Looking at the Cook Political Report's latest competitive race chart is in itself informative - the short version is that of the top 108 competitive races, the following ratios apply:
...but there's some interesting things that can be seen with a little sorting. Below is a chart of competitive seats, sorted by Cook Partisan Rating:
|SD-AL||Stephanie Herseth Sandlin||R+9|
|CA-45||Mary Bono Mack||R+3|
|FL-10||C. W. Bill Young||R+1|
|OH-15||Mary Jo Kilroy||D+1|
As you can see, there are a lot of Democratic incumbents in districts that typically vote Republican in Presidential elections, and almost no Republican incumbents in districts that vote Democratic. For that matter, something like 72% of the total competitive races are in Republican districts... which would sound like bad news for the GOP, except that Democratic-held seats make up 75% of both the total and particularly competitive races surveyed by Cook. The midpoint for that list is at R+3; below that point there are 14 GOP districts held by Democrats, and only 7 Democratic ones held by Republicans.
What does that mean, in terms of the 2010 elections? Well, if you assume that every district held by Democrats that's at R+4 and above gets flipped, every incumbent between R+3 and D+0 keeps his or her seat, and that every Republican in a Democratic district loses his or her seat... the Democrats lose 31 seats next year. Assume that an incumbent needs to at least break even (i.e., has at least a R or D+0), and the number goes up to 48 seats lost by the Democrats. Split the difference, and now you know why Charlie Cook is pessimistic about the Democrats' chances next year.
Moving on, here is a map of the cross-party-held seats:
As the key notes, Red shows states with at least one vulnerable Democrat-held seat; Blue shows states with at least one vulnerable GOP-held; and Purple are states with at least one of each. Again, there's a lot more in the first category than there are in the second and third. The reason that this is important is that the exceptionally wide geographical spread of the territory that the Democrats will have to defend next year should make for some entertainingly hard choices for the White House. Assuming that the administration comes out swinging for the midterms... well, it can either pick its spots (at a guess: NY, OH, & PA) and abandon the rest, or it can try to defend everything (and thus defend nothing). Obviously, neither choice is really optimal.
Conclusion: as of this moment, one year in it looks good for the GOP to get a large number of seats back. One year in.
This will change.
Crossposted to Moe Lane.