How Will the Undecideds Break?

This question obviously has important implications for who wins this election. We can start by dismissing the blanket statement that “undecideds break for the challenger.”  [NB: That phrase was garbled in the original version of this post.] That is not true as a blanket statement.  See Daron Shaw’s chapter on swing votes in Unconventional Wisdom as well as Mark Blumenthal and Nate Silver.

Another take is Bill Galston’s.  He compares 2012 to two other recent elections when the incumbent president was running—1996 and 2004.  Based on when exit poll respondents said they decided, he finds that those who decided in the week before the election were more likely to vote for Dole and Kerry.

I am going to do this differently.  I don’t really trust voters to tell pollsters when they have made a decision.  Much of our decision-making is essentially subconscious.  Instead, I am going to assume that undecided voters will gravitate to the choice that their preexisting opinions would predict.  This assumption is actually a well-known finding in research on campaigns, dating back to the earliest studies.

To preview: I find little evidence that the predicted decisions of undecided voters will benefit either Obama or Romney very much.

I will assume that undecided voters will make a decision that reflects three things: their party identification, their approval of Obama (as a “referendum” model would suggest), and how favorably they feel toward Obama and Romney (the difference between how they feel about each candidate, as a “choice” model would suggest).  I estimate this model on all decided voters and then predict the choices of the undecided voters based on the model’s results.  The model predicts the choices of decided voters correctly 99% of the time, which is no surprise given the factors in the model.

The model assumes, crucially, that each of these factors will be weighted the same way by both decided and undecided voters.  This assumption could be incorrect, of course.  But some such assumption is necessary to estimate the preferences of a group that hasn’t expressed one (yet). The data for the analysis come from the three YouGov national polls conducted since the first presidential debate.  This provides a total of 3,000 respondents, of whom 195 (6.5%) said that they were undecided.

Here’s the first finding. The model predicts that these undecided voters will split almost exactly evenly: 50.1% for Obama and 49.9% for Romney.  There is substantial uncertainty in this estimate, naturally.  The 95% confidence interval for Obama’s predicted vote share is 44% to 56%.

A small number of other respondents (3%) indicated that they will vote for a third-party candidate.  If we include them along with undecided voters—assuming that at least some third-party voters will end up voting for a major-party candidate—then the prediction is 52% Obama and 48% Romney.  Again, the uncertainty means that this is not any real “lead” for Obama.

If we focus on those undecided voters who say they “definitely” will vote—this is about 58% of undecided voters— the balance tilts more toward Romney: 44% Obama vs. 56% Romney.  As in polls more generally, Romney tends to do a bit better among self-described likely voters than among voters as a whole.  (Again, insert caveat about uncertainty.)

However, consider the math here.  Assume that only these likely undecided voters (the 58%) actually make it to the polls.  You’ve got 56% of 58% of 6.5% voting for Romney and the rest for Obama.  This would add at most 0.4 points—that is, less than half of 1 percentage point—to Romney’s margin over Obama nationwide.  That is, if you assume the model is correct, that we can safely ignore the underlying uncertainty, that only these undecided voters will vote, and so on and on.

The lesson I draw from this analysis is: how the undecideds “break” may not be consequential in this election.  As Nate Cohn notes, the mobilization of partisans could be more consequential than the persuasion of undecideds. This is consistent with some of the evidence in Robert Erikson and Christopher Wlezien’s Timeline of Presidential Elections.  They find that in the 1952-2008 presidential elections, the eventual outcome is usually slightly less favorable to the winner than the final national polls suggested.  Why?  It’s mainly about turnout.  The “no-shows”—voters who say they are likely to vote but then don’t show up (see Erik’s post)—tend to favor the winner.

Because there is no clear projected winner in national polls at this point, it is hard to say whether the no-shows would disproportionately hurt Romney or Obama.  Regardless, my analysis, plus this systematic feature of past elections, suggests that late decisions by the undecided voters may not be the, er, deciding factor in 2012.

16 Responses to How Will the Undecideds Break?

  1. Marty October 26, 2012 at 4:26 pm #

    Based on your model variables, I’d call your attention to Rosenstone on third-party candidates. He says that there’s a four-fold threshold for third-party voters: they have to have unfavorable opinions of both parties and both parties’ nominees. Since you already have favorability in there, it’s no theoretical or data-point jump to add Rosenstone in.

  2. Michael Lewyn October 26, 2012 at 4:46 pm #

    To me a more interesting question is: how will undecideds break in Senate races? (More interesting to me because there are more of them).

    For example, let’s look at the Senate race in Florida or Ohio, where some polls show as many as 10-15 percent undecided.

    One way of looking at them is to say: these people are moderates and independents and so will break fairly evenly.

    Another way of looking at them is to assume straight-ticket voting, which means that ultimately these Senate races will break Republican if Romney wins the state.

    Any thoughts?

  3. Matt Jarvis October 26, 2012 at 4:55 pm #

    Um, JS: do you mean that its a myth that “UNDECIDEDS break for the challenger?”

    I haven’t heard of the challenger breaking for the incumbent before, but I wouldn’t put it past Romney!

    • John Sides October 26, 2012 at 6:39 pm #


  4. David Director October 26, 2012 at 5:56 pm #

    It seems to me intuitive that the undecideds (being largely uninformed and unlikely to ever come to a strong preference) will always break somewhere between 50-50, i.e. completely random, and the balance of the decideds, and more likely toward the latter. They will therefore never have more than a minor effect.

  5. Josh October 26, 2012 at 8:49 pm #

    I remember desperately clinging to the “undecideds break to the challenger” myth in 2004 when I worked for the Kerry campaign. I think we all know how well that went.

    • jim October 29, 2012 at 10:54 am #

      Actually the undecideds did break for Kerry on election day. That’s why the polls showed a wider victory for bush than the actual vote did.

  6. matt w October 26, 2012 at 9:05 pm #

    “If we focus on those undecided voters who either have voted or say they ‘definitely’ will vote….”

    I don’t follow this part of the analysis — how can an undecided voter have voted already?

    • John Sides October 26, 2012 at 9:26 pm #

      Sorry, that was a misstatement. I’ve fixed it to refer only to the “definite” voters.

      • Andy Rudalevige October 27, 2012 at 4:10 pm #

        John – maybe in a different section but one paragraph still leads “If we focus on those undecided voters who either have voted…” Of course, they could have blanked the presidential race, having made a brave decision about their call on the state representative line…

        • John Sides October 27, 2012 at 5:51 pm #

          Okay, now this is truly fixed. I hope.

  7. Prof. Toobin October 28, 2012 at 3:24 am #

    I hypothesize that the quadrant of a 50/50 undecided will factor in only if an obtuse angle is added to a linear isotope, I say this only because in 1952 the discovery by Dr. William Castle showed that if election were to be wrote down on a blackboard and the “L” replaced with an “R” it would indeed spell erection, which is how you arrive at what we all are.. screwed.

  8. jim October 29, 2012 at 11:01 am #

    Battleground showed Bush up 4 in their last poll and it was only bush 1.4 or so on election day. Most other showed between a 2-3 point margin for bush.

    In ohio late voters went drastically for kerry. Most expected a relatively easy win for Bush in Ohio…

  9. Kim October 29, 2012 at 6:31 pm #

    John – An assumption in a +/- 1 percent race that the undecided will follow the overall trend will guarantee a 50-50 split. A more likely alternative assumption is that the undecided will follow the overall independent voter trend, which currently does not favor a 50-50 split.

  10. John Carpenter October 31, 2012 at 3:59 am #

    There is a fundamental flaw in this model which it’s primary assumption
    untenable. While it is true that many decisions are made unconsciously, it
    it is equally true that we just as often are not consciously aware of our preferences,
    This, the democrat who feels favorable towards Obama may unconsciously be pro
    life or be highly opposed to Obamacare,

    Of course, that blade cuts both ways but it
    is far more important to decision making than
    party affiliation and perceived likeability of
    a candidate which may be its disdain at an
    unconcsious level.

  11. Ralph November 3, 2012 at 3:09 pm #

    If you let the third party in to the mix, it seems that there is less than 1% undecided. Most of the undecided’s position is ‘none-of-the-above”. There is a faction in some states that has a preferential loss of Romney over Obama where Romney ends up having 3% less and Obama 1% less. Polls including Goode show about a 1 to 2% bleeding from Romney.

    There simply are not enough polls or enough people asked to know. However, it would seem that there are not enough undecided to make a difference, even if they all went to either front runner. None of the polls are correct.