Using Senate and House Races as predictors

Coattails. Political scientists make a big deal about the ability of presidential candidates to drag Senate and House candidates into the legislature behind them. Frankly, presidential coattails have been getting rather short over the past thirty years. Reagan's electoral victory in 1980 brought him a Republican majority in the Senate and dragged a number of conservatives into the House, but after that, presidential candidates have generally underperformed on the coattail metric. Clinton actually had negative coattails in 1992: a net loss of 9 House seats and no gains in the United States Senate. In 1996, Democrats gained 8 House seats but lost 2 in the Senate. And of course, Republicans lost two and four House and Senate seats, respectively, in 2000 despite a Bush victory.

These "split" results are all the more interesting given the propensity of voters to increasingly vote unified tickets since the 1980s. That is, voters are more likely today to cast straight party tickets than they were nearly thirty years ago. Coattails have diminished for other reasons, most notably due to the decline of competitive House seats (likely do to residential self-selection and increasingly sophisticated gerrymandering by the parties).

Given this tidbit, it is interesting to see what's happening in the Senate and House races across the country right now.

At the moment, according to some estimates, Democrats may pick up as many as seven or eight seats in the Senate. And Democrats might pick up 5 to 10 seats in the House.

What might this tell us about the presidential election?

Here's a look at the Senate seats Democrats are poised to pick up (according to the website electoral-vote.com):

Colorado
New Mexico
Virginia
North Carolina
Virginia
New Hampshire
Oregon
Alaska

In 2004, all but Oregon and New Hampshire cast electoral college votes for Bush. Now, Sarah Palin certainly secures Alaska for McCain. But if voters in the other states vote for their Democratic Senate candidates and Obama together, then the election isn't even close. Picking up these states and holding those states that Kerry won in 2004 yields an electoral college vote of 294 for Obama. If only NM and VA flip, Obama is the new president.

I simply do not agree with media assessments that this race will come down to the same battleground states of Ohio, Michigan, Florida, and Pennsylvania. The electoral map has been expanded, and in part, Obama can thank the efforts of both Howard Dean and the Democratic Senatorial Committee. Dean built the Democratic party infrastructure in so-called red states, while Chuck Schumer aggressively courted quality challengers to run for the Senate.

Competitive campaign environments saturate the information environment for voters, making it easier to overcome the costs of voting. This should increase turnout and increasing turnout likely benefits Democrats, given that their natural constituencies are less likely to vote than Republican constituents. Mark Warner will win in November, and if Obama wins Virginia, he very likely will have Mark Warner's campaign to thank. Coattails indeed.