What Political Scientists Can Teach Us About Inequality

When Treasury Secretary Henry M. Paulson recently fought Congress on executive compensation, during discussions of the financial bailout, the curtain opened wider on the troublesome issue of rising income inequality. It’s a phenomenon that has been attracting attention among both economists and political scientists.

Is income inequality due to economics, politics, or some interplay between the two?

“Some pundits are tempted to look inside the Beltway for a cause, but the case is hard to make,” wrote the Harvard economist N. Gregory Mankiw in an April column in The New York Times. “Government policy makers do not have the tools to exert such a strong influence over pretax earnings, even if they wanted to do so.”

The problem, of course, runs much deeper than the symbolic issues of executive compensation and golden parachutes. Economists teach us that a labor-market correlation between increasing income returns and skills and education has separated the fortunes of the skilled from the unskilled. Yes, we live in a new gilded age in which the top 1 percent of taxpayers (earning $383,000 or more per year in 2006 dollars) makes about 23 percent of all annual income. But our new society results from a massive shift in income returns to education.

A sophisticated variation on the theme, which Mankiw was discussing in his column, is the book by his Harvard colleagues Claudia Goldin and Lawrence F. Katz, The Race Between Education and Technology, dubbed by Mankiw “the best diagnosis so far.” Goldin and Katz argue that increasing income inequality arises from a supply-side failure in education, which has failed to keep pace with new knowledge requirements. For most of the 20th century, the United States did something unique: It educated everyone as much as possible, giving us an enormous lead in human capital over every other country. But we stopped giving ourselves that advantage just as it became most essential, in the last quarter of the 20th century, when competitors elsewhere caught up to our educational example and as technological change continued its steady pace. The result has been growing income inequality rooted in the divergence of educational attainment and the constantly evolving skill requirements from technological change. As Goldin and Katz put it, “The sharp rise in inequality was largely due to an educational slowdown.”

Mankiw was right to showcase the Goldin and Katz book, which is magnificent. But economics can’t give us the whole picture. There is, after all, a fundamental political puzzle here. How does it happen that a democratic citizenry will tolerate increasing income inequality without pushing for a systematic response — something more than the recent flood of phone calls and e-mail on Capitol Hill over the Wall Street bailout?

To illuminate that puzzle, we should turn to the Princeton political scientist Larry M. Bartels’s book, Unequal Democracy: The Political Economy of the New Gilded Age. Presidents and parties have escaped accountability for inconsistent management of the economy, Bartels finds. Globalization and increased delegation of macroeconomic policy making to the Fed have diminished presidential influence on the economy in the past quarter-century, he concedes. But if the economy can be compared to a house, the problems Bartels finds are more akin to termites in the basement — they come from the cumulative impact of voter behavior on income distribution. (Disclosure: I commented on parts of his book at the manuscript stage.)

Bartels’s book challenges not just economists like Mankiw who caution against looking inside the Beltway for explanations of income inequality. It also questions the way many political scientists view the politics of macroeconomic performance. Many political scientists have reasoned that because the Fed is independent, macroeconomic performance cannot be politically engineered to decisively help one party or the other. Additionally, voters have the capacity to punish or reward presidents and their parties for apparent failure or success in reducing inflation and unemployment while generating growth in the gross domestic product. That theory — best expressed in The Macro Polity, by Robert S. Erikson, Michael B. MacKuen, and James A. Stimson — holds that in a rational political economy, accountability to the public will exact government competence.

In fact, however, there are biases (as Bartels terms them) in economic accountability. Those are augmented by the great difficulty that ordinary voters have in connecting cause and effect when considering macroeconomic and tax policy.

It’s not that voters are ignorant or irrational. Bartels can easily be misread as having a thoroughly Menckenite view of voters as confused idiots. His perspective is far subtler — rooted in the research on public opinion by the UCLA political scientist John Zaller and, looking further back, by the pioneering opinion research of Philip E. Converse of the University of Michigan. Bartels elsewhere calls his way of thinking about our political system “democracy with attitudes.”

Voters, in this view, are busy with their own lives. They candidly tell surveyors for the American National Election Studies that politics and policy are complicated and hard to understand. Voters do get much of what is going on and also have meaningful attitudes — they are not, for instance, fans of inequality or stupefied admirers of high-profile entrepreneurs like Donald Trump. Americans value equality, and tell researchers in detail how and why.

But voters are also myopic. In judging economic performance, they take into account only quickly accessible information from the very recent past. Democrats regularly deliver gains in income growth for people in the lower half of income distribution in Year 2 of a Democratic administration, which is something that the administrations, after all, promised. In Years 3 and 4, Democrats’ contribution to income growth up and down the distribution of income is attenuated, Bartels writes. But voters completely discount Year 2 and pay attention, instead, to macroeconomic performance in Year 4. So Democrats have lost more elections in the past 50-plus years than they should have.

In contrast, Republicans, who have their own characteristic macroeconomic preoccupations, throw people out of work in Year 2 in order to reduce inflation. That causes a general economic slowdown but also increases incomes disproportionately for the wealthiest. Republicans also happen to preside over strong GDP growth in Year 4. They’re not manipulating the economy; they’re just catching the tail wind from economic recovery after Year 2. That dynamic regularly helps their election prospects, even among lower-income people whom Republicans hurt in Year 2. And so, Bartels reasons, Republicans have won more elections than they otherwise would have.

The cumulative result, then, of party control over macroeconomic performance has been an increase in income inequality. Voters have not voted themselves enough spells of Democratic control to deliver either on keeping income inequality steady or on reducing it.

In highlighting the long-term, powerful influence of voter myopia in assessing macroeconomic performance, Bartels does not intend any derision. Such myopia is perfectly understandable. Can you quickly remember your personal economic circumstances three years ago? But in linking this recurring feature of the electorate’s psychology to the characteristic macroeconomic stances and performances of the two parties, Bartels resolves what is so fundamentally perplexing about increasing income inequality: the seeming disconnect between that phenomenon and the power of a sovereign, democratic citizenry to influence public policy.

There are colleagues of Bartels at Princeton and elsewhere who have unraveled the problem equally cogently but rather differently, with a focus on failures of redistributive social policy. Nolan McCarty, Keith T. Poole, and Howard Rosenthal, in Polarized America: The Dance of Ideology and Unequal Riches, focus on why rational voters do not halt rising income inequality and correct it with redistributive public policy. In their formal model of democracy, the median voter, when his or her income is exceeded by the average income, will prefer higher taxes for redistributive policy (in the real world, think here of Denmark or Sweden). That gap must emerge when income inequality rises — or does it? That is where the U.S. experience with immigration comes into play.

If the increase in income inequality comes from low-wage immigration, then the median voter does not develop a preference for taxes and redistributive policy. That’s because the median-wage earner is not the same person as the median eligible voter. The median-wage earner is either not eligible to vote, or she or he is eligible but usually does not vote.

McCarty and colleagues emphasize the role of immigration in weakening citizen support for unemployment insurance and increased minimum wage. Political inaction is reinforced, they persuasively suggest, by the ideological polarization of the political parties and frequent recurrences of divided government. Government does not correct growing income inequality.

Although Bartels fundamentally agrees with McCarty et al. that the parties and the electorate have allowed income inequality to grow, there are important differences here in emphasis and in starting points. Bartels does not factor in immigration. But he does not seem to think that he needs to — given the extent to which the economic fortunes of lower-income voters and citizens have been stagnating. McCarty and his co-authors connect stalemate in redistributive policy to the preferences of the median voter, who, they claim, is still pretty well off. But Bartels demonstrates that the huge income-distribution effects resulting from the quirks of voters’ decision making are defined by political parties’ behavior and economic performance cycles. In addition, then, to drawing inspiration from Converse and Zaller, Bartels joins the company of cognitive psychologists like Daniel Kahneman and Amos Tversky, and economists with similar research orientations, who have begun to incorporate biased decision making into their models.

Bartels’s “democracy with attitudes” thus allowed for the Bush administration’s Club for Growth brand of Republicanism. Its 2001 and 2003 tax cuts disproportionately rewarded upper-income taxpayers. The cuts are cumulatively costing the Treasury $1.3-trillion. But in a thorough exploration of the relevant survey evidence, Bartels shows that the Bush administration hardly acted against public opinion, as is often claimed. To the extent that there was a view among the public, however uncertain and diffuse, it favored the tax cuts.

How could that be? In a chapter aptly titled “Homer Gets a Tax Cut,” previously published in different form, Bartels distinguishes among those who paid no attention, those who paid some attention, and those who paid a lot of attention and were quite politically well informed. About 40 percent of voters said that they had no idea of what was going on. Higher information levels among the remaining 60 percent of voters strongly interacted with their partisan identification. The best-informed Democrats were intensely opposed to the tax cuts. The same level of information among Republicans effectively made no difference to Republican voters’ support for or opposition to the tax cuts.

The GOP cost the country well over a trillion dollars in a giveaway to the well-off and the very rich, even as the country was at war in Afghanistan. But since it all came with the Republican seal of approval, it was fine with these voters.

Bartels is not being partisan. His analysis clearly implies that any similarly foolish policy enacted by Democrats would very likely be equally acceptable to the well-informed parts of that party’s electorate.

Income inequality became a priority of the American Political Science Association six years ago when Theda Skocpol of Harvard, on becoming president of the association, convened a task force on the matter. Her research agenda is documented in a book she edited with Lawrence R. Jacobs, Inequality and American Democracy: What We Know and What We Need to Learn.

There is also recognition of the problem, I hasten to add, inside the Beltway. Delivering his “State of the Economy” report for 2007, President Bush said, “We have an obligation to help ensure that every citizen shares in this country’s future. The fact is that income inequality is real; it’s been rising for more than 25 years.” In May 2006, the chairman of the president’s Council of Economic Advisers, Edward P. Lazear, speaking at the Hudson Institute, said, “There is little doubt that there has been a 25-year trend of a growing gap, sometimes called income inequality, between the wages of the skilled and the unskilled ... the general picture cannot be disputed. The difference between the earnings of the individuals at the top and earnings of individuals at the bottom has grown.” Ben Bernanke, when he was CEA chairman, told the National Economists Club, in October 2005, that “the ratio of earnings at the 90th percentile of the earnings distribution to earnings at the 10th percentile indicates a fairly steady increase in inequality since 1980.”

But these forthright statements by public officials have all been couched with suggestions that the problem is temporary or less serious than it might seem. Bartels, Goldin and Katz, and McCarty et al., all rightly go further: They emphasize that what has actually happened is a historic reversal.

One can consider the reduction of income inequality as a democratic and economic achievement. There is a famous claim advanced by the economist Simon Kuznets that as countries get richer they become less unequal. Similarly, Goldin and Katz describe America’s reduction in income inequality over the course of the 20th century as an epic national achievement based on what they quite persuasively show is our version of active labor-market policy: mass education.

Americans have had great public reversals before — most visibly, of course, during the Civil War and the Great Depression. In those and other cases, recovery and repair of the polity took considerable time. Similarly it may require a long march in public policy and politics to reduce income inequality and return us to the more democratic position we achieved by the early 1970s.

To be sure, economists like Goldin and Katz are optimistic that local and state governments could recommit to quality mass education and restore national greatness despite problems in the federal political system. Perhaps the march back will not be all that long. Political scientists, in contrast, suggest that a more uncertain, harder road lies ahead. The great instruments of democracy — parties and elections — are far more flawed, they find, than we knew.

In the end, there is more going on here than bragging rights over which discipline better explains rising income inequality or provides more hope. The last time we had as much income inequality as we do today was in the late 1920s. Mainstream political scientists and economists then did not recognize income inequality, as they do today, as an appropriate target for policy. This time around, though, these two disciplines might together have more impact on how politicians and public officials confront the issue — and encourage them to go well beyond tepid acknowledgments of what is in fact a national embarrassment.

Rick Valelly is a professor of political science at Swarthmore College.

http://chronicle.com Section: The Chronicle Review Volume 55, Issue 8, Page B11