I have been working on a model to predict the makeup of the House based on the generic ballot and past margins in individual House districts. It is loosely based on the models used in Britian to translate general polling into seats won or lost. I would appreciate it if anyone would suggest feedback for improvements.
Thank you for your time.
Here is how it works:
1. Subtract the margin from the generic ballot in 2010 polling from the generic ballot margin in 2008. In 2008 the Democrats won 10.68%. The generic ballot average is the GOP +.5%. So the total shift in 2008 would be 11.18%.
2. Subtract the change in the generic ballot margin from the Democratic margin in each House district. So if Democrat had a margin of 12%, this model would predicts a Democratic Margin of .82%.
3. If the member retired, or ran unooposed, use the Cook PVI instead of the margin in the district.
4. Adjust results to reflect cases where there is no opponent in a CD.
The model is here:
http://spreadsheets.google.com…
Right now the model predicts a House of 225 Republicans and 210 Democrats, which seems to low for the Democrats.
Though the results make me a bit ill…..
If I’m understanding right, if the current differences between the ’10/’08 generic ballots hold in all districts
Ds will lose 47 seats in the House.
W/r/t a baseline, anyone can use this and modify by:
1) projecting a D/R generic ballot on election day
2) modifying with obvious factors, e.g. unopposed candidates, clearly safe incumbents, etc.
Thank you for your work, it will help all of us think more rigorously about each of the 435 races (for those of us who have time).
Though it seems like a bit of a blunt instrument. It seems a little hard to believe that Republicans would take control of the house, easily knock off Reps in Blue-ish Districts with weak challengers like Phil Hare and Chellie Pingree, take half of the fucking Massachusetts delegation down in a blaze of glory, manage to hold Joe Cao’s seat by a comfortable margin, BUT still fail to defeat Betsey Markey.
Nonetheless, one interesting thing this model demonstrates is the failure of Republicans to fully capitalize on the climate. Hare and Pingree are good examples of Representatives against whom Republicans should be putting up strong, moderate challengers, if this climate is as good as everybody is saying, anyway. Yet, these two will easily defeat a some-dude teabagger come November.
Massachusetts is just a fluke because the Republican party didn’t bother putting up candidates in most of the districts, right?
which doesn’t sound all that implausible. I’ve got
25 myself, but my analysis is more of a micro, race by race candidatecentric view, so for you to get -36 while looking more at the national macro picture sounds about right.Certainly not every race seems right, like Critz losing in PA-12 just months after cruising to victory, or Niki Tsongas losing in MA-5, but I’m sure we’ll have a few surprises on election night anyway. Good job.
You know Gallup has a model based of the generic ballot? They say at even at 51% of the two party vote the range of seats the Democrats could hold is between 240 and 262.
You have them losing 36 seats based of 51% of the two-party vote. While Gallup would have them losing around 6 seats.
I’m not sure if I fully trust Gallup’s methodology. I think we’re going to lose something like 25ish seats.
Here is an old post from Gallup about it. http://www.gallup.com/poll/124…
And here’s a link Gallup’s expected results based on share of two-party vote (for midterms): http://www.gallup.com/poll/124…