Anxieties of Democracy and Distribution
At least since Tocqueville wrote Democracy in America, it has been routine to see democracy and equality as joined—if somewhat problematically—at the hip. Tocqueville saw democratic politics as the outgrowth of a relentless drive toward equality that he believed had been unfolding for centuries. He also thought that democracy reinforces the egalitarian mores that he believed spawned it, undermining traditional sources of value and commitment in ways that he regarded as lamentable but unstoppable—one of the earliest anxieties of democracy. In later life, Tocqueville came to see that democracy’s connections to equality are more complex than he had thought in the 1830s. His earlier fears about its leveling propensities were replaced by different anxieties as the American system showed itself unable to abolish slavery and, as a result, atrophied as a beacon of hope to flailing European aristocracies. He died in 1859, before the Civil War validated his growing apprehension, but its advent would not have surprised him (Craiutu & Jennings, 2004).
Many subsequent political commentators have probed the relations between democracy and distribution, sometimes prompted by worries about the distributive preconditions for viable democracy and sometimes by anxieties about the perceived consequences of democracy for the distribution of income, wealth, and other social goods. In what follows, I will say something about these anxieties and the expectations that generated them from the perspective of contemporary political science. As with Tocqueville before us, we have learned that there is more complexity to these relations than has often been assumed. It will also become clear that, despite more than a century and a half of sustained study, there is still a good deal pertinent to the connections that we do not know.
Inequality’s impact on democracy
Modernization theorists used to see economic development as a harbinger of democracy, though they disagreed about why (see Lipset, 1960; Apter, 1965; and Moore, 1966). More recently, however, scholars have distinguished transitions to democracy from democratic consolidation, arguing that the former is not predictable by modernization (or anything else) while the latter is well predicted by per capita income (PCI). Specifically, if PCI reaches and remains above $6,055 (in 1985 dollars), an existing democracy will survive indefinitely. But as PCI falls below that threshold, democracy becomes vulnerable—and the further it falls, the more vulnerable it becomes (Przeworski et. al, 2000).
While PCI’s influence on democracy is well documented, the effect of inequality on democracy, if any, is more difficult to pin down. Democracy survives in countries with some of the highest Gini coefficients in the world (e.g., South Africa, Brazil), but it is unclear whether it could endure in less socially and economically diversified unequal countries like Saudi Arabia, were it to be established in them. There are also some outliers, India being the most prominent. But the general finding seems to be robust: PCI matters more for democratic consolidation than do inequality, institutions, culture, or other variables.
Why democracy survives in rich countries is an interesting question, to which answers are mainly speculative. Some suggest that in wealthy countries, both the wealthy and the poor have reasons to maintain democracy—if it exists—over support for a coup. As Lipset (1960: 51) said over half a century ago: “If there is enough wealth in the country so that it does not make too much difference whether redistribution takes place, it is easier to accept the idea that it does not matter too much which side is in power. But if loss of office means serious losses for major groups, they will seek to retain office by any means available.” Benhabib and Przeworski (2006) put it more sharply: if we assume the diminishing marginal utility of income and that people also value the freedom that comes with democracy, the rich will prefer to accept a degree of redistribution in favor of the poor rather than give up their freedom (since it is relatively good bargain). The poor will prefer the substantial utility gains they can achieve as beneficiaries of redistribution over attempting coups—which might fail and will, in any case, cost them their freedom as well.1
There might be something to arguments of this sort, but they are glaringly devoid of mechanisms and they rely heavily on debatable stylized facts. Moreover, they ignore obvious collective action problems among “the rich” and “the poor,” not to mention agency problems between both groups and their political representatives, which are glossed over in Lipset’s formulation. These collective action and agency problems are rife in much of distributive politics (see below).
A more parsimonious and also more political story focuses on office-holders themselves, and turns on the costs to them of giving up office. Democracy depends minimally, but vitally, on the willingness of office-holders to give up power when they lose an election. This, in turn, means that there must be somewhere reasonably appealing for them to go. Poor countries tend to be marked by winner‐take‐all (and, hence, loser‐lose‐all) politics, whereas richer democracies offer other opportunities to formerly employed political elites. When Jimmy Carter lost the presidency in 1980, he could go on to join corporate boards, write books, create foundations, and, had he chosen, get wealthier than he could ever have by remaining in office. Sending the tanks down Pennsylvania Avenue is scarcely worth contemplating in such circumstances. The practice of alternation was by then so deeply entrenched in American politics that it would have been hard to imagine a losing candidate even considering trying a coup, even when—as with the Democrats in 2000—the losers believed that the election had been stolen (see Wells & Levy, 2013).
Because the best predictor of alternation is alternation, this raises the question: how did alternation become established in the U.S. in the first place? It is hard to read Susan Dunn’s (2004) account of the election of 1800 without concluding that prospects for alternation in America’s critical first few elections depended in substantial part on the low appeal that politics held for George Washington, John Adams, Thomas Jefferson, Aaron Burr, and James Madison, especially in light of the more lucrative and more appealing ways they could spend their time. Even at that early stage, the American polity and economy offered elites opportunities that dampened the costs of losing office.2
This reasoning suggests a mechanism for why democracy often struggles to survive in single-commodity export economies (the “oil curse”), where access to the levers of political power is often vital to the continuing wellbeing of economic elites. In such settings, even if PCI reaches levels that should be expected to make democracy sustainable, it might nonetheless be vulnerable to a coup on the part of elites who fear losing power or who find themselves in a position to grab it. On this account, the diversification of the economy is likely to matter more than PCI or relative inequality. What matters, more specifically, is the extent to which everyone’s eggs are in the same basket. If they are, then the incentive to fight over who controls the basket is high. It also suggests that political conflict over divisible goods (like income and wealth) is more compatible with democratic politics than conflict over indivisible goods such as religion, race, or ethnicity. In these latter cases, control of the commanding heights by one group is not easily mitigated; as a result, politics more easily becomes a battle for monopoly control of those commanding heights (see Shapiro 2014: chapter 5).3
Beyond questions of democratic survival, inequality is also studied by political scientists for its impact on the quality of democracy. In the U.S., this subject is dominated by the dreary story of the role of money in politics, ushered in by the Supreme Court’s 1976 equation of money with speech in Buckley v. Valeo.4 Much recent attention has been focused on the Court’s 2010 decision in Citizens United v. Federal Election Commission, which extended Buckley’s logic to corporate entities.5 But once the Buckley move had been made, the rest was only a matter of time. Much of the literature about remedies has focused on public funding of elections and campaign finance proposals that might circumvent the Supreme Court decision. But public funding schemes have to be voluntary, and they have been rendered moot by the billion‐dollar nonpublic presidential campaigns that have been normalized by Barack Obama. Campaign finance reform is little more than a series of exercises in closing the stable doors that the Court opened with Buckley, since only “contributions” and not “expenditures” can be regulated. Water flows effortlessly around that rock. Moreover, money’s impact on politics flows through many channels beyond campaign contributions, the most obvious being political action campaigns and activist think tanks—both of which spend large sums manufacturing ideological ballast and mobilizing effort to support or block candidates and legislative agendas (Graetz & Shapiro, 2005: 85-98). These campaigns and mobilizations, fueled by money, are considered political speech that lies at the core of the First Amendment as interpreted by the Court—untouchable by regulation.
Another way in which inequality influences the quality of democracy concerns the low rates of political participation by the poor. These low participation rates give political candidates in search of the median voter diminished incentives to cater to the interests of the poor. It used to be conventional wisdom in political science that the consequences of this were overstated, inasmuch as the preferences of nonvoters were thought not to deviate much from those of voters—so that increased participation would be unlikely to have much effect on what happens. In recent decades, that conventional wisdom has been called into question (Gerber 1998). This matters less for the elderly poor than for other poor people. High participation rates of the nonpoor elderly, combined with the fact that the programs like Medicare and Social Security are seen as universal social insurance rather than targeted redistribution, protects their interests. This is in line with conventional wisdom about the relative effectiveness of universalist over targeted policies (Brady & Burroway, 2012) and about their greater chances of political survival (Gelbach and Pritchett, 1995; Moene and Wallerstein, 2001; Mkandawire, 2005).
Democracy’s impact on inequality
For a long time, many people thought that universal franchise democracy would lead to the downward redistribution of wealth. A sober Karl Marx in his later years favored the “parliamentary road to socialism” out of the hope that workers, having failed to overthrow capitalism at the barricades, would push for redistribution through the ballot box. Mill’s proposal to blunt the universal franchise (by, for example, giving a second vote to university graduates) was partly motivated by the fear that the masses would do exactly what Marx hoped they would. In the twentieth century, Pareto’s observation—that in capitalist systems, the median income is always below the average income—led many political scientists to predict downward redistribution, on the grounds that the median voter would vote in support of candidates who propose it. This is particularly notable given Jencks’s (1992: 11) contention that reducing inequality of opportunity would be unlikely to have much impact on inequality, which could only effectively be tackled through redistributive politics.
The expectations that drove Marx’s hopes, Mill’s fears, and subsequent predictions based on the median voter theorem have not materialized. The median voter, it turns out, does not vote in any systematic or predictable way for downward redistribution. Voters in capitalist democracies have not embraced Jencks’s (1972: 228-9) income insurance proposal that every family should be guaranteed half the average income (which seemed modest at the time). Notably more modest proposals, such as 1988 presidential candidate Bill Bradley’s proposal to index the minimum wage to the median income, have not gained traction in U.S. electoral politics either. There is plenty of government redistribution in capitalist democracies, but it goes in multiple directions—including, sometimes, flagrantly regressive. The 2001 Bush tax reforms (almost all of which were made permanent in January 2013 with strong bipartisan support in the House and Senate and signed into law by a Democratic president) are the most recent example.
Expectations that majorities in democracies will support downward redistribution fail to take account of the logic inherent in the majority rule divide-a-dollar game. In a three person society, if A and B agree to split a dollar equally, giving C nothing, she can respond in the next iteration by proposing a different split—perhaps sixty/forty—to A or B. The loser can then find a majority partner to upend the new status quo, and so on ad infinitum. Even if all three agree on an equal split, any two can form a majority coalition to defect—dividing the dollar in some way among themselves at the expense of the third. The result is general and carries the implication that every conceivable distribution of a divisible good is potentially unstable under majority rule. From this perspective, the Marxist story fails because it does not acknowledge collective action problems within the group designated as the working class, and the median voter story fails because the median voter can, in principle, always be bribed to switch distributive alliances. This holds even if there is only a single dimension of political competition, which typically there is not. Additional dimensions (such as race, gender, and religion) make the median voter’s coalitional allegiances even less stable, as we saw with Nixon’s Southern strategy and as we see when wealthy women who prefer Republican economic policies vote for Democrats because of their views on abortion.
The 2001 repeal of the estate tax with strong bipartisan support underscored two truths about democracy and distribution. First, it illuminated the reality that the median voter theorem only does well empirically when mass preferences are intense. Unlike Social Security, for example (where mass preferences are intense and the median voter theorem correctly predicts that voters will oppose attempts to cut it), most voters do not have intense preferences about taxes, and particularly not about estate taxes. The polls that showed 70 percent support for estate tax repeal in the late 1990s were all stand-alone exercises in approval voting. When estate tax repeal was put into a list of possible tax cuts, it dropped to the bottom in terms of popular support, and, when paired with cuts to a popular government program such as Medicare prescription drug benefits, support for estate tax repeal evaporated. This polling revealed that no politician stood to win or lose votes based on the position they took on estate tax repeal. In such circumstances, distributive outcomes depend on the activities of organized groups more than on mass preferences (Birney, Shapiro, and Graetz, 2008).6 Unfortunately, there is little systematic data on the intensity of mass preferences and scant knowledge—or even plausible speculation—about what determines the relative intensities of mass preferences or how and why they change.
Second, the story of the estate tax repeal showed that, just because of the potential instability of all distributive coalitions, much distributive politics is about busting coalitions apart, building new ones, and then heading off attempts to break them up. Organized labor, African Americans, gay groups, the life insurance industry, organized charities, and other traditional supporters of estate taxes were effectively convinced, co-opted, or marginalized by the estate tax repeal coalition. Moreover, attempts to split the repeal coalition were effectively parried. The Democrats repeatedly offered to enact thresholds that would exempt most of its members as well as special exceptions for farmers. Holding the coalition together required something more than self-interest: it took a convincing narrative with heroes, villains, and a moral. The activists who did the work to hold the repeal coalition together believed they were on a crusade against an immoral tax—a form of double taxation, a “death tax,” etc. Interviews with them went much like interviews with abortion activists. Without these activists, the venture would have failed (Graetz & Shapiro, 2005: 50-73; 154-67; 221‐8).
The upshot is that it is misguided to think that when people vote against their economic interests, a “wedge issue” involving preferences along some other—perhaps implicit—crosscutting dimension must be involved. There might not be anything the matter with Kansas (see Frank, 2004). Reflecting on the divide‐a-dollar logic reveals that we all vote against some of our interests all the time because there is always a potential coalition in which we might do better than we do at present. The fact that we don’t try to create and sustain that coalition has to be explained by a commitment to something in addition to self-interest: a moral or ideological narrative we find compelling enough to either to keep us in the coalition to which we currently belong or to move us to a different one. All of this depends on intensity of public opinion, however. The moral narrative will not work if the population does not care. As Jencks (1992: 24‐59) pointed out with respect to affirmative action in Rethinking Social Policy, if the moral narrative that underpins a redistributive coalition is not widely endorsed or at least tolerated, the coalition will likely be politically vulnerable.
There are numerous other factors that help explain the dearth of effective pressure for downward redistribution in capitalist democracies. Even if the median voter were available to vote for it, there are many sources of sand in the wheels that can prevent politicians from campaigning on downwardly redistributive policies. The inhibiting role of money in politics, as described above, along with fear of capital flight (and attendant financial crises) on the part of politicians, the large number of veto points in the political system, and the first-past-the-post electoral system all militate against the supply of redistributive policies in political contests over the support of the median voter (Shapiro, 2003: 106-16).
On the demand side, bootstrapping values, beliefs (implausible as they might be) about prospects for upward mobility, and framing effects all play their part. So do the comparisons with others that people make when evaluating their wellbeing and deciding what, if anything, to do about it. Ronald Reagan might have been wrong to assume that the only thing people care about is their own absolute improvement, as in: “Are you better off than you were four years ago?” As Jencks (1972: 5) noted, for most people the cost of living is not the cost of buying a fixed basket of goods and services. Rather, “it is the cost of participating in a social system.” But Marx was also wrong to think that workers would compare themselves to capitalists in assessing their wellbeing and making their demands. People compare themselves to others who are similarly situated. Autoworkers compare themselves to steel workers, doctors to lawyers. This is true up and down the occupational scale. A college professor will be much more troubled to learn that he makes $5,000 less than the professor in the next office than that he makes $500,000 less than the attorney who lives next door (Shapiro, 2003: 116-20). The more illuminating work on this subject takes us out of political science and into sociology, economics, and psychology (see Parkin, 1971; Frank, 1987; and Kahneman, 2011).
Apter, David E. 1965. The Politics of Modernization. University of Chicago Press.
Benhabib, Jess, and Adam Przeworski. 2006. “The Political Economy of Redistribution under Democracy.” Economic Theory 29: 271-290.
Brady, David, and Rebekah Burroway. 2012. “Targeting, Universalism and Single Mother Poverty: A multilevel analysis across 18 affluent democracies.” Demography. 49. 719‐46.
Birney, Mayling, Ian Shapiro, and Michael Graetz. 2008. “The Political Uses of Public Opinion: Lessons from the estate tax repeal.” Ed. by Ian Shapiro, Peter Swenson, and Daniela Donno. New York University Press. 298-340.
Craiutu, Aurelian, and Jeremy Jennings. 2004. “The Third Democracy: Tocqueville’s Views of America after 1840.” American Political Science Review 98: 391-404.
Dunn, Susan. 2004. Jefferson's Second Revolution: The Election Crisis of 1800 and the Triumph of Republicanism. New York: Houghton Mifflin.
Frank, Robert. 1987. Choosing the Right Pond: Human Behavior and the Quest for Status. Oxford University Press.
Frank, Thomas. 2004. What's the Matter with Kansas? How Conservatives Won the Heart of America. New York: Metropolitan Books.
Gelbach, J.B., and L.H. Pritchett. 1995. “Does More for the Poor Mean Less for the Poor? The Politics of Tagging.” World Bank, Washington, DC. Policy Research Working Paper # 1523.
Gerber, Alan. 1998. “Preferences of Voters and Nonvoters.” Yale University.
Graetz, Michael, and Ian Shapiro. 2006. Death by a Thousand Cuts: The Fight over Taxing Inherited Wealth. Princeton University Press.
Jencks, Christopher. 1972. Inequality: A Reassessment of the Effect of Family and Schooling in America. New York: Basic Books.
____________. 1992. Rethinking Social Policy: Race, Poverty, and the Underclass. Harvard University Press.
Kahneman, Daniel. Thinking, Fast and Slow. New York: Farrar, Straus & Giroux.
Lipset, Seymour Martin. 1960. Political Man: The Social Bases of Politics. Garden City, New Jersey: Doubleday.
Mkandawire, Thandika. 2005. “Targeting and Universalism in Poverty Reduction.” United Nations Social Policy and Development Program Working Paper # 23.
Moene, Karl, and Michael Wallerstein. 2001. “Targeting and the Political Support for Welfare Spending.” Economics of Governance 2 (1): 3-24.
Moore, Jr., Barrington. Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World. Boston: Beacon Press.
Parkin, Frank. 1971. Class, Inequality, and Political Order. New York: Praeger.
Przeworski, Adam. 2006. “Self-Enforcing Democracy.” The Oxford Handbook of Political Economy. Ed. by Barry Weingast and Donald Wittman. Oxford University Press. 312-28.
Przeworski, Adam, Michael Alvarez, Jose Cheibub, and Fernando Limongi. 2000. Democracy and Development: Political Institutions and Well‐Being in the World, 1950‐1990. Cambridge, Eng.: Cambridge University Press.
Rae, Douglas, Douglas Yates, Jennifer Hochschild, Joseph Morone, and Carol Fessler. 1981. Equalities. Cambridge, Mass: Harvard University Press.
Sen, Amartya. 1992. Inequality Reexamined. Cambridge, Mass: Harvard University Press.
Shapiro, Ian. 2003. The State of Democratic Theory. Princeton University Press.
____________. 2014. Justice Against Domination. Harvard University Press.
Tawney, Richard H. 1931. Equality. London: Allen & Unwin.
Tocqueville, Alexis de. 1969 [1835, 1840]. Democracy in America. Ed. by J.P. Mayer. New York: Anchor Books.
Wells, Ken, and Ari Levy. 2013. “Gore is Romney‐Rich with $200 Million after Bush Defeat.” Bloomberg. Accessed May 6, 2013. http://www.bloomberg.com/news/2013-05-06/gore-is-romney‐rich‐with-200-million‐after-bush-defeat.html
- Przeworski (2006) speculates that democracy might have better prospects for survival in poor countries if they are egalitarian due to the dearth of wealthy people whose assets could be expropriated in a revolution. Apart from the difficulties attending this reasoning raised in the next paragraph, it fails to account for democracy’s survival in India, which has a comparable Gini coefficient to New Zealand, France, and the United Kingdom. See Global Gini index (Distribution of family income) ranking by country. CIA Factbook for 2010. January 2010. http://www.mongabay.com/reference/stats/rankings/2172.html (03‐27‐2013). ↩
- This perspective suggests a different perspective on some recent democratic transitions. Soviet elites have been attacked as corrupt for seizing state assets on the way out of power, but the price might have been worth paying for the peaceful demise of the U.S.S.R. Good things do not always go together. In the Arab spring, the fates of Muammar Gaddafi and Hosni Mubarak have no doubt been duly noted by leaders in Riyadh, Sanaa, Rabat, and Amman in the response to their own uprisings. ↩
- Jencks (1972: 8) invoked a variant of this reasoning in a broader context when he remarked that “instead of trying to reduce people’s capacity to gain a competitive advantage on one another,” it might be better to “reduce the rewards of…success and the costs of failure.” ↩
- Buckley v. Valeo 424 U.S. 1 (1976). ↩
- Citizens United v. Federal Election Commission 558 U.S. 310 (2010). ↩
- In this instance, the repeal coalition was vastly better funded, organized, and creative than the defenders of the status quo (Graetz and Shapiro, 2005). ↩