Poli-Sci Perspective is a weekly Wonkblog feature in which Georgetown University's Dan Hopkins and George Washington University's Danny Hayes and John Sides offer an empirical perspective on the issues dominating Washington. In this edition, Hayes looks at research showing how harsh language can lead to more polarization. For past posts in the series, head here.

When House Speaker John Boehner told Senate Majority Leader Harry Reid to “go f--- yourself” in the midst of the fiscal cliff negotiations last week, it was but the most recent instance of the breakdown of congressional comity.

House Speaker John Boehner's speech polarizes. (J. Scott Applewhite, File/Associated Press)

In October 2009, Democratic Rep. Alan Grayson called former Vice President Dick Cheney a vampire, replete with “blood that drips from his teeth.” Cheney, for his part, had actually given Boehner his fiscal cliff talking points back in 2004, when on the Senate floor he recommended that Democrat Patrick Leahy “go f--- yourself.”

Big whoop, right? Doesn’t this just tell us that grown men can act like children, something we’re already reminded of every time we see a Judd Apatow movie?

But spats like the Boehner-Reid dust-up might do far more. Some recent research suggests that when politicians engage in PDA – public displays of anathema – Americans’ attitudes toward the party they don’t identify with grow increasingly negative, contributing to mass polarization.

Political scientists for years have debated whether the American public is polarized along ideological lines. On one hand, some research portrays the vast majority of Americans as centrist on most issues. Other work has argued that Republican and Democratic voters’ policy views are far more divided.

But Shanto Iyengar, Gaurav Sood and Yphtach Lelkes suggest in a new article that “affective polarization” may better describe the divide between party identifiers. Instead of focusing on ideology and policy positions, Iyengar and his colleagues draw on a psychological concept called social identity theory.

They argue that simply identifying with a political party, as most Americans do, is enough to generate unfavorable attitudes toward the other side, or the “out-party.” (This idea should feel pretty familiar to Red Sox and Yankees fans.) And a variety of survey evidence shows that in recent decades Democratic identifiers have come to view Republicans increasingly negatively, and vice versa.

First, Iyengar and his colleagues examined partisans’ ratings of the other party on a “feeling thermometer” – a measure that allows a survey respondent to say how “warm” or “cool” they feel toward another person or group.

In surveys in the 1980s, about 40 percent of Americans gave the out-party a rating lower than a neutral score of 50. But that figure climbed to 53 percent by the 1990s, 56 percent in 2004, and 63 percent in 2008. In 2008, the average rating of the out-party on a 0 to 100 scale was around a decidedly chilly 30. By way of comparison, the average score that Catholics and Protestants gave each other was about 66.

Second, the authors show that since 1960, the percentage of partisans who report they would be upset, displeased, or unhappy if they had a child who married a member of the out-party has climbed. In 1960, that number was 5 percent for Republicans, but had risen to 49 percent by 2010. For Democrats, the number increased from 4 percent to 33 percent. Really, who would want their kid to marry a vampire or a dictator?

Finally, Iyengar and his colleagues found that partisans are increasingly likely to describe members of the out-party in unflattering terms. In 1960, 21 percent of partisans said the other party’s supporters were “selfish,” but by 2008, that figure had risen to 47 percent. They also grew less likely to describe the other side as “intelligent,” with that number declining from 27 percent to 14 percent from 1960 to 2008.

To be sure, some of this “affective” polarization stems from the growing ideological differences between the parties. In particular, the authors found that policy preferences on social welfare issues were significantly correlated with how favorably Americans’ rated the out-party in comparison to their own party. This is consistent with Pew Research Center data showing that the largest “values” gap between Republicans and Democrats emerges on issues related to the social safety net.

But Iyengar and his colleagues suggest that issue positions are only part of the story. They argue that exposure to campaign discourse also breeds contempt for the out-party.

They find that in 2004 and 2008 battleground state residents and people in states with more negative advertising were likely to have more polarized attitudes than other Americans. Ultimately, the authors’ data don’t allow them to test this proposition in a thorough way, but their findings are consistent with work showing that cues in the news media can activate social identities in the context of war.

That’s why Boehner’s choice words to Reid are more than just palace intrigue, to use the phrase of the week. When that kind of denigration is widely reported to the public, it may reinforce the social identity that characterizes partisan attachments. And that may ultimately contribute to the corrosive views of the out-party that Iyengar, Sood and Lelkes colleagues document.

Call it the “go f--- yourself” model of polarization.