We've spent a lot of time lately documenting the disappearance of the ideological middle in Congress.
Paging through the latest Vital Statistics on Congress -- the indispensable stat book put out by Norm Ornstein and Thomas Mann (among others) every two years -- we came across even more evidence of the long, slow death of the middle. Here's their chart on ticket-splitting in presidential elections, meaning voters in a congressional district who voted for one party for president and another party for House.
According to the Vital Stats data, ticket splitting peaked in the early 1970s with more than 44 percent of all congressional districts voting for one party for president and the other for Congress. Since the 1972 presidential election, however, the percentage of ticket-splitting districts has declined steadily -- with the lone anomaly being the 2000 presidential election when ticket splitting rose to almost 20 percent of all congressional districts. (NOTE: Only 424 districts are included in the chart above because when Vital Stats went to the publisher, there were still 11 House districts with incomplete presidential results due to the after-effects of Hurricane Sandy.)
In 2012, the percentage of ticket-splitting districts slipped into single digits for the first time in 92 years. Just 26 members of Congress currently represent districts that the presidential nominee not of their party won in 2012 -- 17 Republicans hold districts carried by President Obama in 2012 while nine Democrats hold districts won by Mitt Romney. That's a remarkable erosion of split ticket voting considering that four year prior there were more than three times that number of split-ticket districts.
(Sidenote: Romney won 226 districts in 2012 compared to 209 for Obama. According to Clark Bentsen of Polidata, that marked the first time since 1960 that the candidate winning the presidential election didn't win the popular vote in a majority of congressional districts.)
The decline in split ticket voting, like the disappearance of the political center in Congress, is not directly attributable to a single factor. Redistricting, self sorting and the widening gap between Republicans and Democrats on policy matters all play a role. But, regardless of the reasons, the results are the same: Fewer and fewer Members of Congress represent districts that are closely divided enough along partisan or ideological lines to vote for the opposite party's presidential nominee. And that means less members with any electoral incentive to support -- or even occasionally cooperate with -- the President.