Why can’t members of Congress just get along? Critics of polarization often suggest that a key reason for the decline of bipartisanship is the lack of social connections across party lines. If legislators could get to know each other as people, these accounts suggest, they would be able to put aside their disagreements and work together.

It’s an attractive idea—but one that has little empirical support.

Despite the scant evidence in its favor, the socialization hypothesis is a recurring theme in even high-quality political journalism—most recently, The Boston Globe’s excellent “Broken City” series, which examines how polarization is changing national politics. Back in March, an early installment of the series focused on Bob Dole’s failed 2012 efforts to obtain Senate ratification for a disabilities treaty. Michael Kranish’s article noted the schism within the modern Republican Party and the GOP’s move to the right since Dole served. But it also dwelled on the decline of contact between the parties and suggested it had contributed to polarization:

The Senate of 2011-12, in which the treaty would be voted upon, seemed barely recognizable to those who had witnessed the extraordinary productivity of the one that had convened 22 years earlier. But the partnership of the Mitchell-Dole era had been replaced by the bitter, often-unworkable relationship of majority leader Harry Reid, Democrat of Nevada, and minority leader Mitch McConnell, Republican of Kentucky.

The culture of Washington had shifted dramatically. In the Mitchell-Dole years, many members of Congress lived in the nation’s capital much of the year and socialized with colleagues in the other party. By the time of the 2012 session, fund-raising and home-state demands prompted many members to spend far less time in Washington.

Donald Ritchie, the Senate’s official historian, said some senators don’t have time to know their colleagues. “Someone will come into the room and will ask, ‘Who is that?’ Someone from across the aisle. They just don’t have the kind of opportunities they used to have,” Ritchie said. “One of the few times they get to see each other is when they are on the floor voting.”

The topic figured even more prominently in a May “Broken City” profile of Minnesota Rep. Rick Nolan, who returned to the House in 2011 after leaving office in 1981. The piece is full of his wistful laments for the previous era:

[Nolan] sees a Congress that does not meet as often, where few members linger on Capitol Hill. Lawmakers jet in and out of the city’s airport on a dizzying weekly schedule. Representatives pass in hallways but do not know each other’s names.

…Emblematic is a small change that has become one of Nolan’s pet peeves: The House dining room, where he fondly remembers sitting at all hours of the day with his congressional colleagues, now closes in the early afternoon. Lost is an opportunity to make the personal connections that today’s Congress so sorely lacks.

The profile of Nolan closes with him at a dinner for freshmen members of Congress, talking over pizza and saying, “That’s how it used to be.”

Strikingly, however, the most recent “Broken City” installment directly contradicted this hypothesis. The Dec. 1 story, which was written by Tracy Jan, focused on the latest incarnation of the group No Labels. After struggling to attract public interest, the group has shifted its focus to Congress, where it has sought to bring members from both parties together—most notably, by meeting and socializing with each other more extensively than in the recent past.

But despite these efforts, the No Labels meetings—and other initiatives to encourage members of Congress to meet and socialize across party lines—have had relatively little success. Indeed, Jan concludes, “No Labels has been unable to advance, in any meaningful way, a single item from its relatively modest list of goals.”

Consider, for instance, the experience of the No Labels “working group” that calls itself the “Problem Solvers,” which tried to find common ground at a moment of extreme polarization and gridlock:

Every morning during the 16-day government shutdown, about two dozen members of the Problem Solvers would gather in the Rio Room in a windowless basement at the Tortilla Coast, a Tex-Mex restaurant three blocks from the Capitol.

Some days, they met twice a day over coffee and fruit platters, brainstorming proposals to put forth, how they could help the other party save face, and ways to push leadership to end the gridlock. Smaller groups met in lawmakers’ homes and offices.

But the Problem Solvers could not come to a consensus about a path forward. Some wanted to repeal a medical device tax required by the Affordable Care Act as a compromise to Republican demands to gut the president’s health care law. Others thought that would be giving up too much—or not enough.

At one point during the impasse, Senator Collins, the moderate Maine Republican, was invited to meet with the group to discuss the medical device compromise proposal. She did not seem especially impressed.

“I think No Labels is very well intended, but I have not found them to be particularly effective in achieving their goals,” Collins said. “I think they’re still finding their way

We shouldn’t be surprised that the “Problem Solvers” couldn’t actually solve any problems—Congress is deeply divided as a result of structural factors that go far beyond the hours of the House cafeteria. Though the decline of socialization may play some small role in the rise of polarization and partisanship, there’s no credible evidence suggesting it is a driving force. Indeed, cross-party socialization in Congress is likely to have declined partly because of growing polarization (as well as air travel and fundraising demands) rather than the other way around. As the Harvard political scientist Ryan Enos pointed out, blaming the lack of cross-party socializing for polarization is a classic inferential error in which people confuse the manifestation of a phenomenon with the phenomenon itself.

Unfortunately, this story is hard to tell using the tools of journalism, while the socialization hypothesis facilitates vivid anecdotes and narrative. But we should measure laments about what went wrong when members of Congress stopped drinking and hanging out together against the evidence. As Jan’s story shows, good journalists who report on what actually happens when legislators come together over pizza or fruit platters will find little evidence that social contact is enough to bring Washington together.

Follow @USProjectCJR for more posts from this author and the rest of the United States Project team.

If you'd like to help CJR and win a chance at one of 10 free print subscriptions, take a brief survey for us here.

Brendan Nyhan is an assistant professor of government at Dartmouth College. He blogs at brendan-nyhan.com and tweets @BrendanNyhan.