The rift between Democrats and Republicans feels wider than ever. According to a Pew Research Center study released in June 2016, 70 percent of Democrats consider Republicans “closed-minded,” 47 percent of Republicans think Democrats are “immoral,” and more than half of politically active Republicans and Democrats say that the opposite party makes them feel afraid, frustrated, and angry.

Could women leaders heal the partisan divide? It’s commonly assumed that women have a natural instinct to cooperate, and some people argue that such collaborative spirit is exactly what our gridlocked government needs. As Senator Susan Collins (R-Maine) put it in the final days of 2012, when the United States teetered on the precipice of a so-called fiscal cliff: “I think if [women] were in charge of the Senate and of the administration that we would have a budget deal by now….With all due deference to our male colleagues…women’s styles tend to be more collaborative.”

But is the conventional wisdom true? Labor economist M. Daniele Paserman, a College of Arts & Sciences professor of economics, wanted to find out.

Previous studies comparing men and women in politics have found some differences in how they lead. Women tend to allocate more money for health and welfare programs, according to US and Swiss research, and women in Congress sponsor more bills and deliver more federal dollars to their districts than their male colleagues do. But are these successes really because of an inherent knack for cooperation?

Lab experiments done elsewhere have produced mixed results. Some suggest that women are better than men at cooperating; others don’t. But Paserman tests hypotheses in the real world, not the lab. So he decided to examine the very place that Collins thought needed a major shot of cooperative stimulus: Congress. “I asked, ‘Where can we find a setting where we can have a good measure of working together and cooperating to achieve a common goal?’” he says. “And that led us to examining the behavior of congresspeople.”

Labor economist Daniele Paserman studies gender differences in politics. Photo by Cydney Scott

Working with Italian economist Stefano Gagliarducci of the University of Rome Tor Vergata, Paserman examined bill cosponsorship in the US House of Representatives. Every House bill has one sponsor—its author and cheerleader-in-chief—plus as many cosponsors as the sponsor can persuade to sign on. The number of cosponsors on a bill is a good proxy for the sponsor’s ability to cooperate and forge alliances, Paserman and Gagliarducci reasoned. Furthermore, the number of cosponsors from the opposite party can serve as a stand-in for the sponsor’s skill at building bridges across party lines. So, looking at every bill introduced to the House between 1990 and 2010—more than 61,000 in all—they set out to count how many cosponsors female sponsors recruited, on average, and compare that to the average for male sponsors. They also looked to see what percentage of those cosponsors came from across the aisle. In August 2016, they shared their results in a working paper for the National Bureau of Economic Research.

They found that, on average, women did recruit a larger number of cosponsors than their male counterparts. But when they split the results along party lines, they turned up something curious: the cosponsorship advantage mostly disappeared for Democratic women, and got stronger for Republican women. The results on bipartisanship followed a similar pattern. Republican women attracted more Democratic cosponsors for their bills than their male colleagues did, but Democratic women actually attracted fewer across-the-aisle cosponsors.

“The very stark difference is that among Democrats, women are actually substantially less bipartisan than men,” says Paserman. “The difference in bipartisanship was quite striking,”

The party-line divergence suggests that it isn’t gender that’s driving the difference in cosponsorship. What, then, is the source? Both Republican and Democratic women tend to lean left of their party’s middle, says Paserman. That means that Republican women have policy positions that are closer to those of the average American voter, putting them in an ideal position to attract cosponsors—especially on issues like health, employment, family, and social welfare, where Republican women share a lot of ground with Democrats.

“Ultimately, what seems to drive cooperation are gender differences in policy positions, rather than something inherently about being able to work across the aisle,” says Paserman. “Even if it’s not exactly what we were looking for at the beginning, I think that is in itself an important piece of knowledge.”

Claudia Olivetti, a Boston College professor of economics, says that the study is “careful and clever” and gives firm and quantifiable definition to the often-slippery concept of cooperation. (Olivetti and Paserman are frequent research collaborators, although not on this work.) It is also a good reminder that apparent gender differences may not be quite what they seem. “When we observe that women are more prone to some type of behavior, like cooperation, we shouldn’t stop at gender, but must look at, where is that coming from?” she says, and in this case, “it is not about gender, but about commonality of interest” with other members of Congress.

While the number of women in Congress has risen over the last 25 years, the gains have tilted heavily toward Democrats. Women now make up a third of Democrats in the House and Senate, but are only about 10 percent of Republicans. In fact, Democratic women outnumber Republican women by almost three to one in the House. That means that the moderate Republican women who proved most able to attract bipartisan cosponsors on bills about health and family issues are rare indeed, and the continuing hollowing-out of Congress’ moderate middle may make them even rarer.

Can we gridlock-proof Congress by electing more women? Paserman doesn’t think so. Of course, there are a host of other reasons to work toward selecting leaders who reflect the true diversity of their constituents. “In the end,” he says, “it looks as if by and large both men and women act as agents for the constituents they represent.” For better or for worse, it seems, we get the leaders we deserve.