In Testing Theories of American Politics:
Elites, Interest Groups, and Average Citizens [PDF], a paper forthcoming in Perspectives on Politics by Princeton’s Martin Gilens and Northwestern’s Benjamin Page, the authors analyze 1,779 over the past 20+ years and conclude that policy makers respond exclusively to the needs of people in the 90th wealth percentile to the exclusion of pretty much every one else. Mass-scale intervention from citizens’ groups barely registers, while the desires of the richest ten percent of America dictate practically the entire national policy landscape.
In a summary in the Washington Post, Larry Bartels writes,
Alas, no. In their primary statistical analysis, the collective preferences of ordinary citizens had only a negligible estimated effect on policy outcomes, while the collective preferences of “economic elites” (roughly proxied by citizens at the 90th percentile of the income distribution) were 15 times as important. “Mass-based interest groups” mattered, too, but only about half as much as business interest groups — and the preferences of those public interest groups were only weakly correlated (.12) with the preferences of the public as measured in opinion surveys.
Gilens and Page frame their study as a test of four broad theories of American politics: “Majoritarian Electoral Democracy,” “Majoritarian Pluralism,” “Economic Elite Domination” and “Biased Pluralism.” “Majoritarian Electoral Democracy,” with its emphasis on public opinion, elections and representation, provides the theoretical backbone of most contemporary political science (including mine). The training of most graduate students (including mine) is primarily couched in that framework. But Gilens’s and Page’s work makes that look like a bad scientific bet, wishfully ignoring most of what actually drives American policy-making.
Testing Theories of American Politics:
Elites, Interest Groups, and Average Citizens [PDF] [Martin Gilens/Benjamin I. Page, Perspectives on Politics]
Rich people rule!
[Larry Bartels/Washington Post]
(via Metafilter)