Princeton Election Consortium

Innovations in democracy since 2004

Outcome: Biden 306 EV (D+1.2% from toss-up), Senate 50 D (D+1.0%)
Nov 3 polls: Biden 342 EV (D+5.3%), Senate 50-55 D (D+3.9%), House control D+4.6%
Moneyball states: President AZ NE-2 NV, Senate MT ME AK, Legislatures KS TX NC

House 2012 prediction (final)

November 5th, 2012, 1:48pm by Sam Wang

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) 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. 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.

Tags: 2012 Election · House

108 Comments so far ↓

  • Osso


    I still think that early voting is going to surprise us all and hand over the House as well to the Democrtas.

    Stay tuned!

  • 538 Refugee

    Compared to other possibilities, I can live with two of three for now I guess.

    • Olav Grinde

      And given the right two, President Obama also gets to nominate new justices to fill any vacancies on the Supreme Court. That could prove … supremely important.

    • Ohio Voter

      I wonder who Obama would try to…court…for those vacancies?

    • Craigo

      Diane Wood and Jennifer Granholm were the other frontrunners for the last two spots.

    • 538 Refugee

      Good point about the Supreme Court. I feel Obama made an honest effort to tone down the partisan politics but found out that didn’t work. While I’ve seen signs of him backing off that stance just a tad, he seems to still believe reason will triumph. I wonder if a second term will see those gloves come off?

    • Craigo

      I’m personally interested to see whether Harry Reid does anything about his newfound distaste for the filibuster.

    • Ray Umashankar

      Newscasters and pollsters are already being blamed for sudden shortage of crows likely to ocurr on November 6th in the U.S.

  • securecare

    I too think that the results will surprise.

    I would like the results to be Ds holding a (1) seat advantage in the House but….

    • Osso

      As I wrote before, the National and State Polls are mixing up two distinct events, one for about 35 States with early voting and the other voting only on election day…..

      I think most pollsters will be eating a lot of crows after the elections.

      Stay tuned!

  • Neus

    Dr. Wang, you are a hero—-Many thanks!

    Our house is waiting for your prediction with bated breath.
    Also there will be many “jogs” to the microwave for popcorn tommorow night!
    (Reminds me of the days I used to pace up and down outside the delivery room for my daughters!)


    • Strangeite

      You should pop your corn on the stove. Not only is it FAR cheaper, but tastes much better (especially if you had a tablespoon of bacon grease to the pan).

    • Steve Norton

      I’d wait till after tomorrow night before calling Sam a hero. But I have a feeling Sam won’t be eating any bugs. But if I’m wrong, it’ll taste better cooked on the stove with a little bacon grease.

  • Montykoolaid

    Well Sam, get that bug ready, and Nate Silver, get that check ready.

    Romneys going to win.

    Not because your model is wrong, or your math, or Romney actually got more votes…

    They are going to steal it. Plain and simple. This means too much to millionaires and billionaires.

    I hope I am utterly and completely wrong. I wont sleep until Tuesday Night/Wednesday morning.

    • Vicente

      I’m all for emotional hedging, but can we keep conspiracy talk down until at least after the election?

    • Olav Grinde

      Mr Money&Koolaid: “Romneys going to win. Not because your model is wrong, or your math, or Romney actually got more votes… They are going to steal it. Plain and simple.”

      It I understand correctly, Mr Koolaid, you’re saying that even more votes will be cast for Obama, more Romney votes will be counted?

    • Sebastian Jackson

      Obama is not stupid. He has 2,500 lawyers in Ohio ALONE to blow the whistle on any GOP shenanigans:

      The “koolaid” in your name is very apt, I think.

    • Steve Norton

      I agree with Vicente. I really don’t think it’s gonna happen. Come Wed, I think you’ll be pleasantly surprised (I’m hoping).

    • Montykoolaid

      Olav, a retired NSA agent has apparently found proof that the GOP has been “vote flipping” in larger areas. That is, same number of votes go in, but they “flip” to gop.

      Apparently 10% of the vote can be “flipped” without arousing suspicion.


    • Olav Grinde

      Sebastian: “Obama is not stupid. He has 2,500 lawyers in Ohio ALONE to blow the whistle on any GOP shenanigans:”

      Yes, but do they have software engineers examining the source code in the Romney-owned voting machines? And analyzing the “experimental” software patches Secretary-of-state Jon Husted had installed in 25 counties?

  • James P

    I would like to see the House change to the Democrats, but I realistically do see the Republicans holding on to it. People mostly keep their Congressmen. My hope is that there are enough Republicans in the House to work with the Democrats.

    • RocketDoctor

      I’d like to hope that the House could be flipped in 2014 but as an off-season election with a Dem President I doubt that is very likely either.

    • Craigo

      That’s a common refrain in American public opinion.

      Congress is full of crooks…but not my congressman.

      School are falling apart…but not my kids’ school.

      Crime is out of control…but not in my neighborhood.

      The grass is always greener on our side of the fence, it seems.

    • skmind

      People do keep re-electing their politicians. I had examined this last year when Congress was teetering around at 13%:

    • MoldyOregonian

      yes, here’s to hoping that a decent group of gop congressman will reject the tea party obstructionist stance and start actually doing the job they are paid (well paid) to do.

  • Rafalca Romknee

    I always thought the House was a bridge too far.
    I’ll be happy with the WH and the Senate.
    And Turtle Face McConnell eating a huge pile of crow.

    • Strangeite

      Think of us poor souls that have him and Crazy-Pants Paul as our two Senators.

    • Rafalca Romknee

      But Strangeite….
      I hear you are going to run Ashley Judd against him in ’14…which i support whole-heartedly.

    • Strangeite

      Unfortunately, most high profile Dems now only look at running against McConnell as a platform to garner statewide recognition in advance of a gubernatorial run.

      Be the sacrificial lamp in 2014 against McConnell in order to build a team, develop donor lists, increase name recognition and solidify your image by running a sickingly sweet campaign, although the while knowing you are going to lose.

      Lose in November 2014, so that you can win the primary in May, 2015.

  • bks

    If wild type voters were exposed to too many “99.9% chance of Obama victory” stories, it could depress turnout. However the media insist on going with “it’s balanced on a knife-blade” coverage which should favor a better than average D turnout and thus a better than average result in the House. That’s my story and I’m sticking to it.


  • CoffeeDawg

    It’s my hope that the D’s hold the Senate and Presidency – and make up enough seats in the house to send a message that these shenanigans will cause the Tea Party to lose everything in the next election if they keep this up. I have no idea if the R’s will cave and have to get down to the business of governing or if they’ll dig in their heels and make a stand. If it’s the latter, I look for continued majorities in all 3…

    • Shawn Huckaby

      Maybe so, but in the interim we have votes on the debt ceiling and sequestration at a minimum, so it’s got much more serious ramifications than just posturing for 2014.

  • Khan

    I want this on the record.

    Gallup and Rasmussen are fudging their polls. Both go offline for Sandy and come back with enormous swings that put them inline with other polls now that accuracy matters. It’s absurd.

  • Khan

    Other comment stuck in mod purgatory, so here we go again.

    It’s pure hackery that both Ras and Gal go offline for the hurricane and come back with substantial and inexplicable swings defying their previous trending and it not surprisingly puts them both closer to inline with existing averages.

    • Khan

      Also, Gal now has the President up 3 with registered voters.

    • Craigo

      That’s been SOP for Ras since 2002. Gallup, on the other hand, is probably honestly reporting a shift in their data.

  • counsellorben

    Quick note: the Bayesian prediction totals to 440, rather than 435.

    Also, given the extent to which gerrymandering has rendered so many House districts safe, I would expect the House numbers to come closest to the district-by-district prediction. If the House comes in with more Democrats than the Bayesian median of 205, then it is likely due to underlying polling issues.

  • Robert Dora

    Dear Dr Wang,
    Thanks so much for all of your work.
    You bring sanity to the discussion. It is a welcome break from all of the partisan back and forth.
    Thanks again.

  • E L

    A little Nate tweet for variety: Nate Silver ‏@fivethirtyeight National polls are quite predictable at this point. Everything clustering around Obama +1 +/- the firm’s house effect.

  • eris

    If Sam Wang and Connie Chung had a talk show, they could call it “Men Without Hats.”

  • A New Jersey Farmer

    The Republicans will keep the House. The election of 2010 made sure of that. Going forward, the Democrats need to focus on winning in years when we have the census, not only for Congress but for Governorships and state legislatures. They’re the ones who set the gerrymandered boundaries. On to 2020!

    • Suja P

      The work towards winning the census in 2020 starts now! Kicking out the tea party elements, or at least paring down their numbers is going to be a slow process.

  • Reason

    So FL is back to R again? What a roller coaster. I suppose it was inevitable. I am seeing more and more poll sites getting closer to the 303/235 EV split now. I really do want my state to be one of those 303. Thanks for everything, Sam. It has been real.

  • Nathan Duke

    Dr. Wang,
    I have a question. According to the Washington Times (which I know is a GOP-leaning paper), the reason the Republicans feel so confident of winning is because they claim the GOP has larger party registration among voters in 2012 than Democrats. Of course, they are basing this on numbers from Rasmussen and Gallup. I don’t know if you’ve covered this before, but I thought I’d see if you thought this was valid. Thanks! Just trying to remain relaxed with one day left to go.

  • Reason

    I am predicting D’s get 54 Senatorial seats and gain 7 more in the house. House still remains R, but barely.

  • BrianTH

    Not quibbling with the odds, which I am sure are indeed substantially against a Democratic takeover, but I would place some emphasis on the fact that the final generic poll average has not exactly been nailing the national house vote in recent elections (although it has been off in both directions depending on the year). So in this case, unlike with the Electoral College polling, the limited generic polling we have available would not have to be abnormally inaccurate in order for there to be a surprise outcome.

  • Fred

    So…now national polls matter.??? Until yesterday when the EVropped by like 15 in 1 day with no significant state polls to account for the drastic drop I was taking S. Wang much more seriously. Now I am not so sure what to think. I find that big drop yesterday a bit curious. Could it be he wants to get the EV right even if it means he needs to fudge the numbers?

    • Khan

      No, national polls do not matter. NC swung back to Romney based on a poll.

    • Khan


      And shame on you accusing Doctor Wang of patent dishonesty because YOU do not understand the mathematics.

    • Olav Grinde

      @Fred: Dr Sam Wang, and numerous knowledgeable commenters in the discussion threads, have explained this quite thoroughly. It’s all there if you can be bothered to look.

      Rather than questioning Dr Wang’s intellectual integrity, you need to do something about your own inability/unwillingness to read.

    • Matt McIrvin

      His methods are publicly described and based on the individual polls on Pollster. When Florida’s close to a tossup, large projected EV swings happen because the probability of an Obama win there fluctuates up and down through 50%, which makes 29 electoral votes shift back and forth in the top-line number. You can see what’s going on by looking at the probability distribution.

      Incidentally, right now there are no white tied states on the map. When is the last time that happened?

    • M. Mckenzie

      Translation: you think this math stuff is hard. We get it.

    • skmind

      No, national polls do not matter. Not even now. The POTUS is decided by the winner of the electoral college. Thanks for paying attention

    • Fred

      Ok, now he just just updated it and we are at 310. So I feel a bit better now….lol. Can’t wait for all this to end tomorrow. Then it’s just sit back…relax….and listen to 4 more years of the right whining how Obama stole the election and he must be impeached…

  • Reason

    Just asking. The only recent polls I saw for NC and FL were by PPP. That had NC tied and FL O +1. So was this because older polls fell off the average?

    • 538 Refugee

      Just looking at the polls in NC and Fl leads me to NC Romney, Fl, to close to call. This is based on my gut feelings about house leans on the pollsters I recognize and disregarding the new kids on the block that I don’t have a track record for. Florida would be nice. NC would be the cherry on the Sunday but I’d rather not wait for another Florida recount to know the winner either.

    • Sam Wang

      Agree, those are quite close. At the moment I think Romney will win NC. Running some numbers.

    • Joel

      I interpreted the NC shift as being a lag in accounting for that state’s polling for whatever reason. I think a later update will be R+2.5 when the Gravis poll is rolled in.

    • Fred

      yes, perhaps Wang can explain this. Such a large drop need some explaining don’t you all think? So rather than defend him blindly you should be asking…show me the numbers. I am quite familiar with his method but there was zero explanation for yesterdays big drop….based on 1 poll apparently.

    • 538 Refugee

      My ‘heart’ wants to tell me the Marist poll shows a tilt to Obama but that is as close as I can come to calling Florida.

    • Joel


      In the least charitable interpretation, they’re called bugs. If you don’t like it, do the math yourself. Sam makes it available for anyone.

  • Diogenes

    You guys are going to be so surprised when this imbecile grad student turns out to be wrong.

    Romney will win 50-48 at least and you will all be crying here about how some moron grad student and some daily kos troll named Nate Silver led you wrong.

    • Reason

      “Imbecile grad student”. DR. Sam Wang is a professor at Princeton, specializing in neuroscience. So that first comment shows you are lacking in both brain power and attention span. Also, name me one poll that has Romney ahead by 2 (and actually has a statistical credibility record), genius.

    • skmind

      This is a hunch, but are you a Tea Party Republican?

    • 538 Refugee

      Sam, you’ve arrived. Here come the trolls to shout you down!

    • Khan

      I know I shouldn’t be feeding you people, but I’ll make an exception in this case.

      Dr. Wang is a world renowned neuroscientist, author and Professor of molecular biology and neuroscience at Princeton University.

      He has a Ph.D. from Stanford and graduated with honors from Caltech at age 19.

      What I’m getting at is that this “imbecile grad student” is most certainly more educated, liked, successful, intelligent and CORRECT than you could ever hope to be.

      Now please retreat back under your bridge.

    • J Miller

      Diogenes, it’s people like you that caused me to leave the Republican Party and never look back. “Imbecile”? “Moron?” You sound like an immature, angry teenager.

    • E L

      What would a web site be without a troll? It just didn’t feel right to read only intelligent comments by people who were interested in rational analysis. So… welcome troll. Stick around and read everything available here through this coming Wednesday.

    • Grateful

      Stay classy….

    • Keith

      Even in the unlikely event that your candidate wins, I will be here on Wednesday. However, I get the feeling that you will not return after the president is declared the winner.

    • Sam Wang

      Diogenes, try to make your point without name-calling. Normally I don’t allow this kind of childishness, but several people have already replied. So I’ll leave it.

      Nobody need reply to Diogenes.

    • Fred

      I expect more of this desperately lashing out from the right through out the day today and tomorrow. Then the meltdown tuesday night/wed morning. Don’t know if I will take evil pleasure in it or just find it pathetic. We shall see.

  • Olav Grinde

    In 2012, the Democrats are defending 21 Senate seats, the Republicans just 10 seats. Nevertheless the prospects are that the Democrats maintain control.

    Does anyone know how many seats each party will be defending in 2014?

    • Froggy

      Olav, in 2014 there will be 20 Democratic-held seats up for election, and only 13 seats being held by Republicans. Democrats will be trying to hold seats in Alaska, Arkansas, South Dakota, Montana, and Lousianna.

    • Olav Grinde

      Thanks, Froggy. It sounds like 2016 is the year when more Republican seats are on the line.

  • Steve16748

    Fred, shame on “he wants to get the EV right even if it means he needs to fudge the numbers?” You must eat worms for dinner. Download the models code, download the polls, find us the fudge, or shut up.
    By the way, if I understand things correctly, at the moment, Dr. Wang is telling the world that Romney has 1 chance in 500 of becoming our president. Again, not a person fudging around with his numbers to be safe.

  • Joel

    When it comes to trolls, I wish they would follow Herm Edwards’ advice:

    “Be a man, put your name on it.”

  • F Young

    “But Republicans are unlikely to lose control of the House, to a significant extent because redistricting and incumbency give them a +2.5% advantage.”

    “…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.”

    Why don’t you call a spade a spade?

    Redistricting is when the bountaries of districts are adjusted to reduce excessive population size disparities between districts caused by population shifts.

    When a boundary change increases the probability that the party in power will remain in power, the proper term is gerrymandering.

  • Katman

    I’m confused about how the final Bayesian prediction totals to 540.

  • Duhawk

    I understand that the only way you can believe Mitt will win is if you believe the polls are all wrong (which has never happened) so I shouldn’t be nervous.

    But there is part of me that thinks that not all the polls have to be wrong, just all the polls in Ohio. If Obama loses Ohio, he has to win Virginia & 1 of IA, CO, NH (or CO, IA, & NH and not VA). This is scary.

    In conclusion, I am nervous.

    • Jack Wrong

      Calm down, man. They are getting inside your head. My feeling is that Ohio, Florida and VA all break for the President. Honestly, NH is going Obama, obviously MI, PA, WI and NV are all breaking blue.

  • Montykoolaid

    Hey, quick question.

    My office manager (who is a democrat) just told me her prediction.

    It will be way too close to call tommorow night, that we wont know for a few days, is that possible? Do you guys think that will happen?

    • E L

      I posted a WaPo article on that topic above. It’s the new MSM scare of the day. Trust the nerds not the reporters who need panicky readers.

    • Steve Norton

      Nah, it’ll be over by midnight.

  • Stephanie

    So, if the House is unlikely to lose a majority but be closer to even, is it more likely that the democrats will find a few reasonable republicans to vote with them? I would think that there would be at least 10 who do not wish to completely endanger the country’s credit rating. Also, I was wondering how accurate House Race polling actually is. I have seen the presidential poll and senate poll accuracy, but do House races have the same kind of reliability? From reading the comments, it would appear not so much, but I am definitely not a math guru, and therefore, have no real clue.

    I would also like to congratulate Dr. Wang on all his new-found attention. It is supremely well-deserved. Keep up the good work!

  • Suja P

    This is going around, much to the aid and comfort of my right leaning friends:

    “November 5th – Polls say dead heat. Maybe, maybe not. But 11 reasons to be optimistic: 1) R registration up, D down 2) R’s advantage in voter enthusiasm 3) Mitt drawing much larger crowds than O 4) Decrease in D early voting, increase in R 5) Many more Ds than Rs say they’ll crossover 6) Romney’s lead with independents 7) Mitt campaigning in states O won big in 2008 8) O has lost gender gap advantage 9) O’s bleeding support from most groups, especially Catholics & young people 10) Romney has shown vision & optimism, O no vision and has been petty & negative and 11) Voters view the economy as the number one issue So pollsters wouldn’t be fibbing us, would they? Guess we’ll find out tomorrow.”

    • wheelers cat

      Suja, from my favorite Shakespeare play, Midsummer Nights Dream
      Your right leaning friends are just whistling in the dark, like Shakespeare’s rustics.
      “I see their knavery: this is to make an ass of me;
      to fright me, if they could. But I will not stir from this
      place, do what they can; I will walk up and down here,
      and will sing, that they shall hear I am not afraid.”

      you see…from the organic perspective– this is the extinction event at the K-T boundary for Sailer Strategy conservatism.
      by 2016 they will have to integrate.

    • Froggy

      12) The Redskins lost yesterday.

    • Ross C

      I would contest any number of those statements with actual facts, but it’s just too exhausting at this point chronicling all the deception, misunderstanding and gullibility on the right during this election.

    • Suja P

      Froggy, they trotted that one on NPR yesterday.

  • Steve Roth

    “the true Presidential popular-vote margin is likely to fall near … the Meta-Margin”

    And the question, of course, is whether the House vote margin will be predicted by the popular-vote margin which is (theoretically) predicted by the meta margin…

  • Tim L.

    John // Nov 5, 2012 at 2:09 pm said:
    Well, I want a pony for Christmas, but I’m not getting one.

    I think we should all pitch in and get John a pony for Christmas! Who’s in? I’m good for $50!

  • Ira

    Wow, the 2.5% Republican advantage baseline is actually a little smaller than I’d assumed. Out of curiosity, would the baseline be even odds with no gerrymandering? I have heard in the past that Republicans are overrepresented in the House due to disproportionately representing states with only one seat (and the fact that the size of the House has not increased in 101 years). That is, if the size of the House were increased, the relative influence of the Republicans would decrease due to some of the states with only one Representative continuing to have only one Representative in the new system, even while other, more Democratic-leaning states gained Representatives. Any idea how much truth there is to this, or even how knowable it is?

  • Sandspur

    The republicans seem to think that all the polls are wrong, they are fairly confident of a win. Let’s hope and pray not!!!

  • Richard Vance

    As an Army aviation software engineer safety specialist I find credence in the worry about Ohio’s computer voting software. It is completely open to forgery and in many cases can be changed without a way to verify later. Patches? Please…

    IMHO software based voting should be banned.
    There is nothing wrong with paper ballots and OCR machines. You can easily test the OCR with batches of pre-counted ballots. How do you do that with a touch screen Murphy’s law generator?

  • Kitalaq

    Mr. Wang and Democrats…
    Thank you, Bless you, Good Health, Well Being and Safety in all your endeavors. This website is a pleasure to read both articles and comments although an occassional negative right wing makes itself known of it’s existence.
    It has been very educational experience reading about the difference between the two drastically different ideologies. Democracy vs Plutocracy
    May what is best for 100% of our nation be the victor!!!


  • Kitalaq

    My bad for stating Mr. Wang in my previous statement Dr. Wang…

Leave a Comment