Run Everywhere: Open Seats

I've attempted to build a very crude estimate of Barack Obama's performance on a district-by-district basis. I've given him 95% of the black vote, 60% of the Hispanic vote, 55% of the college white vote, and 35% of the non-college white vote. This gives him 43% of the overall white-plus-Asian/other vote, and just over 50% nationwide. I've then assumed that Black voters and Latino voters exhibit uniform voting patterns, and then computed the Partisan Voting Index among white voters to arrive at a number for Obama in the district This has obvious problems in districts that have a large number of illegal immigrants from Latin America, or in the three Cuban districts, but we can use some common sense to throw out these outliers.

Anywhere that Obama gets 45% of the vote or more, a strong local challenger has a shot at winning the district, especially if it's an open seat.

Districts that are currently on various lists of contested seats (Cook, Rotherberg, CQ, etc.) are in bold.

Unsurprisingly, most of the top open seats already have good challengers. The only real surprises are Dave Weldon's seat in FL-15 (Kissimme, Vero Beach) and Ray LaHood's in IL-18 (Peoria/Springfield), where Obama's home state mojo should provide an extra boost. With the right candidates, LA-6 and NM-2 might end up in play, but those are definitely at the edge of viability.

Comments

[TerryVB]

This is fascinating. Can the same technique be applied at the electoral college level to predict the national outcome?

[jeff bain]

The sole Wyoming seat is being contested by a strong Democrat, Gary Trauner, who lost in 2006 by less than 1%. He will way outperform this profile, and win that seat!!

[absent observer]

I would change the methodology somewhat. A number of us have concluded that the only whites who vote against Obama en masse are Appalachian whites. It appears the old worldview of White/Black/other is radically disrupted by Obama's candidacy. A better methodology would have to involve calculating the percentage of hillbillies in the population.

[Trypticon]

has Cubin declared she's not running in Wyo? That would be big news to me, and probably take the seat out of contention with a new rethug coming in. But I think you got this wrong, that she's still in, and that Trauner (sp?) is again running against her. He lost by like 300 votes last time in a nail biter, and will likely win this time, in part because Barack fired up the Dem party in the Wyo caucus like never before. Cubin runs, Cubin looses. Credit a great candidate in Trauner and Barack. So, one, i don't think this is an open seat, and two, even though Obama will not win Wyo, he might help get a Dem into Cheney's old seat, and get one of the stupidist, most useless members of Congress out, and a great, pragmatic, centrist Dem in. Pay attention to this one!!!

[Nicholas Beaudrot]

absent observer: the white vote is adjusted based on the Presidential partisan index among white voters. This is a decent proxy for the Appalachian phenomenon.

[nimh]

Also, it is not true that "the only whites who vote against Obama en masse are Appalachian whites." Whites in the Deep South have been just as averse to voting for Obama.

It's just that in the Deep South, the reluctance of white voters was balanced out by the eagerness of black voters to vote for Barack. In the Appalachians on the other hand there was no black vote of similar proportions, and so the aversion of whites there was enough to land Hillary big wins.

The states with the lowest white vote for Obama so far have been:

16% - Arkansas
23% - Florida, West-Virginia and Kentucky
24% - South Carolina
25% - Alabama
26% - Tennessee and Mississippi

Then come Oklahoma (29%), Louisiana (30%), and New Jersey as first state outside the greater South (31%). In ten more states the white vote for Obama was under 40%: the three early primary states (IA, NV, NH) and AZ, MO, NC, OH, PA, RI, NY.