« What Republicans Can Learn From Phil Mickelson | The RCP Blog Home Page | Hillary In The Mail »

Thoughts on California 50

I used to be a frequent moviegoer. I am not any more. For some reason, I still really enjoy keeping tabs on what movies come out, and the reviews they receive. The reviewer I read most is Roger Ebert - but not because I think he is the best. Ebert likes far too many movies. He is, after all, a movie buff - and buffs always tend to like the product of their preferred medium more than the average consumer. The great thing about Ebert is his reliability. Once you know that Ebert tends to like too many movies, you can confidently use him as a partial guide. If he says a movie is good, it might still stink. But, whenever he says a movie stinks, you can almost be certain that it does indeed stink. A positive Ebert review, then, is a necessary but insufficient condition for a good movie.

That is how I consider today's special election in CA 50. A win - either today or in June - is a necessary condition for the Democrats to retake the House. But it is not a sufficient condition.

First, the necessary side. The Democrats have to win CA 50 to take the House. If you look at the electoral map, the path to 15 is extremely narrow as it is. In actuality, the real target for the Democrats is probably 17 to 20. In a year when you see many seats switch hands, it is inevitable that the Democrats are going to lose a few. The GOP, for instance, lost 4 seats in 1994. The Democrats lost a few in 1982.

If the Democrats are going to net 15, they are going to have to do very well with the open seats. There are only about 7 of them, including CA 50, that are on the table for the Democrats. Unseating incumbents is a very difficult job these days. They will need to take all 7 of these open seats to take the House - thus, they will have to take CA 50. Losing CA 50 means that they will have to defeat 11 to 14 incumbent Republicans. That sounds easy, but putting together an 11- to 14-person list entitled "Republicans Who Will Definitely Lose" is very hard to do.

This is not to say, however, that CA 50 is a sufficient condition for the Democrats to capture the House. In other words, winning CA 50 is not a sign that the Republican majority is doomed, or even is in serious jeopardy.

CA 50 is Republican, for sure. However, it has become less Republican over the years. It is a district that has skewed Republican in presidential elections by, on average 6.67% in the last 10 years. But the skew has been in decline. It went for Dole by 10% more than the national average, for Bush in 2000 by 6% more than the national average, and for Bush in 2004 by only 4% more than the national average.

For the Democrats to retake the House, they will either have to win more conservative districts than CA 50 or defeat solidly entrenched incumbents. Think of it as a war metaphor. A victory in CA 50 is a victory on Republican ground. However, it is not far enough into Republican territory for a capture of the House. They will have to go deeper: either a victory in a redder open district or a victory against a very secure incumbent.

This is the first reason I do not see CA 50 as being a sign that the Democrats will take the House in the fall: even with a victory in CA 50, there are still harder contests that must be won.

A second, equally important, reason is that you cannot draw any reliable generalizations from a Democratic win in CA 50. The media is prepared to do this, of course. They are still really enjoying their "Democrats Can Take the House!" storyline. If the Democrats take a Republican district like CA 50, they will declare that the GOP majority has gone from "serious jeopardy" to "super-duper serious jeopardy".

The trouble is that CA 50 is not just any Republican district. It is a Republican district whose previous member is wearing an orange jump suit and playing prison softball with George Bluth. It is therefore quite peculiar. This relates to what social scientists call "external validity." To draw an inference about a population (in this case, the 232 Republican-controlled districts) from a sample, the sample must be sufficiently representative of that population. CA 50 is not representative in any sense of the word. It is a lousy proxy for the national political climate. CA 50 is, in statistical terms, an outlier. And you never generalize from the outlier.

Again, it all gets back to Duke. The first rule of thumb in evaluating congressional elections is that partisanship becomes a very poor gauge in districts where there have been or there might be indictments. If a solidly Democratic district like IL 05 could vote out one of the most powerful Democrats in the House in 1994, and TX 22 was well on its way to voting out one of the most effective House leaders in history, anything is possible in CA 50. If the Democrats win CA 50, it might be because the public is so anti-Republican that they will soon install the Democrats in the House. But it might also just be because CA 50's last representative was a Republican and a crook. In other words, from a Democratic victory alone, we could not be able to tell if it is a harbinger of GOP doom.

Personally, I think this district is much more vulnerable than most pundits appreciate. I would not be surprised by an outright Democratic win today. I think that is entirely consistent with a Democratic net of 8-9 seats. Races with indictments in the air are almost never happy occurrences for the incumbent party. Accordingly, I see TX 22, OH 18 and CA 50 as being the three most vulnerable GOP seats. I tend to see the GOP losing 2 of these.