In Uncivil Agreement: How Politics Became our Identity,1 political scientist Lilliana Mason attempts to explain the recent increase in political polarization. In the process, she proposes a theory that homo politicus is tribal. I found this analysis persuasive, but I think she has yet to develop its libertarian implications.

The Theory

In political analysis, it is natural to assume that citizens are rational, so that their political preferences reflect their views on policy. These policy views in turn derive from some combination of self-interest and concern for the general welfare.

But Mason points out that social psychologists find an emotional component in political behavior.

    Primal psychological influences such as motivated reasoning and social identity are capable of shifting and sometimes entirely determining the policies that citizens support.
    … More often than not, citizens do not choose which party to support based on policy opinion; they alter their policy opinion according to which party they support… citizens want to believe that their political values are solid and well reasoned. More often, though, policy attitudes grow out of group-based defense. Partisanship muddies the folk pathway from interests to outcomes, sometimes sending a person in a wrong direction or further down a path than self-interest and values alone would dictate.

The emotional component Mason identifies is based on social identity. Apart from policy preferences, you simply think of yourself as a Democrat, or as a conservative, or what have you. Your sense of political identity is tied in with other elements of your identity, including your race, religion, social class, and so on.

Mason points to classic experiments in social psychology in which people are assigned arbitrarily to different groups and then compete against one another. Within each group, loyalty develops, and people in one group become antagonistic toward people in the other group. The implication is that even if you had become a Democrat purely by random assignment, with no relationship whatsoever to policy, you would become emotionally invested in victories by Democratic candidates, willing to defend the policy positions of Democrats, and be hostile toward Republicans.

The Findings

By carefully and cleverly using survey data, Mason is able to compare emotional factors to policy preferences as drivers of partisanship. For example, one can ask citizens to self-identify as liberals or conservatives and to assess how strongly they so identify. In addition, one can ask them their preferences on issues where there are well-established liberal and conservative opinions. Conducting such a survey in 2011, Mason found that

    … conservative identification was correlated with conservative policy positions at only r = .24, suggesting that there is a weak positive relationship between the two. In other words, those with intense conservative identities do tend to have more intensely conservative policy positions, but one value cannot be precisely predicted from the other. Similarly, liberal identification was correlated with liberal policy positions at r = .25.

Given these low correlations, it is not surprising that Mason finds that many Democrats hold conservative policy positions. Nonetheless, when these conservative Democrats are asked to use a “thermometer” to measure how warmly they feel about Democrats versus Republicans, they strongly prefer Democrats. Similarly, Republicans who agree with Democrats on policy issues nonetheless strongly favor Republicans emotionally.

Partisan identity has grown stronger in recent decades. As Mason puts it,

    American partisans have become more biased, intolerant, angry, and politically active than their policy disagreements can explain.

She writes,

    Among Republicans in 2000, more than 60 percent of them point to conservatives as people most like them. In 1972 and 1992, this figure was only around 25 percent… By 2000, 35 percent of Democrats point to liberals as the people most like them, compared to 11 percent in 1972.

Mason and other political scientists use the term “cross-cutting identities” to describe sources of identity (religion, social class, etc.) that lead people with different partisan inclinations to come in close contact with one another.

    Those people with the most cross-cutting racial, religious, and partisan identities (and average policy attitudes)… place the two parties 14 degrees apart on the feeling thermometer. This is less than half the size of the partisan gap seen among those with the most moderate policy attitudes. A set of cross-cutting identities is much better than moderate issue positions at equalizing feelings toward the two parties.
“We have become only modestly more polarized on issues. But we have become much more invested in partisan victory and much more hostile toward partisan opponents.”

Today, Americans have fewer cross-cutting identities than they did fifty years ago, and this increases the perceived “otherness” of people with different political identities. Mason sees the decline in people with cross-cutting identities as a major factor in the increase in polarization in recent decades. We have become only modestly more polarized on issues. But we have become much more invested in partisan victory and much more hostile toward partisan opponents.


For Mason, the main implication of her analysis is that we need to find ways to reduce the pressures that add to the emotionalism in politics. If we could restore older cross-cutting identities or create new ones, that would help. Although Mason does not say so, it would seem to me that her findings suggest that institutions that become almost exclusively left-oriented (such as universities) or strongly right-oriented (such as the military) are part of the problem.

Mason argues that party leaders could set better examples, by treating leaders of the opposing party with more respect. However, in an environment of active primary voters, this might be futile.

Mason points to research suggesting that low self-esteem is correlated with strongly emotional partisanship. This implies that politics would be more rational if we somehow could increase people’s sense of self-worth.

In seeking to reduce the political anger and blind enthusiasm that have magnified in recent years, Mason has a laudable goal. But I think that she may be missing some larger implications of her work.

Consider the persuasive case she builds that citizens’ political behavior is driven primarily by group emotions and tribal loyalty. This would seem to me to support a libertarian view that a better society is one in which most decisions are kept out of the realm of politics altogether. Making good choices is hard enough even for the most rational of centralized decision-makers. If the underlying political behavior is not even rational to begin with, then the prospects for beneficial government intervention must be even more remote.


[1] Lilliana Mason, Uncivil Agreement: How Politics Became Our Identity. University of Chicago Press, 2018.

*Arnold Kling has a Ph.D. in economics from the Massachusetts Institute of Technology. He is the author of several books, including Crisis of Abundance: Rethinking How We Pay for Health Care; Invisible Wealth: The Hidden Story of How Markets Work; Unchecked and Unbalanced: How the Discrepancy Between Knowledge and Power Caused the Financial Crisis and Threatens Democracy; and Specialization and Trade: A Re-introduction to Economics. He contributed to EconLog from January 2003 through August 2012.

Read more of what Arnold Kling’s been reading. For more book reviews and articles by Arnold Kling, see the Archive.