Presidential election wash-up

Scattered observations on the state of polling and forecast models following the rude shock of the presidential election result.

Donald Trump’s victory was a close-run thing, achieved by the grace of winning margins of 1.2% in Pennsylvania, 1.0% in Wisconsin and 0.3% in Michigan. The New York Times is still projecting that Hillary Clinton 0.2% lead on the raw popular vote will resolve at 1.2%, which presumably has something to do with an incomplete vote in Washington state. Nate Silver has offered a penetrating counter-factual to illustrate how different the result and its implications would look if things had played out just a little bit more happily for Clinton.

Assuming the New York Times projection is right, the crude average of national poll results conducted by RealClearPolitics will prove to have exaggerated her margin by 2.0%, which is quite a bit less bad than the news media consensus would have you believe. Not a few individual pollsters have questions to answer, but the shellacking the industry as a whole is copping is out of proportion to its failure. A similar error in the context of a landslide result would barely have attracted notice, but in this instance it had the effect of encouraging a misapprehension about the likely result.

Of the aggregators, Nate Silver emerged with the most dignity intact because it was his model that had the highest uncertainty factored in. So far as the predicted electoral map, Silver’s model was identical to that of Sam Wang of the Princeton Election Consortium, which fatally estimated Clinton’s chances of winning at 98%. All it took to make a mockery of this was a regional failure of polling in the rust belt states, where the fierce enthusiasm for Trump among normally disengaged white working-class voters was under-reported. Pennsylvania, Michigan and Wisconsin produced 62 polls of presidential voting intention between the start of October and polling day, and only two had Trump in the lead (both conducted on the eve of the election by the Republican-aligned Trafalgar Group).

My own poll aggregate did no better or worse than the others, having won brownie points by tipping Florida correctly, but lost them again by giving Nevada to the Republicans (by insignificant margins in each case). Its findings are of use in ascertaining where the state polls did best and worst, as detailed in the list below showing the size of the error in favour of Trump (so negative results mean the polls were biased to Clinton). This is limited to states where three or more polls were published in October and November. Two patterns emerge: the noted under-measurement of the Trump earthquake in the industrial mid-west, and a tendency for both sides’ winning margins to be underestimated in their strongest states.

Crikey subscribers can enjoy more of my thoughts on these matters here.

Missouri -10%
Utah -9.7%
Indiana -9.3%
Iowa -6.8%
Minnesota -6.1%
Wisconsin -6.1%
Ohio -5.6%
Michigan -5.4%
Pennsylvania -5.2%
Louisiana -4.7%
Maine -3.8%
North Carolina -3.8%
New Hampshire -2.2%
Arizona -1.5%
Florida -1.2%
Colorado -0.8%
Georgia -0.5%
Virginia -0.4%
Texas -0.1%
Massachusetts +0.4%
New Jersey +1.3%
New York +2.3%
Illinois +2.4%
Nevada +2.5%
New Mexico +3.2%
Oregon +4%
Washington +4.8%
California +6.4%

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

26 comments on “Presidential election wash-up”

  1. William
    Thanks for this very helpful analysis. Like many people ai am still trying to understand the extent of this error, as much as the real nature of who the USA has voted for. Overall, to me, the problem seems to be largest in the polling, not in the aggregation. Different aggregators using different methodologies all predicted Clinton by varying margins. There was a problem with the input data. If all the polls goong in suggest Clinton will win, then that is how any aggregator will come out.

    While the national polls do not look too bad, those state level polls are out by margins well beyond MOE, especially in the rust belt. Why is that?

    In my own line of work (transport planning) there has been increasing difficulty getting accurate and representative travel behaviour data, because it is increasingly hard to get people to answer interview surveys. Samples from time poor age groups tend to be small and heavily weighted. Trip data from sources like blur tooth and mobile phone records may be more accurate. Of course, that does not help with voting intentions.

    Were betting markets and internet search predictors equally far out?

  2. I can’t help but wonder if every media (except Fox, Breitbart, etc), and very poll predicting a win for Clinton actually discouraged voters not to turn out. Why bother going through all that hassle if it’s a foregone conclusion?

    On the other hand, did all the predictions of a Clinton victory encourage more Trump voters to vote?



  3. If Trump bullies the GOP into running up the debt to invest in infrastructure, he will become more popular and difficult to dislodge from the Whitehouse. It is ironic that only Trump could do this – a Democrat president would be blocked in the House as was Obama.
    Australia should be taking advantage of low interest rates to fund the energy transition to a low carbon future and a decent NBN.

  4. If Trump bullies the GOP into running up the debt to invest in infrastructure

    His crooked tax plan will run up the debt all on its own, no infrastructure investment required. His policies, if he’s allowed to implement them as promised, will bankrupt the U.S. as sure as he’s bankrupted his own businesses in the past.

    The question is, will the fiscal conservatives (and the perennial doomsayers on Fox News et. al.) call him out on it, or will they go quietly along because it’s their party that Trump has co-opted?

  5. So America chooses a racist biggot over someone who is not.

    No different to Australians though we elected Abbott, Turnbull with the likes of Nationals, One Nation and LNP QLD senators George and others like Brandis.

  6. Borrowing to build infrastructure is not a silly plan. Unless the infrastructure is a dodgy Campbell Newman toll road tunnel, most results in a net increase in economic activity, and then tax income. To be honest, Trump turning his back on the whole austerity-don’t-borrow nonsense and borrowing to build the things that are needed is a very sensible break from the worst policy features of his predecessors.

    The only people who benefit from government’s not borrowing are private bankers. And guess who influences policy?! Seriously, we can criticise Trump for a LOT of bad things, but ditching the silly debt caps is one thing he is right about.

  7. Thanks for your analysis William. Can I draw your attention to one other aspect which strikes me as important?
    This is the trend in total votes for both parties. The popular vote (in millions) for President the last two elections was
    Democrat Republican
    2012 65.9 60.9
    2016 60.1 59.8
    I have no doubt that Trump did appeal to many disaffected blue collar whites. However, his increase there seems to have not quite offset losses in other parts of the Republican vote, which went down overall by 1 million.

    In marked contrast, with Clinton the Democrat vote went down by almost 6 million. Quite a few people did leave the Democrat fold, and she was unsuccessful in inspiring enough replacements.

    I’ve has a similar look at a couple of the key states. They all have slightly different stories, but Wisconsin is particularly interesting:
    Democrat Republican
    2012 1.62 1.40
    2016 1.31 1.39
    Trump did not do more here than hold the Republican vote – while Clinton lost 20% of the Democrat base

  8. Interesting that preferential voting is getting a look in in Maine? Its would be an interesting way to shake up their system, and could provide ways for the more “disengaged and dissatisfied” voters to be brought more into the process.

    If the Americans are going to repair their democracy and deal with what i see as frustration issues in the electorate i think this is one method they really need to have a serious look at.

  9. Clinton lost 20% of the Democrat base

    Who interestingly didn’t go over to Trump but rather just didn’t show up at all (or voted for a third-party option).

  10. Is there a good site that slices and dices the voting patterns that are being reported (such as 70% of white college educted men voted for …., etc).

    I see individual demographics here and there, and some specific cross classified results, but can’t find any site that provides a good set of results like that.

  11. I see everyone is still missing the biggest lesson in all of this. The main reason that so many were shocked with the result, was the media. You were all conned into thinking Clinton was vastly superior to Trump in all facets, particularly morality, because of the unashamedly bias portrayal of the two protagonists throughout the campaign. Thankfully, a sufficient number of the US voters were not so easily fooled.

    Now, as I’ve mentioned previously, I can understand how the layperson could be deceived in this way, but I’m astounded that so many with such political insight, and significant expertise in exposing media agendas, especially in a domestic capacity, so blindlessly accepted everything they were fed in this regard.

    The real lesson in all of this, is that the media has no integrity whatsoever. One cannot trust a word that they say. You all know this already, but for some baffling reason, you all forgot about that this time.

    So pull yourselves together, the world won’t come to an end over this, and above all, for all of our sakes, please make sure, you never ever forget that the media is your enemy, both here and abroad, and don’t accept anything they tell you on face value.

  12. a r @ #14 Friday, November 11, 2016 at 3:53 pm

    Clinton lost 20% of the Democrat base

    Who interestingly didn’t go over to Trump but rather just didn’t show up at all (or voted for a third-party option).

    Or it is also possible (indeed I suspect more likely) that some previous Democrat voters stayed home, some previous Democrat voters crossed to Trump, and some previous Republican voters stayed home.

  13. Thank you William,

    That “counter-factual” from Nate Silver was excellent. Psephology is an inexact but worthy science. I also found the NYTimes “Live Predictor”, recommended by Kevin Bonham, to be an excellent predictor. Belatedly Based on Anthony Green’s idea of matching “booths”(god the Yanks are backward), It told me what I didn’t want to hear well in advance.

  14. I’m still struggling to make sense of all this. Lot’s of anger and disbelief on my part.

    Of course, Trump claims to represent this demographics interests, but in reality he’ll be an anarchist and a wrecker at best and a vengeful despot at worst. Great job everybody!

    I’ve already seen a lot of American commentators try to recontextualise Trump. Arguing he didn’t really mean the things he said in his campaign and was being hyperbolic. People who are likely to believe this crap would do well to read this article.

  15. Tradition dictates that departing presidents write a letter and leave it in the oval office for their successor. I think Obamas should just say

    “Good luck!”

  16. from Mytym:

    “You were all conned into thinking Clinton was vastly superior to Trump in all facets, particularly morality, because of the unashamedly bias portrayal of the two protagonists throughout the campaign. Thankfully, a sufficient number of the US voters were not so easily fooled.”

    She was. And they were.

  17. On current figures, Trump and Gary Johnson polled more than half of the total vote between them, while Clinton’s total increases by less than one percent when Jill Stein’s vote is added to hers. Would it be fair to say that Trump would have won the popular vote under a preferential voting system?

  18. Can’t look at total number of votes yet.
    There are millions of votes left to be counted. A great many of them are actually in safe Democratic states (California, Wash., etc) where most people don’t vote on the day nor early in their precinct.
    Hillary’s popular vote gap will continue growing, and possibly reach well over a million.

Leave a Reply

Your email address will not be published. Required fields are marked *