FiveThirtyEight extends a hearty welcome to John Sides, a political scientist at George Washington University, who will be writing a series of posts for this site over the next month.
Mr. Sides is also the founder of the outstanding blog The Monkey Cage, which was named the 2010 Blog of the Year by The Week magazine.
John Edwards’s $400 haircut. Senator John McCain’s apparently uncountable houses. President Obama’s vacation in Martha’s Vineyard. Most recently, Mitt Romney’s home renovations. These things suggest that many, if not most, politicians at the federal level come from the upper social classes. Certainly they are much wealthier than the average American. But does the social class of elected leaders actually affect how they vote?
Nicholas Carnes, a political scientist at Duke University, finds that it does. In this forthcoming paper, he studied the connection between the occupational backgrounds of members of Congress from 1901 to 1996 and their voting behavior. Occupational backgrounds have proven to be stronger predictors of many political attitudes than other markers of class, like education and income. (The data he draws upon are here.) He uses a simple seven-category typology: farm owners, businesspeople, other private-sector professionals (like doctors), lawyers, politicians, service-based professionals (like teachers), and workers (industrial, farm and union).
As one might imagine, people from working-class backgrounds were only a small minority during this period. In fact, except for the growing representation of politicians at the expense of lawyers — a shift that derives in part from changes in how the data was coded rather than anything substantive — the occupational backgrounds of members of the House have been remarkably stable.
How does this matter? Using one measure of the ideology of House members — DW-NOMINATE scores — Mr. Carnes estimates that the gap between representatives who entered Congress as workers and those who entered as businesspeople is approximately 10 points on a 100-point scale. A measure of voting behavior, scores awarded by the A.F.L.-C.I.O., showed even larger occupational disparities. On a 0-100 scale where 0 indicates preferences closer to the A.F.L.-C.I.O.’s, workers scored about 20 on average, while service-based professionals scored about 35 and businesspeople about 55.
Furthermore, these disparities were largely stable across decades and are apparent among many subgroups of legislators — Democrats, Republicans, men, women, representatives in safe seats and swing seats, representatives who represent larger and smaller numbers of working-class people, and so forth.
Tracing the policy implications of these disparities is tricky. But Mr. Carnes suggests that the effects could be meaningful:
For instance, while the average member of Congress in my sample voted with the A.F.L.-C.I.O. approximately 56 percent of the time … if the class composition of Congress were identical to that of the nation as a whole, the average member would have supported the A.F.L.-C.I.O.’s position between 59 percent … and 69 percent of the time … which translates into approximately one to three more major progressive economic policies in each Congress (assuming that the A.F.L.-C.I.O. supports approximately two dozen bills every two years).
In Larry Bartels’s book Unequal Democracy, he finds that the voting behavior of senators is much more closely tied to the views of their upper-class constituents than their middle-class or working-class ones. One reason could be that upper-class people are more likely to vote, donate to politicians and contact them to express their opinions. But Mr. Bartels’s results persist even when controlling for the views of actual voters and those who have contacted their representatives. Mr. Carnes’s findings suggest another explanation: members of Congress vote as upper-class people would want them to because members of Congress tend to be upper class themselves.
Increasing the representation of lower and working classes in Congress is no easy feat. As Mr. Carnes notes, their under representation precedes the growth of campaign spending and existed when parties largely controlled the nominations process and when they had less influence. But here is one possibility. The literature on the underrepresentation of women in political office centers on their lack of political ambition and the failure of political elites to recruit them. It may be time to study the underrepresentation of certain social classes in the same way.