Is the source of polarization simple demographics — in particular, the effects of where Democrats and Republicans live and how that affects congressional representation?

(Green Film Company)

That’s the case that Nate Silver makes in an important post today about House districts, in which he demonstrates two things: There are far, far, fewer swing House districts now than  20 years ago, and that with the decline of split-ticket voting fewer House districts vote for the candidate from the minority party in that district. Moreover, he shows, as others have, that the effects of gerrymandering are relatively minor in producing now-common lopsided districts.

Silver doesn’t mention it, but the same thing is true at the state level, with fewer competitive states now than there had been. That’s important, because we see polarization in both the House and the Senate, so any explanation for partisanship in Congress shouldn’t rely very much only on House districting.

Some caution is in order. Measuring the real partisan swing of any state is notoriously difficult. Silver uses the presidential vote, which is a reasonable choice, but it’s still true that both candidates and campaigns can be particularly popular or unpopular in different districts — even when, as Silver shows, there’s something close to uniform swing from one election to the next (remember, after all, that Barack Obama is a constant in the 2008 and 2012 elections). Moreover, as far as I know, we don’t really know the mechanism by which same-party citizens have been clumping together; we have good guesses, but it could just be a phase that will fade away soon, either because housing patterns change or because the parties find new issues to change partisan loyalties.

We know more, however, about the disappearance of districts with GOP presidential majorities and Democratic members of Congress, which is largely a consequence of the end (after 1992) of the last of the Southern conservative Democratic Party — and that isn’t going to change. 

As Silver says, partisanship and polarization haven’t ever been constant in American history. We’re at a high point now, but I’d be careful about projecting too much into the future. It’s also the case that we don’t really know how much of congressional polarization is a consequence of constituent pressures and how much is caused by other internal party mechanics; for example, the conservative marketplace appears to create perverse incentives in which some Republicans profit financially from extremism, especially with a Democrat in the White House. 

Still, constituencies matter, and to the extent that fears of primary challenges are more immediate for many members of Congress than fears of general-election defeat, we’re likely to see high levels of polarization.