Do politicians need to be friends to cut a deal?

President Obama recently spoke about the value of developing personal relationships in politics, something that a bunch of political scientists have been writing about recently (examples below). Jonathan Bernstein wrote about this in his blog and backed up the President by arguing that scholars of the presidency “have been telling us for decades that schmoozing just isn’t all that important.”  However, other recent political science research has investigated the value of social networks among political actors and draws some conclusions that contradict the President, sort of.

Mr. Obama’s argument is not wrong, but it is incomplete.  He argues that deals get worked out (or not) in Washington not because of who knows whom, or who likes whom, but because of ideology and policy preferences. This is true. An examination of roll call votes in Congress would reveal that political party, our best and simplest way of measuring a legislator’s ideology, explains more than 90 percent of the variance across all votes.  But how do we know that one’s party (or ideology or policy preference) has not been informed by social interactions?  And what about the remaining 10% of votes that aren’t explained by these primary factors?

Politicians’ personal social networks inform their thinking, and their voting.  Research shows that legislators’ personal friendships with one another can affect their voting patterns, or at least that legislators develop strategic relationships with one another for the purposes of forging future coalitions or, in the least, to have a sounding board against which to “check” one’s preferences.   Other scholars (and here) have argued that by examining patterns of cosponsorship in the U.S. Congress we observe the effects of consequential social networks that ultimately affect legislative voting.

We also have good evidence that voters’ social networks affect patterns of political activity in the civic arena (see here, here and here).

The evidence is mounting and compelling: social networks play an important role in human behavior. What is not yet entirely clear is how these networks matter and to what extent they matter. In a forthcoming book with University of Michigan Press, Nils Ringe and I argue that legislators form social relationships with one another strategically through Legislative Member Organizations (such as caucuses in the U.S. Congress). The relationships forged through LMOs cross typical partisan and institutional divides in ways that are ultimately productive for lawmaking.

My view of the relative importance of social networks in political decision making is heavily influenced by the conclusions we draw in this research.  That is, social networks are consequential, but most likely in an indirect way.  Who you know, who you talk to, how frequently political actors socialize with one another—these are things that may affect political events on the margins, or at the agenda setting stage where effects are more difficult to discern.   

But sometimes it’s the margins that count. President Obama noted that he enjoyed playing golf with Speaker Boehner, but that it didn’t help them reach a deal on the fiscal cliff.  However, the deal came together because of negotiations between Vice President Biden and Senate Republican leader Mitch McConnell, who have been colleagues and friends for more than 20 years. Biden and McConnell have successfully helped the nation avert crises three times in recent years (including the deal about the 2010 expiring Bush tax cuts, and in August 2011 regarding the debt ceiling). Explaining why Biden and McConnell seem so successful at these deals, one Republican staffer explained, "It's a buddy thing." So, with all due respect to the President, it may not entirely be true that social relationships had no impact on finding a deal on the fiscal cliff.