If you had one thing, and one thing only, to predict which Democratic House incumbents would lose their seats in 2010, what would you take? The amount of money they raised? Their TARP vote? Their health care vote? Whether they had a Tea Party opponent? A Nazi reenactor opponent?
The best predictor by far is none of those. It is simply how Democratic their district is. In our dataset, Eric McGhee and I measure that with the percentage of the vote that Obama received in 2008 (see “here”:http://www.cqpolitics.com/cq-assets/eap/campaigns/girouxgems/2008PresidentialVotebyCD-CQP.pdf). It’s not a perfect proxy, but it’s a pretty good one (see “this paper”:http://jackman.stanford.edu/papers/download.php?i=13 by Matthew Levendusky, Jeremy Pope, and Simon Jackman). Brendan Nyhan has a nice graph in “this post”:http://www.brendan-nyhan.com/blog/2010/11/a-first-take-on-election-2010.html showing that district partisanship matters. Our analysis elaborates on his.
In all 402 contested House elections, the 2008 presidential vote in that district would explain 83% of the variation in the Democratic House candidate’s vote share. Nothing else in our dataset comes close.
Focusing on the Democratic incumbents who lost shows how crucial the partisanship of their district was. Consider the graph below, which plots the Democratic candidate’s share of the vote against Obama’s share in that district. The blue dots are the districts where Democrats won. The red dots are the districts where Democratic incumbents lost. The pink dots are any other districts where Republicans won. I’ve labeled the points for Gene Taylor and Chet Edwards just for illustration.
The strong relationship is obvious. And it’s also clear how many losing Democratic incumbents represented districts that either favored Republicans or were narrowly divided “swing” districts. Obama won an average of 67% of the vote in districts where Democratic incumbents won in 2010. He won an average of 49% in those districts where Democratic incumbents lost. Using only the 2008 presidential vote, you can predict which Democratic incumbents won and lost with 85% accuracy.
None of this is surprising, of course. People knew that Democrats in Republican-leaning districts were particularly threatened. But sometimes post-election analysis tends to under-emphasize this. For example, this “New York Times story”:http://www.nytimes.com/2010/11/04/us/politics/04region.html?_r=1&hp frames the election around GOP gains in the Midwest. But the region where a district resides is largely secondary to its partisan breakdown. Similarly, it’s not necessarily striking that the GOP gained seats all across the country, as Pete Sessions claimed “here”:http://www.nytimes.com/2010/10/24/us/politics/24campaign.html. There are swing districts all across the country.
GOP gains had much more to do a simpler fact: when the political winds are blowing against a party, it’s the incumbents in the swing districts that are most likely to be blown out of office.
(Yes, that metaphor is terrible. But at least it wasn’t a water metaphor playing on “wave election”! “Wolf Blitzer, I’m looking at you.”:http://www.theatlantic.com/politics/archive/2010/11/wolf-blitzers-displaced-water-synonym-bingo/65684/)
NB: Our data for House election outcomes is obviously not completely finalized, although we have no reason to suspect that the final version would affect our conclusions here.