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

After the storm

November 7th, 2012, 9:42am by Sam Wang

We shall not cease from exploration
And the end of all our exploring
Will be to arrive where we started
And know the place for the first time. -T.S. Eliot

Good morning! The day after the election is always a bit of a relief for me. We’re still waiting on a few races, but here’s a preliminary look back at how our polls-only approach did. All in all, extremely well. As Randall Munroe sent me last night: “BREAKING: Numbers continue to be best tool for determining which of two things is larger.”


In the races called thus far, pre-election polling medians were correct in 50 out of 50.

There is one race remaining, Florida, where Obama leads Romney by 49.91% to 49.36%, with 100% of votes counted. It hasn’t been declared yet, partly because the threshold margin for a recount is 0.5% – they are probably still working that out. Pre-election polls were a near-perfect tie. I suspect I am about to lose that coin toss. Update, 11/15: final Florida result, Obama 50.0% Romney 49.1%.

The two-candidate popular vote share is Obama 51.1% to Romney 48.9%. This exactly matches my prediction, which was derived from state polls with a little Bayesian help from national polls. Update, 11/15: popular vote Obama 51.4% Romney 48.6%.

Bottom line: I will not have to eat a bug.

Senate: Of the closest races, election returns match polling medians in 10 out of 10. Of particular note is the North Dakota race, where Heitkamp (D) leads Berg (R), 50.5-49.5%. That’s one where I had Heitkamp based on polls, and Nate Silver had Berg based on polls plus other factors. We are also waiting on the Montana race, where Tester (D) leads Rehberg, 49-45%. The upper chamber appears headed right for the median that we predicted, 55 D/I to 45 R.

House: This one will take some time to sort out. From the Republicans’ current majority of 242 seats, I predicted losses of 2-22 seats. It looks like the losses will be toward the low end of that range.

I am very interested in whether Democrats win the national House popular vote, which would mean a mismatch between the vote and the seat count. This is due in part to redistricting. It would be only the second time since World War II that it’s happened, and is antidemocratic with a small “d.”

I’ll give a more detailed look later, especially regarding our prediction challenge. For now, I think we can safely say that the following people had a good outcome: President Obama, Congressional Democrats, and quantitative poll analysts.

Finally, a trip down memory lane: the Princeton Election Consortium long-term predictor, as of August 3rd:

Back in a bit with more wrap-up…

Tags: 2012 Election · House · President · Senate

177 Comments so far ↓

  • Paul Griner

    Thanks again Sam. Great to see this site once again be right. A victory in so many ways. Congratulations

    • JV

      Hey Paul,

      How’s it going? Small world (or small internet, anyway). Good to see another English-major-y type with a thirst for hard statistical data.

      See you around campus!
      John Vance (from UofL; I’m not the JonV posting on these threads, though).

      PS: Paul Griner is a dang good writer and creative writing teacher, just so everyone knows.

  • shma

    Don’t forget the popular vote: right now Obama is +2.2, exactly where PEC predicted it (+2.2 +/- 1.0)

    There are still votes to be counted, but it looks like it will end up in the middle of the predicted range.

    • Sam Wang

      How could I forget? Thank you. The main issue there is that my error bar was a bit too large.

      Take that, national-poll worshippers!

    • Matt McIrvin

      Nate Silver needs to seriously consider the possibility that he is estimating his confidence intervals far, far too timidly.

    • Paul

      Nate Silver definitely needs to think about its confidence intervals. He does have one good argument on his side, though: the electorate is changing, polling methods are changing, and it’s entirely reasonable to think that likely voter models will miss in a future election. Do we really believe that there’s only a 1% or 2% chance of that happening? That seems over-optimistic to me.

      If Nate’s error bars correspond to the odds of systematic poll failure, he may be on to something.

      I take PEC’s prediction as “this is what the polls say if you take them at face value” (and so far, that’s been spot on by Election eve). I take 538 as “this is what happens if you try to account for every source of uncertainty.” And I take Votamatic as “this is what happens if this election is not radically surprising” (which nailed the outcome (except •maybe• FL) four months in advance, with later waver in the interim).

    • BrianTH

      I think there is still room for debate about what probability we should assign to the scenario in which the pollsters end up systematically off by enough to swing an election of this sort due to fresh challenges to their methodologies.

      However, I for one would note that pollsters have been facing fresh challenges for a while now, and although individual pollsters may struggle in one way or another, it appears in the aggregate they are adjusting their methods well enough to avoid such systematic biases from occurring.

      In other words, I think the evidence is mounting that the polling industry as a whole has the ability to respond to such challenges well enough such that the probability of such a systematic bias appearing in any given Presidential election really is quite low.

  • securecare

    Thank your family for us since they gave you up all the time that you spent on this.

    Great Job one and all.

  • SahibSpeaks

    With all the talk about Republican enthusiasm this year, it looks like their total popular Presidential vote will fall short of 2008. True for the Democrats, too, but then they were not touted to be fired up vs. 2008.

  • Stephen Coltrane

    Anyone know why they can’t just call Florida? Are they double-checking to make sure the margin really is over 0.5%?

    • Stephen Coltrane

      It’s OK, answered my own question. Some absentee ballots still to be counted. Official result later today (Wednesday).

  • dhogaza

    Congrats on your fine work this election cycle.

  • Neal

    Congrats! You were on top of it and flawless.

  • Jonathan

    Does anyone know how total voter turnout compares to 2008? The current numbers are (59.6 million for Obama, 57 for Romney) are not final, right? Since they’re still counting in some places?

    I’m a European but very interested in this election. A lot of people online are talking about 10 million less voters for Obama compared to 2008, but aren’t they much to fast in saying this?

  • Lee

    PEC did fantastic! Well done.

    The only thing I could possibly find in need of improvement for 2016 is that it looks like PEC’s error bars were too wide. That is, the 2012 votes came in closer to predictions than was pedicted!

  • Scott Tribe

    Hi Sam:

    Read your site with interest from Canada. Congrats on the predictions and not havign to eat any insects.

    I was wondering if you were considering doing a pollster comparison to see whose polls were the most (and least) accurate. That would be interesting.

    (I’m thinking specifically of Mason-Dixon predicting Romney by 6 in Florida and saying the networks will call it soon after polls close, if they do the #’s correctly. What crow that will taste like).

  • Amitabh Lath

    Numbers beat “gut feelings”. Quelle suprise. I hope the George Wills of the chattering classes have the decency to admit they never really believed their astounding claims, it was just a cynical ploy to put out a narrative they so hoped would swing things Romney’s way.

    It didn’t of course. Any more than jailing Galileo made the moons of Jupiter disappear.

    And insects are considered a delicacy some places.
    Klingons love ’em. You wouldn’t want to offend anyone by putting a negative spin on bug-eating.

    And finally, Nevada. Holy split tickets. Big Obama win, but senate seat goes red.

    • wheelers cat

      But Amitabh those people actually believed Romney would win. He did too.
      It wasnt cynical.
      Sure some of the media deliberately pimped the horse race.
      And people like Sean Trende followed the paycheck.
      But the true believers, Barone, Rove, Rasmussen, Palin….they are so blinded by their loathing for Obama that they couldnt do otherwise.
      Chait has a really good explanation.
      Republicans bet everything, and lost it all.
      game theory rules.
      I guess my systematic error in the legacy polling hypoth for the nat’l polls is fail.
      I am going to get 332 EV though.
      In July I had EV at 297 on Nates blog.
      Dr. Wang improved my prediction with the meta-margin.

      Does this feel a little bit like the end of the summer at election camp? we are all exchanging contacts and compliments and packing up our skull furniture for next time.

      my favorite Eliot
      “To arrive where you are, to get from where you are not,
      You must go by a way wherein there is no ecstasy.
      In order to arrive at what you do not know
      You must go by a way which is the way of ignorance.
      In order to possess what you do not possess
      You must go by the way of dispossession.
      In order to arrive at what you are not
      You must go through the way in which you are not.
      And what you do not know is the only thing you know
      And what you own is what you do not own
      And where you are is where you are not.”
      ― T.S. Eliot

    • Amitabh Lath

      Hi Cat. Whatever the Republican blowhards are, here is some Eliot for them:

      We are the hollow men. We are the stuffed men. Leaning together. Headpiece filled with straw.

    • E L

      Fox News ends:
      “not with a bang
      but a whimper.”

  • mediaglyphic, extremely well done. kudos.

    1) the house popular vote issue is one that needs to be broadcast LOUDLY.
    2) the current congress cannot believe that they have a mandate to stop all legislation.

    You have built an amazing platform, what will you do with it now? (the platform includes the math, plus the followers and bloggers)

  • Khan

    Just so we’re clear on things here…

    Gallup had the wrong candidate winning with a majority of the vote. The wrong candidate.

    • Froggy

      And as I pointed out in the last thread, Rasmussen had Romney winning in Virginia, Colorado, Iowa, and Florida, and tied in Wisconsin and Ohio. So much for the conservatives’ “most accurate” pollster.

    • Shawn Huckaby

      Amazingly, Charles Krauthammer actually used both Gallup and Rasmussen to explain how the hurricane cost Romney lost in the last week. Since both polls did CYA corrections at the last minute to get closer to the real consensus, to him that indicated the popular vote actually declined 5 points in one week. Unbelievable.

    • Shawn Huckaby

      Sorry, should read “cost Romney the election in the last week”.

    • Khan


      What’s more ridiculous is that he claims to be a psychiatrist.

    • Shawn Huckaby

      “Physician, heal thyself!”

    • mediaglyphic

      as a canadian i apologize for Charles Krauthammer. He is totally out to lunch on every issue i can think of plus doctors don’t need to add.

  • shawnthesheep

    Having sent my absentee ballot from Mauritius. I’m guessing mine might be one of the ones still to be counted. Since neither I nor my partner voted 3000 times for Romney, I think Obama’s lead in Florida is pretty safe.

    I find a great deal to be happy about in this election cycle, but the one thing that will drive me crazy is that the pundits will pretend that they weren’t horribly, awfully, completely wrong. There will be a hundred “The Hurricane that Gave Obama a 2nd Term” stories written. Of course, those of us who trust things like “numbers” know that Obama led in the polls wire to wire. He never trailed. There was no big story needed to shift the race back to Obama.

    • Sam Wang

      Hmmm, better knock down that one soon.

    • JD Newman

      I see you’re mucking about with those who do not bleat. Perhaps one day, we’ll find a way to meet.

      I am so sorry for this reply. I couldn’t help myself. I just couldn’t

    • Matt McIrvin

      Sandy (not Benghazi) is going to be the Iran hostage crisis of this election cycle, the thing that gets blamed for the result even though it probably would have happened anyway.

  • Adam

    Great work!

    I really look forward to seeing your analysis of the House of Representative popular vote and any mismatch with the seats won. I fully agree that this is antidemocratic.

  • Noam

    Great job Sam and team. Thank you!

  • Khan

    I’ll do the work for everyone here on Rasmussen.

    Wisconsin – Missed by 7 points
    Colorado – Missed by 7 points
    Iowa – Missed by 7 points
    Virginia – Missed by 5 points
    Ohio – Missed by 2 points

    • Olav Grinde

      Please do send that list and more to newspapers and broadcasters, and politely suggest a dire need for feature stories on Rasmussen and Gallup — that they have some explaining to do. Oh, you can add numbers that evidence the inexplicable “movement” of these pollsters in the closing week of the election.

  • maye

    My heartfelt and enthusiastic thanks for the work you do. Nerds forever!

  • Olav Grinde

    Dr Wang, you calculated the GOP gerrymandering advantage to 2.5%. When your time permits, I wonder whether you would be so kind as to calculate what the Democratic/Republican split would have been on a level playing field?

    Are the results a clear warning to the Democrats, that they need to lay a long-term strategy for winning control at the state level? Is this the only way to prevent unreasonable gerrymandering, curtailing of early vote, unethical election directives by partisan secretary-of-states, partisan governors’ refusal to extend voting when voters are still waiting in line, and other abuse-of-power?

    • Michael

      @Olav — check the next page. In Virginia, Republicans won 8 of 11 house seats with 51.6% of the house vote. According to my thumb-nail calculations, anyway. Now there’s a skew, huh?

      I’m not a big Al Sharpton fan, but he did have one of the best lines of the night: “People don’t wait in line for five hours to try to vote with fake IDs.”

  • InmanRoshi

    With all the (well deserved) talk of how successful Nate/Sam/Drew etc. are with their models, I think what maybe doesn’t get talked about enough is what an amazing job the professional polling collective does at projecting what the voting electorate is going to look like, particularly at the state level.

    Kudos to the pollsters for providing such sound data.

  • MAT

    It appears Scott Rasmussen this morning is starting to admit the gig is up:

    “This race was very likely the last presidential election of the telephone polling era. While the industry did an excellent job of projecting last night’s election, entirely new techniques will need to be developed before 2016. The central issue is that phone polling worked for decades because that was how people communicated. In the 21st century, that is no longer true.”

    I’m hoping the conversation at PEC in the future turns towards how polling will, could or should evolve in the future. Do we see a polling arms race to drive the aggregator narrative? How do they handle the falling response rates? Lots of cool discussion topics for the Numerati.

    • Craig

      Well, Scott, you could actually try calling cell phone only voters. Everybody else is doing it.

    • wheelers cat

      thats going to be his excuse for why he failed.
      he knew this in 2010 when he whiffed on CO and NEV.

      i consign him to the ninth hell, The Betrayers.
      he betrayed the math to fuel a narrative.

      “Just so the livid dead are sealed in place
      up to the part at which they blushed for shame,
      and they beat their teeth like storks. Each holds his face
      bowed toward the ice, each of them testifies
      to the cold with his chattering mouth, to his heart’s grief
      with tears that flood forever from his eyes.”

      thats the thing about a catholic girls’ school education– great for the classics.

      ima miss you guys.

    • wheelers cat

      Sandy knocked that ratbastard off the air so he couldn’t do his usual trick of converging on the true outcome.
      I just hate him so much.

    • Froggy

      @wheelers cat: I have Rasmussen in the Eighth Circle, Bolgia 4, among the sorcerers, astrologers, and false prophets. Bolgia 8 is another possibility. And I can see your point about the Ninth Circle.

      The situation brings to mind a passage from Larry Niven’s novel Inferno:
      “Lawyers.” Minos laughed. “I have problems with lawyers. There are so many places appropriate to that breed.”

    • Ben

      And here I was, hoping the lawyers would escape this discussion. Oh, well.

    • Matt McIrvin

      Hmm, I wonder why every major pollster other than Rasmussen and Gallup basically did all right.

    • securecare

      “…hoping the lawyers would escape this discussion…”

      I think Shakespeare had an approach to dealing with that.

      “…The first thing we do, let’s kill all the lawyers…”

    • E L

      Gonna miss Cat, Froggy, Ohio, Ms, Jay and many, many others. I’m so old I have a small thing called fear that this might have been my last rodeo. I’m from Shane country and the sod busters beat the cattle barons again. Tester v Rehberg.

  • mediaglyphic

    @MAT — take a look at Seth stephens-davidowitz work on using google insight to forecast what will happen. He predicted 2008 like turnout and that obama would win, without a poll!!

    • MAT

      @media – Thanks! Did a little looking into it, seems like an area with some promise.

  • NPR

    You, Nate Silver, Drew Linzer, Simon Jackman et al are precisely the reason we need to focus on math and science in the education system. Kudos!!

  • Jay Bryant

    Excellent work, Sam. I now await your analysis of whether the popular vote for the house is out of whack with the members in each party. Small “d” is the “democratic” that matters most.

  • DB

    I’ve been lurking for months–just wanted to add my voice to the chorus of thank-yous. You helped keep this antsy liberal sane (and I even learned a few things, to boot!)…thanks again and keep up the good work!

  • mlhradio

    Woke up to one more (small) surprise this morning: One of the media darlings of the conservative right, Mia Love (UT-4), unexpectedly loses to Jim Matheson:

    • Froggy

      That’s a shocker — I expected that Mia Love was headed for bigger things.

      It looks like Allen West lost (but he hasn’t conceded), although Michelle Bachmann is still around.

    • wheelers cat

      we learned that evo theory of culture.
      race trumps religion.

    • TrentB

      Yep, despite all the recent gerrymandering here in utah to ensure the removal of Matheson, the RNC still had to pump millions into that race. LOL!
      Interesting to see the white christian male voter err, the mormon voter losing grip in such a deep red state as Utah.

      It just befuddles me that the GOP continues to ignore the rising tide of the majority-minority.
      21st century WHIGS.

      …in closing

      Strong work Sam Wang!

  • bks

    Now what are we going to obsess over?


  • Steve

    Thank you, Sam and your consortium.

  • TycheSD

    Thank you Sam, for helping to keep me sane these past few months.

  • JD Newman

    I’m really interested to see how you build the model to look at popular voting #s in the House.

  • Tapen Sinha

    Congratulations Sam and your team. Once again it proves Occam’s Razor is a good model – or as some people call it the KISS principle. See you in 2014 and beyond.


  • piktor


    I am in awe of your 51.1 – 48.9 prediction.

    I was in awe of your 2008 prediction.


  • Daniel Wallace

    This site has been the moral high point of the election. Today, I told my fb friends:

    “Sam Wang has not simply built a model that correctly predicted the end result (months ago). More importantly, he has tried to show what the existence of such a model means: that much of political commentary, horserace claims, flashing countdowns is simply a ruse to keep us bewildered, nervous, and deceived. On his blog, Wang has been demonstrating a different kind of political journalism, transparent, informative, and calm. Take a look.”

    • Olav Grinde

      Amen! And may it, in due time, displace as much of the noise-based and noise-making punditry as possible. May it also help define a new baseline for political journalism, where fact-checking, research and intellectual rigor once again become pre-requisites.

    • wheelers cat

      its the rise of the Third Culture.
      we evolve.

    • Suja

      Hope you don’t mind if I steal.

  • Suja

    Amazing job, Sam! I told my parents yesterday what your predictions were, before the results started rolling in. They are shocked (SHOCKED!!!) that these things can be predicted with such accuracy.

    The Pew guy was on NPR this morning, positively glowing. Kudos to the pollsters (except the ones fudging the numbers to suit their narrative) as well.

    Note to self: Listen to Sam. And other math-y types. Ignore MSM, pundits and feeeeewings.

  • ASC

    Great site, great work, which really cuts through the crap.

    Not that Hurricane Sandy impacted key results, but something worth throwing into the mix for thinking about the years ahead:

    “Geometry deceives; only the hurricane is accurate.” -Victor Hugo

  • Pat

    I think an interesting to note from the results we have so far is that Ohio actually was not the tipping point state!
    Contrary to expectations, Colorado (margin so far 4.5 points) and Virginia (so far O+3) gave a wider margin than Ohio (currently just O+1.9).
    And surpisingly, Pennsylvania ended up being closer than many hard fought states (the margin is now smaller than in Nevada, Wisconsin, Iowa and NH !)
    The tipping point state is Colorado, with VA, OH and FL providing some extra icing on the cake, but being a bit closer.

    • Ben

      Pat, even though Obama appears to have won PA with a smaller margin than some of the swing states, the “inelasticity” of PA (due to the high proportion of straight ticket voters for both parties) meant PA was not in play. The closeness is illusory, which is perhaps why Republicans always express confidence that PA can be flipped in a presidential election.

  • Pat

    States ranked by Obama margin (yes things can change, especially in CO, but this is what we have so far)
    MI: O+8.3
    MN: O+7.6
    NV: O+6.6
    WI: O+6.5
    IA: O+5.6
    NH: O+5.3
    PA: O+5.1
    CO: O+4.5 >>>>> so far already 272 EV
    VA: O+3
    OH: O+1.9
    FL: O+0.5

    • Andy Manning

      Surprised that PA ended up with a smaller margin than IA and NH. What do the good folks here attribute that to? Lack of focus from the candidates (not seen as a swing state), D leaning areas most affected by Sandy?

      Also, congrats to Sam and team on the accuracy of their prediction. I raised a glass last night to you and all the others on the side of math and science, defiant in the face of, frankly, idiots.

  • Mike P

    I am also in awe . . . will never, ever doubt the numbers guys again! Bravo, Sam and team, for the most intelligent and engaging analysis on the planet!

    • securecare

      For the sake of sanity remember to trust but verify.

      Anyone can make a mistake at any time.

  • Khan

    Jon Tester just won :)

    Also, I have my final analysis. Scott Rasmussen missed the 12 potential swing states by an average of 5.2% in favor of Romney every time.

    • wheelers cat

      then my Artifacts of Legacy Polling hypoth is false, and its game theory all the down.
      Rasmussen and Gallup were deliberately manipulating the polling.

    • Khan

      @wheelers grrll cat

      You may very well be right on the manipulation.

    • Mike M

      It matters not whether Rasmussen and Gallup were manipulating results or simply bad at their job. Regardless this is their “Dewey defeats Truman” moment. They no longer matter.

  • Nora

    The victory for quantitative poll analysts implies another victory as well: the relative success of all who fought against Republican efforts to suppress the vote. As a Pennsylvania voter who for the first time ever was asked but not required to provide ID, I was acutely aware of the tactics Republicans were employing to confuse voters from traditionally Democratic demographics.

  • Jason

    Thank you and congratulations! I learned and enjoyed your work and appreciate the groundedness amidst the insanity of the election.

  • paul griner

    Wait, Amitabh. Jupiter has moons? Since when?

    And Eliot, quoted on a political website?

    Great. But how about Yeats?

    • wheelers cat

      i think this is appropriate

      “Turning and turning in the widening gyre
      The falcon cannot hear the falconer;
      Things fall apart; the centre cannot hold;
      Mere anarchy is loosed upon the world,
      The blood-dimmed tide is loosed, and everywhere
      The ceremony of innocence is drowned;
      The best lack all conviction, while the worst
      Are full of passionate intensity.”
      ― W.B. Yeats, The Collected Poems

      Courtesy, Immaculate Conception School for Girls

    • Matt McIrvin

      I won’t look through that telescope thing, there’s some kind of witchery going on in there!

  • Shawn Huckaby

    Dr. Wang, I know this isn’t your day job, but would you consider lending your talents toward other scientific topics where the actual data is frequently spun to sneak political/corporate agendas past a mathematically illiterate public? Specifically I was thinking about climate change, but I’m sure there are others…

    In any case, I’ve really enjoyed the interactions here, and would love to see this community continue.

  • Pat

    So far about 119 million votes have been counted. Estimates (e.g. by Prof. McDonald) were putting the total vote at 132 million.
    Is it likely that so many votes remain to be counted? (or was turnout a bit lower than these estimates?)
    Does anyone remember which fraction of the vote was typically missing on Wednesday after the election 2008? (could it also have been that an extra 10% of the total vote was tabulated afterwards?)

    • Michael

      Yes, there are lots of votes still to be counted. The west coast states have a lot of mail-in votes and take their time tabulating them. I remember watching the 2008 totals slowly increase for several weeks after election day.

    • Pat

      Which site were you watching the 2008 totals increase? I remember most news networks like CNN didn’t bother updating anymore for the hardcore geeks after a few days…

    • Michael

      @Pat — not sure I even remember to tell you the truth. Might have been Yahoo’s dashboard, but I also was following several state BOE sites directly because of some very closely contested recounts (including my own CD).

  • Olav Grinde

    I would love to be a fly on the wall during Karl Rove’s post-election meeting with his “investors”.

    • Shawn Huckaby

      That Rove 1/2 hour on Fox last night was one for the time capsule. As he fought with Fox’s own analysts I could actually hear Lee Atwater rolling in his grave.

    • Suja

      @Shawn: Fox News should be re-dubbed Comedy Central. Karl was apparently up WAY past his bed time.

  • Pete

    It’s interesting that 3 smart guys (Sam Wang, Nate Silver and Drew Linzer) all took different approaches and all ended up in the same place. Nailing 50/50 states. Nice convergence :-)

  • Michael S

    Well done, Sam. See you in early 2014.

  • Andy

    Just wanted to say thanks for the amazing work! I was new to the site and to this level of analysis. The last few weeks felt a little like ‘Moneyball,’ only it was real. And the stakes were higher. And no Jonah Hill.

    Beautiful job…

  • Barb

    Great job Sam! Thanks to you I was mostly sane this time around. Obsessive but sane.

    Regarding gerrymandering, I’m proud of my state Ohio in many ways this morning, but we had a chance to go to bipartisan redistricting and voted it down. There was so much misinformation advertised about Issue 2. Oh well, work to be done.

  • paul griner

    Yes, Wheeler, exactly!

Leave a Comment