I am delighted that Perspectives on Politics has decided to devote a symposium to Jacob Hacker and Paul Pierson's Winner-Take-All Politics. I cannot think of any other book that so successfully synthesizes and popularizes political science research while at the same time articulating original insights and a distinctive approach to the study of American politics. Hacker and Pierson set out to provide a political explanation of why the rich did so much better than anyone else over the last 30 years. Downplaying electoral politics, and hence public opinion, their account of the rise of inequality emphasizes the growing influence of organized business interests over American public policy.
My commentary is that of a comparativist. Though Winner-Take-All Politics is a book about the United States, comparisons with other advanced capitalist countries play an important role in Hacker and Pierson's argumentation. In companion articles published by Politics and Society, Footnote 1 moreover, the authors affirm that their approach builds on the tradition of comparative political economy. In their words, “comparative scholarship has long insisted that conflict among organized interests is central to explaining the enormous cross-national and longitudinal variation in political and policy outcomes.”Footnote 2 By contrast, “too many economists and political scientists have treated the American political economy as an atomized space, and focused their analysis on individual actors, from voters and politicians to workers and consumers.”Footnote 3
Hacker and Pierson's juxtaposition of comparative and Americanist scholarship ignores the recent trend among comparative political economists, myself included, to theorize about individual preferences and voting behavior and to engage in empirical analyses that link individual-level and macro-level variables. In what follows, I begin by spelling out what I consider to be the main contributions of this book and then present some critical reflections on its arguments and analysis, from the perspective of comparative political economy. In due course, I will also present some comparative evidence suggesting that public opinion deserves more attention than Hacker and Pierson's analytical perspective allows.
Challenges for Comparative Political Economy
Implicit in Winner-Take-All Politics is an important critique of recent comparative literature on the politics of inequality and redistribution in the advanced capitalist countries. Almost all of this literature conceives of “redistribution” in terms of the effects of taxes and transfers on the distribution of income and treat this as the crucial outcome to be explained. For purposes of quantitative analysis, this is a convenient way of framing the research question. As Hacker and Pierson remind us, however, the existing literature is limited, indeed problematic, to the extent that it conceives the “pre-fisc” distribution of income simply as a function of market forces or, in other words, as “prepolitical.”
The comparative literature often recognizes, as an aside, that taxes and transfers may affect the pre-fisc distribution of income, but we have yet to tackle the question of “second-order effects” in a systematic fashion. In addition, Hacker and Pierson demonstrate persuasively that government regulatory activities have important distributive consequences. In the American case, politics has obviously played an important role in the decline of union density, and successive governments have failed to maintain the real value of the minimum wage. As Hacker and Pierson emphasize, financial deregulation has also contributed to the dramatic increase of income inequality that has occurred since the mid-1970s.
The main challenge that the book poses to comparative political economy, then, is to tackle head-on the question of how politics shapes the distribution of “market income,” as well as the distribution of disposable income. I am tempted to say that the distribution of “market income” should be put back on the agenda, for I and others have written on the determinants of earnings inequality.Footnote 4 However, this earlier literature entirely misses the dramatic rise of top-income shares that Hacker and Pierson's analysis of the American case emphasizes, drawing on the comparative top-income database developed by Anthony Atkinson, Thomas Piketty, and Emmanuel Saez.Footnote 5 Regarding the rise of top-income shares, it is not the description of the phenomenon that poses a challenge for comparative political economy, but rather Hacker and Pierson's compelling claim that the political dynamics behind the rise of top-income shares differ fundamentally from the political dynamics behind other forms of (rising) income inequality.Footnote 6
I am also very much persuaded by the book's argument that to appreciate the role of politics in the rise of inequality in the United States, we need to pay attention to what successive administrations and Congresses did not do, as well as to the tax cuts and other changes in public policy that they introduced. The concept of “policy drift,” referring to “systematic, prolonged failures of government to respond to the shifting realities of a dynamic economy” (p. 43), strikes me as highly relevant for students of comparative political economy and public policy. In this respect, the perspective of Winner-Take-All Politics provides an important corrective to the literature (including Pierson Reference Pierson1996) that emphasizes the resilience of mature welfare states during the 1980s and 1990s. In most countries of the Organization for Economic Co-operation and Development (OECD), social spending grew relative to GDP and, with the United States and the Netherlands as notable exceptions, social spending actually became more redistributive over these two decades. Yet the distribution of disposable income among working-age households also became more unequal in all but one of the countries for which we have comparative data covering this period.Footnote 7 New policy initiatives would have been required to maintain the status quo of the 1970s, and it is not only the United States that failed to undertake such initiatives.
Critique of Winner-Take-All Politics
There is an obvious tension between Hacker and Pierson's emphasis on politics as the source of rising inequality and their emphasis on policy drift. To the extent that the US story is indeed a story of policy drift, it would seem more appropriate to characterize public policy as permissive and accommodating than to cast it as the “driver” of changes in the distribution of income. The question then becomes: What are the forces behind the “rapidly changing economic realities” to which elected officials chose not to respond? The authors argue strenuously against the standard story of skill-biased technological change, on the grounds that the US story of dramatic increases of top-end inequality is unique, while similar technological changes have occurred across all advanced capitalist countries. However, they provide very little discussion of other social and economic developments that might be relevant to the rise of inequality.Footnote 8
I suspect that many comparative political economists who read this book, particularly those of an historical-institutionalist bent, will share my feeling that it would have been still better had Hacker and Pierson engaged in a more systematic analysis of the changing dynamics of the American economy and its relationship to the global economy, including sectoral changes and new corporate practices. Though highly critical of economists' explanations of rising inequality, they seem content to leave the analysis of economic change to others and to focus their attention on the politics of “organized combat” among parties and interest groups in Washington, DC.
My second criticism is that Hacker and Pierson exaggerate the uniqueness of the American case. As noted, rising inequality, in disposable incomes as well as gross earnings, represents a common trend across most OECD countries. The authors would presumably retort that this observation misses what is truly unique about the US experience, namely, the dramatic rise of top-income shares. Even in the top-income data that they feature in their book, however, we observe big increases of top-income shares in Australia, Canada, Ireland, New Zealand, and the UK as well as the United States, from the mid-1970s to the late 1990s (see p. 39). Hacker and Pierson's suggestion that rising top-income shares in these other countries is essentially a spillover from the United States, due to competition over (English-speaking) executive talent, strikes me as a less-than-compelling explanation.Footnote 9 A deeper analysis of the distinctive dynamics of “liberal market economies” and, in particular, the expansion of financial services in these economies would seem to be in order.
It is also noteworthy that virtually all OECD countries cut top personal income tax rates in the period 1985–2005 and that many countries did so more drastically than the United States.Footnote 10 This suggests some need to qualify Hacker and Pierson's interpretation of pro-rich tax cuts as an expression of the growing political influence of the business community in the context of the peculiarities (conservative bias) of American political institutions. Arguably, structural power (increasing capital mobility) is more important to the political influence of business than the organizational factors that the authors emphasize, and its political effects are less contingent on political institutions than their account suggests.
Finally, I want to question Hacker and Pierson's provocative claim that only the very rich benefited, in any meaningful way, from the new American economy of the 1980s and 1990s. In arguing against skill-biased technological change as an explanation of rising inequality, they note that “within-group inequality,” that is, inequality among college graduates and people without college degrees, has risen as much, perhaps more, than “between-group inequality” (p. 36). To my mind, this observation does not fit so well with the claim that only the rich have benefited, for it points to the pervasiveness of rising inequality. Inequality has increased among college graduates and among those who never obtained a college degree. We also observe significant increases of inequality among women and minorities.
Real income growth in the fourth and fifth quintiles of income distribution looks paltry by comparison to the top 1% of households, but it looks pretty good by comparison to the bottom two quintiles (see p. 23). Furthermore, households in the fourth and fifth quintiles have arguably benefited from declining relative wages in the bottom half of the income distribution, as personal services have become relatively less expensive. Along these lines, a plausible case can be made, I think, that the pro-rich politics described by Hacker and Pierson have enjoyed broader societal support, motivated by material interests, than they suggest.
Middle-Class Preferences and Redistribution in Comparative Perspective
In a recent article in the American Political Science Review, Noam Lupu and I argue that the structure of inequality shapes the politics of redistribution.Footnote 11 Focusing on the policy preferences of middle-income voters, our argument illustrates the “individualist turn” of comparative political economy over the last 10 years and apparently runs counter to the thrust of Hacker and Pierson's approach to the politics of inequality. We stipulate that when the bottom half of the income distribution is more compressed than the top half, middle-income voters feel affinity with the poor and also believe that they might become poor. Hence, the middle-income voters will lean in favor of a pro-redistribution coalition with the poor. Conversely, middle-income voters will feel affinity with the rich and believe in upward mobility, hence lean in favor of an anti-redistribution coalition with the rich, when the top half of the distribution is more compressed than the bottom half.
Lupu and Pontusson (Reference Lupu and Pontusson2011) measure the structure of inequality by dividing the ratio of earnings in the 90th percentile to median earnings by the ratio of median earnings to the earnings in the 10th percentile. This variable, which we refer to as “earnings skew,” takes on values greater than zero when the upper half of the earnings distribution is more dispersed than the lower half. Pooling observations from 18 OECD countries over the period 1970–2004, we find that earnings skew is strongly associated with redistribution among working-age households (% change of Gini coefficient brought about by taxes and transfers). We obtain these results, even though one of our cases, the United States, does not fit the argument very well. Ignoring the tail ends of the distribution, earnings inequality in the United States is at least as skewed as that of many European countries, yet redistribution is much lower. And increasing skew of the US earnings distribution since the 1970s has certainly not resulted in more redistributive politics.
For the purposes of this essay, the interesting question is exactly how the United States departs from the average OECD pattern. Adjusted from Lupu and Pontusson (Reference Lupu and Pontusson2011), Figures 1 and 2 address this question in a preliminary fashion. With US observations as empty circles, Figure 1 plots redistribution against the percentage of middle-income respondents who agree with the statement that “it is the responsibility of the government to reduce the differences in income between people with high incomes and those with low incomes.” Figure 2, in turn, plots middle-income support for redistribution against our measure of earnings skew.Footnote 12
Figure 1 Actual redistribution and middle-income support for redistribution
Source: Lupu and Pontusson (Reference Lupu and Pontusson2011).
Figure 2 Middle-income support for redistribution and earnings skew
Source: Lupu and Pontusson (Reference Lupu and Pontusson2011).
Winner-Take-All Politics strongly suggests that what distinguishes the United States, from a comparative perspective, is a sharp disconnect between what middle-income voters want and what governments deliver and that this disconnect has become more pronounced over time. With the US data series ending in 2000, Figure 1 does not seem to bear out this vision of American exceptionalism. With the notable exception of 1996, all of the US observations are clustered in the lower left corner of this figure; that is, the United States is distinguished by low middle-income support for redistribution, as well as low redistribution.Footnote 13 On the other hand, Figure 2 suggests that the US case is exceptional in the sense that middle-income support for redistribution is lower than the structure of earnings inequality would lead us to expect.
Why are middle-income Americans apparently less responsive to the structure of inequality than their counterparts in other advanced capitalist democracies? The concentration of racial/ethnic minorities among the poor must surely figure prominently in any attempt to answer this question.Footnote 14 Another piece of the puzzle may be that Americans know less about what the distribution of income looks like and are also poorly informed about the distributive implications of public policy.Footnote 15
Final Remarks
I am very sympathetic to Hacker and Pierson's emphasis on organized interests, but I believe that they go too far in their efforts to differentiate their approach from that of Larry Bartels (Reference Bartels2008) and others who explore the politics of inequality through the lens of public opinion and voting, creating a false divide between parallel lines of inquiry. Hacker and Pierson's own argumentation suggests that parties and interest groups mobilize voters and shape public opinion in order to influence public policy.Footnote 16 This perspective represents an important corrective to the “individualist turn” in comparative political economy (including Lupu and Pontusson Reference Lupu and Pontusson2011), but it surely does not follow that public opinion is irrelevant.
For students of American as well as comparative political economy, the challenge is to integrate the “demand side” and the “supply side” more effectively. In so doing, we must go beyond the notion of “prepolitical” voter preferences being aggregated by vote-maximizing or policy-seeking parties. Hacker and Pierson usefully remind us that firms as well as individuals have policy preferences. They also remind us that organized interests as well as parties matter, and that these political actors do not simply aggregate preferences but also mobilize voters and influence their opinions. On the other hand, it goes without saying, I think, that successful efforts to mobilize and influence voters must somehow resonate with their material circumstances.
