We've been careful to use the world "projections" rather than "predictions" because there's one element we don't try to predict: what the national balance will be in popular support for the two major parties. You can adjust that "two-party" vote in a single cell in our nifty Monopoly Politics spreadsheet to determine how projections change, but we don't try to predict that national divide.
As of now, however, it's quite clear that if the election were held today, the split would not be 50-50 -- it would tilt strongly toward Democrats, perhaps even larger than the 54%-46% advantage they had in November 2006. The evidence is in the five special elections that have taken place this spring to fill vacancies: two in Louisiana and one each in California, Mississippi and Illinois.
Last fall, Republicans held four of those five seats. That's been reversed to 4-1 Democratic, with Democrats picking up a third seat yesterday in Mississippi when Travis Childers defeated Republican Greg Davis. Those three Democratic pickups were in districts where our projection for Republicans in an open seat in a 50-50 year would be:
* Mississippi 1:Republican with 61%
* Louisiana 6: Republican with 58%
* Illinois 14: Republican with 54%
Despite the fact that since 1992 open seats have overwhelming gone to the party with the partisan advantage, Democrats won all three, including winning with 54% yesterday in a district where they were projected to win only 39% -- meaning fully 15% above their projected share of the vote despite this race being heavily financed on both sides, with very high turnout for a special election.
It's too early to say that this Democratic advantage will continue to November, but Republicans are right to be scrambling.
Looking beyond 2008, however, here's an early warning for Democrats: 2010 could be a year to fear. Democrats already can only win a majority of the U.S. House by winning many seats that a Republican presidential candidate will win in a 50-50 year -- in such a year, Democratic support is more concentrated, resulting in a Republican presidential candidate carrying some 40 to 50 more House seats than a Democrat.
Already, then, the current Democratic majority is more vulnerable to voters locking into their presidential party preference when voting in House races than are Republicans. If Democrats win the White House and pick up more House seats in "red" terrain this fall, they better hope to either be remarkable incumbents or avoid the midterm backlash that so often happens after one party has a big presidential win. If not, and the fiield tilts toward the 54%-46% advantage Republicans had in 1994, literally dozens of Democratic incumbents will be washed out with the partisan tide.