Politicians are not ignoring you

If you're registered to vote in the United States and you're not among the richest of the rich, political scientist Peter K. Enns has a message for you: Your voice still matters.

So does data analysis methodology.

In new research, Enns, professor of government in the College of Arts and Sciences and in the Brooks School of Public Policy, uses statistical analysis to contradict a widespread conclusion that government policy disregards the preferences of the median voter.

"Correcting the record matters because when people are incorrectly told that politicians ignore them, they may decide that it's not worth it to be involved with politics or they may become angry with the political system, increasing support for populist candidates," Enns said. "In other words, the incorrect finding that average citizens are not represented may have actually harmed the political system."

His paper, "How a Seemingly Innocuous and Intuitive Methodological Choice Confused a Generation of Research on Policy Responsiveness," published May 4 in Sociological Science.

In his study, Enns identifies problems with the method at the heart of research first published in 2005 and built upon with papers in 2012, 2014 and 2017, concluding that federal government policy is not at all responsive to the median-income voter, especially when affluent voters disagree. Meanwhile, the original study claimed, economic elites, measured as the top 20% of the income distribution, have substantial impacts on U.S. government policy.

To reach these conclusions, author Martin Gilens (later joined by Benjamin Page) analyzed survey data that includes measures of how much different groups supported 1,779 potential policies between 1981 and 2002 and whether those policies were enacted.

The initial finding has been cited more than 8,250 times in academia, and sparked articles in The New York Times, the BBC and other popular media outlets; its authors even appeared on "The Daily Show."

It caught so much attention because it's extreme, Enns said. "No responsiveness is a shocking claim," he said. But it also resonated with voters who often feel, regardless of actual policy, like their voices aren't being heard by their representatives.

But Enns was uneasy about the conclusion that voters in the economic middle have zero influence. He took a closer look at the public opinion surveys Gilens used - and how he had used them.

Policy preference among different income groups in the U.S. overlap substantially; many policies popular among middle-income group have similar support from those in the 90th income percentile. Gilens chose to eliminate this overlap by only analyzing policies where groups' preferences diverge by more than 10 percentage points. Enns found that the conclusions were skewed by their selective use of policies, which fed into a statistical phenomenon called Simpson's paradox. This effect shows up when there are patterns in underlying groups of data; but when you pool the data together, that pattern reverses.

"Only analyzing policies where a large preference gap exists has become a widely accepted strategy for studying whose opinions policymakers consider," Enns wrote in a blog post that details his analysis. "Unfortunately, this approach has led researchers to incorrectly conclude that policy ignores huge numbers of Americans."

Gilens and Page have drawn attention to important questions about political influence, Enns said. Now, he hopes to advance scholarship on this question in new directions.

"There are political inequalities, certainly. Outcomes are inequal in the U.S.," Enns said. "As social science researchers trying to understand this, if we're mislabeling the problem, we're not understanding the real factors."

The truly important gap, Enns said, is not between middle-income and affluent, but rather between all these and the superrich, the top 0.01% or even 0.001%. He suggests new avenues of study on the country's billionaires and their political influence, which appears in forms far beyond voting - in networking, campaign contributions and exclusive access.

In Enns' analysis, lower- and middle-income voters (marked in the Gilens study as 10th percentile and 50th percentile) are closely aligned on most policies with the affluent in the 90th percentile.

"'Affluent' means well-off, but these are not the superrich," Enns said. "In a lot of school districts, a family with two public school teachers would be in the 90th income percentile, meaning they meet Gilens' definition of affluent. But the family with two public school teachers and two kids, they're not thinking, 'I love it that politicians keep listening to me and ignoring everyone else.' We're missing where the real imbalances in influence are."

Kate Blackwood is a writer for the College of Arts and Sciences.