Democrats and Republicans are tied in the generic Congressional ballot. But Republicans are unlikely to lose control of the House, to a significant extent because redistricting and incumbency give them a +2.5% advantage. A nongeneric, district-by-district look also favors maintained Republican control. Based on these two lines of evidence, Democrats appear likely to gain 2-22 seats, but their takeover probability is only 13-23%. This estimate does not take into account one unknown influence: in the last two weeks, President Obama has taken the lead in national surveys, moving up by about 1.5%. This raises the possibility of hidden House strength for Democrats in the form of Presidential coattails.
The generic national Congressional ballot, a loose predictor of the national House vote, is tied (D+0.0+/-1.1%, n=3, Nov. 1-4). However, this is unlikely to be enough for Democrats to take control of that chamber of Congress. As I pointed out several weeks ago (“Bush v. Gore times five,” October 16), partisan redistricting of an unusual intensity has created a situation in which one side, the GOP, has an advantage of 2.5% before any votes are counted.
For my final snapshot (and prediction), I do what I did two weeks ago (“House: prediction update and GOTV advice“, October 24): combine (a) my prior analysis of redistricting/incumbency and (b) Pollster.com district-by-district data. The two pictures are somewhat convergent.
Since my last update, undecideds have dropped from 10% to 5%, either because voters are more aware of their commitment or because pollsters are pushing them harder. I estimate the popular vote outcome as D+0.0+/-3.0%. Republican retention of the House is very likely. Based on these numbers, the House should be more closely divided.
Pollster.com also gives a breakdown of individual races. Individual district polls, where available, give a similar picture. They have 17 tossup races, 14 leaning D, and 21 leaning R – a total of 52 races in play. The remainder are relatively safe, 171 D and 212 R. These numbers are a slight improvement for Republicans compared with two weeks ago.
From these numbers come the following predictions:
- Generic ballot-based: 210 +/- 9 D, 225 +/- 9 R.
- District-by-district: 194 +/- 17 D, 241 +/- 17 R.
- Democratic takeover probability: 13-23%.
- Combined Bayesian prediction: 205 +/- 10 D, 230 +/- 10 R.
The two approaches both point toward Democratic gains – but retained Republican control.
There is one remaining factor that could push things a little bit. In the last two weeks, the national Presidential numbers have moved toward Obama by about 1.5%. Here is an average of national polls, separated day-by-day as I have done previously.
Keep in mind that in the last three elections, national poll aggregates have only come within 0.3-2.4% of the final outcome. So all we can get out of national polls is that that relatively, opinion has moved in the President’s favor.
However, that is not our only source of information on the Presidential horserace. On Saturday (“How likely is a popular vote/electoral vote mismatch?“, November 3) I pointed out that the true Presidential popular-vote margin is likely to fall near a state-poll-based measure, the Meta-Margin. Tonight I’ll use that (with some input from national polls) to give a final popular-vote estimate.