Political Polarization

Polidata has published an interesting table showing presidential election results by House district for the last four presidential elections. What is interesting are the third and fourth rows of the table which count the number of House districts where the party that won the House election differed from the party that won the presidential, e.g. a Democrat won the House seat but the Republican candidate for president won the presidential vote in that district. These are called split-ticket districts. There is a clear trend that the number of such districts is declining.

Year	Split-ticket districts
1992	103
1996	110
2000	86
2004	59

The Washington Post has written on these findings. The declining number of split-ticket districts reflect the ongoing polarization of American voters.

The steady decline in districts where voters pick different parties to represent them in the White House and Congress reflects in part the effects of the redistricting process, which has created more and more strongly Republican or strongly Democratic districts. But the trend also underscores what political scientists and party strategists have known for several years, which is that party identification is now a powerful indicator of how someone votes in national elections.

Also of note is that, even with the declining numbers, the split-ticket is a reliable predictor of election results. In three of the four elections covered, the winner also won the split-ticket race. Only Clinton in 1992 failed to win this race, and that's probably more a measure of how dissatisfied everyone was with both candidates in that particular election. That was the only election listed in which both candidates scored high in split-ticket districts, reflecting the difficulties both Bush and Clinton had getting support within their own parties.