process Democracy  241 Elections also plausibly strengthen some of the well-​documented causes of political trust. As noted in c­ hapter 2, citizens tend to trust systems more when they have been involved in making decisions in that system or, at least, believe that they have.20 Margaret Grimes argues that “perceived ability to exert influence in a decision-​making process affects views of a group authority and its decisions because it engenders a feeling of control and being able to secure a more favorable decision outcome.”21 Elections are a form of citizen involvement, and citizen involvement is part of procedural fairness. Second, one of the components of quality of governance is voice, which concerns people’s perceptions about “the extent to which a country’s citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media.”22 So, insofar as electoral rights promote aspects of procedural fairness and quality of governance, elections promote political trust. One complication with the data is that political trust is affected by which party is in power. People tend to trust democracy more when their preferred candidates are in office. Marc Hetherington has argued that, in the United States, “political trust has polarized by party identification.”23 Listhaug and Jakobsen find “a significant impact of ideological distance on political trust,” and that “citizens who vote for parties who win elections are more likely to support the political system than those who vote for parties that are on the losing side in elections.”24 Uslaner argues that “polarized politics makes people less trusting” as a result.25 So, while electoral mechanisms probably increase political trust, they do so more effectively in political orders with less diversity of opinion, or where moral and political disagreements are less strongly correlated with the ideological commitments of political parties.26 This is because citizens will often “evaluate their regime and its institutions by their own core values.”27 There are data suggesting that some institutional reforms could help electoral mechanisms bolster political trust. In particular, certain electoral mechanisms could reduce perceived ideological distance between parties and thereby mitigate 20 Grimes 2017, pp. 259–​62. 21 Ibid., p. 257. 22 Van Ham et al. 2017, p. 163. 23 Hetherington 2015, p. 446. 24 Listhaug and Jakobsen 2018, p. 562. 25 Uslaner 2015, p. 363. 26 As in the United States in the middle of the twentieth century. For discussion of the low level of divergence in this period, see McCarty 2019, pp. 70–​1. 27 Van Ham et al. 2017, p. 154. Interestingly, filtering government action through ideology leads to complex results in policy preferences:  “Voters’ support for broad government intervention in society depends on social trust: higher trust levels are strongly associated with less interventionist preferences.” But “low-​trust voters who have more confidence in government actors than in market actors will demand more government intervention, while high-​trust voters with more confidence in market actors than government actors will rationally demand less interventionist regulations” (Bjørnskov 2019b, pp. 637–​8).

242  Trust in a Polarized Age divergence. Examples include proportional voting systems and ranked-​choice voting systems, both of which allow for the formation of more moderate political coalitions.28 We can illustrate by recalling the dramatic difference between divergence in the United States and in Germany.29 There is some evidence to suggest that proportional systems create more political trust than majoritarian systems, but the evidence is limited.30 And new research backs this up: “Political support is higher in more proportional electoral systems and lower in closed-​list electoral systems;”31 It is also “higher in political systems with multiparty governments; higher in political systems with unicameral legislatures and shorter terms; higher in political systems with more members of parliament per head of the population; and higher in more decentralized political systems.”32 While the correlation between elections and social and political trust is modest, there are two important mitigating factors worth addressing. First, democracy may increase social and political trust indirectly through promoting trust-​creating institutions, and these effects may be hard to detect. For instance, if electoral processes help to protect legal property rights and reduce corruption, then they should increase social trust. If electoral processes improve economic performance and procedural fairness, then they should increase political trust. And there is good reason to think that electoral processes do help to create and sustain trust-​creating institutions.33 Second, the mixed relationship between democracy and political trust may be the result of rising expectations. For instance, in North America, while “people have become less trustful of government, other opinion surveys show continued and widespread attachment to democracy and its ideals, which even may have strengthened in recent decades.”34 It may be that, as economic performance and wealth increase in democratic countries, citizens raise their expectations of government and expect it to perform additional tasks. For instance, as citizens grow wealthier, they start to care more about procedural fairness, which arguably includes fairness in the electoral process itself. And since more educated citizens are often more discerning, rising education levels may make citizens more likely to take note of dysfunctional and institutional failures that poorer, less educated people may not have the luxury or the capacity to track. Russell Dalton points out that “between 1952 and 2000 the trust levels of the better educated decreased at 28 McCarty 2019, pp.  125–​31. This probably occurs because centrist parties can gain political power by being the second or third choice of most citizens. 29 Bélanger 2017, p. 243. 30 Van der Meer 2017, p. 275. 31 Van Ham et al. 2017, p. 13. A closed list electoral system is one where voters only vote for political parties, and not candidates. 32 Ibid. 33 See not only Achen and Bartels, but the data mentioned in ­chapter 2. 34 Dalton 2017, pp. 381–​2.

Elections and Process Democracy  243 a steeper rate than for the less educated. This reinforces the notion that changing expectations, rather than changing performance, is eroding political trust.”35 Political scientists sometimes call these richer, more disgruntled citizens “dissatisfied democrats,” because they are low political trusters but strong democrats. Ronald Inglehart has argued that, as wealth increases, “public priorities are broadening to include new postmaterial values that stress autonomy, self-​actualization, and a more assertive political style” and that the citizens of wealthier publics are somewhat “libertarian” in that they question authority, which leads them to be less trustful of political institutions.36 These engaged, postmaterialist citizens still support democracy, especially protecting equality and protecting minority rights, even though they’re low in political trust. It is worth spending time on this potential explanation because it provides hope that democracy as such is not trust-​reducing. The main idea is that the highly educated adopt higher expectations of democracy, which explains why trust seems lower. Now, importantly, many argue that public support for democracy has not actually eroded in most countries, including the established democracies.37 But when support erodes, it looks like an education effect. While factors like “self-​expression values, social trust, and associational activism” all increase democratic aspirations, “Only the effects of education . . . actually widened the democratic deficit,” that is, created lower trust in democracy. Satisfaction with the “performance of democracy diverges from public aspirations”38 because of a number of factors.39 But probably the biggest reason is that “the higher educated are more likely to be morally troubled by [poor] institutional quality than the lower educated.”40 So falling trust in democracy in some established democracies is probably due to having a more educated populace that can notice and track untrustworthy political behavior more effectively. This is not a reason, then, not to have democracy: higher public expectations from education may provide an incentive for democratic governments to become more trustworthy, allowing trust in democracy to lag behind, but not too far behind, expectations, and perhaps also because people of this sort would distrust an undemocratic system even more. Moreover, where we find falling political trust, it is not because democracy irrationally reduces political trust. People simply come to have higher expectations, and so their expectations of trustworthiness are harder to satisfy.

35 Ibid., p. 390. For an examination of a related theory involving “critical citizens,” see Norris 2011, pp. 4–​15. Political performance might also matter quite a bit in post-​communist societies as well. See Závecz 2017, p. 453. 36 Inglehart 1990; Dalton 2017, p. 387. 37 Norris 2011, p. 4. Also see p. 241. 38 Norris 2011. 39 Ibid., p. 243. 40 Van der Meer and Hakhverdian 2017, p. 86.

244  Trust in a Polarized Age In this way, democracy may reduce social and political trust because it works well, since it leads people to increase their expectations and may thereby create a large gap between these higher expectations and citizens’ assessment of their government’s performance. This means that political trust can fall even if government functioning improves! So, if we could factor out citizens’ expectations of democracy from the political trust data, we might find a positive correlation between elections and political trust that would justify the claims that democracy increases political trust and that elections play a central role in that causal connection. However, there is another possible explanation for falling political trust in established democracies. Perhaps what matters most for political trust is not elections so much as other institutions that accompany democracy, such as institutions that manifest “equal treatment [of citizens] by the judiciary and the executive.”41 Following Jacques Thomassen, Rudy Andeweg, and Carolien van Ham, we can call these Rechtsstaat institutions.42 Some recent work indicates that Rechtsstaat institutions are intimately tied to measures of procedural fairness and corruption, so they may play a larger role in creating political trust than even economic performance.43 Consequently, making electoral institutions more effective “might have only weak or indirect effects on strengthening citizen perceptions of [state] legitimacy” and so on strengthening its trustworthiness.44 We won’t increase political trust with more democracy but with “better quality of governance” that derives from a “well-​established Rechtsstaat.”45 So the dissatisfied-​democrat literature may only reveal that rights practices like equal treatment under the law46 are what really matter for social and political trust. Of course, if elections preserve and strengthen a Rechtsstaat, then elections will produce social and political trust indirectly, but it is hard to establish that connection in the data. I should point out that, while democracy may not do much for trust, social trust can do quite a bit for democracy. For instance, recent data show “a strong and positive association between trust and institutional quality in societies with stable, democratic institutions and no clear association in nondemocracies.”47 Thus, social trust may lead to better institutions in democratic societies but not in nondemocratic societies. And when democratization is successful, long-​ lasting, and arrived at peacefully, then high-​trust democracies better facilitate 41 Hayek also uses this term. See Hayek 2011, pp. 287–​307. 42 Rechtsstaat can be translated as “state under law” or “rule of law,” and in this context, it refers to a state that strongly adheres to the rule of law. 43 Thomassen, Andeweg, and van Ham 2017, p. 522. 44 Ibid. 45 Ibid., p. 31. 46 Which we explored in ­chapters 2 and 7. 47 Bjørnskov 2019, p. 635.

Elections and Process Democracy  245 economic growth and security than otherwise.48 Further, in high-​trust societies, we may end up with better policy choices because changing preferences are more easily expressed in democratic institutions when people trust one another more.49 Next, in high-​trust societies, democracies may function better because politicians and political parties that are not trustworthy may be “forced to mimic the behavior of trustworthy actors to gain votes whenever their overall trustworthiness is lower than average.”50 Perhaps most importantly, however, is the “centrality of trust to democratic continuation,” because it facilitates voluntary compliance: “Citizens must trust their legislators to have their best interests in mind and citizens must trust each other to abide by democratically established regulations.”51 Overall, then, we can embrace the modest claims that electoral institutions and elections increase political trust in established democracies and that trust-​ increasing effects may be masked by the presence of critical citizens. And insofar as electoral institutions cause the determinants of social and political trust, the exercise of electoral rights creates social and political trust in the real world. So, if electoral rights are publicly justified, they do just what the thesis of this book says liberal institutions, such as elections, do: create and sustain trust for the right reasons and slow the distrust-​divergence process.

2.  Primary Procedural Rights* Recall that primary rights are rights that all persons with a rational plan of life or conception of justice would want for themselves and would be willing to extend to others on reciprocal terms. Moderately idealized members of the public will see procedural rights as essential for protecting their more fundamental primary rights of agency, association, and jurisdiction, and for promoting an effective public policy process. The main reason to think that a primary right to vote is publicly justified is that moderately idealized members of the public can see that democratic systems of government are better guarantors of rights and producers of public goods than authoritarian, nondemocratic regimes.52 Given that all members of the public will want their rights protected and the vast majority will want government to produce at least some public goods and services, there 48 Eberhardt 2019, p. 1. 49 Bjørnskov 2019, p. 635. 50 Ibid. 51 Lenard 2008, p. 328. 52 This includes the tendency of democracies to prevent famine, avoid going to war, produce stable growth, protect a range of human rights, and not kill their own people at a mass scale. Reviews of the data include the following: Brennan 2016b;, Sen 2000, p. 178; Halperin, Siegle, and Weinstein 2005, 2010; Acemoglu et al. 2019; Acemoglu and Robinson 2013; Christiano 2011.

246  Trust in a Polarized Age is reason to believe that they will affirm primary rights that allow others to outvote or control them by electing competing officials. So long as the reach of democracy is limited to the protective and productive functions of the state, then, it makes sense for members of the public to rank equal democratic voting rights above less egalitarian decision-​making processes, and to defeat authoritarian decision-​making. Let’s call the protection of the more fundamental primary rights and the provision of public goods the instrumental values supporting democracy, since they justify democracy on the grounds that democracy produces the right effects. Members of the public will also see electoral rights as reflecting the values of equal treatment and respect for persons. Equal voting rights, for instance, are nearly universally seen as a proxy for equal treatment and equal consideration of interests. Electoral rights must be equal if for no other reason than any departure from equal rights would invariably be seen as privileging one social or ideological group over another.53 It is possible that unequal rights will help to produce more publicly justified outcomes, but given that restricting electoral rights will privilege certain persons and that members of the public will not agree about who should have that status, members of the public will instead insist on equal political access. Let’s call these the proceduralist values supporting democracy, since they justify democracy on the grounds that democracy is a fair procedure that embodies respect for persons. Even democracies that follow proper procedures can often violate primary rights of agency, association, and jurisdiction. So, if electoral rights systematically lead to violations of these more basic primary rights, then that is a good reason to restrict the reach of democratic choice. One might take this point to imply that members of the public will be pure instrumentalists about democracy, embracing democracy only when it generates rights-​respecting outcomes. However, most members of the public will give some independent weight to proceduralist considerations. That means they will be prepared to sacrifice some instrumental value to secure proceduralist values like equal treatment.54 For this reason, most members of the public will likely embrace a weighting function that determines the trade-​off rate between instrumentalist and proceduralist values. Evaluative pluralism suggests that members of the public will place different

53 Unless access is somehow restricted to persons that all moderately idealized agents would regard as generally superior producers of publicly justified outcomes. 54 As we have seen throughout the book, political trust judgments are rooted in large part in concerns about whether governments act procedurally fairly, which suggests that insofar as people see democracy as part of procedural fairness, they place a lot of value on democratic input, not just policy output. This helps to support my claim that many members of the public are bound to value democracy for deontic, noninstrumental reasons. People approve of their governments in terms of their policy outputs and political inputs.

Elections and Process Democracy  247 weights on these two types of considerations and so may favor somewhat different democratic institutions as a result. Classical liberal members of the public will follow the instrumentalist tendency of libertarian and neoliberal political economists and philosophers by placing more weight on instrumental considerations; they will also be prepared, if not eager, to limit what citizens can vote over in order to achieve liberal outcomes.55 In contrast, progressive members of the public will tend to give proceduralist considerations relatively more weight. There is nothing unreasonable about either weighting, so electoral rights will be grounded in a multiperspectival balance of instrumentalist and proceduralist considerations. In this way, moderately idealized members of the public will adopt a dual justification for electoral rights. Given the evidence that democracies produce outcomes superior to those in nondemocratic regimes and the weight that many members of the public will assign to the proceduralist considerations that favor electoral rights, these rights should be publicly justified for most members of the public. Importantly, there is wide agreement, even among skeptics who argue that democracy is not responsive to the interests or preferences of citizens, that democratic elections produce several enormous goods. Here is the list provided by Achen and Bartels: (i) elections generally provide authoritative, widely accepted agreement about who shall rule; (ii) in well-​functioning democratic systems, parties that win office are inevitably defeated at a subsequent election because citizens tend to favor parties less in proportion to how long they remain in power, a cost of time in power effect; (iii) electoral competition also provides some incentives for rulers at any given moment to tolerate opposition; (iv) democratic citizens have potential benefits for the development of human character (though Achen and Bartels admit that here the evidence is weaker); and (v) politicians seeking re-​election in well-​functioning democracies will strive to avoid being caught violating widely supported ethical norms in their society.56 Moderately idealized agents should be able to recognize the truth of most of these claims because there is a strong correlation between democratic institutions and these effects and there are social scientific models that can explain the correlation. Thus, these instrumentalist considerations should go a long way toward publicly justifying electoral rights for most members of the public. One novel implication of my position is that epistocratic regimes, where experts have greater influence in policy choice than the average citizen, are not in principle defeated.57 If there are empirical data demonstrating that restricting 55 I discuss neoliberal attitudes toward democracy, as found in F. A. Hayek, Milton Friedman, and James Buchanan, in Vallier 2020b. 56 Ibid., pp. 317–​8. 57 Estlund 2008; Brennan 2016a.

248  Trust in a Polarized Age electoral rights increases, say, rights protections, epistocracy can be publicly justified for many members of the public because, on their evaluative standards, the instrumental goods of epistocracy outweigh the proceduralist values favoring democracy. However, given that democracy will be widely seen as realizing proceduralist values better than epistocracy, the epistocrat will need to empirically demonstrate, in accordance with the standards of policy epistemology, that epistocracy will have considerable instrumental advantages over democracy.

3.  The Value of Equality and Participation Electoral rights are publicly justified because they generate important social goods and treat others fairly. Citizens will weight these two considerations differently, but the aggregate social ranking should be sufficient to demonstrate that the public will rank the possession of equal electoral rights over the alternatives of no electoral rights, weak rights, or unequal rights. We now turn to questions about how to institutionalize electoral rights. These are questions of institutional design. I will appeal to the political values of deliberation, participation, and equality as articulated by James Fishkin.58 These values help to illustrate why designing democratic institutions requires that we make trade-​offs between different values, which will help to motivate process democracy. Fishkin understands deliberation as “the process by which individuals sincerely weigh the merits of competing arguments in discussions together.”59 Deliberation looks to achieve a certain level of quality, such that good legislative and policy options can be identified and decided upon. The goal of deliberation is to realize five values:  information, substantive balance, diversity, conscientiousness, and equal consideration.60 Information should be reasonably accurate; arguments should be offered from different perspectives to achieve balance; viewpoints should be diverse to represent diverse views among the public; participants should conscientiously evaluate arguments; and each person’s arguments should be given equal consideration. Deliberation is, at least, an epistemic value. It helps democracies to make informed decisions. Deliberation may have intrinsic moral value, but in my model, deliberation is important primarily for helping legislative processes avoid legislative error, that is, failures to pass publicly justified laws and failures to reform or repeal defeated



58

Fishkin 2009, pp. 46, 60. Ibid., p. 33. 60 Ibid., p. 34. 59

Elections and Process Democracy  249 laws. I assume here that high-​quality deliberation among political officials will at least somewhat increase their legislative success rate. For Fishkin, participation occurs when the bulk of the population engages in the political process.61 Political participation can be understood more specifically as “behavior on the part of members of the mass public directed at influencing, directly or indirectly, the formulation, adoption, or implementation of governmental or policy choices.”62 Participation includes many activities, including writing letters to one’s representative, marching in demonstrations, signing petitions, publishing editorials, and organizing get-​out-​the-​vote efforts. Participation, like deliberation, seems to both generate important social goods and treat persons with equal respect. In my model, participation is required to ensure that political processes are at least widely responsive to the interests of citizens. Low participation means that political officials are less likely overall to track the interests of citizens. Fishkin defines political equality as “the equal consideration of political preferences.”63 Each person’s preference or choice should be given the same weight. Political equality is typically realized through a one-​person, one-​vote standard, where each member of the public is given one and only one vote over the relevant issues and has an equal opportunity to affect the political process. But Fishkin stresses the indeterminacy of the idea of political equality because equality is neutral between one person, one vote and “microcosmic experiments” where deliberative bodies are chosen through a random-​selection mechanism from the public, such that the public’s views are addressed in discussion among representative persons. I will understand political equality in a general way as the equal representation of interests or viewpoints in politically competitive conditions where persons have roughly equal influence on political decisions.64 Thus, political equality as such doesn’t require that each person have a vote, just that his or her interests or perspectives be considered and treated equally. Equality can, however, be realized by primary procedural rights like the right to vote and participate in elections, as well as by assigning political officials relatively equal legislative power. Ideally, we want democratic procedures that fully realize deliberation, participation, and equality. But in nonideal theory of the sort I pursue in this book, we must weigh these political values against one another in designing democratic institutions. The primary reason for this is that ordinary people have limited time to reflect and deliberate about how to vote.



61

Ibid., p. 45.

63

Ibid., p. 43. Ibid., p. 44.

62 Ibid. 64

250  Trust in a Polarized Age If we want everyone to participate in the democratic process, we will have to forgo high-​quality deliberation, given that most people are unable and often unwilling to engage in the complex learning process required to deliberate well. If we want high-​quality deliberation, we must forgo mass participation and instead accept that deliberation will occur primarily among a small subset of the population, organized under special circumstances. We may be able to design deliberation so that deliberators are at least representative of the population, and so equally represent and consider everyone’s interests, but participation is still sacrificed. The history of democratic practice illustrates the trade-​off between participation and deliberation; as Fishkin argues, there has “never . . . been an institution that reliably delivered political equality, deliberation, and mass participation simultaneously.”65 Fishkin argues that overcoming this limitation is likely to require extremely expensive, even infeasible measures, such as holding national Deliberation Days.66 It might be possible to realize all three values through compulsion, but the necessary compulsion would be so great that it would conflict with personal liberty.67 Compulsory voting may not be excessively restrictive, but compulsory deliberation is far more invasive. And less politically involved members of the public are likely to have defeaters for such coercion. To combine adequate participation and deliberation, we might consider giving up political equality. One way this might be done is to give everyone an incentive to take part in deliberative forums and then, even if only a tiny, unrepresentative minority shows up, allow them to make political decisions. Fishkin calls this “enclave deliberation,” and he rightly worries that such deliberation will “by its very nature attract partisans or even true believers” and so is unlikely to be representative of public opinion.68 The deliberating minority is especially unlikely to represent underprivileged and marginalized groups who, due to a history of oppression, may feel sufficiently discouraged by the political process that they do not participate. Because my approach to public justification appeals to equality in a number of ways, I will take for granted that political equality is the most essential of the three values. Unequal systems lack even the most basic justification and legitimacy, whereas systems with lower-​quality participation and deliberation are flawed but raise fewer fundamental complaints. Participation and deliberation conflict, and we must choose between them. Equality, however, is nonnegotiable.

65 Ibid., p. 47. That said, there is some recent evidence that Brazil’s practice of mass consultation in forming policy realizes both participation and deliberation. Pogrebinschi and Samuels 2014. 66 Fishkin 2009, p. 46. 67 Ibid., p. 54. 68 Ibid., p. 53.

Elections and Process Democracy  251 If we accept that institutions must make trade-​offs between participation and deliberation, we can envision how democratic systems that stress participation will differ from democratic systems that stress deliberation. Participatory democracies will focus on encouraging people to vote, under equal conditions, but will not expend great effort to increase the quality of deliberation in the choice of officials or policies. Such a democracy will instead focus on directly and frequently consulting as many citizens as possible and ensuring that their input informs decisions. Consultations will both shape the substance of which representatives and policies should be chosen and help citizens formulate and change their views.69 Participatory processes will have the moral attraction of being based on an attempt to establish mass consent. We can also hope that participatory democracy will educate ordinary citizens as well as consult them. Deliberative democratic systems realize political equality through representative deliberation rather than mass participation.70 Participatory democracy is vulnerable to the fact that rich and highly educated citizens participate most often, and we typically do not want to compel voting and discussion in order to compensate.71 But deliberative systems can avoid this problem with microcosmic deliberation, where, again, persons are selected at random from the populace and then placed in face-​to-​face discussion.72 These individuals are given information about a policy choice and engage in structured deliberation; they are then given equal voting rights and vote on a proposed public policy. If citizens are randomly selected and have equal voting rights, they should tend to represent a cross-​section of the public and not just the most informed and wealthy citizens; the resulting policy choice can be expected to mirror what the populace would have chosen under similarly informed and reflective circumstances. A random sample of voters can serve as a proxy for the preferences of the public as a whole. Some may worry that citizens are treated unequally when only a small number of citizens participate in public deliberation; on this view, equality cannot be realized without mass participation. But everyone still has voting rights in the scenario I’ve outlined, so all are allowed to participate. Everyone already has the power to influence the democratic process, so minipublics will not violate equality, provided that we do not restrict electoral rights and that choosing deliberators through random sampling leads to representing an appropriate cross-​section of public opinion. 69 Ibid., p. 77. 70 Deliberative democracy has many advocates. For what continues to be a nice range of views, see Bohman and Rehg 1998, especially the articles by Joshua Cohen, Rawls, and Habermas. For a recent update on the state of the theory of deliberative democracy, see Parkinson and Mansbridge 2012. 71 Fishkin 2009, p. 80. 72 Ibid, p. 82.

252  Trust in a Polarized Age As I have noted, two of the main stages of the democratic process are choosing officials and choosing policies. These stages involve different groups. Officials are chosen by the public as a whole, by sectors of the public, or other officials, whereas the officials themselves are usually the only people involved in policy choice. Both stages of political choice can prioritize participation or deliberation. In the case of selecting officials, participation means that a large number of citizens voted in favor of electing an official. Deliberation means that some group of representative citizens deliberated about which officials to elect. In the case of policy choice, participation means that a large number of officials voted in favor of a public policy. Deliberation means that some group of officials deliberated about which policies to impose. Democratic input and output can be participatory, or both can be deliberative. Alternatively, democratic input could be deliberative and output participatory, or vice versa. Process democracy, as we will see, combines a participatory public with a deliberative legislative body.

4.  Participation, Representation, and Public Justification* One reason to support electoral democracy is that it represents the will or interests of citizens. So, in this section, I want to say more about how citizen participation leads officials to represent them. I will not provide a formal analysis of the concept of representation.73 I will instead adopt a working account of representation as occurring when representatives are positively responsive to the justificatory reasons of the people they represent. This account will help address Achen and Bartels’s recent argument that democracy is not responsive to citizens.74 My public justification-​based account of representation can sidestep their criticisms of much democratic theory. Here I will follow John Thrasher’s “thin” account of representation, which is neutral between various theories of appropriate or good representation. Thrasher defines representation in terms of scope and substance.75 The scope of representation identifies the set of citizens to be represented. This will vary depending on whether the set of citizens is equivalent to a nation as a whole, a state, or a locality. We can use the term “constituency” to denote the relevant scope. The substance of representation refers to the “feature or features of the constituency the representative should take into account when acting as a representative.”76 An attractive conception of representation combines conceptions of scope and substance. 73 Rehfeld 2018 provides a developed account of the concept of representation, as does Mansbridge 2003. 74 Achen and Bartels 2016. 75 Thrasher 2016, p. 618. 76 Ibid.

Elections and Process Democracy  253 This means that if political officials have a duty to represent their constituents, then they must represent both their scope and substance. Thrasher argues that there are two main ways of specifying the substance of representation:  the delegate and trustee models.77 The delegate model holds that the substance of representation is the aggregation of the preferences of the constituency, whereas the trustee model holds that the substance of representation is something like the interests of constituents. Importantly, preferences and interests can come apart, as some may base their preferences on mistaken views of their interests and some may prefer to sacrifice their interests for others. So, the delegate and trustee models may sometimes conflict. My account of the substance of representation differs from the delegate and trustee models.78 Representatives should represent neither the preferences nor the interests of citizens but their justificatory reasons. Thus, representatives must be responsive to what their constituencies endorse based on their constituents’ justificatory reasons at a moderate level of idealization.79 A representative satisfies the ideal of representation when she expends adequate effort to align the policies she supports with what policies her constituency has sufficient reason to support or oppose. To understand the idea roughly, representatives should strive to advance policy based on their constituents’ best arguments and values. This public justification model of representation also provides an account of political trustworthiness. A  representative is fully politically trustworthy only when her political behavior tracks what is publicly justified for citizens. Even though a trustworthy representative sometimes ignores what citizens support in the real world, she can always defend her actions by arguing that she has taken their constituents’ concerns into account by adopting policies supported by their best arguments. In this way, the representative is not hoodwinking her constituents but respecting them as free and equal. Representatives will not always be appropriately responsive, and they will respond to some of their constituents more than others. But so long as citizens can determine that their representative has tried to track what they have sufficient reason to support or oppose, that is a basis for thinking that the representative has acted out of concern for her constituents, since there is a general social norm that representatives should so act. In that case, representatives merit their constituents’ trust. Politicians will sometimes track what is justified for citizens without having any moral motivation to do so; they may be solely concerned, say, with

77 Here Thrasher follows Fox and Shotts 2009. 78 I assume here, without defense, that the scope of representation includes all members of the public. 79 I develop an ethic of representation in Vallier 2014a, chap. 6.

254  Trust in a Polarized Age re-​election. In that case, I maintain that trust has decayed into reliance. Citizens who believe that politicians only care about remaining in office will tend to trust those politicians less, and reasonably so. Such a politician or representative does not have the interests or concerns of her constituents at heart. Importantly, the public-​justification account of representation implies that good representatives will often be morally required to lose elections if advancing policy based on the best arguments of citizens conflicts with citizens’ preferences during the electoral process. For instance, representatives might be required to implement carbon mitigation policies because she thinks that her constituents’ best arguments support it even if they deny this when asked. I grant that this recommendation conflicts with how politicians behave, a problem I address later.80 We can now scale up and explain the conditions under which democracy as a whole represents citizens and trust in democracy is well grounded. A democracy is generally understood to be responsive when “the democratic process induces the government to form and implement policies that the citizens want” or, at least, when there is something like policy adoption, where politicians as a whole “fall in line with citizens’ sentiments.”81 On my account of democratic representation, a democracy is representative when its representatives track what is publicly justified for their constituents and voting rules are structured so that the normal behavior of those representatives tends to yield policies publicly justified for citizens in the aggregate. Trust in democracy is appropriate, therefore, when the democratic process meets some threshold of representativeness and there is at least a sense that policy adoption is based on the recognition that most democratic officials are morally motivated to choose those policies. The most powerful challenge to my model of representation is that representatives are both unable and unwilling to respond to what is publicly justified for their constituents. They are unable because determining what moderately idealized citizens want is beyond the ordinary cognitive ken of representatives or, perhaps, anyone else. And they are unwilling because following what moderately idealized citizens want, when it contradicts what real citizens are willing to vote on the basis of, may lead them to lose elections. The objection from inability is plausible because even citizens themselves may not know what their moderately idealized counterparts will endorse. But I assume that a representative is not wholly in the dark when it comes to what citizens may want when moderately idealized. Moderately idealized agents share the values and deep commitments of their constituents, but not every passing

80 We might be able to encourage politicians to represent their citizens’ best arguments rather than their current preferences through secret ballots for representatives. For a nice discussion of when secret voting and public reasoning might support one another, see Chambers 2004. 81 Esaiasson and Wlezien 2017, pp. 701–​2.

Elections and Process Democracy  255 whim they have. There are ways of determining the stability of the preferences of citizens and the values they endorse. Some preferences, and many moral and political values, are quite stable over time, though we will see that Achen and Bartels contradict this claim. My hope is that minipublics with access to prediction markets will yield the arguments that most citizens would accept under conditions of moderate idealization; if, in fact, minipublics uncover the reasons of citizens, then representatives are bound to follow their positive recommendations on the whole, or if the minipublics uncover defeater reasons for a policy, to not implement it. This will be especially well facilitated if the minipublics list the arguments they found compelling in their reports of their discussions. The duty of representation does not require that representatives endlessly collect, organize, and model data. They can proceed, for instance, by taking well-​designed polls as a prima facie proxy that future research—​say complex deliberative polls—​might overturn. Further, I only require that representatives be sensitive to information about what moderately idealized members of the public want when that information is available. When the representative cannot track down the relevant information, she is freed from the duty of representation. The objection from unwillingness is serious because the public is often misinformed about simple facts that moderately idealized agents will quickly acknowledge. When this is the case, I think representatives should take some risk of losing elections if they are genuinely attempting to represent their citizens. In the meanwhile, a representative can try to convince her constituents that she has in fact represented them well. That is a common feature of many political campaigns and other forms of politicking, so it does not appear excessively demanding. My accounts of responsiveness and representation help explain why we should have democratic policy choice, not just democratic elections. Note that in ­chapter 7, I did not argue that all legislators must be equals in the sense that each has a constituency of the same size and a vote of the same weight. That was intentional: the arguments in c­ hapter 7 are not sufficient to establish equality among legislators. The primary rationale for legislators having equal power is derived from the equal electoral rights of citizens. Insofar as representatives are equally apportioned to their constituents, in that each legislator represents a roughly equal number of citizens, they should have roughly equal power in order for citizens to have roughly equal power. Further, democratic government is more likely to be equally responsive to all citizens if each representative has one vote and represents the same number of constituents. If Congressman John represents one million people and Congresswoman Reba represents one thousand people, then assigning them equal voting rights will mean that Reba’s constituents have

256  Trust in a Polarized Age far more power than John’s constituents and that government may be more responsive to Reba’s constituents than John’s constituents. One complication is that assigning representatives unequal power may help to ensure that laws and policies are publicly justified. Imagine, for instance, a society with sharp cultural and ideological differences between provinces that are greatly unequal in size, which is in this way much like the United States. In that case, giving representatives of each province equal votes would make policy more likely to consider the interests of every cultural or ideological group, which may help citizens of those provinces to have stable, ongoing social trust between them. This was arguably one of the rationales for assigning two senators to each of the original thirteen states in the United States. However, the fact that Wyoming senators have as much sway as California senators raises real moral concerns because it gives each Wyoming citizen more influence than each citizen of California. While it has not raised complaints in many periods of American history, perhaps because the House of Representatives is better apportioned to the citizenry, serious problems can arise. For instance, today in the United States the greater influence of rural citizens relative to urban citizens arguably exacerbates divergence and political distrust.82 The fact that Donald J. Trump won the presidency in 2016 despite receiving three million fewer votes than Hillary Clinton arguably undermines the rational basis for political trust in the government because it is led by someone who is not representative of a majority of the populace. That said, it may be better to have different principles of representation, so long as at least one part assigns citizens equal voting power by means of equally apportioned representatives that each have one vote. And it may be reasonable for another part (like the US Senate) to apportion representation by community, independent of size. Another branch could apportion representation by community, independent of size. So, the case for equal voting power among legislators is in some respects incomplete since there are other ways to institutionalize the equality of citizens. In sum, electoral rights should cover the election of officials responsible for tracking what is publicly justified for their constituents, which should help to generate more publicly justified laws than if politicians ignored their constituents or followed unreconstructed mass opinion. That will help to ensure the public justification of electoral rights as well, since structuring them in this way makes it more likely that elections will ultimately yield publicly justified outcomes.



82

Wilkinson 2019.

Elections and Process Democracy  257

5.  Empirical Challenges to Democratic Responsiveness My public-​ justification account of representation can be summarized as follows: Political officials have a duty to try to advance policy that they reasonably believe is publicly justified for their constituents. While this is challenging, politicians can roughly approximate what is publicly justified by trying to act on what they think are their citizens’ best arguments for endorsing certain policies. Their best arguments will be a combination of their personal values and beliefs, as well as their empirical views, and the official must do her best to harmonize those considerations to determine how to form and vote on legislation. Politicians merit our trust when they act in this way; representing citizens’ best arguments is what makes them politically trustworthy. Similarly, a democracy is representative to the extent that policy is made based on the input and best arguments of citizens, and a democracy is trustworthy to the extent that it tends to produce policy that can be justified to citizens on their own terms. Two worries about my account of representation are that democracies are not actually responsive to the reasons of citizens and that they are responsive in pernicious ways. Since I want to show that electoral democracy creates trust for the right reasons and I claim that democracy is only trustworthy when it is responsive to its citizens, I need democracy to in fact be responsive. So, in this section, I will address Achen and Bartels’s well-​known recent argument that democracy is not responsive to citizens, as well as how their work challenges public reason approaches to democracy generally. I will also assess Gilens’s argument that democracy is only responsive to the interests of the affluent, in part because Achen and Bartels’s findings challenge his argument.

5.1.  Achen and Bartels’s Challenge The thesis of Achen and Bartels’s Democracy for Realists is that elections don’t produce responsive government. Instead, “Voters, even the most informed voters, typically make choices not on the basis of policy preferences or ideology, but on the basis of who they are—​their social identities.”83 The worry is that “if voting behavior primarily reflects and reinforces voters’ social loyalties, it is a mistake to suppose that elections result in popular control of public policy.”84 Instead, election outcomes are simply “erratic reflections of the current balance of partisan loyalties in a given political system.”85 I have already addressed

83

Achen and Bartels 2016, p. 4.

85

Ibid., p. 16.

84 Ibid.

258  Trust in a Polarized Age the possibility of using small-​scale deliberative bodies to reveal the preferences and reasons of citizens in c­ hapter 7, but Achen and Bartels argue that empirical studies of these small political bodies “seem to be less relevant for understanding democratic politics on a national scale.”86 For the purposes of this section, I will assume that microcosmic deliberation cannot serve as a substitute for national elections. Part of Achen and Bartels’s argument is that citizens lack enough information and motivation to engage in retrospective voting, voting based on the past performance of political parties and candidates. In general, Achen and Bartels claim to cast “considerable doubt on the view that citizens can reliably form and act upon sensible retrospective judgments at election time.”87 Retrospective voting is complicated by three factors: low political knowledge,88 mistaken assessments of one’s own well-​being,89 and a limited time horizon for economic accountability where citizens only attend to their economic condition in proximity to a national election.90 Achen and Bartels embrace a group theory of politics. This means that “the primary sources of partisan loyalties and voting behavior . . . are social identities, group attachments, and myopic retrospections, not policy preferences or ideological principles.”91 This is so in part because extra information doesn’t make citizens more accurate; instead, “Party identification colors the perceptions of the most politically informed citizens far more than the relatively less informed citizens.”92 Citizens are not generally responsive to information about political parties, candidates, and the like, but rather filter the information to empower their group and disempower out-​groups. Achen and Bartels think their research has profound implications for democratic theory. It leaves democratic theory “in a shambles” because all “the conventional defenses of democratic government are at odds with demonstrable, centrally important facts of life,” specifically human limitations in information and motivation.93 I think they overstate their case. One reply is that voting based on group membership can be an informational shortcut insofar as one’s group allegiance leads one to support the candidates one’s reasons would lead one to support. For instance, if black Americans vote based on how leading black legislators tell them to vote, that may indeed be a reasonable heuristic that helps protect them from serious harms by more powerful social groups. Thus, voting based

86 87 88 89 90 91 92 93

Ibid., p. 2. Also see p. 29. Ibid., p. 175. Ibid., p. 40. Ibid., p. 93. Ibid., p. 158. Ibid., p. 267. Ibid., p. 279. Ibid., p. 306.

Elections and Process Democracy  259 on group membership, while sometimes problematic, may not be irrational. We can only show that it is irrational if we can show that following the views of the leaders of one’s group leads one to embrace more objectionable policies than one might otherwise accept. Second, Achen and Bartels need to argue that because we determine our policy preferences by group membership, those preferences are unjustified. But what makes our beliefs rational is not how they came about (the causes of the beliefs) but whether the beliefs are justified by good reasons at present (the sustaining justifications of the beliefs). Thus, to show that our political preferences are problematically arbitrary, it is not enough to show that the beliefs are caused by group membership; it has to be shown that the influence of group membership shows that political preferences are irrational. After all, most of our beliefs are heavily affected by all kinds of nonrational factors, but this fact does not by itself undermine the rationality of our beliefs. Achen and Bartels can reply that policy preferences and related beliefs change so erratically and quickly that it is hard to imagine that these beliefs could be justified while they’re held. Rational beliefs should be more stable, since the reasons to support various policies do not change erratically and quickly. However, if it turns out that making political judgments is very difficult even for highly rational agents, then the evidence we have might not force us to hold one position or even limit us to a handful. If we move between reasonable, incompatible beliefs, even unpredictably and quickly, then that may merely mean the issue at hand is hard to resolve, not that the public’s preferences are irrational. The point is that while the causes of a belief or position might be arbitrary and tribal, the belief or position might still be a reasonable one. But the best way to answer Achen and Bartels is to examine how their arguments challenge particular approaches to justifying democracy. So, I will now address how their views might undermine my public-​justification approach. The basic defense I offer is that Achen and Bartels at best show either that most people have bad reasons to support, or that most people have no reasons to support, whatever policies they support—​that most people do not have genuinely justifying reasons. But neither claim is a good reason to abandon a public-​ justification approach.

5.2.  Achen and Bartels versus Public Reason* On my view, elections are grounded in primary procedural rights that are publicly justified, and constitutionalizing them proceeds in three stages:  demonstrating the public justifiability of electoral rights as primary rights, examining the public justifiability of the laws we can expect them to generate in their normal

260  Trust in a Polarized Age operation, and showing that they produce the requisite kinds of stability. Achen and Bartels’s work most strongly challenges electoral rights at the second stage of constitutional justification. Let’s see why their results do not challenge these rights at the first and third stages. At the first stage, electoral rights are publicly justified as primary rights for several reasons other than the responsiveness of democratic institutions. Democratic societies generally protect rights of agency, rights of association, and jurisdictional rights, like the right to private property, better than nondemocratic ones. They have a number of instrumental benefits that Achen and Bartels themselves cite, and which I listed previously. And members of the public will likely still find systems with equal electoral rights morally superior to systems with unequal voting power. So Achen and Bartels’s findings should not challenge the public justifiability of electoral rights as generic primary rights. At the third stage, we should ask whether constitutionalizing electoral rights will generate durability and immunity better than nondemocratic regimes, regimes with unequal political power. I think there is an argument for elections here as well, given that authoritarian societies generally have less durability and immunity and so have less of the right kind of stability. They have less durability because citizens lack assurance that government officials will be responsive to their interests, and they have less immunity because legislatures are more likely to be dominated by extractive rent-​seeking groups that citizens cannot resist through the electoral process.94 Democracies have loads of rent-​seeking, but authoritarian societies can be entirely predatory, as predation is often led by political leaders. Authoritarian regimes also incentivize the creation of gray and black markets, which lead to self-​interested, low-​trust behavior, though these markets can generate trust between market participants in some respects. But Achen and Bartels’s work does challenge the second stage of constitutional justification, since their data can be used to argue that a right to participate in elections will not tend to produce publicly justified laws. Thus, exposing lawmaking to elections may yield fewer publicly justified laws and policies than rules that limit the range of laws and policies that elections affect. This suggests that a publicly justified polity should rely more heavily on epistocratic mechanisms to insulate the selection of political officials from direct citizen input. Let us call laws and policies that depend in some direct way on electoral outcomes democratically exposed. Laws might be democratically exposed through referenda or in directly elected legislative bodies. Achen and Bartels’s arguments can be used to challenge my claim that democratically exposed laws and policies will track the reasons of moderately idealized agents. This is



94

Again, for an excellent analysis of extractive institutions, see Acemoglu and Robinson 2013.

Elections and Process Democracy  261 because, given extreme group loyalty, democratic outcomes will be sensitive to arbitrary factors that undermine any connection to what moderately idealized agents support. Call this the tracking objection. What’s worse, the data suggest that many citizens may not have justificatory reasons at the right level of idealization because they have no settled views about what to support or oppose. Thus, democratically exposed laws and policies will tend to reflect arbitrary and biased preferences and group loyalties. Call this the bad-​reason objection. The tracking objection can be answered by conceding that citizens’ political preferences and behaviors can be affected by their group while insisting that political parties may nonetheless be good proxies for what moderately idealized versions of their members would support.95 One reason for this is fairly obvious:  if citizens are committed to valuing what their in-​group values, and their in-​group is their political party, then what the parties favor is directly tied to what citizens favor. Thus, insofar as moderately idealized citizens are committed to their party’s success and effectiveness, their good reasons will be determined by what the parties favor. Compare a citizen’s commitment to the scientific community’s findings. If a person has the reason “Generally believe what the scientific community believes, because scientists are rational people, and I’m a rational person too,” then that person has sufficient reason to defer to the findings of the scientific community. In the same way, if a person has the reason “Generally believe what my party believes,” then that person has sufficient reason to track what the party believes. One might respond that moderately idealized agents will give less weight to the opinions of their group vis-​à-​vis good arguments for choosing policies and political officials. Moderately idealized agents will endorse laws and policies on the basis of decent information and arguments, and those determinations are unlikely to mirror what their group affirms. And yet, while this is so, many political issues are sufficiently complicated that boundedly rational agents will not be able to form opinions on all issues, so they may nonetheless want to use party or political group support as a heuristic. Along the same lines, parties are not always biased in favor of power and a desire for oppression. In many democratic societies, parties track political ideologies and the moral values of large swaths of the electorate. They also articulate platforms and arguments that are loosely responsive to evidence. So citizens may not act irrationally in embracing party platforms, and their endorsements may well track their justificatory reasons. Their motivation for adopting the party platform may be largely tribal, but they may yet have good reason to endorse that platform all the same.

95 For an excellent attempt to show how public reason liberals should approach the ethics of political parties, see Bonotti 2017.

262  Trust in a Polarized Age A second objection is that, in societies with significant partisan divergence, especially when parties are more divergent than the general public, parties may be subject to greater biases than ordinary people. In that case, the heuristic “Generally believe what my party believes” may lead people to epistemically worse beliefs. Consequently, the heuristic may not help people’s opinions track their justificatory reasons but, rather, make them less likely to recognize what they have good reason to believe even by their own lights.96 But this leads into the bad-​reason objection. The bad-​ reason objection says that, since people typically have tribal preferences, moderately idealized versions of tribal people will endorse not good reasons but reasons that we ordinarily would not want to have figure into public justification. They may, for instance, change their views on the appropriate foreign policy with respect to Russia based on the fact that President Trump says good things about Vladimir Putin. This implies that we should not base laws on the reasons that moderately idealized persons will typically and normally endorse. I have two replies to this objection. First, even if many citizens have merely tribal reasons to support a coercive policy, that merely means at worst that we cannot have that policy. And if the law is not coercive, or if persons regard the process for selecting the law as legitimate, then we can permissibly impose the law on them because they are too flighty to have sound objections. We must of course ensure that the law is publicly justified for persons who do embrace good reasons. But the mere fact that most members of the public lack good reasons to favor or oppose many laws and policies does not make a public-​justification standard inappropriate. It just means that the public justification of the law depends primarily on the reasons of that minority of the public whose reasons should figure into public justifications. Second, it is implausible to think that persons will have no good reasons at a moderate level of idealization simply because they exhibit arbitrary and easily altered preferences in the real world. Deliberative polls, for instance, often yield stable judgments, which even Jason Brennan acknowledges have promise.97 And deliberative polls are arguably proxies for moderate idealization since they subject opinions to critical scrutiny and proffer better information. Moreover, plenty of empirical work in psychology demonstrates “value stability,” the empirical finding that the values that persons embrace are remarkably stable over the course of their lives.98 Thus, insofar as citizens can trace inferential routes from their stable values to their support of or opposition to particular

96

I am grateful to Scott Simmons for this objection. Brennan 2016a, pp. 66–​7. 98 For recent work demonstrating value stability, see Milfont, Milojev, and Sibley 2016. 97

Elections and Process Democracy  263 laws and policies, there is a basis for ascribing good reasons to them because laws and policies can clearly advance or undermine their stable moral values. For this reason, I conclude that a public reason liberal case for democracy is not in shambles, despite Achen and Bartels’s important work. Their work does challenge the position I defend, but their objections can be answered. Perhaps in future work, Achen and Bartels, or other democratic theorists, will tease out the way in which their findings challenge public reason liberalism, but for now we can proceed without too much worry.

5.3.  Gilens I now turn to a second challenge to democratic responsiveness, one that I have already partially addressed, namely Gilens’s empirical work showing that democracy only responds to the preferences of the top 10 percent of the income distribution. If Achen and Bartels are right, democracy isn’t responsive to anyone’s preferences in the sense we ordinarily expect, including the top 10 percent. So, then, who is right? If we want to side with Achen and Bartels, then we need to explain how policy is correlated with the opinions of the top 10 percent without their having much causal impact. This could occur in two ways: either parties influence the top 10 percent, or politicians tend to hail from the top 10 percent and do not change their opinions much when in office. In the first case, parties may care a lot about shaping the opinions of the top 10  percent for various reasons, like garnering campaign contributions. The top 10 percent may also be more easily influenced because they have greater access to education and information, so they pay more attention to what parties say, and parties know this. In the second case, since most politicians come from the top 10 percent, they may simply share the views of the group in which they originated and may tend to respond to new information in much the same way as other members of that group. If either hypothesis is true, it explains the correlation Gilens finds without relying on a causal connection. Thus, we do not need to worry much about the correlation Gilens uncovers.99 We should instead worry about whether parties pass laws that are publicly justified and reform or repeal laws that are undermined by citizens’ defeater reasons. Still, the public reason liberal should worry that parties have little or no incentive to respond to the reasons of most citizens even if, as Gilens notes, the policy preferences of the rich and the poor often converge. If political officials do 99 Also recall McCarty 2019, pp. 150–​2. Gilens’s correlation may be explained by inequality, which creates divergence and gridlock, thus generating less responsiveness. And conversely, more divergence and gridlock may lead to both less responsiveness and more inequality.

264  Trust in a Polarized Age not care about a group’s policy preferences, they will probably not care about the reasons that group would endorse when moderately idealized, that is, the group’s best arguments. Fortunately, I  think the legislative reforms I  have explored should help fix the problem. If, for instance, we use minipublics and prediction markets to make policy recommendations, the influence of the poor and middle class should increase, since they will have access to good data and have relatively more influence in discussion and debate than without these institutions. This should improve responsiveness to their reasons and arguments, given that deliberative polls and prediction markets should lead minipublics to approximate moderate idealization. It is plausible that inequality-​compressing policies will make democracy more responsive to the preferences of the nonrich. Unfortunately, however, inequality of influence does not seem to vary much with the increases in economic inequality in the United States. It is not clear from the data, for instance, whether increasing the top tax rate by 20 percent would make democracy more responsive to most people. It may be that the rich have methods of influence that do not depend on their absolute income levels, such as high social status. Even if that is so, however, we can nonetheless push for coercion-​reducing policies that compress inequalities: such policies can, as we saw, be publicly justified on the grounds that they reduce coercion.100 My concern is that greater political influence might be subject to a ratchet effect. Perhaps previous economic inequalities enabled the economically advantaged to establish stable chains of influence, such that they are not much affected by increased redistribution of wealth. Once politicians begin to fraternize with and hail from the richest classes, the revolving door will persist even with large amounts of redistribution. This may be pessimistic, but there may be ways around the problem, such as moving to a system where representatives are chosen by lot.101

6. Selecting Officials The democratic process typically begins with the nomination and selection of political officials, such that the selection of officials is the political process at the furthest causal distance from legislative outcomes. Subsequent stages of the political process can neutralize or alter public input.

100 That said, there is evidence that material inequalities are usually translated into status inequalities, in which case compressing material inequalities would reduce status inequality. See Harkness 2017. 101 Guerrero 2014.

Elections and Process Democracy  265 I contend that political equality and participation are the most important values to consider in designing a process for selecting officials; deliberation is important, but it can be sacrificed for the other two values as long as deliberation is institutionalized later in the democratic process. And I already argued in the last chapter that deliberation can be institutionalized during policy choice. Political equality is critical for justifying the selection of political officials for two reasons. First, politically equal choice procedures treat persons with equal respect, and second, politically equal choice procedures convey officials’ judgments about the equal worth of citizens. The first argument begins by recognizing that citizens have publicly justified equal primary rights to vote and participate in the political process. Citizens have equal rights to attempt to cause the selection of the officials that govern them, even if their contributions are likely to be miniscule. If officials are not selected by a system that respects equal rights to influence the political process, then that system cannot be said to respect our considered judgment that citizens have equal rights to vote. Perhaps we can allow for restrictions of equal rights later in the democratic process because other values are more important. But those restrictions are plausible primarily because political equality is guaranteed at the beginning of the democratic process. The second argument for the importance of political equality is that unequal democratic systems can convey to some citizens that their government does not believe that their interests are of equal importance to those of more privileged citizens. Political inequality can even send the message that political officials and the politically privileged classes do not believe that persons are moral equals. Consequently, violations of political equality send the message that the marginalized count for less than others. As noted above, and as Christiano argues, democratic decision-​making should not only realize equality but do so publicly because democratic decision-​making treats persons as equal in ways that all can appreciate.102 Citizens must be able to see that their government treats them all as equal. Some democratic theorists may argue that political equality is undermined by congressional representation on the theory that political equality means that persons must have direct causal impact on policy formation. But since this book is a work of nonideal theory and real citizens typically lack the time to formulate and discuss policy positions, feasibility considerations justify representation as a heuristic for implementing policies that most people want. The selection of political officials must also realize the value of participation. We ordinarily care more about everyone having input into the selection process



102

See Christiano 2010, p. 95.

266  Trust in a Polarized Age than about generating informed, quality choices, even though we care about quality choices a great deal. We care more about participation for two reasons. First, participation helps to make political institutions responsive to the interests of everyone, at least with regard to the factors endorsed Achen and Bartels. And second, participation helps to sustain descriptive legitimacy by giving all citizens a sense that their input into the political process matters. While the responsiveness of political institutions to public input is far from perfect, political institutions may be more responsive to the input of citizens if political officials cannot afford to ignore large swaths of the public. Systems that suppress or fail to encourage participation will be less responsive to the collective will of their citizens. If we prevent some people from participating in the democratic process, or if people believe that their participation is ineffective, then stressing the need for democratic participation will also undermine descriptive legitimacy.103 If citizens are prohibited or discouraged from participating in the selection of officials and are instead represented by elite deliberators, or some other subset of the public, the chosen officials will not be seen as politically legitimate. Democratic citizens will believe, perhaps rightly, that their basic political rights have been nullified even if someone else represents their views. Citizens may feel as though they have no voice. This felt loss of empowerment could easily undermine the stability of the political order, as citizens will see relatively little reason to abide by legislation passed by officials chosen by someone else. Restrictions on participation may even convey that political elites believe that their input has greater significance than that of excluded citizens, which can further sow distrust.104 There are three main challenges to my claim that the selection of officials can realize the values of participation and equality: concerns about (i) the moral significance of unreconstructed, raw, mass public opinion, (ii) the coherence of public opinion, and (iii) the choice of a voting rule that reflects the will of the people. The first objection holds that people should not be coerced as a result of a process that is based primarily on uninformed, irrational, and bigoted groups of people, like much of the voting populace.105 This problem, I think, is the easiest to solve because the main reason to discount raw, mass public opinion is that public opinion might lead to the election of an evil or tyrannical or incompetent official. But with proper constraints later in the political process, such as rights protections, the power of such officials can be sufficiently limited to make 103 I  understand descriptive legitimacy as the property a democratic state possesses when its subjects believe it to be legitimate in the sense that they believe that the state has the right to rule. 104 This point is complemented rather than challenged by the fact that democratic policy outputs may matter more for justification and legitimacy than democratic inputs. 105 For a review of public political ignorance, see Brennan 2011, pp. 161–​78.

Elections and Process Democracy  267 the problem manageable. So even if a bad official supports policy that cannot be publicly justified, further constrains on his power can circumscribe the effect of poor voting. Of course, bad political officials can find some way around the constraints placed upon them by other branches of government or the constitution. This is true, though it is less of a problem in some democratic systems than in others. In the United States, the primary culprit is arguably the office of the presidency, which has accumulated incredible power over the course of the last one hundred years; those powers are easily and frequently abused.106 But a better reply is that abuses of power can happen under any system where social trust and descriptive legitimacy break down to the point where the constraints in the constitution are no longer effective, or at least not adequately so. My contention is that political trust, descriptive legitimacy, and stability are better maintained when everyone feels that they have some causal input into the political process than by a system where representative persons make decisions for them, even if the decision-​making quality of elite deliberators would be greater. My primary evidence for this is drawn from work suggesting that democratic governments tend to generate more trust in political institutions over time, reviewed in ­chapter 2. Again, a participatory system should be more stable and legitimate than a system with low participation and high-​quality deliberation because citizens will regard a system where they have causal input as morally superior to systems that prevent or discourage them from participating. Further, political officials, even presidents, can be hemmed in by social norms if citizens and elite political officials expect them to follow certain kinds of public norms; presidents wield less influence by violating these norms. So policies that increase political trust can also limit the power of malign political officials. A third reply derives from Achen and Bartels, who argue that democracy is not responsive to the preferences of most citizens, in which case the fact that the people are uninformed, irrational, bigoted, and so on, is not a very potent threat to policy outcomes. We should instead examine whether the opinion of political parties is of the same low quality.107 The potential incoherence of mass opinion is a harder problem. A  huge amount has been written on the subject, so I  will limit myself to a sketch of some prominent positions. William Riker and his followers have insisted that real-​world social preferences contain cycles, where individuals have transitive 106 Healy 2009. 107 However, the only literature I’m aware of suggests that elected officials do not have more political knowledge than most people and, in some cases, have less. But the survey does not distinguish between local and national elected officials, nor does it look at the role of these elected officials in the formation of group statements like party platforms (Intercollegiate Studies Institute 2008). I thank Bryan Caplan for the reference.

268  Trust in a Polarized Age preferences (if a > b and b > c, then a > c) but society has intransitive preferences (a > b and b > c, but c > a) over some list of candidates or issues.108 This means that group decision-​making requires violating some basic moral or rational conditions on collective choice. But their attempts to demonstrate the existence of cycles in the real world have been forcefully challenged on empirical grounds. Gerry Mackie has done more than anyone to muddy the Rikerian waters, arguably demonstrating that Riker’s examples of important cycles, such as the US election of 1860, are controversial at best and distortions of the truth at worst.109 I think Mackie shows that we have yet to empirically vindicate the Rikerian claim that “on the very most important subjects, cycles may render social outcomes meaningless.”110 Mackie claims, in contrast, that “the cycles that are alleged to make democracy meaningless are rare,” such that “the question is not one of logical possibility but rather one of empirical probability.”111 Cycles don’t happen much. Through his intense scrutiny of common claims of cycles, Mackie has shown that, while Riker could in principle turn out to be correct about the pervasiveness of cycles, there is no decisive, or even weighty, evidence in Riker’s favor. I therefore set the issue of the meaningfulness of mass public opinion aside as a challenge to my approach.112 The more serious empirical concerns about the value of consulting mass opinion are that the public is systematically biased.113 The case for systematic voter bias is based in part on the already-​mentioned extensive empirical data demonstrating vast public ignorance on nearly all matters of political importance. There are a number of theories that purport to explain the apparent problematic bias of voters by identifying what voters think they are doing in the ballot box. One influential theory, advanced by Geoffrey Brennan and Loren Lomasky, is that “voting in large-​scale elections is disconnected in a fundamental way from citizen preference over electoral outcomes” and that voters in fact vote expressively, that is, to express their affiliation with a value system or some group.114 Importantly, since citizens vote expressively, they are much more likely to vote based on ethical considerations, for “expressive behavior will reflect various kinds of ethical and ideological principles that [are] suppressed in the market setting”; this means that expressive political behavior “gives much freer range 108 Riker 1988. 109 Mackie 2003, pp. 258–​80. 110 Riker 1988, p. 128. 111 Mackie 2003, p. 17. 112 Mackie’s critique may be mistaken, but even Riker thinks that consulting mass opinion is important, since mass opinion should be meaningful when it comes to controlling exceptionally bad officials. This consideration should allow democratic theorists tempted by Riker’s arguments to nonetheless support choosing officials in a participatory fashion. 113 See Caplan 2007. 114 Brennan and Lomasky 1993, p. 1.

Elections and Process Democracy  269 to ethical considerations.”115 Yes, voters are ignorant (because they have little incentive to become informed), but an ignorant citizen might vote on factors that could nonetheless have moral and ideological import, such as whether a candidate’s expressed policy preferences and political principles match the voter’s own values. Such voting may not be problematically biased at all, since much documented political bias concerns certain nonnormative beliefs about how policies function; moral beliefs are typically, though not always, another matter. And, again, if parties set the agenda and determine some of the preferences of voters, this provides a different kind of check on the influence of uninformed mass opinion. We should also ask if there are good ways to use deliberation to improve mass opinion. To answer the question, deliberation must be understood in real-​world terms. We are not concerned with highly formalized models of deliberation but, rather, with data about how real-​world deliberation functions in real democracies. This means that we need to look at evidence that real-​world publics that deliberate in certain ways choose systematically better policies (where “better” is understood as more justified or justifiable) than nondeliberating publics. Unfortunately, the only systematic data that we have on improving opinion through deliberation are from minipublics. Fishkin himself has gathered much of this evidence, which has led him to a moderately favorable view of minipublics, but even he acknowledges that these deliberative bodies may be a poor proxy for public opinion in some cases.116 That means we have few data on how to improve the quality of mass opinion. And the prospects for gathering such data are dim because carrying out controlled experiments over the whole public is practically impossible. To compensate for this lack of data, we might appeal to a priori models of the potential benefits of deliberation, such as the model of deliberation found in Landemore’s work that I discussed in c­ hapter 7.117 I worry that Landemore’s model is also too far removed from contemporary circumstances to be useful in determining whether democratic procedures for official selection can be justified. However, as I argued in ­chapter 7, I think her model is helpful when trying to determine how policy formation should work. In particular, we can see that an adequately diverse public may be able to arrive at high-​quality decisions. That said, there is reason to worry whether the benefits of diversity will survive when a small number of people are randomly selected from a large, diverse group. Another option is to try to set out and impose ethical principles to govern public deliberation, like Rawls’s duty of civility.118 These principles often require

115

Ibid, p. 16. See Fishkin 2009, pp. 194–​6. 117 Landemore 2013. On her numbers-​trump-​ability theorem, see pp. 103–​4. 118 Rawls 2005, p. 217. 116

270  Trust in a Polarized Age that citizens discuss important political issues in terms of shareable or accessible reasons that are at least candidates for surviving public scrutiny. If the public can adequately internalize these principles and compliance will have the effect their advocates hope, this approach may have promise. However, exercising this sort of restraint—​especially when this means not advocating one’s real views or expressing the real values behind them—​can attach considerable costs to political participation, which may discourage many citizens from contributing to political life.119 Many citizens implicitly understand that their causal impact on the political process is miniscule and that makes sincere and informed voting difficult. Placing burdens on political participation could therefore have significant costs. In my view, democratic theorists should instead focus on imposing an ethic on political officials, who have a more substantial causal impact on the political process and who can be held accountable more easily.120 A third concern about the selection of political officials is that determining which voting rules are best for reporting the preferences of citizens is too difficult. As Riker notes, “Social choice depends not simply on the wills of individuals, but also on the method used to summarize these wills.”121 We also know, from social choice theory, that any voting rule that is rational and fair has flaws and is potentially subject to manipulation. Here are two ways concerns about voting rules might be addressed. First, we might combine voting rules so that one rule can compensate for the weaknesses of another. For instance, we could conduct ranked-​choice voting and compute the results using both the Borda count and the Condorcet rule. In this combined system, each candidate would receive a sum total based on his or her ranking, in accordance with the Borda count, and each candidate would also be ranked by who wins head-​to-​head votes, in accordance with the Condorcet rule. If one candidate won in both voting schemes, she would win the election. This is a normatively attractive result, since each voting rule compensates for the decision-​ theoretic flaws in the other. Borda allows for violations of independence of irrelevant alternatives, but Condorcet does not, whereas Condorcet allows for violations of transitivity, but Borda does not.122 When the Borda and Condorcet winners differ, we could have an instant-​run-​off rule between the Borda winner and the Condorcet winner.

119 As I argued in Vallier 2014a, chap. 2. 120 Ibid., chap. 6. 121 Riker 1988, p. 31. 122 Following Gaus 2007, the basic idea behind independence of irrelevant alternatives is that a person or a society’s ranking of preferences x and y should not be affected by their ranking of z. Preferences are transitive when, if a is preferred to b and b is preferred to c, then a is preferred to c. See p. 155.

Elections and Process Democracy  271 Second, we might insist on a system of proportional representation, found in some parliaments, where citizens vote for a party that they like, and then seats of parliament are divided up according to the parties’ shares of the vote. This might be an effective alternative to the system of first-​past-​the-​post voting rules, as it would allow for the formation of robust third parties, increasing the diversity of members of the public who see their views represented in the legislature, and lending itself to a more coalition-​based model of government, which could compromise on issues that under first-​past-​the-​post lead to acrimony.123 Moreover, if citizens vote for parties rather than individual candidates, this may lead people to focus more on issues: a system that focuses on parties simplifies choices for citizens by focusing on basic aims more than personalities. And parties can negotiate with one another better if they can rely on their collective capacities, rather than the capacities of individual officials.124 Compound social choice rules and voting for parties could comprehensively represent the will of the public, because they would enable citizens to express complex preferences about the relative worth of candidates, parties, and policies. By providing outlets for expressing third, fourth, and fifth choices, elections can produce governments that reflect diverse perspectives.

7.  Process Democracy against Deliberative, Aggregative, and Agonistic Approaches Process democracy, the approach to democratic theory in this chapter, has two parts:  a theory of institutions and a theory of legitimacy. Regarding institutions, process democracy stresses participation and political equality in selecting officials and deliberation and political equality in selecting policies. Thus, it recommends familiar democratic electoral processes in choosing political officials, but it stresses newer forms of democratic organization, such as minipublics and prediction markets, in selecting public policies. Regarding legitimacy, process democracy adopts my account of public justification. Based on that approach, process democracy protects the primary rights of agency, association, and jurisdiction. It also embraces primary electoral rights, both as a means of protecting other primary rights and because members of the public will regard electoral rights as procedurally fair and instrumentally valuable in generating publicly justified public policies.

123 For a helpful discussion of public reason and proportional representation, see Bonotti 2017, p. 143. 124 Christiano 1996, pp. 222–​3.

272  Trust in a Polarized Age We can illustrate the uniqueness of process democracy by contrasting it with the three main strands of democratic theory of legitimacy in the literature: deliberative democracy, aggregative democracy, and agonistic democracy. Deliberative democracy resembles mainstream public reason views, holding that democratic decisions are legitimate when they are the outcome of certain structured, egalitarian forms of deliberation.125 Aggregative democracy ascribes legitimacy to democratic decisions when the judgments of citizens are aggregated in accordance with rationally and morally attractive principles, like the Pareto principle.126 Agonistic democracy rejects aggregation and is not optimistic about the prospects for consensus following deliberation.127 Democratic decisions are, it therefore holds, supposed to balance the differing, and invariably contradictory, interests of different social groups vying for power.

7.1.  Contrasted with Deliberative Democracy Process democracy differs from deliberative democracy by limiting reliance on deliberation among citizens and not imposing an ideal or principle of restraint that is meant to structure the deliberation of citizens. It also rejects consensus as an ideal outcome of the political process. Instead, it requires that democratic institutions track what is publicly justified for citizens. But process democracy does appeal to deliberation in the political process, especially with respect to policy governing the production of public goods. Deliberation is important for increasing the quality of political decisions and, insofar as deliberation helps to realize public justification, it has great instrumental value. Yet public deliberation is simply one process among many that may help to create and maintain a publicly justified polity. It should not be the central focus

125 Classic works of deliberative democracy are many, but some of the main monographs and articles can be found in the following works: Habermas 1998; Bohman and Rehg 1998; Gutmann and Thompson 2004; Freeman 2000; Bowman and Richardson 2009; Parkinson and Mansbridge 2012. 126 Rather, aggregative democrats appear to ascribe legitimacy on this basis, though in reality, they tend to say little if anything about their normative assumptions. In fact, many people believed to be aggregative democrats are critics of more ambitious democratic theories on the grounds that aggregations are subject to problems. Theorists often associated with aggregative democracy include Schumpeter 1942; Downs 1957; Riker 1988; and Dahl 1989. Riker (1988, pp. 116–​7) articulates some principles of “fairness” that include the Pareto principle and the universal admissibility of different preference orderings. I thank David Estlund, Alex Guerrero, Enzo Rossi, and Joshua Miller for discussion about how exactly one comes to be called an aggregative democrat. 127 Ernesto Laclau, Chantal Mouffe, and William Connolly are among the best-​known contemporary agonists (Laclau and Mouffe 2014; Connolly 2002; and Honig 1993). For a recent articulation of agonistic democracy, see Wenman 2013. For discussion of the distinctiveness of agonistic democratic theory and the ways in which it is distinct from deliberative democracy, see Gürsözlü 2009. I thank Alan Reynolds for helping me to understand contemporary agonistic democracy.

Elections and Process Democracy  273 of a democratic political order.128 Further, deliberation is subject to a variety of flaws even under relatively controlled conditions, so we must also focus on other aspects of constitutional rules to determine which laws we should have. In particular, a publicly justified polity will often appeal to political bargaining and adjudication as methods of establishing law and policy. Constitutional rules allowing, but governing, bargaining and adjudication may be publicly justified as a result.

7.2.  Contrasted with Aggregative Democracy Citizen input into process-​democratic institutions roughly corresponds to the approach of aggregative democrats. Both kinds of democracy take raw mass opinion and then use rationally and morally attractive voting rules to choose candidates that are in some sense chosen by the people. Deliberative democrats have often complained that aggregative democrats take citizen judgments as given, rather than as something shaped by the political process.129 But this criticism does not damage process democracy. Preferences are formed by an open process within the moral, legal, and political orders. Preferences therefore will evolve and change. Further, deliberation is critical for process democracy, as deliberation is a central governing value of the productive function of the state. Citizens’ input should affect the policy process through their deliberation in minipublics. I indeed use an aggregative approach in converting mass opinion into the selection of candidates and officials, but process democracy does not rely upon aggregation alone.

7.3.  Contrasted with Agonistic Democracy Agonistic democracy, in Chantal Mouffe’s terms, views the political domain as having a critical and inevitable “dimension of antagonism.”130 On the agonistic view, political foes are understood not necessarily as enemies but as “adversaries,” which are defined as “persons who are friends because they share a common symbolic space but also enemies because they want to organize this common symbolic space in a different way.”131 On this view, “Power is constitutive of social 128 I focus on drawing out the deliberation-​justification distinction in Vallier 2016. For a discussion of exit mechanisms as well, see Vallier 2018a. 129 For an early version of the criticism, see Cohen 1991, though I think this misfires as a criticism of Dahl, who is not obviously an aggregative democrat. For an overview of the deliberative democratic narrative, see Bohman and Rehg 1998, pp. ix–​xiii. 130 Mouffe 2009, p. 11. 131 Ibid., p. 13.

274  Trust in a Polarized Age relations” because the values at the heart of liberal democracy are in ineradicable conflict; this means politics is invariably built on exclusion.132 In one way, my entire approach to liberal democratic theory is a repudiation of the philosophical foundations of agonism, given that agonists typically insist that democratic politics is invariably a power struggle. However, while I reject the agonistic approach to democracy, agonistic critiques of deliberative democracy have some force for the process democrat. Agonists emphasize that political groups are not going to come to consensus on a regular basis, as deliberative democrats often hope. I agree. Agonistic democrats are also correct to emphasize that politics invariably involves disagreement and contestation and that our votes will reflect those divisions. And process democracy agrees that politics involves contestation between diverse groups based on irreconcilable evaluative differences. However, unlike the agonistic democrat, the process democrat denies that we must conceive of politics as a power struggle. If we pursue public justification, process democracy holds, we can limit political war. Given that I allow different democratic principles to govern different stages of the political process, and given that my view possesses the strengths of the three dominant democratic theories—​deliberative, aggregative, and agonistic—​I think the process-​democratic approach has much to commend it. Despite the attractions of process democracy, other constitutional structures, in particular exit mechanisms like federalism, are required to ensure that laws and constitutional rules are publicly justified. Sustaining a system of social trust might require a political and moral order that distributes power across different levels of government. However, while I think I can demonstrate that federal organization is publicly justified, there are not enough data demonstrating even a correlation between federalism and social or political trust to make the case that federal arrangements help to establish moral peace between diverse persons in the real world. For this reason, I only focus on the voice mechanism of democratic elections.133

8.  Democracy Creates Trust for the Right Reasons This chapter has argued that the exercise of electoral rights creates trust for the right reasons. First, I argued that democratic elections can promote real political trust and, perhaps, social trust, albeit obliquely. Second, I argued that electoral

132 Ibid., p. 98. Also see Laclau and Mouffe’s claim that “social objectivity is constituted through acts of power” and that “every order is ultimately political and based on some form of exclusion” (2014, pp. 99–​100). 133 I discuss how to publicly justify exit mechanisms in Vallier 2018a.

Elections and Process Democracy  275 rights are publicly justified, creating the right reasons for trust. Electoral rights are publicly justified because they are procedures that treat persons as equals and because they promote publicly justified outcomes by making government responsive to the interests and perspectives of all citizens. The resulting process-​ democratic approach to justifying democracy combines the virtues of three different approaches to democratic theory and strengthens the book’s thesis.

Epilogue How Is Trust Restored?

Liberal institutions, and only liberal institutions, can establish trust for the right reasons among diverse persons. A liberal society, therefore, can solve the problem of political war despite the challenges posed by diverse perspectives. It can do so through five liberal rights practices: freedom of association, markets and private property rights, social welfare programs, democratic constitutionalism, and electoral democracy. Democratic constitutionalism seems to create the most social and political trust, with markets in second, social insurance in third, freedom of association in fourth, and electoral democracy in last place. But they all sustain trust, and they do so for the right reasons. And since liberal institutions sustain trust for the right reasons, we can conclude, after much argument, that liberal politics is not war. We may also draw two further lessons. First, liberal politics has the internal resources to interrupt the divergence-​distrust feedback loop. By increasing social and political trust, liberal rights practices, without going outside of the liberal policy toolkit, dampen partisan divergence and—​with the right liberal policies—​ may even be able to reverse it. Second, we can more clearly see the deficiencies of alternative hypotheses. Recall that direct theories of divergence identify a range of causes that contribute to partisan divergence directly, such as economic inequality. Group theories focus on features of our tribal mentality and behavior in increasing divergence. And inherent-​corruption theories tell us that a loss of social cohesion and an increase in conflict is natural to liberal political order. The arguments of this book provide good reason to reject inherent-​corruption theories, since liberal order has the internal resources to stymie distrust. The distrust-​divergence connection helps to explain why some phenomena are correlated with divergence: for instance, we can expect falling social trust to increase divergence and economic inequality at the same time. Finally, the connection between trust and divergence helps to explain why our tribal mentalities are activated in a way that leads people to identify with tribes within their society, rather than seeing their society as a whole as their tribe. If national trust falls, people will tend to be suspicious of other groups within that nation.

278 Epilogue I suspect most readers who follow American politics will remain skeptical of my arguments. After all, we are a liberal society, and yet we are experiencing increasing partisan divergence and lower trust. My response is that we can address the problem by adopting laws and policies that strengthen the liberal rights practices we already have. If we can limit corruption, create more even-​handed and professional judicial systems and police forces, and stably protect everyone’s legal property rights, we may regain some measure of social trust. We could do even better by compressing economic inequality through coercion-​reducing policies. And if we increase quality of governance and procedural fairness through minipublics and prediction markets, limit rent-​seeking, and increase broad-​based economic growth, we should regain a measure of political trust. We should also consider electoral reforms aimed at either reducing divergence or controlling its deleterious effects.1 And since social and political trust seem to cause one another, measures to improve each may bring us greater increases in both. The trust-​promoting policies I have explored are not politically infeasible. Far from it. The hope then is increasing trust will take a large bite out of affective polarization and sorting, since people will become more trusting of their others and of their government when the other party is in power. Issue-​based polarization and sorting becomes much less problematic under conditions of trust, since people will see those who offer different policies as persons of goodwill, even if they disagree with their political agenda. The democratic rotation of power can continue with much less friction. The challenge is how to get the process of reform in motion. That requires something of each of us. First, we must fight our own distrust of our political opponents. Otherwise, we are likely to interpret everything they do as evil and wicked and take that as further evidence that we were right about them all along.2 To begin to trust, we must reflect on our own intellectual limitations and adopt an attitude of humility. None of us have all the answers, and every one of us is likely to be wrong on some important political issues. The other side just might help us get at the truth, at least from time to time. Second, we must allow ourselves to recognize that our nonpolitical interactions with our political opponents are positive and give us reason to trust them. Perhaps you are hermetically sealed in your ideological bubble, but if not, you probably regularly interact with some people whose politics differs from your own. You may not even have noticed that their politics are different. And the ways they have treated you have given you no grounds to mistrust them. Third, we should try to take a global perspective on 1 For helpful reflection about which electoral reforms lead to less divergence, see McCarty 2019, pp. 125–​31. 2 Karen Jones (2019) calls this looping affect and distrust phenomenon “affective looping.”

Epilogue  279 our problems. Few developed liberal democratic societies face our levels of divergence and distrust, at least to the same degree. Our condition is not terminal. But here’s what won’t work:  populism, right-​wing or left-​wing. Right-​wing populism proposes a kind of American nationalism that has served as a basis for social unity in the past, typically during wars. You can see this just by tracking the spikes in social and political trust over time. Nationalism indeed creates social and political trust by defining the nation against some enemy or threat. Wars bind us together, and those bonds help our societies survive. But we don’t want to base trust on constant warfare. And in times of peace, nationalist leaders tend to search for a new foe, typically an enemy within the nation, people who do not fit the interpretation of national identity held by those with political power. These efforts may increase trust among persons who fit the nationalist’s definition of the nation, but they will decrease trust between the nationalist’s in-​group and his out-​group. This is not the way to create trust among all Americans. When a US president leads a public chant to send a female, Muslim congresswoman of color back to her country of origin, it may create fellow feeling among the president’s supporters, but it frustrates trust among people with different genders, ethnicities, and faiths. This much is probably uncontroversial for the average reader of a philosophical monograph. But I  think left-​wing populism also carries dangers. It may not risk racism, sexism, or the oppression of immigrants, but it engages in its own form of tribalism by demonizing not merely the rich, but more privileged members of society. And in some cases, left-​wing populists denounce people who aren’t especially privileged, like white working-​class males without college degrees, or very religious citizens. Left-​wing populism also bases social unity on an extremely large state with great power to interfere with people’s economic lives. That approach to economic policy would likely reduce economic growth and assign the distribution of economic gains to the discretion of well-​connected bureaucrats and professional rent-​seekers. And advocating it alienates right-​ wing citizens who believe such policies are disastrous and unjust. Left-​wing populism, therefore, is likely to hurt social and political trust too. In Why Liberalism Failed, Patrick Deneen argues that “to call for the cures of liberalism’s ills by applying more liberal measures is tantamount to throwing gas on a raging fire. It will only deepen our political, social, economic, and moral crisis.”3 But if I am right, we really do need more liberalism. If we protect central liberal rights and support those who exercise them, we can overcome the mistrust and conflict that some claim liberalism engenders by nature.



3

Deneen 2018, p. 4.

280 Epilogue I am alive to the concerns that nonliberals raise that liberalism as an ideology has deleterious social effects. A liberal ethos can become sectarian and authoritarian in beating up on those with more traditional social and political views, and it can often be socially atomizing, destroying older social bonds that are hard to replace. But I have not defended liberalism as an ideology or ethos in this book; indeed, I reject that ethos. I have instead defended liberal rights practices, and in my view liberalism’s contemporary critics have not shown that we can only have liberal institutions if we have a liberal ethos. We should be proud of liberal civilization, and we have good reason to hope that liberal orders have the internal resources to end the cold civil war we are currently waging against one another. May we soon return to the irritating, exhausting, and deeply flawed politics of liberal democracy. Through liberal reforms, we can, paraphrasing Sigmund Freud, turn our hysterical, polarized misery into common political unhappiness.4



4

Freud and Breuer 2004, p. 306.

Works Cited Aas, Sean. 2019. You Didn’t Build That: Equality and Productivity in a Complex Society. Philosophy and Phenomenological Research 98, 69–​88. Abrams, B. A., and M. A. Schmidtz. 1984. The Crowding Out Effect of Government Charitable Contributions: Cross Sectional Evidence. National Tax Journal 37, 563–​68. Acemoglu, Daron, Suresh Naidu, Pascual Restrepo, and James Robinson. 2019. Democracy Does Cause Growth. Journal of Political Economy 127, 47–​100. Acemoglu, Daron, and James Robinson. 2013. Why Nations Fail: The Origins of Power, Prosperity, and Poverty. Random House. Achen, Christopher, and Larry Bartels. 2016. Democracy for Realists: Why Elections Do Not Produce Responsive Government. Princeton University Press. Ahn, T. K., Elinor Ostrom, David Schmidt, and James Walker. 2003. Trust in Two-​Person Games:  Game Structures and Linkages. In Trust and Reciprocity:  Interdisciplinary Lessons for Experimental Research, edited by Elinor Ostrom and James Walker, 323–​51. Russell Sage. Alesina, Alberto, and Dani Rodrik. 1994. Distributive Politics and Economic Growth. Quarterly Journal of Economics 109, 465–​90. Algan, Yann, and Pierre Cahuc. 2010. Inherited Trust and Growth. American Economic Review 100, 2060–​92. Allport, Gordon. 1954. The Nature of Prejudice. 25th Anniversary ed. Basic Books. Anderson, Christopher, and Yuliya Tverdova. 2003. Corruption, Political Allegiances, and Attitudes toward Government in Contemporary Democracies. American Journal of Political Science 47, 91–​109. Anderson, Elizabeth. 1999. What Is the Point of Equality? Ethics 109, 287–​337. Anderson, Elizabeth. 2010. The Imperative of Integration. Princeton University Press. Anderson, Elizabeth. 2017. Private Government: How Employers Rule Our Lives (and Why We Don’t Talk about It). Princeton University Press. Andeweg, Rudy, and David Farell. 2017. Legitimacy Decline and Party Decline. In Myth and Reality of the Legitimacy Crisis: Explaining Trends and Cross-​National Differences in Established Democracies, edited by Carolien van Ham, Jacques Thomassen, Kees Aarts and Rudy Andeweg, 76–​94. Oxford University Press. Anomaly, Jonathan. 2015. Public Goods and Government Action. Politics, Philosophy and Economics 14, 109–​28. Anomaly, Jonathan. 2017. Trust, Trade, and Moral Progress:  How Market Exchange Promotes Trustworthiness. Social Philosophy and Policy 34, 89–​107. Ansolabehere, Stephen, John Figueiredo, and James Snyder. 2003. Why Is There So Little Money in U.S. Politics? Journal of Economic Perspectives 17, 105–​30. Arneson, Richard. 2013. Egalitarianism. In Stanford Encyclopedia of Philosophy, edited by Edward N.  Zalta. Summer 2013 ed. http://​plato.stanford.edu/​entries/​ egalitarianism.

282  Works Cited Auten, Gerald, and David Splinter. 2017. Income Inequality in the United States: Using Tax Data to Measure Long-​Term Trends. November 12. http://​davidsplinter.com/​ AutenSplinter-​Tax_​Data_​and_​Inequality.pdf. Bächtiger, André, and Simon Beste. 2017. Deliberative Citizens, (Non)Deliberative Politicians: A Rejoinder. Daedalus 146, 106–​18. Baier, Kurt. 1958. The Moral Point of View:  A Rational Basis of Ethics. Cornell University Press. Banerjee, Abhijit, Rema Hanna, Gabriel Kreindler, and Benjamin Olken. 2017. Debunking the Stereotype of the Lazy Welfare Recipient: Evidence from Cash Transfer Programs. World Bank Research Observer 32, 155–​84. Bargsted, Matías, Nicolás Somma, and Juan Carlos Castillo. 2017. Political Trust in Latin America. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 395–​417. Edward Elgar. Barrett, Jacob. 2020a. Review of Must Politics Be War? Journal of Moral Philosophy. Forthcoming. Barrett, Jacob. 2020b. Social Reform in a Complex World. Journal of Ethics and Social Philosophy 17, 103–​32. Barrett, Jacob, and Gerald Gaus. 2020. Laws, Norms, and Public Justification: The Limits of Law as an Instrument of Reform. In Public Reason and Courts, edited by S.  A. Langvatn, M. Kumm, and W. Sadurski, 201–​28. Cambridge University Press. Bartels, Larry. 2010. Unequal Democracy: The Political Economy of the New Gilded Age. Princeton University Press. Bauer, Paul, and Markus Freitag. 2018. Measuring Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 15–​36. Oxford University Press. Baumgartner, Frank. 2010. Interest Groups and Agendas. In The Oxford Handbook of American Political Parties and Interest Groups, edited by Sandy Maisel, Jeffrey Berry, and George Edwards, 519–​33. Oxford University Press. Beck, John. 1994. An Experimental Test of Preferences for the Distribution of Income and Individual Risk Aversion. Eastern Economic Journal 20, 131–​45. Beito, David. 2000. From Mutual Aid to the Welfare State: Fraternal Societies and Social Service, 1890–​1967. University of North Carolina Press. Bélanger, Éric. 2017. Political Trust and Voting Behavior. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 242–​55. Edward Elgar. Benkler, Yochai, Robert Faris, and Hal Roberts. 2018. Network Propaganda: Manipulation, Disinformation, and Radicalization in American Politics. Oxford University Press. Berggren, Niclas, and Christian Bjørnskov. 2017. The Market-​Promoting and Market-​ Preserving Role of Social Trust in Reforms of Policies and Institutions. Southern Economic Journal 84, 3–​25. Berggren, Niclas, and Henrik Jordahl. 2006. Free to Trust: Economic Freedom and Social Capital. Kyklos 59, 141–​69. Berggren, Niclas, and Therese Nilsson. 2013. Does Economic Freedom Foster Tolerance? Kyklos 66, 177–​207. Bergh, Andreas. 2016. Sweden and the Revival of the Capitalist Welfare State. Edward Elgar. Bergh, Andreas, and Christian Bjørnskov. 2014. Trust, Welfare States, and Income Equality: Sorting Out the Causality. European Journal of Political Economy 35, 183–​99. Bergh, Andreas, Christian Bjørnskov, and Kevin Vallier. 2021. Social and Legal Trust: The Case of Africa. In Social Trust, edited by Kevin Vallier and Michael Weber, forthcoming. Routledge.

Works Cited  283 Bergh, Andreas, Henrik Jordahl, and Richard Öhrvall. 2019. Does Welfare State Universality Promote Social Trust? July 7. http://​www.oru.se/​globalassets/​oru-​sv/​ institutioner/​hh/​seminarieserien-​nek/​seminarieserie-​nek-​-​-​henrik-​jordahl-​190509. pdf. Bergh, Andreas, and Richard Öhrvall. 2017. How Persistent Is High Social Trust When Institutional Quality Is Low? Unpublished. Berning, Carl, and Conrad Ziller. 2017. Social Trust and Radical Right-​Wing Populist Party Preferences. Acta Politica 52, 198–​217. Bertrand, Marriane, and Esther Duflo. 2017. Field Experiments on Discrimination. In Handbook of Economic Field Experiments, vol. 2, edited by Abhijit Banerjee and Esther Duflo, 309–​93. North Holland. Bicchieri, Cristina. 2006. The Grammar of Society. Cambridge University Press. Bicchieri, Cristina, Erte Xiao, and Ryan Muldoon. 2011. Trustworthiness Is a Social Norm, but Trusting Is Not. Politics, Philosophy, and Economics 10, 170–​87. Bjørnskov, Christian. 2006. Determinants of Generalized Trust:  A Cross-​ Country Comparison. Public Choice 130, 1–​21. Bjørnskov, Christian. 2009. Social Trust and the Growth of Schooling. Economics of Education Review 28, 249–​57. Bjørnskov, Christian. 2010. How Does Social Trust Lead to Better Governance? An Attempt to Separate Electoral and Bureaucratic Mechanisms. Public Choice 144, 323–​46. Bjørnskov, Christian. 2011. Combating Corruption:  On the Interplay between Institutional Quality and Social Trust. Journal of Law and Economics 54, 135–​59. Bjørnskov, Christian. 2018. Social Trust and Economic Growth. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 535–​55. Oxford University Press. Bjørnskov, Christian. 2019. The Political Economy of Trust. In The Oxford Handbook of Public Choice, vol. 1, edited by Roger Congleton, Bernard Grofman, and Stefan Voigt, 628–​48. Oxford University Press. Bjørnskov, Christian. 2020. Social Trust and Patterns of Growth. Aarhus University. Unpublished. Blevins, Sue. 1995. The Medical Monopoly:  Protecting Consumers or Limiting Competition? Policy Analysis 246, 1–​36. Boberg-​Fazlić, Nina, and Paul Sharp. 2013. Does Welfare Spending Crowd Out Charitable Activity? Evidence from Historical England under the Poor Laws. EHES Working Paper 49. Bohman, James, and William Rehg, eds. 1998. Deliberative Democracy: Essays on Reason and Politics. MIT Press. Bohnet, Iris, Bruno Frey, and Steffen Huck. 2001. More Order with Less Law: On Contract Enforcement, Trust, and Crowding. American Political Science Review 95, 131–​44. Bolton, Gary, and Katok Elena. 1998. An Experimental Text of the Crowding Out Hypothesis:  The Nature of Beneficent Behavior. Journal of Economic Behavior and Organization 37, 315–​31. Bonotti, Matteo. 2017. Partisanship and Political Liberalism in Diverse Societies. Oxford University Press. Bou-​Habib, Paul. 2015. Gaus on Coercion and Welfare-​State Capitalism:  A Critique. Political Studies, 64, 651–​65. Bowman, James, and Henry Richardson. 2009. Deliberative Democracy, Liberalism, and Reasons That All Can Accept. Journal of Political Philosophy 17, 1–​22.

284  Works Cited Boyd, Robert, and Peter Richerson. 2005. The Origin and Evolution of Cultures. Oxford University Press. Branham, J. Alexander, Stuart Soroka, and Christopher Wlezien. 2017. When Do the Rich Win? Political Science Quarterly 132, 43–​62. Brennan, Geoffrey, and Loren Lomasky. 1993. Democracy and Decision: The Pure Theory of Electoral Preference. Cambridge University Press. Brennan, Jason. 2011. The Ethics of Voting. Princeton University Press. Brennan, Jason. 2013. Epistocracy and Public Reason. In Democracy in the Twenty-​First Century, edited by Ann Cudd and Sally Scholz, 191–​204. Springer. Brennan, Jason. 2016a. Against Democracy. Princeton University Press. Brennan, Jason. 2016b. Democracy and Freedom. In The Oxford Handbook of Freedom, edited by David Schmidtz and Carmen Pavel, 335–​49. Oxford University Press. Brennan, Jason, and Peter Jaworski. 2016. Markets without Limits:  Moral Virtues and Commercial Interests. Routledge. Bryan, Benjamin. 2016. Natural Rights and Convention. PhD diss., Bowling Green State University. OhioLINK Electronic Theses and Dissertations Center. Buchanan, James. 1999. Taxation in Fiscal Exchange. In The Collected Works of James M. Buchanan, vol. 1: The Logical Foundations of Constitutional Liberty, 133–​49. Liberty Fund. Buchanan, James, and Gordon Tullock. 1962. The Calculus of Consent. University of Michigan Press. Caplan, Bryan. 2007. The Myth of the Rational Voter:  Why Democracies Choose Bad Policies. Princeton University Press. Cassar, Alessandra, Giovanna d’Adda, and Pauline Grosjean. 2014. Institutional Quality, Culture, and Norms of Cooperation:  Evidence from Behavioral Field Experiments. Journal of Law and Economics 57, 821–​63. Cawvey, Matthew, Matthew Hayes, Damarys Canache, and Jeffrey Mondak. 2018. Biological and Psychological Influences on Interpersonal and Political Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 119–​47. Oxford University Press. Chambers, Clare. 2008. Sex, Culture, and Justice: The Limits of Choice. Pennsylvania State University Press. Chambers, Simone. 2004. Behind Closed Doors: Publicity, Secrecy, and the Quality of Deliberation. Journal of Political Philosophy 12, 389–​410. Chambers, Simone, and Jeffrey Kopstein. 2001. Bad Civil Society. Political Theory 29, 837–​65. Chambers, Simone, and Jeffrey Kopstein. 2006. Civil Society and the State. In The Oxford Handbook of Political Theory, edited by John Dryzek, Bonnie Honig, and Anne Phillips, 363–​81. Oxford University Press. Christiano, Thomas. 1996. The Rule of the Many:  Fundamental Issues in Democratic Theory. Westview Press. Christiano, Thomas. 2010. The Constitution of Equality:  Democratic Authority and Its Limits. Oxford University Press. Christiano, Thomas. 2011. An Instrumental Argument for a Human Right to Democracy. Philosophy and Public Affairs 39, 142–​76. Christiano, Thomas. 2017. Review of Jason Brennan’s Against Democracy. Notre Dame Philosophical Reviews, July 9.  http://​ndpr.nd.edu/​news/​must-​politics-​be-​war-​ restoring-​our-​trust-​in-​the-​open-​society/​.

Works Cited  285 Coffey, Bentley, Patrick McLaughlin, and Pietro Peretto. 2016. The Cumulative Cost of Regulations. Mercatus Working Paper. Cohen, Joshua. 1991. Review of Robert Dahl’s Democracy and Its Critics. Journal of Politics 53,  221–​5. Cohen, Marc. 2015. Alternative Conceptions of Generalized Trust (and the Foundations of the Social Order). Journal of Social Philosophy 46, 463–​78. Cohn, Alain, Michel André Maréchal, David Tannenbaum, and Christian Lukas Zünd. 2019. Civic Honesty around the Globe. Science 365, 70–​73. Cole, G. D. H. 1920. Social Theory. Frederick A. Stokes. Connolly, William. 2002. Identity/​Difference: Democratic Negotiations of Political Paradox, Expanded ed. University of Minnesota Press. Cooper, Christopher, H. Gibbs Knotts, and Kathleen Brennan. 2008. The Importance of Trust in Government for Public Administration:  The Case of Zoning. Public Administration Reform 68, 459–​68. Cordelli, Chiara. 2017. Democratizing Organized Religion. Journal of Politics 79, 576–​90. Cowen, Tyler. 2018. Stubborn Attachments: A Vision for a Society of Free, Prosperous, and Responsible Individuals. Stripe Press. Cox, J. C. 2004. How to Identify Trust and Reciprocity. Games and Economic Behavior 46, 260–​81. Dahl, Robert. 1989. Democracy and Its Critics. Yale University Press. Dalton, Russell. 2017. Political Trust in North America. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 375–​94. Edward Elgar. Dalton, Russell, and Christian Welzel. 2014. The Civic Culture Transformed:  From Allegiant to Assertive Citizens. Cambridge University Press. Dasgupta, Partha, and Ismail Serageldin. 2000. Social Capital: A Multifaceted Perspective. World Bank. Davidai, Shai, and Martino Ongis. 2019. The Politics of Zero-​Sum Thinking:  The Relationship between Political Ideology and the Belief That Life Is a Zero-​Sum Game. Science Advances 5, 1–​10. Delhey, Jan, Kenneth Newton, and Christian Welzel. 2011. How General Is Trust in “Most People”? Solving the Radius of Trust Problem. American Sociological Review 76, 786–​807. Deneen, Patrick. 2018. Why Liberalism Failed. Yale University Press. Dinesen, Peter Thisted, and Kim Mannemar Sønderskov. 2018. Ethnic Diversity and Social Trust:  A Critical Review of the Literature and Suggestions for a Research Agenda. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 175–​204. Oxford University Press. Dinesen, Peter Thisted, and Kim Mannemar Sønderskov. 2018b. Cultural Persistence or Experiential Adaptation? A Review of Studies Using Immigrants to Examine the Roots of Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 205–​29. Oxford University Press. Dow, Gregory. 2003. Governing the Firm:  Workers’ Control in Theory and Practice. Cambridge University Press. Downs, Anthony. 1957. An Economic Theory of Democracy. Harper. Dryzek, John. 2010. Foundations and Frontiers of Deliberative Governance. Oxford University Press.

286  Works Cited Eberhardt, Markus. 2019. Revisiting the Causal Effect of Democracy on Long-​ Run Development. Vox EU.  April 28. voxeu.org/​article/​revisiting-​causal-​effect-​ democracy-​long-​run-​development. Edmundson, William. 2017. John Rawls: Reticent Socialist. Cambridge University Press. Enns, Peter. 2015. Relative Policy Support and Coincidental Representation. Perspectives on Politics 13, 1053–​64. Enoch, David. 2015. Against Public Reason. In Oxford Studies in Political Philosophy, vol. 1, edited by David Sobel, Peter Vallentyne, and Steven Wall, 112–​44. Enos, Ryan. 2017. The Space between Us:  Social Geography and Politics. Cambridge University Press. Esaiasson, Peter, and Christopher Wlezien. 2017. Advances in the Study of Democratic Responsiveness: An Introduction. Comparative Political Studies 50, 699–​710. Estlund, David. 2008. Democratic Authority. Princeton University Press. Farrant, Andrew, and Edward McPhail. 2010. Does F.A. Hayek’s Road to Serfdom Deserve to Make a Comeback? Challenge 53, 96–​120. Faulkner, Paul. 2018. Finding Trust in Government. Journal of Social Philosophy 49, 626–​44. Fishkin, James. 2009. When the People Speak:  Deliberative Democracy and Public Consultation. Oxford University Press. Foa, Roberto Stefan, and Yascha Mounk. 2016. The Danger of Deconsolidation:  The Democratic Disconnect. Journal of Democracy 27, 5–​17. Fouirnaies, Alexander. 2018. When Are Agenda Setters Valuable? American Journal of Political Science 62, 176–​91. Fouirnaies, Alexander, and Andrew Hall. 2018. How Do Interest Groups Seek Access to Committees? American Journal of Political Science 62, 132–​47. Fox, Justin, and Kenneth Shotts. 2009. Delegates or Trustees? A  Theory of Political Accountability. Journal of Politics 71, 1225–​37. Frankfurt, Harry. 2015. On Inequality. Princeton University Press. Fraser, Nancy. 1994. After the Family Wage: Gender Equity and the Welfare State. Political Theory 22, 591–​618. Freeman, Samuel. 2000. Deliberative Democracy: A Sympathetic Comment. Philosophy and Public Affairs 29, 371–​418. Freiman, Christopher. 2017. Unequivocal Justice. Routledge. Freud, Sigmund, and Josef Breuer. 2004. Studies in Hysteria. Penguin. Friedman, Benjamin. 2005. The Moral Consequences of Economic Growth. Knopf. Friedman, Milton. 2002. Capitalism and Freedom. University of Chicago Press. Frohlich, Norman, Joe Oppenheimer, and Cheryl Eavey. 1987. Laboratory Results on Rawls’s Distributive Justice. British Journal of Political Science 17, 1–​21. Fukuyama, Francis. 1995. Trust:  The Social Virtues and the Creation of Prosperity. Free Press. Fuller, Lon. 1969. Two Principles of Association. In Nomos XI: Voluntary Associations, edited by John Chapman and J. Roland Pennock, 3–​23. Atherton Press. Ganghof, Steffen. 2013. Does Public Reason Require Supermajoritarian Democracy? Liberty, Equality, and History in the Justification of Political Institutions. Politics, Philosophy and Economics 12, 179–​96. Gaus, Gerald. 2007. On Philosophy, Politics, and Economics. Wadsworth. Gaus, Gerald. 2009. The Idea and Ideal of Capitalism. In The Oxford Handbook of Business Ethics, edited by George Brenkert and Tom Beauchamp, 73–​99. Oxford University Press. Gaus, Gerald. 2011. The Order of Public Reason. Cambridge University Press.

Works Cited  287 Gaus, Gerald. 2012. Property. In The Oxford Handbook of Political Philosophy, edited by David Estlund, 93–​114. Oxford University Press. Gaus, Gerald. 2018. Self-​Organizing Moral Systems:  Beyond Social Contract Theory. Politics, Philosophy and Economics 17, 119–​47. Gaus, Gerald. 2019. What Might Democratic Self-​Governance in a Complex Social World Looks Like? San Diego Law Review 56, 967–​1012. Geurkink, Bram, Andrej Zaslove, Roderick Sluiter, and Kristof Jacobs. 2020. Populist Attitudes, Political Trust, and External Political Efficacy: Old Wine in New Bottles? Political Studies 68, 247–​67. Gilbert, Michael. 2017. Entrenchment, Incrementalism, and Constitutional Collapse. Virginia Law Review 103, 631–​72. Gilens, Martin. 2012. Affluence & Influence: Economic Inequality and Political Power in America. Princeton University Press. Gilens, Martin, and Benjamin Page. 2014. Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens. Perspectives on Politics 12, 564–​81. Glaeser, Edward, David Laibson, and Bruce Sacerdote. 2002. The Economic Approach to Social Capital. Economic Journal 112, 437–​58. Goodin, Robert, and David Schmidtz. 1998. Social Welfare and Individual Responsibility. Cambridge University Press. Gould, Eric, and Esteban Klor. 2019. Party Hacks and True Believers: The Effect of Party Affiliation on Political Preferences. Journal of Comparative Economics 47, 504–​24. Gramlich, John. 2019. “Young Americans Are Less Trusting of Other People—​and Key Institutions—​Than Their Elders,” Fact Tank, Pew Research Group, August 6. https://​ www.pewresearch.org/​fact-​tank/​2019/​08/​06/​young-​americans-​are-​less-​trusting-​of-​ other-​people-​and-​key-​institutions-​than-​their-​elders/​. Grimes, Marcia. 2017. Procedural Fairness and Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 256–​69. Edward Elgar. Guerrero, Alexander. 2014. Against Elections: The Lottocratic Alternative. Philosophy and Public Affairs 42, 135–​78. Gürsözlü, Fuat. 2009. Agonism and Deliberation—​Recognizing the Difference. Journal of Political Philosophy 17, 356–​68. Gutmann, Amy, and Dennis Thompson. 2004. Why Deliberative Democracy? Princeton University Press. Habermas, Jürgen. 1998. Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. MIT Press. Hall, Robert, and Charles Jones. 1999. Why Do Some Countries Produce So Much More Output Per Worker Than Others? Quarterly Journal of Economics 114, 83–​116. Halperin, Morton, Joe Siegle, and Michael Weinstein. 2005. Why Democracies Excel. Foreign Affairs 83, 57–​71. Halperin, Morton, Joe Siegle, and Michael Weinstein. 2010. The Democracy Advantage: How Democracies Promote Prosperity and Peace. Routledge. Hanson, Robin. 2013. Shall We Vote on Values, but Bet on Beliefs? Journal of Political Philosophy 21, 151–​78. Harcourt, Bernard. 2011. The Illusion of Free Markets. Harvard University Press. Hardin, Russell. 2004. Trust and Trustworthiness. Russell Sage. Harkness, Sarah. 2014. The Consequences of Interpersonal Rewards. Sociology Compass 8, 219–​31. Harkness, Sarah. 2017. Spread of Status Value:  Rewards and the Creation of Status Characteristics. Social Science Research 61, 98–​111.

288  Works Cited Hasnas, John. 2013. Is There a Moral Duty to Obey the Law? Social Philosophy and Policy 30, 450–​79. Hayek, F. A. 1945. The Use of Knowledge in Society. American Economic Review 35, 519–​30. Hayek, F. A. 1973. Law Legislation and Liberty, vol. 1:  Rules and Order. University of Chicago Press. Hayek, F. A. 1978. Law, Legislation and Liberty, vol. 2:  The Mirage of Social Justice. University of Chicago Press. Hayek, F. A. 1979. Law, Legislation and Liberty, vol. 3: The Political Order of a Free People. University of Chicago Press. Hayek, F. A. 2007. The Road to Serfdom. Definitive ed. Routledge. Hayek, F. A. 2011. The Constitution of Liberty. Definitive ed. University of Chicago Press. Healy, Gene. 2009. The Cult of the Presidency, Updated: America’s Dangerous Devotion to Executive Power. Cato Institute. Hedges, Chris. 2002. War Is a Force That Gives Us Meaning. PublicAffairs. Heinemann, Friedrich, and Benjamin Tanz. 2008. The Impact of Trust on Reforms. Journal of Economic Policy Reform 11, 173–​85. Helliwell, John, Shun Wang, and Jinwen Xu. 2014. How Durable Are Social Norms? Immigrant Trust and Generosity in 132 Countries. NBER Working Paper No. 19855. Henrich, Joseph, Robert Boyd, Samuel Bowles, Colin Camerer, Ernst Fehr, Herbert Gintis, Richard McElreath, Michael Alvard, Abigail Barr, Jean Ensminger, Natalie Smith Henrich, Kim Hill, Franscisco Gil-​White, Michael Gurven, Frank W. Marlowe, John Q. Patton, and David Tracer. 2005. “Economic Man” in Cross-​ Cultural Perspective: Behavioral Experiments in 15 Small-​Scale Societies. Behavioral and Brain Science 28, 795–​855. Hetherington, Marc. 2005. Why Trust Matters. Princeton University Press. Hetherington, Marc. 2015. Why Polarized Trust Matters. The Forum 13, 445–​58. Hetherington, Marc, and Jonathan Weiler. 2018. Prius or Pickup? How the Answers to Four Simple Questions Explain America’s Great Divide. Houghton Mifflin Harcourt. Hirst, Paul. 1997. From Statism to Pluralism. UCL Press. Hjerppe, Reino. 1998. Social Capital and Economic Growth. Government Institute for Economic Research (VATT). Honig, Bonnie. 1993. Political Theory and the Displacement of Politics. Cornell University Press. Honore, A. M. 1961. Ownership. In Oxford Essays in Jurisprudence, edited by A. G. Guest, 107–​47. Oxford University Press. Hooghe, Marc. 2018. Trust and Elections. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 617–​31. Oxford University Press. Hosner, Larue. 1995. Trust: The Connecting Link between Organizational Theory and Philosophical Ethics. Academy of Management Review 20, 379–​403. Hsieh, Chang-​ Tai, and Enrico Moretti. 2015. Housing Constraints and Spatial Misallocation. NBER Working Paper No. 21154. Hutchison, Marc, and Kristin Johnson. 2017. Political Trust in Sub-​Saharan Africa and the Arab Region. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 461–​87. Edward Elgar. Inglehart, Ronald. 1990. Culture Shift in Advanced Industrial Society. Princeton University Press.

Works Cited  289 Inglehart, Ronald, C.  Haerpfer, A.  Moreno, C.  Welzel, K.  Kizilova, J.  Diez-​Medrano, M. Labos, P. Norris, E. Ponarin, and B. Puranen (eds.). 2020. World Values Survey: Round Seven–​Country-​Pooled Datafile 2017–​2020. Madrid: JD Systems Institute. http://​www. worldvaluessurvey.org/​WVSDocumentationWV7.jsp Intercollegiate Studies Institute. 2008. Our Fading Heritage: Americans Fail a Basic Test on Their History and Institutions. Intercollegiate Studies Institute. Iyengar, Shanto, and Sean Westwood. 2015. Fear and Loathing across Party Lines: New Evidence on Group Polarization. American Journal of Political Science 59, 690–​707. Jankauskas, Vidmantas, and Janina Seputiene. 2007. The Relationship between Social Capital, Governance and Economic Performance in Europe. Business:  Theory and Practice 8, 131–​8. Jennings, M. Kent, and Laura Stoker. 2004. Social Trust and Civic Engagement across Time and Generations. Acta Politica 39, 342–​79. Jones, Karen. 2019. Trust, Distrust, and Affective Looping. Philosophical Studies 176, 955–​68. Kenworthy, Lane. 2013. Progress for the Poor. Oxford University Press. Kerr, Gavin. 2017. The Property-​Owning Democracy:  Freedom and Capitalism in the Twenty-​First Century. Routledge. Kinder, Donald, and Cindy Kam. 2010. Us against Them:  Ethnocentric Foundations of American Opinion. University of Chicago Press. Klein, Ezra. 2020. Why We’re Polarized. Avid Reader Press. Knack, Stephen, and Paul Zak. 2002. Building Trust: Public Policy, Interpersonal Trust, and Economic Development. Supreme Court Economic Review 10, 91–​107. Knight, Jack, and James Johnson. 2011. The Priority of Democracy: Political Consequences of Pragmatism. Princeton University Press. Kogelmann, Brian. 2017. Justice, Diversity, and the Well-​Ordered Society. Philosophical Quarterly. Kogelmann, Brian. 2021. Secret Government. Cambridge University Press. Kramer, Roderick. 2018. Ingroup-​Outgroup Trust: Barriers, Benefits, and Bridges. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 95–​115. Oxford University Press. Krugman, Paul. 2012. End This Depression Now. Norton. Kruse, Douglas, Richard Freeman, and Joseph Blasi. 2011. Shared Capitalism at Work: Employee Ownership, Profit and Gain Sharing, and Broad-​Based Stock Options. University of Chicago Press. Kukathas, Chandran. 2007. The Liberal Archipelago. Oxford University Press. Kumlin, Staffan, and Alte Haugsgjerd. 2017. The Welfare State and Political Trust: Bringing Performance Back In. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 285–​301. Edward Elgar. Kumlin, Staffan, and Bo Rothstein. 2005. Making and Breaking Social Capital: The Impact of Welfare-​State Institutions. Comparative Political Studies 38, 339–​65. Kumlin, Staffan, Isabelle Stadelmann-​Steffen, and Alte Haugsgjerd. 2018. Trust and the Welfare State. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 385–​408. Oxford University Press. Laborde, Cécile. 2017. Liberalism’s Religion. Harvard University Press. Laclau, Ernesto, and Chantal Mouffe. 2014. Hegemony and Socialist Strategy. Verso. Lafont, Cristina. 2017. Can Democracy Be Deliberative and Participatory? The Democratic Case for Political Uses of Mini-​publics. Daedalus 146, 85–​105.

290  Works Cited Lamprianou, Iasonas, and Antonis Ellinas. 2017. Institutional Grievances and Right-​ Wing Extremism:  Voting for Golden Dawn in Greece. South European Society and Politics 22, 43–​60. Landemore, Hélène. 2013. Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many. Princeton University Press. Layman, Geoffrey, Thomas Carsey, and Juliana Horowitz. 2006. Party Polarization in American Politics:  Characteristics, Causes, and Consequences. Annual Review of Political Science 9, 83–​110. Leech, Beth. 2010. Lobbying and Influence. In The Oxford Handbook of American Political Parties and Interest Groups, edited by Sandy Maisel, Jeffrey Berry, and George Edwards, 534–​51. Oxford University Press. Leech, Beth. 2011. Lobbying and Interest Group Advocacy. In The Oxford Handbook of the American Congress, edited by Eric Schickler, Frances Lee, and George Edwards, 598–​ 617. Oxford University Press. Leibrecht, Markus, and Hans Pitlik. 2015. Social Trust, Institutional and Political Constraints on the Executive and Deregulation of Markets. European Journal of Political Economy 39, 249–​68. Lenard, Patti Tamara. 2008. Trust Your Compatriots, but Count Your Change: The Roles of Trust, Mistrust, and Distrust in Democracy. Political Studies 56, 312–​32. Letki, Natalia. 2018. Trust in Newly Democratic Regimes. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 335–​56. Oxford University Press. Levi, Margaret, and Laura Stoker. 2000. Political Trust and Trustworthiness. Annual Review of Political Science 3, 475–​507. Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. Penguin. Levy, Jacob. 2015. Rationalism, Pluralism, and Freedom. Oxford University Press. Lindsey, Brink, and Steven Teles. 2017. The Captured Economy: How the Powerful Enrich Themselves, Slow Down Growth, and Increase Inequality. Oxford University Press. Lissowski, Grzegorz, Tadeusz Tyszka, and Woldzimierz Okrasa. 1991. Principles of Distributive Justice: Experiments in Poland and America. Journal of Conflict Resolution 35, 98–​119. Lister, Andrew. 2013. Public Reason and Political Community. Bloomsbury Academic. Listhaug, Ola, and Tor Georg Jakobsen. 2018. Foundations of Political Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 559–​77. Oxford University Press. Liu, Christopher, and Dietlind Stolle. 2017. Social Capital, Civic Culture, and Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 338–​52. Edward Elgar. Lomasky, Loren. 1987. Persons, Rights, and the Moral Community. Oxford University Press. Luttmer, Erzo, and Monica Singhal. 2014. Tax Morale. Journal of Economic Perspectives 28, 149–​68. Mackie, Gerry. 2003. Democracy Defended. Cambridge University Press. Maddison, Angus. 2007. Contours of the World Economy: 1–​2030 AD. Oxford University Press. Magness, Phillip. 2019. Detecting Historical Inequality Patterns: A Replication of Thomas Piketty’s Wealth Concentration Estimates for the United Kingdom. Social Science Quarterly 100, 1566–​76. Mankiw, Gregory. 2011. Principles of Microeconomics. Cengage Learning.

Works Cited  291 Manning, Colin, and Kevin Vallier. 2021. On the Individuation of Issues in Public Reason. Bowling Green State University. Unpublished. Mansbridge, Jane. 2003. Rethinking Representation. American Political Science Review 97, 515–​28. Margolis, Michele. 2018. From Politics to the Pews:  How Partisanship and the Political Environment Shape Religious Identity. University of Chicago Press. Marien, Sofie. 2017. The Measurement Equivalence of Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W.  G. van der Meer, 89–​103. Edward Elgar. Martin, Gregory, and Ali Yurukoglu. 2017. Bias in Cable News:  Persuasion and Polarization. American Economic Review 107, 2565–​99. Marx, Karl, and Friedrich Engels. 2011. Capital: A Critique of Political Economy. Vol. 1. Dover Publications. Mason, Lilliana. 2018. Uncivil Agreement: How Politics Became Our Identity. University of Chicago. Mayne, Quinton, and Armen Hakhverdian. 2017. Education, Socialization, and Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 176–​96. Edward Elgar. McCarty, Nolan. 2019. Polarization:  What Everyone Needs to Know. Oxford University Press. McCloskey, Deirdre. 2016. Bourgeois Equality:  How Ideas, Not Capital or Institutions, Enriched the World. University of Chicago Press. McLaren, Lauren. 2017. Immigration, Ethnic Diversity, and Inequality. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W.  G. van der Meer, 316–​37. Edward Elgar. Meade, James. 1964. Efficiency, Equality and the Ownership of Property. George Allen and Unwin. Melki, Mickael, and Andrew Pickering. 2020. Polarization and Corruption in America. European Economic Review 10, 1–​32. Mellers, Barbara, Philip Tetlock, and Hal Arkes. 2019. Forecasting Tournaments, Epistemic Humility, and Attitude Depolarization. Cognition 188, 19–​26. Messick, David, and Roderick Kramer. 2001. Trust as a Form of Shallow Morality. In Trust in Society, edited by Karen Cook, 89–​117. Russell Sage. Milfont, Taciano, Petar Milojev, and Chris Sibley. 2016. Values Stability and Change in Adulthood:  A 3-​Year Longitudinal Study of Rank-​Order Stability and Mean-​Level Differences. Personality and Psychology Bulletin 42, 572–​88. Mill, John Stuart. 1963–​. Collected Works of John Stuart Mill. 33 vols. University of Toronto Press. Minozzi, William. 2011. A Jamming Theory of Politics. Journal of Politics 73, 301–​15. Moehler, Michael. 2019. Replies to Gaus, Van Schoelandt and Cooper:  Prudence, Morality, and the Social Contract. Analysis 79, 140–​53. Moller, Dan. 2019. Governing Least: A New England Libertarianism. Oxford University Press. Mondak, Jeffrey, Matthew Hayes, and Damarys Canache. 2017. Biological and Psychological Influences on Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 143–​59. Edward Elgar. Mouffe, Chantal. 2009. The Democratic Paradox. Verso. Mueller, Dennis. 2003. Public Choice III. Cambridge University Press.

292  Works Cited Muldoon, Ryan. 2018. Understanding Norms and Changing Them. Social Philosophy and Policy 35, 128–​48. Muñiz-​Fraticelli, Victor. 2014. The Structure of Pluralism. Oxford University Press. Murphy, Liam, and Thomas Nagel. 2002. The Myth of Ownership:  Taxes and Justice. Oxford University Press. Mutz, Diana. 2002. The Consequences of Cross-​ Cutting Networks for Political Polarization. American Journal of Political Science 46, 838–​55. Nannestad, Peter. 2008. What Have We Learned about Generalized Trust, If Anything? Annual Review of Political Science 11, 413–​36. Narveson, Jan. 2013. Comment on Tilo Wesche:  On Property-​Owning Democracy. Analyse & Kritik 35, 113–​9. Neblo, Michael. 2015. Deliberative Democracy between Theory and Practice. Cambridge University Press. Necker, Sarah, and Andrea Voskort. 2014. Politics and Parents:  Intergenerational Transmission of Values after a Regime Shift. European Journal of Political Economy 36, 177–​94. Nelson, James. 2013. Conscience, Incorporated. Michigan State Law Review 5, 1565–​620. Nelson, James. 2015. The Freedom of Business Association. Columbia Law Review 115, 461–​513. Newton, Ken. 2017. Political Trust and the Mass Media. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 353–​72. Edward Elgar. Newton, Ken, Dietlind Stolle, and Sonja Zmerli. 2018. Social and Political Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 37–​56. Oxford University Press. Norris, Pippa. 2011. Democratic Deficit:  Critical Citizens Revisited. Cambridge University Press. Norris, Pippa. 2017. The Conceptual Framework of Political Support. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W.  G. van der Meer, 19–​32. Edward Elgar. Norris, Pippa and Ronald Inglehart. 2019. Cultural Backlash: Trump, Brexit, and Authori­ tarian Populism. Cambridge. North, Douglass, John Joseph Wallis, and Barry Weingast. 2013. Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History. Cambridge. Nozick, Robert. 1974. Anarchy, State and Utopia. Basic Books. Nussbaum, Martha. 2002. Rawls and Feminism. In The Cambridge Companion to Rawls, edited by Samuel Freeman, 488–​520. Cambridge University Press. O’Neill, Martin. 2009. Three Rawlsian Routes towards Economic Democracy. Revue de philosophie economique 17, 29–​55. O’Neill, Martin, and Thad Williamson. 2012a. Free Market Fairness: Is There a Moral Case for Free Markets? Boston Review, November 5. http://​bostonreview.net/​books-​ ideas/​free-​market-​fairness. O’Neill, Martin, and Thad Williamson. 2012b. Property-​Owning Democracy: Rawls and Beyond. Wiley-​Blackwell. Olson, Mancur. 1971. The Logic of Collective Action:  Public Goods and the Theory of Groups. Harvard University Press. Olson, Mancur. 1984. The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities. Yale University Press.

Works Cited  293 Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. Oz Yalama, Gamze, and Erdal Gumus. 2013. Determinants of Tax Evasion Behavior:  Empirical Evidence from Survey Data. International Business and Management 6, 15–​23. Paluck, Elizabeth, Seth Green, and Donald Green. 2018. The Contact Hypothesis Re-​ evaluated. Behavioural Public Policy 3, 129–​58. Parkinson, John, and Jane Mansbridge. 2012. Deliberative Systems:  Deliberative Democracy at the Large Scale. Cambridge University Press. Pavić, Željko, and Antun Šundalić. 2016. A Comparison of Online and Offline Social Participation Impacts on Generalized Trust. Italian Sociological Review 6, 185–​203. Paxton, Pamela. 2007. Association Memberships and Generalized Trust:  A Multilevel Model across 31 Countries. Social Forces 86, 47–​76. Paxton, Pamela, and Robert Ressler. 2018. Trust and Participation in Associations. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 149–​72. Oxford University Press. Pettigrew, Thomas. 2000. Reducing Intergroup Conflict with Optimal Intergroup Contact. In Diplomacy and Psychology, edited by Tommy Garling, Gunnel Backenroth-​ Ohsako, Bo Ekehammar, and Michael Wessells, 301–​7. Psychology Press. Pettigrew, Thomas, and Linda Tropp. 2006. A Meta-​analytic Test of Intergroup Contact Theory. Journal of Personality and Social Psychology 90, 751–​83. Pettit, Philip. 2013. Taking Back the Economy:  The Market as a Res Publica. Open Democracy, January 31. http://​w ww.opendemocracy.net/​en/​opendemocracyuk/​ taking-​back-​economy-​market-​as-​res-​publica/​ Pettit, Philip. 2019. Social Norms and the Internal Point of View: An Elaboration of Hart’s Genealogy of Law. Oxford Journal of Legal Studies 39, 229–​58. Pharr, Susan, and Robert Putnam. 2000. Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton University Press. Piketty, Thomas. 2014. Capital in the 21st Century. Belknap Press of Harvard University Press. Pincione, Guido, and Fernando Tesón. 2006. Rational Choice and Democratic Deliberation: A Theory of Discourse Failure. Cambridge University Press. Pitlik, Hans, and Martin Rode. 2019. Radical Distrust: Are Economic Policy Attitudes Tempered by Social Trust? WIFO Working Papers 594. Pogrebinschi, Thamy, and David Samuels. 2014. The Impact of Participatory Democracy: Evidence from Brazil’s National Public Policy Conferences. Comparative Politics 46, 313–​32. Public Trust in Government: 1958–​2019. http://​www.pewresearch.org/​politics/​2019/​04/​ 11/​public-​trust-​in-​government-​1958-​2019/​ Putnam, Robert. 1994. Making Democracy Work:  Civic Traditions in Modern Italy. Princeton University Press. Putnam, Robert. 1995. Bowling Alone: Democracy in America at the End of the Twentieth Century. Journal of Democracy 6, 65–​78. Putnam, Robert. 2001. Bowling Alone: The Collapse and Revival of American Community. Simon & Schuster. Putnam, Robert. 2007. E Pluribus Unum: Diversity and Community in the Twenty-​First Century—​The 2006 Johan Skytte Prize Lecture. Scandinavian Political Studies 30, 137–​74.

294  Works Cited Ramey, Adam. 2017. Filibusted:  The Mixed Effects of “Going Nuclear.” July 9. http://​ dx.doi.org/​10.2139/​ssrn.3065237. Randolph, Thomas. 2017. Political Trust as a Heuristic. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 197–​211. Edward Elgar. Rawls, John. 1971. A Theory of Justice. Oxford University Press. Rawls, John. 1999. A Theory of Justice. Rev. ed. Belknap Press of Harvard University Press. Rawls, John. 2001. Justice as Fairness: A Restatement. Universal Law Publishing. Rawls, John. 2005. Political Liberalism. Columbia University Press. Raz, Joseph. 1970. The Concept of a Legal System. Oxford University Press. Raz, Joseph. 1986. The Morality of Freedom. Oxford University Press. Reeskens, Tim, and Marc Hooghe. 2008. Cross-​Cultural Measurement Equivalence of Generalized Trust from the European Social Survey (2002 and 2004). Social Indicators Research 85, 515–​32. Rehfeld, Andrew. 2018. On Representing. Journal of Political Philosophy 26, 216–​39. Richardson, Grant. 2006. Determinants of Tax Evasion: A Cross-​Country Investigation. Journal of International Accounting, Auditing, and Taxation 15, 150–​69. Riker, William. 1988. Liberalism against Populism: A Confrontation between the Theory of Democracy and the Theory of Social Choice. Waveland Press. Robbins, Blaine. 2011. Neither Government nor Community Alone:  A Test of State-​ Centered Models of Generalized Trust. Rationality and Society 23, 304–​46. Robbins, Blaine. 2012. A Blessing and a Curse? Political Institutions in the Growth and Decay of Generalized Trust: A Cross-​National Panel Analysis, 1980–​2009. PLoS One 7,  1–​14. Roemer, John. 1994. A Future for Socialism. Harvard University Press. Rognlie, Matthew. 2015. Deciphering the Fall and Rise in the Net Capital Share. Brookings Papers on Economic Activity,  1–​54. Rosenblum, Nancy. 1998. Membership and Morals. Princeton University Press. Rosenfeld, Sam. 2018. The Polarizers: Postwar Architects of Our Partisan Era. University of Chicago Press. Rothbard, Murray. 1973. For a New Liberty: The Libertarian Manifesto. Collier Books. Rothstein, Bo. 2011. The Quality of Government: Corruption, Social Trust, and Inequality in International Perspective. University of Chicago Press. Rothstein, Bo, and Dietlind Stolle. 2008. The State and Social Capital: An Institutional Theory of Generalized Trust. Comparative Politics 40, 441–​59. Rougier, Eric, Francois Combarnous, and Yves-​Andre Faure. 2018. The “Local Economy” Effect of Social Transfers: An Empirical Assessment of the Impact of the Bolsa Familia Program on Local Productive Structure and Economic Growth. World Development 103, 199–​215. Sabatini, Fabio. 2007. The Role of Social Capital in Economic Development. AICCON Cultura Cooperazione. July 9. http://​dx.doi.org/​10.2139/​ssrn.901361. Sabl, Andrew. 2015. The Two Cultures of Democratic Theory:  Responsiveness, Democratic Quality, and the Empirical-​Normative Divide. Perspectives on Politics 13, 345–​65. Sapienza, Paola, Anna Toldra-​Simats, and Luigi Zingales. 2013. Understanding Trust. Economic Journal 123, 1313–​32. Satyanath, Shanker, Nico Voigtländer, and Hans-​ Joachim Voth. 2017. Bowling for Fascism: Social Capital and the Rise of the Nazi Party. Journal of Political Economy 125, 478–​526.

Works Cited  295 Scanlon, Thomas. 2011. Libertarianism and Liberty:  How Not to Argue for Limited Government and Lower Taxes. Boston Review, October 19. http://​bostonreview.net/​ forum/​libertarianism-​and-​liberty. Schemmel, Christian. 2015. How (Not) to Criticise the Welfare State. Journal of Applied Philosophy 32, 393–​409. Schmidtz, David. 2011. Nonideal Theory: What It Is and What It Needs to Be. Ethics 121, 772–​96. Schultz-​Bergin, Marcus. 2020. Does Convergence Liberalism Risk Anarchy? Journal of Ethics and Social Philosophy, forthcoming. Schumpeter, Joseph. 1942. Capitalism, Socialism and Democracy. Harper and Brothers. Schyns, Peggy, and Christel Koop. 2010. Political Distrust and Social Capital in Europe and the USA. Social Indicators Research 96, 145–​67. Scott, James C. 2009. The Art of Not Being Governed. Yale University Press. Sen, Amartya. 2000. Development as Freedom. First Anchor Books. Shiffrin, Seana. 2005. What Is Really Wrong with Compelled Association? Northwestern University Law Review 99, 839–​88. Shorten, Andrew. 2018. May Churches Discriminate? Journal of Applied Philosophy 36, 709–​17. Singer, Abraham. 2019. The Form of the Firm. Oxford University Press. Siu, Alice. 2017. Deliberation and the Challenge of Inequality. Daedalus 146, 119–​28. Skyrms, Brian. 1996. Evolution of the Social Contract. Cambridge University Press. Smith, Graham, Robert Richards, and John Gastil. 2015. The Potential of Participedia as a Crowdsourcing Tool for Comparative Analysis of Democratic Innovations. Policy and Internet 7, 243–​62. Smith, Matthew, Danny Yagan, Owen Zidar, and Eric Zwick. 2019. Capitalists in the Twenty-​First Century. Quarterly Journal of Economics 134, 1675–​745. Smith, Tom W.  Michael Davern, Jeremy Freese, and Stephen Morgan. General Social Surveys, 1972–​2018, Sponsored by the National Science Foundation. NORC 2018 at the University of Chicago. Data accessed from the GSS Data Explorer Website at gssdataexplorer.norc.org. Southwood, Nicholas. 2016. Does ‘Ought’ Imply ‘Feasible’? Philosophy and Public Affairs 44,  7–​45. Southwood, Nicholas. 2018. The Feasibility Issue. Philosophy Compass 13, 1–​13. Southwood, Nicholas, and David Wiens. 2016. ‘Actual’ Does Not Imply ‘Feasible’. Philosophical Studies 173, 3037–​60. Stanford, P. Kyle. 2018. The Difference between Ice Cream and Nazis:  Moral Externalization and the Evolution of Human Cooperation. Behavioral and Brain Sciences 41, 1–​49. Stevenson, Betsey, and Justin Wolfers. 2008. Economic Growth and Subjective Well-​ Being: Reassessing the Easterlin Paradox. Brookings Papers on Economic Activity 39,  1–​102. Stigler, George. 1971. The Theory of Economic Regulation. Bell Journal of Econoomics and Management Science 3, 3–​18. Sugden, Robert. 1993. Normative Judgments and Spontaneous Order: The Contractarian Element in Hayek’s Thought. Constitutional Political Economy 4, 393–​424. Sunstein, Cass. 2002. The Law of Group Polarization. Journal of Political Philosophy 10, 175–​95. Sunstein, Cass. 2011. Going to Extremes:  How Like Minds Unite and Divide. Oxford University Press.

296  Works Cited Sunstein, Cass, and Stephen Holmes. 1999. The Cost of Rights: Why Liberty Depends on Taxes. Norton. Sutch, Richard. 2017. The One Percent across Two Centuries: A Replication of Thomas Piketty’s Data on the Concentration of Wealth in the United States. Social Science History 41, 587–​613. Swope, Kurtis, John Cadigan, Pamela Schmitt, and Robert Shupp. 2008. Social Position and Distributive Justice:  Experimental Evidence. Southern Economic Journal 74, 811–​18. Talisse, Robert B. 2017. New Trouble for Deliberative Democracy. Les ateliers de l’éthique 12, 107–​23. Talisse, Robert B. 2019. Overdoing Democracy. Oxford University Press. Tamir, Yael. 1998. Revisiting the Civic Sphere. In Freedom of Association, edited by Amy Gutmann, 214–​38. Princeton University Press. Tanner, Michael. 1996. The End of Welfare:  Fighting Poverty in the Civil Society. Cato Institute. Tanner, Michael. 2018. The Inclusive Economy. Cato Institute. Tetlock, Philip. 2006. Expert Political Judgment:  How Good Is It? How Can We Know? Princeton University Press. Tetlock, Philip, and Daniel Gardner. 2016. Superforecasting:  The Art and Science of Prediction. Broadway Books. Thomas, Alan. 2017. Republic of Equals: Predistribution and Property-​Owning Democracy. Oxford University Press. Thomassen, Jacques, Rudy Andeweg, and Carolien van Ham. 2017. Political Trust and the Decline of Legitimacy Debate: A Theoretical and Empirical Investigation into Their Interrelationship. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 509–​25. Edward Elgar. Thompson, Douglas. 2019. Review of Must Politics Be War? Restoring Our Trust in the Open Society. Notre Dame Philosophical Reviews, July 9.  http://​ndpr.nd.edu/​news/​ must-​politics-​be-​war-​restoring-​our-​trust-​in-​the-​open-​society/​. Thorpe, Andrew. 2001. A History of the British Labour Party. Palgrave Macmillan. Thrasher, John. 2016. The Ethics of Legislative Vote Trading. Political Studies 64, 614–​29. Tomasi, John. 2012. Free-​Market Fairness. Princeton University Press. Torcal, Mariano.  2017. Political Trust in Western and Southern Europe. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W.  G. van der Meer, 418–​39. Edward Elgar. Treisman, Daniel. 2000. The Causes of Corruption: A Cross-​National Study. Journal of Public Economics 76, 399–​457. Tullock, Gordon. 1972. The Purchase of Politicians. Western Economic Journal 10, 354–​5. Tullock, Gordon. 1993. Rent Seeking. Edward Elgar. Uslaner, Eric. 2002. The Moral Foundations of Trust. Cambridge University Press. Uslaner, Eric. 2015. Congressional Polarization and Political Trust. The Forum 13, 361–​73. Uslaner, Eric. 2017. Political Trust, Corruption, and Inequality. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 302–​15. Edward Elgar. Uslaner, Eric. 2018. The Study of Trust. In Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 3–​14. Oxford University Press. Vallier, Kevin. 2014a. Liberal Politics and Public Faith: Beyond Separation. Routledge.

Works Cited  297 Vallier, Kevin. 2014b. Review of Andrew Lister’s Public Reason and Political Community. Notre Dame Philosophical Reviews, July 9. https://​ndpr.nd.edu/​news/​public-​reason-​ and-​political-​community/​. Vallier, Kevin. 2015. A Moral and Economic Critique of the New Property-​Owning Democrats: On Behalf of a Rawlsian Welfare State. Philosophical Studies 172, 283–​304. Vallier, Kevin. 2016. Public Justification versus Public Deliberation: The Case for Divorce. Canadian Journal of Philosophy 45, 139–​58. Vallier, Kevin. 2018a. Exit, Voice, and Public Reason. American Political Science Review 112, 1120–​24. Vallier, Kevin. 2018b. Public Reason in the Open Society. Cosmos and Taxis 5, 38–​46. Vallier, Kevin. 2019. Must Politics Be War? Restoring Our Trust in the Open Society. Oxford University Press. Referenced with MPBW. Vallier, Kevin. 2019. Rawls, Piketty, and the Critique of Welfare-​State Capitalism. Journal of Politics 81, 142–​52. Vallier, Kevin. 2020a. Associations in Social Contract Theory:  Toward a Pluralist Contractarianism. Political Studies 68, 483–​503. Vallier, Kevin. 2020b. Neoliberalism. Bowling Green State University. Unpublished. Vallier, Kevin. 2020c. In Defense of Idealization in Public Reason. Erkenntnis, forthcoming. van der Meer, Tom W. G. 2017. Democratic Input, Macroeconomic Output, and Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 270–​84. Edward Elgar. van der Meer, Tom W. G. 2018. Economic Performance and Political Trust. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 599–​615. Oxford University Press. van der Meer, Tom W. G., and Armen Hakhverdian. 2017. Political Trust as the Evaluation of Process and Performance: A Cross-​National Study of 42 European Democracies. Political Studies 65, 81–​102. van Deth, Jan. 2017. Compliance, Trust, and Norms of Citizenship. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W.  G. van der Meer, 212–​27. Edward Elgar. van Ham, Carolien, Jacques Thomassen, Kees Aarts, and Rudy Andeweg. 2017. Myth and Reality of the Legitimacy Crisis:  Explaining Trends and Cross-​National Differences in Established Democracies. Oxford University Press. van Wietmarschen, Han. 2019. Political Testimony. Politics, Philosophy, and Economics 18,  23–​45. Vanek, Jaroslav. 1970. The General Theory of Labor-​Managed Market Economies. Cornell University Press. Vigoda-​Gadot, Eran, and Shlomo Mizrahi. 2014. Managing Democracies in Turbulent Times. Springer. Ward, Benjamin. 1958. The Firm in Illyria:  Market Syndicalism. American Economic Review 48, 566–​89. Warren, Anne Marie, Aninin Sulaiman, and Noor Ismawati Jaafar. 2014. Social Media Effects on Fostering Online Civic Engagement and Building Citizen Trust and Trust in Institutions. Government Information Quarterly 31, 291–​301. Warren, Mark. 2001. Democracy and Association. Princeton University Press.

298  Works Cited Warren, Mark. 2017. What Kinds of Trust Does a Democracy Need? Trust from the Perspective of Democratic Theory. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 33–​52. Edward Elgar. Warren, Mark. 2018. Trust and Democracy. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 75–​94. Oxford University Press. Watson, Lori, and Christie Hartley. 2018. Equal Citizenship and Public Reason: A Feminist Political Liberalism. Oxford University Press. Wenman, Mark. 2013. Agonistic Democracy: Constituent Power in the Era of Globalization. Cambridge University Press. Westwood, Sean, Shanto Iyengar, Stefaan Walgrave, Rafael Leonisio, Luis Miller, and Oliver Strijbis. 2018. The Tie That Divides: Cross-​National Evidence of the Primary of Partyism. European Journal of Political Research 57, 333–​54. White, Stuart. 1998. Trade Unionism in a Liberal State. In Freedom of Association, edited by Amy Gutmann, 330–​56. Princeton University Press. Whiteley, Paul. 2002. Economic Growth and Social Capital. Political Studies 48, 443–​66. Wilkes, Rima, and Cary Wu. 2018. Trust and Minority Groups. In The Oxford Handbook of Social and Political Trust, edited by Uslaner, 231–​50. Oxford University Press. Wilkinson, Will. 2019. The Density Divide:  Urbanization, Polarization, and Populist Backlash. In Niskanen Center Research Papers, 1–​79. Niskanen Center. Williamson, Thad. 2013. Constitutionalizing Property-​Owning Democracy. Analyse & Kritik 1, 237–​53. Wilson, James Lindley. 2017. Review of Republic of Equals: Predistribution and Property Owning Democracy. Notre Dame Philosophical Reviews, July 17. http://​ndpr.nd.edu/​ news/​republic-​of-​equals-​predistribution-​and-​property-​owning-​democracy/​. Wilson, Rick. 2018. Trust Experiments, Trust Games, and Surveys. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 279–​303. Oxford University Press. Wilson, Rick, and Catherine Eckel. 2017. Political Trust in Experimental Designs. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 125–​ 39. Edward Elgar. You, Jong-​sung. 2012. Social Trust: Fairness Matters More Than Homogeneity. Political Psychology 33, 701–​21. You, Jong-​sung. 2018. Trust and Corruption. In The Oxford Handbook of Social and Political Trust, edited by Eric Uslaner, 473–​96. Oxford University Press. Závecz, Gergõ. 2017. Post-​communist Societies of Central and Eastern Europe. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 440–​ 60. Edward Elgar. Zmerli, Sonja, and Ken Newton. 2008. Social Trust and Attitudes toward Democracy. Public Opinion Quarterly 72, 706–​24. Zmerli, Sonja, and Ken Newton. 2017. Objects of Political and Social Trust: Scales and Hierarchies. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 104–​24. Edward Elgar. Zmerli, Sonja, and Tom W.  G. van der Meer. 2017. The Deeply Rooted Concern with Political Trust. In Handbook of Political Trust, edited by Sonja Zmerli and Tom W. G. van der Meer, 1–​15. Edward Elgar.

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. Figure are indicated by f following the page number Achen, Christopher, 19, 177–​78n.21, 237, 247, 252, 254–​55, 257–​63, 265–​66, 267 agents, 28n.14, 28, 28n.15, 29, 32–​33, 38, 54, 73, 122–​23, 134–​35n.43, 147, 214, 218, 226–​ 27, 259, 261 conditionally cooperative, 38, 38n.37 moderately idealized, 27, 28–​29, 28n.14, 32, 35, 47–​48n.50, 54, 130, 229, 234, 246n.53, 247, 254–​55,  260–​61 moral, 21n.2, 24, 27–​28, 31–​32, 52 morally responsible, 21n.2, 23–​24, 53–​54,  102–​3 aggregative democracy, 237, 271–​74, 272n.126 agonistic democracy, 237, 271–​74, 272n.127 Allport, Gordon, 90, 92 anarchist, 119n.43, 148, 235–​36 Anderson, Elizabeth, 72–​73n.111, 148n.24, 197–​98, 198n.79 antipoverty, 138, 167–​68, 169, 171 association membership, 88, 92–​93, 94n.27 associations, 17, 46, 77–​78, 86, 88, 89–​94, 90n.9, 94n.28, 94n.30, 95, 95–​96n.34, 96–​97, 98–​100, 101–​6, 107, 107n.73, 107n.75, 108, 109, 109n.86, 110, 110n.88, 111, 112, 113, 114–​15, 116–​17,  238–​39 bonding, 91–​92, 93n.26 bridging, 91–​92, 93n.26, 110, 118 civic, 77–​78, 77–​78n.131, 89–​90, 97–​100, 104–​5, 115,  197–​98 commercial, 89, 99–​101, 108, 114–​15 moral, 89, 95–​97, 101–​2n.55, 103, 104–​6, 108, 109, 126–​27 and real trust, 89–​94 and trust for the right reasons, 118 voluntary, 89–​90,  92–​93 authoritarian regimes, 47–​48, 47–​48n.50, 75, 77, 204n.100, 260   bad-​reason objection, 260–​61, 262 balkanization, 89 objection, 109

and private tyranny, 107–​10 Bartels, Larry, 19, 177–​78n.21, 237, 247, 252, 254–​55, 257–​63, 265–​66,  267 bicameralism, 225, 225n.60 Bolsa Familia (BF), 139, 145–​46, 147, 167–171 Borda count, 270 Bou-​Habib, Paul, 199–​200, 201, 201n.87 Brazil, 139, 169–​71, 250n.65 Brennan, Jason, 120–​21n.2, 215n.24, 245–​ 46n.52, 262, 266–​67n.105 Bryan, Ben, 113 Buchanan, James, 225n.60, 225–​26, 225–​26n.61, 247n.55 bureaucracies, 139–​40, 143, 144, 162, 163–​67, 192–​94,  195 bureaucracy, 74, 123–​24, 139–​40, 142–​43, 154–​ 55, 164, 189, 195, 223n.44   capital, 38–​39, 45–​46, 47, 67, 68, 102, 114n.103, 119, 121, 129, 137, 169–​70, 173–​74n.2, 177, 180, 180n.31, 181–​82, 186–​90, 189n.59, 191–​92, 193–​94, 194n.68, 195, 198–​99, 201, 202–​4, 206, 222, 228–​29. See capitalism; social capital capitalism, 119, 120–​21, 120–​21n.3, 128–​29, 130–​33, 134–​35, 136–​37, 137n.48, 139–​40, 180n.31, 181–​82, 186, 187, 192, 196–​97, 202–​3, 207. See welfare state capitalism defined,  120–​21 Cassar, Alexandra, 62, 72, 122–​23 Catholic, 94, 98–​99n.44, 174, 197, 207–​8 Catholic church, 94, 94n.31, 97, 98, 174, 197 causal channels, 7–​8, 69–​70, 89, 92 Chambers, Clare, 111 Christian, 4, 26, 40–​41, 97, 112n.97 Christiano, Thomas, 265 citizen involvement, 72, 75, 241 civil society, 97–​98, 99, 105–​6, 107, 116–​17, 151n.34, 151–​52, 155

300 Index classical liberal, 158n.53 Clinton, Hillary, 1, 256 coercion, 104, 110, 135–​37, 148, 151–​54, 155, 158, 165n.68, 166–​67, 176, 179n.29, 188, 198–​99, 198–​99n.80, 201, 202, 206, 222, 223–​24, 228, 250 justification of, 104, 135–​36, 152–​53, 154, 155–​56, 164, 166–​67, 179, 182, 185, 188, 198, 207, 226, 264 legal, 110 POD-​based,  196–​97 political,  136–​37 redistributive, 133–​34, 163, 179–​80, 183,  185–​87 state, 101, 201 unjustified, 135, 148, 164, 166, 222 workplace, 173, 197–​201, 205 coercion-​increasing policies, 165, 166, 177, 185, 201, 222, 223–​24 coercion-​reducing policies, 166–​67, 167n.71, 168, 171, 176, 181, 182, 185, 264, 278 combinatorial complexity, 165–​66 communism, 7, 63, 204–​5 communist, 63, 186, 204–​5 congress, 4, 14, 53–​54, 220, 224, 225–​26, 232–​33 Congressional Budget Office (CBO), 160–​61 consensus, 6–​7, 41, 60, 69–​70, 127, 134, 160–​62, 170–​71, 222–​23, 272, 274 conservative, 4, 5n.27, 12, 18–​19, 53n.15, 64n.64, 147, 148, 151–​52, 154n.44, 159–​60, 165, 165n.68, 200, 215, 221, 224 constitutional rules, 31, 38, 39, 48, 115, 127–​28, 140, 163–​64, 176, 177, 182–​83, 210, 211, 212, 218, 219, 219n.34, 219n.36, 220, 221, 235, 272–​73, 274 publicly justified, 39, 89, 210, 216–​17, 219, 221, 274 constitutional stability, 38 contractarian, 17, 113n.99, 117, 117n.113, 130. See social contract pluralist, 113, 116n.108 contracts, 101, 108, 127 conventionalist challenge, 113–​14, 115, 127–​29 convergence, 41, 111–​12, 112n.96 corruption, 2, 9n.52, 38, 51, 53, 55, 61, 73, 78n.140, 79, 85, 86, 123–​24, 139, 142, 144n.10, 146n.21, 163–​64n.66, 168, 195–​ 96, 213–​14, 222, 230–​31, 235, 240, 242. See inherent-​corruption theories control of, 73, 123n.13, 144, 212 in democracies, 79, 86, 195–​96, 240 and inequality, 72, 141–​42 in legal systems, 84, 204–​5

and legal trust, 68–​69 limiting, 211, 212, 213–​14, 278 perceived, 8, 85, 123, 141–​42, 144, 184, 194, 212 political,  141–​42 and political trust, 8, 8n.111, 72–​73, 142, 145, 186n.48, 194, 211–​12, 213–​14, 222, 244 public, 62, 79, 141, 144–​45, 184, 186, 194, 204, 222 and the rule of law, 61–​63 and social trust, 8, 69, 72, 84, 123n.13, 123–​ 24, 144, 186n.48, 194, 211–​12, 222 and the welfare state, 139, 144–​47 crowding out, 36–​37n.32, 151–​52, 153 culturalist approach, 64 cultural-​persistence hypothesis, 65   Dalton, Russell, 79–​80, 79n.147, 242–​43 decision procedures, 96–​97, 217–​18, 219 defeaters, 135, 148, 170–​71, 170–​71n.81, 174, 175, 196–​97, 205, 208, 250 deliberation, 106, 229, 230, 231, 232, 233, 234, 248–​52, 250n.65, 265, 269–​70, 271, 272–​ 73n.128, 273. See deliberative democracy high-​quality, 229, 233, 234, 248–​49, 250, 267 microcosmic, 164, 230, 251, 257–​58 public, 229, 230, 251, 269–​70, 272–​73 structured, 222–​23, 231, 234, 251, 272 deliberative democracy, 237, 251, 251n.70, 271–​73, 272, 272n.127, 273n.129, 274. See deliberation deliberative polls, 262 deliberative practices, 230 deliberators, 230–​31, 230n.74, 232, 233, 234, 234n.94, 250, 251, 266, 267 democracy, 2–​3, 10–​11, 10n.55, 13, 47–​48n.50, 51, 59–​60, 74, 77–​78, 79, 79n.147, 80, 85–​ 86, 93, 104–​6, 137n.48, 144–​45, 173, 176, 177–​78, 177–​78n.21, 214–​15, 238, 240, 242–​48, 246–​47n.54, 247n.55, 251, 254, 257–​58, 263, 268 aggregative (see aggregative democracy ) agonistic (see agonistic democracy) and authoritarianism, 75 deliberative (see deliberative democracy) electoral, 252, 257, 277 and falling trust, 10, 11, 75, 243 liberal, 10, 10n.55, 18–​19, 20, 109, 273–​74,  280 and political trust, 1, 2, 7, 18, 71, 75, 80, 81, 85–​86, 93, 237–​39,  242–​45 priority of, 104–​6, 128 process (see process democracy)

Index  301 property-​owning democracy (see property-​ owning democracies [PODs]) publicly justified, 239–​40, 245–​46, 260, 264,  274–​75 representativeness of, 77–​78, 188, 257 responsiveness, 181, 252, 254, 257, 263, 264, 267 satisfaction with, 58–​59, 74–​75, 141, 144–​45,  243 and social trust, 2, 18, 237–​40 trust in, 1–​2, 53, 70–​71, 75, 80, 237–​39, 241, 242–​45, 254, 257, 274–​75 workplace, 191n.64, 193 democratic constitutionalism, 18, 49, 236 public justification of, 49, 211 relation to trust, 216, 236, 277 democratic elections, 47–​48, 210, 255–​56,  274–​75 Democratic Party, 7, 11–​12, 53–​54 democratic procedures, 137, 249, 269 democratic process, 18, 80, 173–​82, 184, 185, 207, 210–​11, 237, 238, 250, 251–​52, 254, 265, 266 Deneen, Patrick, 10–​11, 10n.55, 279 deservingness, 133–​34, 150, 153 discrimination objection, 110–​13 distrust, 1–​2, 16, 21, 25n.9, 50n.4, 53–​54, 53n.15, 63, 71, 75, 78, 82, 85–​86, 90n.11, 103–​4, 138, 142, 143, 183, 184–​85, 238–​ 39n.11, 240, 243, 266, 277, 278–​79n.2. See distrust-​divergence hypothesis of media, 76–​77 of officials, 71, 81, 238 political, 43–​44, 71, 256, 278–​79 propensity to, 11, 35, 89 distrust-​divergence hypothesis, 3–​10, 11, 13, 16, 18, 78, 86–​87, 94, 126, 147, 208–​9, 216, 245, 277 divergence, 3–​4, 11–​12, 13–​14, 69–​70, 173–​74, 178–​79, 230n.74, 241–​42, 241n.26, 256, 263n.99, 277–​78, 278n.1 direct theories of, 11–​12 group theories of, 13–​15 partisan, 11–​12, 13, 15, 18, 78, 86–​87, 231, 262,  277–​78 publicly justified, 16 and trust, 11–​12, 16–​17, 69–​70, 78, 256, 277 diverse social order, 85–​87 diversity, 5, 11, 21, 59–​66, 71, 78, 78n.134, 85, 173, 184, 192, 197, 231, 241, 248, 269, 271   economic equality, 2, 43–​44, 61, 69, 173, 174, 179, 183, 184–​85

economic growth, 2, 17–​18, 67–​68, 75, 85–​86, 119, 130–​33, 131n.34, 146, 162–​63, 169, 186, 198, 198n.79, 198n.4, 201, 211, 213–​ 14n.12, 222, 226–​27, 235–​36, 235–​36n.97, 278, 279 and political trust, 10, 17–​18, 125 and social trust, 17–​18, 67, 71, 124–​25 economic inequality, 5, 9, 11–​12, 13, 14, 61, 69–​ 70, 72, 78, 122, 131, 132, 140, 145, 175, 179, 183, 184, 185, 277 economic performance, 67, 67n.76, 71, 74, 75, 85–​86, 124, 125, 125n.23, 141, 145, 186n.48, 211, 212n.4, 213–​14, 242–​43, 244 economic regime, 119, 152–​53, 173, 186 economic security, 9, 18, 52–​53, 78, 139, 206–​7,  235–​36 education, 46, 55, 67, 68, 69, 77, 77n.130, 92–​93, 124–​25, 142–​43n.7, 155, 167–​68, 178, 187–​ 88, 189–​90, 242–​43, 263 efficiency, 36–​37n.33, 57, 67, 104–​5, 130, 151, 163–​64, 202, 205 economic, 67, 120, 164, 171, 188–​89 egalitarian, 129, 133n.39, 135, 148, 148n.24, 174, 177, 205, 207–​9, 214–​15, 245–​46, 272 conceptions of justice, 139–​40, 148, 196–​97,  207 egalitarianism, 174, 174n.5 luck, 148, 174, 174n.3 egalitarians, 18–​19, 20, 114–​15, 129, 140, 148, 152–​54, 165, 174, 182 elections, 2–​3, 5–​6, 11–​12, 16–​17, 18, 47–​48, 209, 210, 237–​45, 247, 249, 254, 255–​56, 257–​58, 259–​60, 268–​69, 271,  274–​75 and political trust, 18, 74, 86, 145, 175, 194–​95,  237–​45 electoral rights, 237–​38, 241, 245, 246, 247–​48, 256, 259–​60, 271, 274–​75 public justification of, 18, 210, 237, 245, 247, 256, 259–​60,  274–​75 Environmental Protection Agency (EPA), 52, 53–​54, 53n.15 equality, 11–​12, 13, 81, 111–​12, 116, 133n.39, 181, 221, 243, 248–​52 democratic, 176n.12 economic (see economic equality) political, 249, 250, 251, 265, 271 publicly justification of, 250 and trust, 184 value of, 248, 249, 250, 266 (see participation) equilibrium, 115, 158–​59n.55, 201 high-​trust, 31, 44, 69 low–​trust, 31, 69

302 Index evaluative pluralism, 21, 31–​32, 59, 127, 167n.72,  203–​4   feasibility, 15, 32–​33, 34, 44 federalism, 16–​17, 109n.86, 165, 213–​14n.14, 274 feminism,  110–​13 firms, 100, 101, 115, 116, 126–​27, 128–​29, 130, 131, 135, 137, 177, 181–​82, 189, 191–​92, 193–​94, 203, 204, 205, 206–​7, 224, 225, 261 private, 104–​5, 181 publicly traded, 100, 100n.51 First Amendment, 220–​21. See also rights Fishkin, James, 230, 231, 231n.78, 234, 248, 249, 250, 269 Foa, Roberto, 80 freedom, 7, 17, 38–​39n.39, 40–​41, 45–​46, 47, 73, 86, 87, 88–​89, 101–​4, 105–​6, 107–​8, 111, 112, 116–​18, 122–​23, 129, 144, 195–​ 96n.73, 198n.79, 202–​3, 206–​7, 208, 212, 214, 220, 241, 277 of association, 16, 31, 73, 88–​89, 92–​118, 93n.26, 109n.86, 110, 110n.88, 197–​98, 212, 241, 277 of contract, 120–​21, 206 of expression, 73, 241 of press, 16–​17, 220 of religion, 16–​17, 40–​41, 94n.31 of speech, 16–​17, 45–​46, 220 free market, 115, 120–​21, 125, 192–​93 Friedman, Benjamin, 132, 132n.37 Friedman, Milton, 147, 170–​71, 247n.55 Fuller, Lon, 95–​96, 95–​96n.34 functional independence, 156, 157, 158n.53, 158 functionally independent policy units, 158   Gaus, Gerald, 35n.28, 36–​37n.34, 41–​42n.44, 47–​48n.49, 104, 111–​12n.95, 117, 127, 221n.37, 271n.122 Gilens, Martin, 175, 176, 176n.12, 177n.19, 182, 185, 237, 257, 263n.99, 263–​64 Glaeser, Edward, 66 governance, 10, 18, 71, 72, 73, 74, 78, 86, 197, 205, 206, 212–​13, 214, 216, 237, 239, 239n.12, 241, 244, 278 Great Enrichment, 131, 131n.34 Green, Donald, 91 Green, Seth, 91 Grosjean, Pauline, 62, 69 gross national product (GDP), 68, 124, 144, 162 group theories, 10, 13, 14, 15, 277

Harcourt, Bernard, 115, 121 Hayek, F. A., 117, 117n.113, 130, 134, 170–​71, 193–​94, 202n.90, 202–​3, 203nn.94–​97, 204, 247n.55 Head Start, 139, 167–​68 Hetherington, Marc, 81, 241 high-​trust orders,  123–​24 Hooghe, Mark, 240   idealization, 27–​29, 31, 41–​42, 43, 113, 131–​32, 197n.75, 234, 253, 254–​55, 262,  263–​64 identities, 3–​4, 14, 92, 99–​100, 230–​31, 258 illusion of culpable dissent, 22 immigrants, 7, 16–​17, 65, 279 immigration, 16–​17, 216 incentive, 59, 62, 91, 98–​99, 100, 123, 125–​26, 130, 142, 150, 168, 169, 181, 188–​91, 194–​95, 196, 199, 202, 206, 222–​23, 225, 229–​30, 234, 247, 250 inefficiency, 143, 149, 153 inequality, 5, 7–​8, 10, 11–​12, 13, 14, 43–​44, 61, 62, 69, 72, 74–​75, 78, 79, 107, 119n.6, 122, 139, 141–​42, 145, 169, 173–​74, 175, 176, 177, 181, 182, 183, 190, 193–​94, 205, 227–​28, 263n.100, 264, 264n.100, 265, 277, 278 corruption and, 72, 145, 176 and the democratic process, 173–​82 income, 5, 13, 173–​82, 183 and public justification, 175 and trust, 9, 78, 145, 182–​85 Inglehart, Ronald, 80, 243 inherent-​corruption theories, 10, 11, 15. See corruption institutionalist approach, 64 institutional rules, 34–​35, 43, 52–​53, 53n.15, 54n.17 institutions, 5, 11, 13, 16, 16–​1 7n.67, 18, 20–​2 1, 22, 31, 32–​3 3, 34–​3 5, 36, 41–​4 2n.44, 46, 51, 52n.13, 52–​5 4, 53n.15, 58–​5 9, 62, 65, 67, 71, 76, 77, 84, 85–​8 6, 92–​9 3, 94n.28, 97–​9 8, 98–​9 9n.44, 101–​2 , 104–​5 , 109, 112, 116–​1 7, 121, 124–​2 5, 128–​2 9, 137, 144, 151, 163–​6 4, 166, 176, 178–​7 9, 186–​8 7, 200, 203, 213–​1 4, 216n.27, 227, 240, 241, 244, 251, 271 adjudicative, 36 commercial, 84, 114 democratic, 7, 40–​41n.42, 67, 74, 81, 105–​6, 106n.69, 124–​25, 173, 174–​75, 174–​75n.6,

Index  303 177–​78, 179–​80, 204, 237, 244–​45, 260, 272, 273 economic, 59–​60, 62, 66, 85, 109, 137, 138, 173, 198–​99,  207–​9 electoral, 86, 178–​79, 237–​39, 240, 244, 245, 246–​47, 248, 249 extractive, 183–​84, 260n.94 financial, 228 for-​profit,  99–​100 government,  55–​56 inclusive,  183–​84 judicial, 85, 115n.104, 144 law-​applying,  36 legal, 31, 51, 62, 63, 95, 123, 125   238–​39 liberal (see liberal institutions) non-​profit,  89 order, 51, 238–​39 political, 7, 20–​21, 31, 51, 52, 54, 55, 62, 63, 66, 67, 80, 89–​90, 95, 109, 113, 124–​25, 141, 144–​45, 212–​13, 214–​15, 238–​39, 243, 254–​55, 265–​66,  267 private, 151 publicly justified, 43, 44, 144, 198–​99, 216 real-​world,  34–​35 Rechtsstaat, 244 religious, 97 social, 20, 24, 46, 52, 59–​60, 66, 68, 98, 109, 148n.26 and trust, 16, 20–​21, 23, 31, 43, 44, 51, 52–​53, 66, 71, 86, 92–​93, 94n.28, 122–​23, 125, 144, 207–​9, 214–​15, 216n.27, 216–​17, 238–​39, 242, 243, 244–​45 trust in, 51, 52–​53, 55–​56, 81, 216n.27,  238–​39 welfare state (see welfare state institutions) instrumentalist,  246–​47 intellectual property laws, 177, 184 intergroup contact theory, 90–​91   Jakobsen, Tor George, 213–​14n.14, 241 Jankauskas, Vidmantas, 67, 67n.76 Johnson, James, 105, 128 Jones, Charles, 50n.4, 67 Jordahl, Henrik, 122–​23, 142n.5 justification, 16, 38n.36, 43, 87, 101–​2, 113n.100, 133, 139–​40, 142–​43n.7, 205, 216–​17, 259–​60, 266n.104, 272–​73n.128. See public justification   Knight, Jack, 105, 128

Laborde, Cecile, 105 law, 36–​37, 41, 90, 95, 95–​96n.33, 98, 104, 108n.81, 108, 109, 110, 112, 116, 127–​28, 135, 146, 154n.45, 156, 216–​17, 217n.30, 218n.33, 219, 220, 224, 244, 262, 272–​73. See functional independence; laws; legal rules; legislation; rule of law administrative, 146, 193–​94 bad, 48, 219 constitutional, 9–​10, 39 defined, 36 emergence of, 117 enforcing, 51–​52, 136 of group polarization, 14 liability, 190 liberal-​democratic,  10 patent, 228 public justification of, 36, 37–​38, 37n.35, 38n.36, 262 state, 9–​10, 106, 127–​28 tort,  154–​55 unjustified, 38 welfare state of, 203 laws, 36–​37, 38, 57, 74, 94, 97, 137, 153, 156, 157, 158, 177, 184, 198–​200, 201, 207, 210, 217–​18, 221–​22, 234, 256, 260–​61, 262–​63, 270, 272–​73, 278. See law; legislation defeated, 219, 248–​49, 263 defined,  36–​37 individuation of, 158 property, 177, 184 publicly justified, 36–​37, 37n.37, 39, 111–​12, 137, 149, 156, 211, 219, 248–​49, 256, 259–​ 60, 263–​64, 274 rights-​protecting, 120–​21,  221 rights-​violating,  220 unjustified, 219 zoning, 176, 177, 177n.18, 178n.24, 184 left-​wing, 7, 47, 173–​74, 179–​80, 182, 191, 279 legal justification principle, 37n.37, 217 legal rules, 37, 38, 95–​96, 100, 104–​5, 115, 122, 129, 129n.41, 131–​32, 135, 154, 179–​80, 196, 201, 207, 217–​18, 219n.36 publicly justified, 39, 216–​17 legislation, 12, 109, 128, 129, 157, 160, 161, 175, 201, 210, 211, 217n.30, 221–​22, 223–​24, 225, 226–​27, 229, 230, 234, 235, 235n.96, 257, 266 legislative process, 210, 211, 223, 248–​49 Levy, Jacob T., 89, 107, 107n.75, 116–​18 liberal institutions, 10, 15–​16, 17, 35, 42, 43–​44, 144, 158n.53, 245, 277, 280

304 Index liberalism, 10–​11, 116–​18, 279, 280 pluralist,  116–​18 public-​reason, 117–​18, 161, 197, 263 rationalist,  116–​17 liberal regime, 15, 28–​29, 40–​41, 236 liberal rights, 9–​10, 11, 16–​17, 18, 22, 23, 30, 35, 40–​41, 42, 43, 48, 49, 85, 86, 87, 114, 118, 210–​11, 277, 278, 279, 280. See liberal rights practices derivation of, 40–​42 publicly justified, 48 and trust, 41, 42, 43, 44 liberal rights practices, 10, 17, 18, 22, 23, 35, 43, 44n.47, 49, 85, 277, 280 liberal socialism, 173, 186–​87, 202–​5, 207,  208–​9 liberal socialist, 186, 202 libertarian, 18–​19, 114, 134–​37, 138, 139, 140, 148, 149, 153–​54, 159–​60, 164, 165, 168, 170–​71, 177, 200, 207–​8, 226, 227, 227n.64, 243, 247 bias,  227–​29 property rights, 136, 138 libertarianism, 114, 134, 136–​37, 207 libertarians, 18–​19, 134–​35, 136, 138, 140, 148n.26, 149, 149n.36, 164, 165n.68, 167n.71, 177, 200, 207–​8 radical, 148, 165n.68 Lindsey, Brink, 176n.13, 227–​29 lobbying, 195, 223, 224, 225 Locke, John, 6, 64, 116–​17, 128–​29 Lomasky, Loren, 114n.103, 117, 117n.113,  268–​69   Mackie, Gerry, 223, 268, 268n.112 Manning, Colin, 156, 157 market economy, 17–​18, 119, 121, 123–​24, 126, 130, 131–​33, 138, 139, 152–​53, 155, 173,  207–​8 market orders, 123, 186, 208 markets, 17–​19, 86, 105, 115, 119, 121–​24, 125–​ 26, 127–​29, 130, 133–​34, 138, 140, 155, 165–​66, 187, 189, 192, 193–​94, 195–​96, 205, 208, 211, 215, 232, 254–​55, 260, 263–​ 64, 271, 278 creating trust for the right reasons, 138 defined, 121, 193–​94 free, 91, 115, 195–​96 prediction, 164, 211, 213, 215, 232, 235–​36, 254–​55, 263–​64, 271, 278 and trust, 121–​26, 277 and the welfare state, 139–​40 Marx, Karl, 132–​33, 136–​37 McCarty, Nolan 3–​4, 11–​12, 78

media, 5, 7, 11–​12, 13, 64, 73, 76, 77, 80, 97, 116–​ 17, 124–​25, 142–​43, 201, 232, 235, 240, 241 Mill, John Stuart, 1, 107, 131, 133, 256 minipublics, 211, 215, 229–​30, 232–​34, 235, 251, 254–​55, 263–​64, 269, 271, 273, 278 monarchy, 64 moral rules, 20–​21, 24–​25, 30, 35–​37, 49–​51, 52, 56, 82, 85, 94–​97, 98, 104, 108, 115, 118,  216–​17 publicly justified, 35–​39, 94–​95, 99, 104, 109, 118,  216–​17 publicly recognized, 6, 24, 109 and standing, 94–​95 and trust, 23–​26, 50, 98–​99 Mouffe, Chantal, 10–​11, 10n.55, 272n.127, 273–​ 74, 273–​74n.132 Must Politics Be War? (MPBW), 11, 15, 17, 18, 20–​24, 22–​23n.5, 24n.6, 26, 27n.11, 27n.12, 27–​28n.13, 28n.15, 32–​33, 36–​37nn.33–​34, 38n.37, 38–​39n.39, 40, 45, 47–​48n.49, 50n.6, 58n.40, 65n.66, 101–​2n.54, 102, 108n.84, 112n.19, 130n.36, 131–​32n.36, 137, 151n.34, 154n.42, 156, 163–​64, 210n.1, 217 n.29, 217–​ 18n.32, 219, 219n.35   Nagel, Thomas, 114 negative income tax, 147, 193 Nelson, James, 99–​100 Newton, Ken, 51, 214–​15, 238–​39 nondemocratic, 80, 204–​5, 238, 244–​46, 247, 260 norm erosion, 2–​3, 8, 20   O’Neill, Martin, 186–​87n.50, 187–​88, 190, 194, 195 objections to the principle of social insurance,  149–​53 objections to the public justification model of representation,  254–​55 officials, 9n.52, 53–​54, 58–​59, 71, 72–​73, 75, 80, 81, 97, 135, 142–​45, 150n.29, 163–​64, 211, 213, 218, 222, 223, 229, 236, 238–​39, 252, 268n.112 democratic, 53, 202–​3, 254 elected, 51–​52, 72–​73, 81, 202–​3, 245–​46, 252, 256, 267n.107 government, 47–​48, 51–​53, 72–​73, 75, 93, 144–​45, 210, 212, 224, 260 legal, 37, 68–​69, 144 legislative,  230–​31 local, 213–​14n.14 political, 6, 38, 48, 55–​56, 72–​73, 81, 81n.56, 135, 144, 211, 214–​15, 215n.18, 218, 222–​23, 229, 235, 237–​38, 248–​49, 252–​53, 257, 260, 263–​71 selection of, 237, 251, 252, 264–​71, 273

Index  305 Olson, Mancur, 163–​64, 226–​27 open society, 22 organizations, 76, 90, 95, 97, 100, 102–​3, 109, 111–​12,  151–​52 bureaucratic, 194 charitable, 97 civic, 30n.31 commercial, 89, 91n.110, 97, 99–​101, 108, 110–​11,  114–​15 economic, 126–​27 (see unions) ethnic, 90 for-​profit,  101 fraternal, 106 governmental, 112 hierarchical, 107, 108 media, 97 nonprofit, 100 political, 101 private, 113, 151–​52 professional, 98, 101 religious, 97, 98, 112 service, 46 social, 98   Page, Benjamin, 175, 176, 182, 186, 244–​45 Pareto principle, 272, 272n.126 participation, 77–​7 8, 122–​2 3, 186n.48, 202, 229, 231, 237, 240, 248, 249, 250–​5 2, 250n.65, 265–​6 6, 267, 269–​7 0,  271 association, 77–​78, 86, 88, 89, 91, 93 and representation and public justification,  252–​56 and the value of equality, 248–​52 Paxton, Pamela, 90, 91, 93, 110 personal economic improvement, 71 Pettigrew, Thomas, 90, 91, 92 Pettit, Philip, 36n.31, 114 Piketty, Thomas, 173–​7 4nn.1–​2 , 180–​8 1,  182 pluralist contractarianism, 113 polarization, 2, 3–​4, 9n.47, 13n.60, 14, 14n.64, 230, 231, 231n.78 defined, 3 issue-​based, 3–​4,  278 and trust, 3–​10 (see distrust-​divergence hypothesis) policy epistemology, 131–​32, 139, 155–​56, 155–​ 56n.49, 158–​63, 164–​65, 166, 167, 170–​71, 173, 176, 179, 226n.62, 226 idealization,  170–​71 and minipublics, 229–​34 and policy individuation, 155–​63 and public justification, 158, 167, 188

standards of, 155–​56, 158–​59, 161–​62, 164, 167, 168, 180, 182, 183, 185, 196, 199, 207, 222, 223, 232, 235 policy experimentation, 165, 165n.68, 166, 166n.70 policy individuation, 155–​58 policymaking, 18, 161, 166, 215, 221–​22 political equality, 249, 250, 251, 265, 271 political justification principle, 217 political obligation, 211, 218, 219 a theory of, 216–​20 political trust, 1, 2–​3, 5, 7, 8, 9–​10, 14, 15, 16, 17, 18, 20–​21, 31, 39, 41, 43, 49, 51–​52, 55–​56, 58–​59, 65–​66, 70, 71, 72, 73, 74, 75, 76, 77–​ 78, 79–​80, 85–​87, 89, 91–​92, 93, 118, 119, 125, 126, 132, 138, 141, 142–​43n.7, 145, 147, 184, 186, 194, 204–​5, 212, 213–​14, 215, 216, 222, 237, 238–​39, 240, 241–​43, 253, 267, 278, 279 causal relationship with social trust, 6, 9–​10, 79, 211, 214–​15, 238–​39, 278 causes of, 70–​79 consequences of, 79–​81 and corruption, 8, 8n.111, 72–​73, 142, 145, 186n.48, 194, 211–​12, 213–​14, 222, 244 creating, 16, 17, 18, 30, 43–​44, 125, 138, 141, 144, 146–​47, 211, 212, 213–​15, 214–​15n.18, 216, 236, 241–​42, 244, 245, 277, 279 decreasing, 2, 7, 9, 10, 14–​15, 16–​17, 54, 71–​72, 74–​75, 78, 79, 80, 141, 144–​45, 181, 186n.48, 196, 204, 227–​28, 239, 242–​43 and democracy, 1, 2, 7, 18, 71, 75, 80, 81, 85–​ 86, 93, 237–​39, 242–​45 described, 20–​21, 30, 51–​52, 53–​54, 54n.16 determinants of, 78, 141, 212, 213–​14, 245 and divergence, 5, 9 and elections, 237–​45 increasing, 16–​17, 71, 72–​73, 74–​75, 77, 123–​24, 125, 141, 147, 211, 212–​13, 216, 237–​38, 241–​42, 244, 245, 267, 277 measurement of, 6, 54–​59, 72–​73n.111, 144–​45,  238 promoting, 17, 18, 75, 86, 118, 119, 212,  241–​42 rational basis for, 121–​22, 139, 256 real, 44, 211, 274–​75 in the real world, 16, 119, 237–​38, 245 for the right reasons, 30, 39, 125–​26, 227, 236 sustaining, 16–​17, 18, 30, 39, 49, 85, 139, 141, 211, 213, 227, 237 undermining, 9, 35, 37, 54, 78, 85, 118, 142, 143, 163–​64, 204 and the welfare state, 140, 141, 146

306 Index populism, 7, 279 power-​and-​status veil, 40,  41–​42 prediction markets, 164, 165–​66, 211, 232, 235, 254–​55, 263–​64, 271, 278 predistribution, 185, 187, 190, 190n.61 preferences, 6, 7, 111, 176, 215, 221, 224, 230, 234n.94, 241n.27, 249, 251, 253, 254–​55, 259, 260–​61, 262, 263, 264, 267–​68, 270n.122, 271, 273 of citizens, 247, 253, 254n.80, 254–​55, 257–​ 58, 261, 267, 270 policy, 177–​78n.23, 241n.27, 257–​58, 259, 263–​64,  268–​69 presumption on behalf of civil society, 151 primary rights, 40, 41, 47–​48, 119, 151, 175, 208, 210, 210n.1, 211, 220, 221–​22, 235, 245–​47, 259–​60, 265, 271 agency,  45–​46 associational, 17, 45–​46, 88, 102–​3 convergence on, 41 jurisdictional, 46–​47, 119, 127, 260 procedural rights, 245–​48, 259–​60 protection of, 121–​22, 179, 210, 211, 212, 220, 221 publicly justified, 39, 45–​46, 48, 102–​3 and trust, 39, 210, 211–​16 principle of social insurance, 149, 153, 161, 168, 203, 205n.98, 207 objections to, 149–​53 public justification of, 148–​49 principle of sustainable improvements, 131–​32, 163, 188, 196, 198n.79, 205, 222 private property, 17–​18, 47, 113, 119, 120, 121–​ 22, 126–​30, 132–​34, 135, 137, 138, 139–​40, 141–​42, 148, 153, 168, 200, 206, 210–​11, 260, 277 public justification of, 126–​29, 131n.34, 132,  133–​34 private property rights, 17–​18, 47, 113, 114, 119, 120, 126–​29, 132–​33, 137, 138, 139–​40, 168, 206, 277 as jurisdictional rights, 119, 260 as primary rights, 121–​22, 126–​29 protection of, 135, 141–​42, 200 private tyranny, 107–​8, 107n.75, 116–​17 problem of political war, 20–​22, 34, 35, 277 procedural fairness, 18, 72–​73, 211, 212, 213–​ 14, 241, 242–​43, 246–​47n.54, 278 proceduralist,  246–​47 process democracy, 237, 252, 271–​73, 274 productive policy, 211, 222–​23, 227, 229,  234–​35 progressives, 20, 53n.15, 152–​54, 165n.68

property ownership, 123, 128–​29 property-​owning democracies (PODs), 186–​88, 186–​87n.49, 192–​96, 194–​95n.70, 195n.71, 198–​200 inability to publicly justify, 196–​97 incentive problems in, 189–​92 inefficiency of, 188–​96, 188n.58 property rights, 36, 47, 73, 113–114, 115, 118, 119, 120–​24, 125–​30, 126–​27n.26, 131n.34, 132–​33, 134, 134–​35n.44, 135n.46, 135–​ 36, 137, 138, 139–​40, 141–​42, 148, 153, 168, 179n.29, 182, 183, 187, 198, 200–​1, 206, 208, 277, 278 extensive, 114, 128–​29, 130, 137, 138, 198 intellectual, 134–​35n.44, 177 legal, 62, 122–​23, 186, 204, 211, 212, 242 and markets, 115, 119–​20, 121, 125–​26 as prepolitical, 113, 127–​28 as primary rights, 126–​27, 137, 182, 208 private (see private property rights) protections, 123–​24, 129, 135, 166, 201 publicly justified, 123, 127–​28, 132, 135, 201 and the welfare state, 139–​40 protective policy, 211, 220, 221–​22n.41 public goods, 139, 147, 153, 155–​56, 163–​64, 167, 171, 185, 187, 193, 211, 245–​46,  272–​73 justifying the provision of, 153–​55 public justification, 15–​16, 30, 35–​36, 40, 43, 94–​95, 98, 101–​2n.54, 103, 108, 111–​12, 113, 116, 126, 131–​32n.36, 132–​33, 148, 153, 154n.42, 156, 160, 175, 182–​83, 185, 198, 198n.79, 205, 207, 210, 210n.1, 216–​17, 221, 224, 227, 237, 239–​40n.13, 250, 252, 253, 262, 272–​73. See public justification principle and coercion, 36, 148, 153–​54n.41, 154, 179n.29, 185, 198, 202 consensus and convergence views, 40, 111–​12 of constitutional rules (see constitutional rules) defined, 30, 35 of electoral rights (see electoral rights) of freedom of association, 101–​4 idealization, 30, 35, 113, 262 individuation,  156–​58 of law, 36, 37, 38n.35, 149, 262 of legal rules, 140, 216–​17 of moral rules, 35–​36, 94–​95, 216–​17 of policy, 139, 149, 158, 167, 198 of the principle of social insurance, 148–​49 of private property (see private property) of process democracy, 271, 274

Index  307 relation to legal systems and constitutions,  35–​39 relation to participation and representation,  252–​56 relation to trust and trustworthiness, 30, 36, 43 of the welfare state, 139–​40, 167 public justification principle, 35, 217 public reason, 36, 47–​48n.49, 95, 99, 108, 116, 117–​18, 127, 130, 156, 158n.53, 173, 175, 179n.29, 206–​7, 213–​14n.14, 227–​28, 254n.80, 257, 271n.123, 272 Achen and Bartels against, 257–​59 consensus and convergence views (see public reason: consensus and convergence) and democracy, 263, 272 idealization (see idealization) incompatibility with PODs, 197 (see property-​owning democracy [PODs]) individuation,  156–​58 liberal(s), 109, 111–​12, 157, 174, 210–​11,  263–​64 liberalism, 117, 197, 263 literature, 47–​48n.49, 156 and moral rules, 99, 109 pluralist,  116–​18 publicly justified, 153 and trust, 173 and the welfare state, 140   Rand, Ayn, 202 Rawls, John, 38–​3 9, 40, 40n.40, 41–​ 42n.44, 43n.46, 105–​6 n.68, 139–​4 0, 155–​5 6, 181–​8 2, 186–​8 7, 186–​8 7n.50, 193–​9 4, 195, 196–​9 7n.74, 202n.89, 202n.90, 202–​3 , 204n.102, 251n.70,  269–​7 0 Rawlsian, 38–​39, 101–​2n.55, 186–​87n.50, 187n.51, 188n.58, 193–​94, 196–​97, 202–​3 form of welfare state capitalism, 187 ideal theory, 186 real trust, 16, 30, 34, 43, 88–​89, 92, 103–​4, 125–​ 26, 172, 210, 212, 236. See political trust; social trust; trust for the right reasons and associations, 89–​94 creating, 44, 92, 118, 139, 172 relevance of, 42–​44 for the right reasons, 16, 48, 103–​4 sustaining, 94, 212 and the welfare state, 140–​43 redistribution, 7–​8, 43–​44, 61, 81, 134, 134n.42, 141–​42n.4, 142–​43, 146, 148, 151, 168, 173, 178, 179–​81, 182, 183n.41, 184–​85,

187, 188, 189, 189–​90n.61, 190, 192, 228–​29,  264 coercion-​reducing,  185 coercive, 148, 174, 179–​80, 182, 188 income, 11–​12, 69–​70,  181–​82 preemptive, 185, 190 (see predistribution) and public justification, 146, 179–​80, 182–​83,  207 and public reason, 153, 207 and trust, 61, 182–​83, 184–​85 of wealth, 132n.37, 148, 180 reforms, 6–​7, 69–​70, 171, 206–​7, 214, 224–​25, 239, 280 campaign finance, 5–​6, 177n.19 economic, 6, 161–​62, 214 electoral, 278, 278n.1 institutional,  241–​42 legislative,  263–​64 and liberal rights, 9–​10, 18 and trust, 6–​7, 239 regulations, 47, 139, 140, 146, 153–​54, 155–​ 56, 158, 159–​60, 161, 162, 163–​64, 171, 177, 188, 192, 193–​94, 207, 228–​29, 241n.27,  244–​45 economic, 47, 162 public justification of, 153–​55, 158–​63, 164, 207 regulatory capture, 154–​55, 163–​64, 195, 195nn.71–​72 rent seekers, 194, 225–​27, 279 rent-​seeking, 110n.88, 135–​36n.65, 168n.74, 168, 171, 177–​78, 182, 185, 189n.59, 194–​ 96, 206, 212, 223n.44, 226–​27, 228–​29, 260 and corruption, 38, 85, 146, 163–​64n.66, 168, 186, 204, 222, 223, 235 discouraging,  223–​29 insulating bureaucracies from, 163–​67 in PODs, 188–​89, 195–​96 (see property-​ owning democracies) reducing or restricting, 211, 212, 222, 225, 226–​27, 228–​29,  278 undermining trust, 163–​64 unjustified, 185, 222 in welfare states, 139, 195–​96 representation, 249, 253, 253nn.78–​79 and participation and public justification,  252–​56 proportional, 11–​12n.57, 271, 271n.123 public justification account of, 252, 253, 254, 257 Thrasher’s “thin” account of, 252–​53 Republicans, 1, 4, 11–​12, 55–​56, 81n.157 responsibility, 33, 33n.24, 100, 206

308 Index responsiveness, 174–​75, 176n.12, 185n.46, 185, 221, 237, 255–​56, 263n.99, 263–​64, 266 democratic, 178–​79, 185, 263 of democratic institutions, 174–​75, 260 empirical challenges to democratic, 257–​64 lack of, 175, 178–​79 and representation, 255–​56 (see representation) rights, 16, 41–​42, 44–​45, 63, 102–​3, 108, 114, 119, 120–​21, 126, 127–​28, 129, 134–​35, 136–​37, 140, 156, 176, 198, 205, 206, 210, 212, 220, 221, 245–​47, 245–​46n.52, 248 agency (see primary rights) associational rights (see primary rights) basic, 2, 9–​10, 22, 38, 39, 102, 108, 126–​27, 211, 220, 266 capital, 119, 187–​88, 187–​88n.56, 194, 195, 222 claim-​, 88, 120, 126 constitutional, 48, 63 democratic, 185, 210, 245–​46 economic,  17–​18 electoral (see electoral rights) equal, 13, 41, 44–​45, 63–​221, 238–​39, 251, 265 First Amendment, 220 individual, 46, 85–​86 jurisdictional (see primary rights) justified,  45–​46 liberal (see liberal rights) minority, 80, 243 natural, 114, 153 political, 47–​48, 175, 176, 177, 179, 266 primary (see primary rights) procedural, 45, 47–​48, 245–​48, 249, 259–​60 property (see private property rights; property rights) publicly justified, 47–​48, 101–​2, 136 schemes, 40–​42,  102–​3 and trustworthiness, 44 unequal, 246, 248 voting (see electoral rights) welfare, 17–​18, 46, 138, 145–​46, 149, 171, 212 women’s, 113 rights practices, 9–​10, 17, 18, 22, 23, 35, 41, 43, 44, 45, 48, 49, 59, 85, 127, 244, 277, 280. See liberal rights practices trust through, 44–​48 right-​wing, 7, 7n.42, 111n.97, 279 Riker, William, 267–​68, 268n.112, 270, 272n.126 Robbins, Blaine, 122–​23

Rosenblum, Nancy, 89–​90, 89–​90n.4, 91, 107, 109 Rothstein, Bo, 51–​52, 67, 81, 81n.152, 238–​39 rule compliance, 31 rule of law, 18, 73, 86, 122–​23, 144, 146, 202–​3, 212, 217n.30, 214–​15, 244n.42 and corruption, 61–​63 rules, 15, 20–​21, 28–​29, 31, 31n.18, 37n.37, 40, 41, 43, 47–​48n.49, 73, 84, 95–​96, 97, 98–​99n.44, 105n.62, 105–​6, 115, 117, 132, 145–​46, 157, 201, 204n.101, 212, 213, 217n.30, 221, 226–​27, 260 constitutional (see constitutional rules) historic preservation, 171 institutional, 34–​35, 43, 53, 53n.15, 54n.17 justified, 28–​29, 132, 221, 235 legal (see legal rules) moral (see moral rules) nonmoral,  96–​97 political, 35 procedural, 54, 136 property, 114–​15, 126–​27, 133–​34, 136, 137, 153, 176 publicly recognized, 72–​73 social, 31–​61, 31n.18, 115 social choice, 271 supermajority, 220, 221, 225, 227, 235–​36, 235n.96 taxpayer-​funded, legislative, 115 and trust, 31–​32, 41, 43, 84 unjustified, 104, 108 voting, 235, 254, 270–​71, 273 satisfaction, 58–​59, 80, 136, 157, 212, 238–​39 with democracy, 51, 58–​59, 74–​75, 141, 144–​ 45, 238, 243 measures, 58–​59, 238 and trust, 58–​59, 238 Scanlon, Thomas, 114, 114n.102 schooling, 69, 112n.97, 148n.27, 166–​67 segregation, 13, 60, 86 self-​interest, 23–​24, 24n.7, 41–​42, 57, 66, 73, 82–​84, 82–​83n.160, 85, 92–​93, 146, 222, 260 Singer, Abraham, 205, 206–​7 social capital, 66, 67, 67n.73, 78n.134, 80 social contract, 16, 28–​29, 40, 40n.40, 43, 43n.46, 113n.99, 116–​18, 206 social insurance, 22, 47, 74–​75, 125, 141, 142–​ 43n.7, 147–​48, 149–​50, 151, 153, 161, 168, 168n.76, 188, 203, 207, 212n.4, 221–​ 22, 235–​36, 277. See Bolsa Familia (BF); Head Start; principle of social insurance described,  94–​149

Index  309 and markets, 86 provision of, 139–​40, 187, 198–​99 public justification of, 148–​49, 167–​71 and trust, 142–​43, 277 and the welfare state, 153 socialist, 7, 47, 132–​33, 137, 138, 186, 187, 191, 202n.89, 202–​3, 204, 205, 207–​8 order, 202–​3, 204 socialists, 126–​27n.26, 132–​33, 137, 137n.48, 138, 186, 187, 202, 205, 207–​8 social media, 5, 77 social norms, 10, 22, 24, 31n.18, 36n.31, 36–​ 37n.34, 52, 63, 65, 106, 111–​12, 112n.96, 195–​96, 205, 219, 219n.34, 267 social rank, 178, 248 social security, 142–​43, 142–​43nn.7–​9, 146 social trust, 6–​9, 7n.42, 10, 11, 14, 23–​24, 25, 30, 36, 36–​37n.32, 37, 40n.41, 42, 43–​44, 49–​ 52, 50n.6, 53–​54, 55, 55n.22, 58, 58n.40, 59–​63, 62n.55, 62n.57, 64–​66, 65n.66, 65–​ 66n.71, 67n.73, 67, 68–​71, 76, 79, 86–​87, 89–​90, 91–​93, 122–​23, 123n.13, 125, 132, 132n.37, 137, 140–​41, 142, 142–​43n.7, 144, 147, 173, 179–​80, 182, 183, 184–​85, 186, 186n.48, 204–​5, 208, 211, 212, 214–​15, 238–​40, 243, 244–​45, 256, 267, 274–​75 causal relationship with political trust, 6, 9–​ 10, 79, 211, 214–​15, 238–​39, 278 causes of, 10, 41, 49, 62n.55, 124, 184–​85, 214–​15,  244 consequences of, 49, 66, 79–​81 and corruption, 8, 69, 72, 84, 123n.13, 123–​ 24, 144, 186n.48, 194, 211–​12, 222 creating, 17–​18, 31, 44, 50, 82, 86–​87, 122, 126, 138, 144, 147, 182, 211, 212, 214–​15, 214–​15n.18, 236, 237, 245, 277 decreasing, 2, 6, 7, 8–​9, 10, 11, 13, 14–​15, 60, 62, 69–​70, 75, 86–​87, 122, 144, 183, 186, 196, 204–​5, 227–​28, 239, 244, 277 and democracy, 2, 18, 237–​40, 244–​45 described, 1, 5, 6, 23–​26, 49–​51, 53–​54, 55n.22, 79 determinants of, 64, 70, 212, 244, 245 and elections, 237–​45 increasing, 43–​44, 68, 74–​75, 86–​87, 122–​23, 138, 144, 147, 182, 184, 211, 238–​39, 242, 277 measures of, 6, 34, 54–​59, 278 promoting, 17–​18, 63, 68, 86, 118, 119, 122,  274–​75 rational basis for, 227–​28 real, 44, 212 in the real world, 147, 237, 245

for the right reasons, 30, 31, 39, 125–​26, 236, 237 sustaining, 9–​10, 16, 18, 21, 31, 39, 49, 59–​66, 85, 86–​87, 126, 137, 139, 140, 144, 211, 212, 237, 274 sustaining causes of, 59–​66 undermining, 35, 43, 85, 107, 118, 137, 163–​ 64, 204, 227–​28 and the welfare state, 139, 140, 146 social trust literature, 17, 59–​60, 64, 124, 142–​43n.7,  182 social welfare programs, 18–​19, 22, 139, 147, 277 sorting, 3–​4, 9n.47, 14, 230n.73, 231n.78, 278 southern realignment, 5, 11–​12, 13 Southwood, Nic, 32–​33, 34, 34n.26, 34–​35n.27 Special interest, 146, 183, 220, 223, 224–​25, 227 group, 110n.88, 146, 223–​24 standard trust question, 55, 55n.21 standing, 33, 94–​95, 94–​95n.32, 96, 97, 98–​ 99n.44, 105–​6, 198n.79 and moral rules, 88 state of nature, 101–​2n.54, 151n.34, 217–​18, 219n.36 Stolle, Dietland, 51–​52, 67, 238–​39 Sunstein, Cass, 14, 14n.64, 231, 231n.78   Talisse, Robert, 3n.13, 13, 231n.78 taxation, 62n.55, 114, 135, 140, 141, 148, 150, 151, 152, 153, 173, 174, 179, 184, 185, 196, 201, 208–​9,  225–​26 Teles, Steven, 176n.13, 227–​29 Tetlock, Philip, 161, 232 Thomas, Alan, 90, 114, 176, 180, 181–​82, 183n.41, 204n.100, 204, 244 Thrasher, John, 252–​53 tracking objection, 260–​61 Tropp, Linda, 90, 91, 92 trust for the right reasons, 16, 21–​22, 22–​23n.5, 23, 25, 30, 31–​32, 39, 42–​43, 44, 48, 49, 62n.58, 87, 88, 103–​4, 119, 125–​26, 139, 182–​83, 186, 208–​9, 211, 227, 235, 236, 237. See political trust; social trust egalitarian economic institutions do not create,  207–​9 and trustworthiness, 23–​26 trustworthiness, 15–​16, 22–​23n.3, 23, 25–​26, 25n.9, 35, 41, 44, 55, 57, 59, 62, 64, 65–​ 66, 69, 83, 91, 92, 103, 122–​23, 182–​83, 243, 244 evidence of, 22–​23n.3, 65–​66, 84, 88–​89,  103 intelligible reasons for, 26–​27

310 Index trustworthiness (cont.) internal, non-​exclusive deontic reasons for,  29–​30 justification of, 28–​29 (see public justification) moderately idealized reasons for, 27–​29 of others, 25n.9, 29–​30, 41, 65–​66, 83–​84, 92, 123 (see untrustworthiness) political, 253 reasons that morally motivate, 28 relation to trust, 23–​26, 44 right reasons to be trustworthy, 23–​30 social, 25, 30, 62 touchstone of, 22–​23, 30, 31, 103 Tullock, Gordon, 178, 225n.60 Type-​1 errors, 223–​24, 225 Type-​2 errors, 223–​24, 225   unfairness, 61, 72, 79, 140–​41, 183, 184 unions, 46, 99–​100, 101, 126–​27, 137, 146, 197–​ 98, 198–​99n.80, 199 universal welfare programs, 142–​43, 146, 147, 204n.102, 240 untrustworthiness, 25n.9, 82–​83, 84, 88–​89,  214–​15 Uslaner, Eric, 9, 9n.47, 50n.6, 61n.50, 72, 78, 90, 132n.37, 145, 182, 238–​39, 241   values, 8, 20–​21, 24, 24n.7, 26, 28n.14, 35, 40–​ 41, 42, 45, 53n.15, 55, 88, 99–​100, 103, 176, 237, 240, 241, 243, 246, 248, 250, 253, 254–​ 55, 261, 262–​63, 265, 266, 273–​74 democratic, 80, 237 feminist, 111, 112 home, 228 incompatible, 40–​41, 89 instrumental,  245–​46 moral, 23, 158, 158n.54, 261, 262–​63 personal, 257, 268–​69 political, 248, 249, 254–​55, 265 postmaterial, 80, 243 proceduralist,  246–​48 sectarian, 15, 15n.65 shared, 20, 92, 126–​27, 254–​55 Van Der Meer, Tom, 71n.100, 125n.25, 213–​14, 240 veil of ignorance, 40, 40nn.40–​41, 41–​42n.44, 102–​3,  126–​27 voting, 4, 8, 9, 11–​12, 11–​12n.57, 78, 216, 220, 221–​22, 224, 232–​33, 235–​36, 241–​42, 250, 251, 254n.80, 256, 257–​59, 260, 268–​69,  271 retrospective, 258

rules, 235, 254, 266, 270–​71, 273 winner-​take-​all, 75,  145 voting rights, 245–​46, 251, 255–​56   Warren, Mark, 89–​90, 239–​40, 239–​40n.13 welfare state, the, 106, 138–​39, 142–​43n.7, 147, 151n.32, 153, 173, 173–​74n.2, 187, 188, 188n.58, 199–​200, 202n.89, 206–​7. See welfare states; welfare state institutions; welfare state capitalism of administration, 203–​4 and basic justification, 139–​40 and corruption, 144–​47 creating trust, 144, 171–​72, 208–​9 described,  139–​40 of law, 203 and political trust, 139, 140, 141–​42, 146, 147 programs,  141–​42 public justification of, 139–​40, 167 and public reason, 140 Rawls’s criticism of, 204n.102 Rawlsian, 187 and real trust, 139, 140–​41 redistribution in, 184, 186–​87 and social trust, 139, 140, 146, 147 welfare state capitalism, 119, 139–​40, 192, 196–​97,  201 critiques of, 180n.31, 181–​82 defined, 207 liberal democratic, 186 modest, 207 welfare state institutions, 125, 138, 139, 140–​41, 142, 147–​48, 172, 181–​82 publicly justified, 138, 171 and trust, 140–​41, 147–​48, 171–​72 welfare states, 106, 147, 153, 187, 189–​91, 192, 194–​96, 198–​200 compared to PODs, 189–​200 ideal, 195, 195n73 liberal, 204 modest, 171–​72, 173 rent-​seeking in, 195 Williamson, Thad, 186–​87n.50, 187–​88, 190,  194–​95 Wilson, Rick, 57, 58, 94n.27, 216n.27 World Values Survey, 55, 70   You, Jong-​sung, 95, 122–​23, 144, 144n.11,  238–​39   Zmerli, Sonja, 51, 71n.100, 210, 214–​15n.18,  238–​39 zoning laws, 176, 177, 177n.18, 178n.24, 184