Speaking at the Conservative Political Action Conference last Friday morning, Senate Minority Leader Mitch McConnell had this to say about factionalism within the Republican Party:
"The mainstream media loves nothing more than to sow division among conservatives. They love it when we take shots at each other. It gets more coverage than a D.C. snowstorm."
As a member of the media, my Illuminati overlords prohibit me from acknowledging media bias. But I might point out that McConnell’s own crew of Senate Republicans have divided themselves quite ably without anyone sowing anything, based purely on how they have voted in 2013.
To find out how united or divided both Republicans and Democrats are in the current senate, I started with a simple concept: For every member, I calculated which other senators voted the same way at least 75 percent of the time. In effect, this organizes the senate as a mini-Facebook of 100 users, in which any given pair of senators are friends if they meet this 75-percent threshold. When visualized, the picture looks like the final stages of cell division when a Paramecium reproduces, in which a formerly unified body has nearly split into two distinct creatures.
Visualizations like this one work by treating the senators as particles that repel one another, and treating the connections between them as springs that hold them together. Because the Democrats vote so cohesively, with few defectors, they are held together by a large number of springs. Because Republicans are less organized and less disciplined in their opposition—or more bipartisan, depending how you want to cast it—their side of the network is more diffuse.
The 75-percent figure I’m using to draw connections is a bit arbitrary, so you can adjust it with the slider at the top to see how connections materialize or disintegrate based on that threshold of bipartisanship. As you drag the slider to the right, you’ll see that the Republican coalition breaks up far faster than the Democratic cluster. In fact, 22 Democrats have voted together in all 37 votes so far in 2013.
If 37 votes doesn’t seem like enough information to draw any conclusions, just select a previous congress from the dropdown menu in the upper right corner. It’s the same picture for every session of congress.
You may also notice that voting discipline seems to be function of who’s in charge, not an intrinsic value of liberalism. (Anyone who has seen Democratic politics from the inside knows this in their bones.) If you select the Republican-controlled 108th Congress, you can see a similar effect in reverse. At a 75 percent threshold, there is not a single connection between the clusters. As you drag the slider to the right, you see the blue dots scatter much faster than the red dots.
Alexander Furnas, a political scientist and fellow at the Sunlight Foundation who examined this data for me, suggests this is because of the control the majority party holds over what comes to a vote.
“The majority is unlikely to bring things to vote unless they can garner 60 votes to be filibuster proof. Given that usually the majority does not have a filibuster-proof supermajority, that means that they are bringing things to vote that they know can peel off a few votes from the minority,” he said. “It is likely that that process contributes to the lower voting cohesion in the minority party.”
At high thresholds, the factions begin to emerge. If you select Florida Republican Marco Rubio in the 113th congress and drag the bar to the 80 percent mark, a purely organic image of the tea party emerges in the mini-network of like-minded voters: Rubio, Rand Paul (R-KY), Ted Cruz (R-Texas), James Risch (R-ID), Mike Lee (R-Utah) and Tim Scott (R-SC).
There are dozens of ways to slice and dice lawmakers based on their voting behavior, most of which place members of Congress on a linear scale of partisanship. Congressional Quarterly, for example, measures the percentage of the time a member votes with his or her party, counting only votes in which a majority of Democrats opposed a majority of Republicans. Political scientists tend to prefer more complex methods like NOMINATE, which is more statistically rigorous but less intuitive.
But I do not think these measures fully capture what the Senate is: not so much two homogenous sides as a series of affiliated cliques. In other words, more like the lunchroom scene in Mean Girls than the gym scene in West Side Story. McConnell’s denial that there are divisions within the party, in fact, has clear echoes of Regina George herself:
Can I just say that we don't have a clique problem at this school? And some of us shouldn't have to take this workshop because some of us are just victims in this situation?
Regina wasn’t correct about that either.