The presidential candidates might not agree on much, but on the stump, in their advertisements, and in debates they all have made it clear they agree on one thing: The other party is the Devil.

It isn’t just this year. Politicians increasingly pillory and oppose the other party’s ideas. As Alan Abramowitz of Emory University has shown, partisan voters have followed their lead, increasingly disdaining the other party, as well.

Politicians and interest groups think that lambasting the other party helps win elections. They may be right. But, in a recent article, we show how these attacks might also have long-term consequences that make it harder for them to represent their districts and craft successful policies.

Why? When citizens come to think of the other side’s politicians as evil incarnate, many give up telling these politicians how they feel. The result? Politicians hear more and more from constituents of their own partisan stripe — often the strongest hard-liners — and not constituents of the other party who might encourage compromise.

We show this dynamic at work in a series of studies reported in a forthcoming article. Republican voters are more likely to contact their elected official if the official is also a Republican. Democratic voters are more likely to contact an elected official if the official is a Democrat. As a result, even politicians sincerely hoping to represent public opinion may unwittingly skew their positions in favor of the voters who share their partisan leanings — the ones most comfortable reaching out to them.

We document this basic pattern in several different ways.

Who you gonna call?

In one set of studies, a progressive advocacy organization emailed thousands of Americans, asking them to contact their senator to urge opposition to the FISA Improvement Act. Seventeen states are represented by one Democratic and one Republican senator. In those states, the organization randomly assigned which of the two senators the email recipient would be encouraged to contact.

Citizens were 6.4 percent more likely to sign a petition if it was addressed to the senator of their own party. They were 13.3 percent more likely to make a phone call to a senator of their own party.

In a separate study, we examine congressional districts that had narrowly decided House congressional races. More than a year after the election, constituents in these districts were 27 percent more likely to say they contacted their congressional representative if the winner was in the “their” party.

We’re good. They’re evil.

Why do citizens prefer to contact representatives who share their party affiliation? One potential explanation is that, on divisive topics, citizens expect to be ignored by the opposing party.

But their hesitation appears less explicitly political than that; in another study, we find that Americans are just as uneasy about asking an elected official of the opposing party for help with casework and constituent services as to express an opinion on legislation.

Our best explanation for citizens’ hesitancy draws on work showing that citizens think about their own party’s politicians as fellow group members, and as the other party’s politicians as enemies — exactly how both parties’ presidential candidates (and so much else in the partisan public discourse) are encouraging their partisan compatriots to feel.

Political scientists are finding lately that citizens think about political parties as that kind of ingroup-outgroup dynamic. More and more, people say that thinking about the opposing party makes them angry or unhappy, even when partisans agree with each other, that it influences hiring decisions unrelated to politics, and that partisans care more about how their political side fares than about losses or victories on specific issues.

But a classic finding of social psychology is that people feel more comfortable talking with their own group’s members and avoid communication with non-group members. This is exactly what we find voters doing with regards to political party. Our results suggest that partisan social identification is even spilling into what opinions elected officials hear.

Talk amongst yourselves

So what? These patterns matter because hearing from constituents does affect what politicians do. Political science research finds that, because they lack better information, members of Congress are highly responsive to constituents’ phone calls, letters, petitions and probably even emails. That feedback can let them know the risks and rewards attached to supporting or opposing a given bill.

There’s been concern lately that Americans increasingly live in echo chambers of reinforcing views. So far, worry has focused on citizens receiving biased information — say, getting a one-sided stream of information online. But it could go both ways. Citizens may be sending (or not sending) biased information to politicians — reinforcing each party’s tendency to govern through an ideologically tinted lens.

There’s a way out

But there’s an optimistic corollary. Evidence consistently finds that one of the most effective ways to reduce biases is to make people aware of them.

If elected officials realize that they’re hearing only from their own party and not the rest of their constituents, they may be able to resist the temptation to think they’re getting the whole picture. And if citizens make a conscious effort to talk to political opponents, no matter how uncomfortable that may seem, the other side just might listen.

Timothy J. Ryan is an assistant professor of political science at the University of North Carolina at Chapel Hill. David E. Broockman is an assistant professor at the Stanford Graduate School of Business.