Partisan rancor rooted in policy disputes, study finds — not in mere tribalism

The vitriolic political spats erupting daily on social media and cable news and in Congress can leave the impression that Republicans and Democrats blindly hate each other. But a new study by Yale political scientists suggests that policy disagreements, not raw tribalism, drive partisan animosity in the United States.

The study, published in the American Journal of Political Science, draws on a set of novel experiments to show that partisan divisions are largely rooted in conflict over prominent policy issues, such as gun control, immigration, and access to abortion.

“Preventing and mitigating the worst effects of partisan hostility requires understanding its root causes,” said Lilla V. Orr, a graduate student in Yale’s Department of Political Science and the study’s lead author. “We found that people generally don’t innately dislike others for belonging to an opposing political party, but rather that their animosity is grounded in disagreement over contentious policy issues.”

Recent scholarship has demonstrated that partisan affiliation directly influences people’s attitudes toward one another. For example, a 2012 study showed that parents are less excited about their child marrying a member of a political party opposed to their own and that the magnitude of this kind of political discrimination has increased since 1960.

Orr and her co-author, Gregory A. Huber, the Forst Family Professor of Political Science at Yale, designed a series of survey experiments to gauge the extent to which policy conflict fuels this rising partisan animosity.

In one set of experiments, survey respondents were asked to read a vignette describing a person. The party affiliation of the individual described was randomized. The researchers also randomized whether the party affiliation was presented alongside information that offered a social cue about the individual’s political beliefs, such as their occupation or home state, and their stance on a policy issue. Some vignettes also featured individuals who took policy positions but did not have a party affiliation.

For example, a vignette might describe a waiter from Colorado who is a Republican. Another might depict a sports fan from South Carolina who is a Democrat and opposes gun control. Yet another would describe a personal injury lawyer who likes rock music and opposes adoption by same-sex couples. Respondents were asked to rate the warmth of their feelings toward each individual.

The researchers found that respondents evaluated individuals more positively when they shared a party affiliation than when they belonged to an opposing party. But when the vignette mentioned party affiliation or a policy position in isolation, sharing a political party was only about three-quarters as effective in generating warm feelings as agreeing on a policy position. When the two factors were presented together, party affiliation had less than half the effect of policy beliefs. Overall, people responded more positively to people from an opposing party with whom they shared a policy position than with members of their own party with whom they disagreed on a policy issue.

“This doesn’t mean that party identification is irrelevant, but it appears that it is a less important source of partisan animosity than people’s positions on the salient issues of the day,” Huber said. “If an individual is passionate about strengthening gun control laws, then that becomes the lens through which they evaluate your politics regardless of whether you are a Republican or a Democrat.”

A second survey experiment explored the traits that people consider most important for positive social interactions. Respondents were asked to imagine that they were assigning seats at a friend’s wedding reception and to consider what information they most wanted to know about the guests. Respondents most frequently asked to learn about how well guests knew the couple, a personality trait, and their hobbies. Gender and sexuality, political partisanship, and religion were requested less often, each by only about 20% of respondents.

A random subset of respondents had the opportunity to learn about guests’ policy preferences. Requests to learn about people’s party affiliation dropped substantially in this group, according to the study, implying that people are more worried about policy disagreement than partisanship as a source of conflict. 

Yale’s Institution for Social and Policy Studies and its Center for the Study of American Politics supported the study.