On May 13, 2010, Larry M. Bartels was inducted as the Robert A. Dahl Fellow of the Academy at the Newseum in Washington, DC. When inducting him, Alan B. Krueger, Assistant Secretary for Economic Policy and Chief Economist for the U.S. Treasury, cited Bartelsâs extensive work on American electoral politics, public opinion, and political accountability. âHe uses empirical research to examine, in his words, âwhether democracy works as advertised,ââ Krueger said. âLarryâs work runs from the theoretical to the applied, from evaluating quasi instrumental variables to discovering uncomfortable facts that donât fit with popular beliefs. He is also one of the few political scientists I know who attends the econometrics workshop at Princeton.â The following is a transcript of Larry Bartelsâ remarks.
âI am very grateful for the honor you are bestowing on me this evening, and especially for my designation as a Robert A. Dahl Fellow. I had the privilege of being a student in one of Professor Dahlâs seminars 33 years ago, and it was an inspiring experience. His commitment to rigorous social scientific analysis of big questions about democracy impressed me then, and it impresses me even more now that I have learned by trial and error how difficult such analysis can be.
My most recent book, Unequal Democracy, focuses on the political causes and consequences of economic inequality in contemporary America. The first sentence quotes the first sentence of Dahlâs classic Who Governs?, which was published 50 years ago: âIn a political system where nearly every adult may vote but where knowledge, wealth, social position, access to officials, and other resources are unequally distributed, who actually governs?â
In reflecting on how my own scholarly work has mattered, or might matter, in the world beyond academia, I think it may be instructive to trace the fate of two of the empirical claims presented in Unequal Democracy.
First, through a simple tabulation of patterns of income growth under Democratic and Republican presidents over the past 60 years, I showed that middle- and low-income Americans have generally fared much better economically under Democratic presidents than they have under Republican presidents. On average, middle-class families have experienced twice as much real income growth under Democrats as they have under Republicans; working poor families have experienced six times as much real income growth under Democrats.
This finding, appearing six months before the 2008 presidential election, attracted a good deal of attention. On the campaign trail, Barack Obama cited âa book thatâs come out right now, by a prominent economistâirrefutableâlooking at the evidence showing that when Democrats have been in charge of the economy, the economy has grown faster and itâs also been fairer in the sense that everybody benefits.â Many other people with similar partisan inclinations touted the finding, from Dani Rodrik and Alan Blinder to James Carville and Bill Clinton. Meanwhile, Republicans mostly ignored it or dismissed it as a coincidenceâwhile displaying very little interest in the chapter full of statistical analyses buttressing the causal argument.
Another chapter of Unequal Democracy was devoted directly to addressing Robert Dahlâs big questionââwho actually governs?â Building on decades of statistical research on political representation, it related the roll call votes cast by U.S. senators to the political views of their constituents. However, rather than simply assuming that every constituentâs opinion gets equal weight, I examined the differential responsiveness of senators to the views of affluent, middle-class, and poor constituents. The results suggested that the views of affluent constituents are fairly influential, especially for Republican senators. The views of middle-class constituents matter rather less, while the views of constituents in the bottom third of the income distribution have no apparent effect on their senatorsâ roll call votes. Parallel work by my Princeton colleague Martin Gilens, employing a very different research design, produced similar results.
If these findings are correct, they seem to me to represent a gigantic embarrassment for American democracy. As Dahl put it in another classic work, âa key characteristic of a democracy is the continued responsiveness of the government to the preferences of its citizens, considered as political equals.â So where were the New York Times and CNN this time? By my admittedly impressionistic estimate, the analysis of unequal responsiveness got about one-hundredth as much attention as the analysis of partisan patterns in income growth.
Leaving aside the possibility that nobody actually reads the last 50 pages of a 300-page book, I can think of three possible (and mutually reinforcing) explanations for this disparate attention. First, neither political party has anything to gain from calling attention to evidence of shameful unresponsiveness to the preferences of low-income Americans. (Indeed, my analysis suggests that both Republican and Democratic senators were utterly unresponsive to the views of low-income constituents.) Poor people themselves might have something to gain from calling attention to that evidence, but they are not heavily represented among journalists and political commentators.
Second, while a putative âprominent economistâ may be a respectable source for claims about patterns of income growth, political scientists writing about political topics are much less newsworthy. Journalists generally assume they know everything political scientists know and then some. In many cases they are probably right; but when they are wrong they are unlikely to recognize that fact.
Finally, the âpracticalâ world of journalists and policy-makers is generally much more interested in this weekâs campaign controversy or legislative maneuver than in broad, enduring patterns of politics and policymakingâand much more interested in conclusions than in the evidence supporting those conclusions.
Perhaps it is naĂŻve to wish that serious people paid more serious attention to big, enduring questions about the workings of American democracy. Perhaps it is equally naĂŻve to wish that academics spent less time spinning toy theories and more time adducing the basic facts we need to shed light on the workings of American democracy. In both respects, I am unrepentantly naĂŻve.
