It’s not that ordinary Americans never get what they want from the political system. It’s just that they only get what they want if the elites want the same thing.
A recent paper from Princeton’s Martin Gilens and Northwestern’s Benjamin Page is making a splash with the claim that they’ve assembled a data set (download) that shows the opinions of ordinary Americans have only a limited influence on actual policy outcomes. Here’s the key thesis (emphasis ours):
Multivariate analysis indicates that economic elites and organized groups representing business interests have substantial independent impacts on US government policy, while average citizens and mass-based interest groups have little or no independent influence.
Such a proposition would seem to punch a major hole in America’s democratic bona fides. Can these academics back it up?
Gilens and Page based their findings on a study of 1,779 policy issues—dating from 1981 to 2002—for which polling data existed that was broken down by income level for respondents. Researchers looked for issues with clear yes/no answers and specific issues pertaining to the federal government policy. (They also looked for questions that were phrased categorically, rather than conditionally.)
They then assessed responses by income level, using the answers for those in the 90th income percentile as a proxy for opinion among the affluent. These guys aren’t necessarily super rich, by the way. The 90th percentile corresponded to an income level of about $146,000 in average household income. But the writers argue that the difference between the opinions of the average citizen and the super rich are likely to look like a more extreme version of the difference between the average citizen and the merely affluent.
As a proxy for opinion among business elites, researchers looked at Fortune’s Power 25 lists. They also added groups from 10 key industries that reported the highest lobbying expenditures. Then they started searching through the public record—newspapers, press releases, Congressional Quarterly articles—to find out where these groups stood on the policy questions for which they had polling data. They coded the business groups’ position as “strongly favorable,” “somewhat favorable,” “somewhat unfavorable,” or “strongly unfavorable” to the change. And then they created a kind of index based on the responses… Read more from Quartz.