I have no idea who’s going to do what in the 2010 midterm elections. The mere fact that Democrats hold most of the House seats suggests to me that the odds favor the Republicans picking some up. But this logic from Karl Rove predicting big things for the GOP seems badly flawed:
[T]he 2010 Census could allocate as many as four additional congressional districts to Texas, two each to Arizona and Florida, and one district to each of a number of (mostly) red-leaning states, while subtracting seats from (mostly) blue-leaning states like Michigan, New York, Ohio and Pennsylvania and, for the first time, California. Redistricting and reapportionment could help tilt the playing field back to the GOP in Congress and the race for the White House by moving seven House seats (and electoral votes) from mostly blue to mostly red states.
But of course in the House we vote by Congressional District and not by state. Yes, New York will probably lose a House seat. But at the same time, the New York State Senate flipped from Republican to Democratic control. With the state overwhelmingly Democratic in its electorate, and the state government under total Democratic control, the lost seat will almost certainly be a lost Republican seat. And by the same token, the population growth in Texas, Arizona, and Florida is being driven by growth in the Democratic-leaning Hispanic population. In all these cases, you have to look at the demographics and the redistricting process in detail, not just ask whether it’s a blue state or a red state.
Again, the GOP was unusually unpopular in 2008 and so it seems likely they’ll be more popular and do better in 2010. But the main demographic trends are against the Republicans — they’re strong with old people and weak with minorities.