Ask someone to describe the U.S. political climate, and they are likely to say “polarized.” Across a broad swath of policy areas, the two dominant political parties appear to be intractably opposed. Considered across the usual liberal-conservative economic dimension, the ideological distance between the two parties has been rising for decades and has reached its widest point in more than 150 years.

But political scientists have also noted that the ideology of members of Congress can differ along a second dimension that is distinct from their liberalism or conservatism on economic matters. This second dimension does not have a well-accepted name, but, historically, it has been associated with disagreement on social issues.

In their paper, “The Changing Polarization of Party Ideologies: The Role of Sorting,” Philadelphia Fed Vice President and Economist Satyajit Chatterjee and Senior Economic Advisor and Economist Burcu Eyigungor examine the history of interparty differences along both dimensions and find a surprising pattern: The two ideological distances tend to be inversely related. In periods when the ideological distance along the social dimension widened, the ideological distance between the two parties along the economic dimension tended to shrink, and vice versa.

Chatterjee and Eyigungor argue that this inverse association “might not be a historical accident” and could emerge if each party’s policy platform responds to the preferences and composition of that party’s voter base and if each party’s base in turn responds to its party’s platform.

Their explanation casts voters as the central actors in each party’s ideology-formation process, with each political party balancing the desires of its base against the preferences of the nation at large because “a party can come to power only if it wins the general election.”

To understand their argument, it helps to step back and consider an example that incorporates issues outside of the political realm but that nevertheless illustrate their argument’s basic framework. Take, for instance, a political party that strongly favors shopping for groceries in bulk, while its opposing party adamantly supports buying in small batches. Shoppers who feel strongly about the size of their purchases will be attracted to this polarization and will affiliate with political parties accordingly. But what about voters who care more about a second issue, such as whether to carry groceries in plastic bags or in paper bags? Unless these voters’ opinions are closely correlated to their opinions about buying in bulk, the two political parties will moderate their positions on shopping bags, thereby increasing their chances of appealing to a larger number of voters overall. Stated differently: Each party could certainly choose to keep a firm stance on the “paper or plastic” issue, but only at the risk of alienating voters who don’t share that same stance.

To see why parties try to minimize this risk, it’s worth remembering the endgame they have in mind: securing enough votes, across a broad voter base, to win a general election. Chatterjee and Eyigungor show that their theory of ideology formation gives a good quantitative account of the historical evolution of the ideological distances between the Republican and Democratic parties along the two dimensions.

Their theory offers a thought-provoking path for understanding fluctuations in political ideologies. They describe, for instance, how their model could explain the ideological shifts in the latter half of the 20th century. Leading up to the Civil Rights Act of 1964, the ideological distance between the Democratic and Republican parties widened along the social dimension. “According to our theory,” they write, “this widening can account for the narrowing of ideological distance along the economic dimension during the same period.” The relationship worked the same way in the decades after 1964, but with party preferences moving in the opposite direction: A narrower ideological distance along the social dimension played a significant role in creating a wider gap along the economic dimension.

What about the present? Does a narrowing of differences along the second, social dimension explain the historically high polarization along the first, economic dimension? Chatterjee and Eyigungor say no: “During the last three decades, changes in the ideological distance along the second dimension have been small and do not imply significant changes to polarization along the first dimension….” Thus, as they note, their model does not explain why polarization along the economic dimension has risen over the last 30 years. Perhaps, as other studies have suggested, changes in income inequality, changes in campaign contribution laws, and the advent of social media as a news source are to blame. But if voters and parties become polarized enough on a social issue, the authors predict that ideological distance along the economic dimension will shrink.

  1. The views expressed here are solely those of the author and do not necessarily reflect the views of the Federal Reserve Bank of Philadelphia or the Federal Reserve System.