A Bit More on How Undecided Voters Might Break

In my earlier post, I modeled the choices of decided voters and then looked to see what that model would predict about the remaining undecided voters.  The apparent answer was that these voters will break pretty evenly for Obama and Romney.

My co-author Lynn Vavreck suggested a different approach.  In our data for The Gamble, we have interviews from a December 2011 YouGov survey with approximately 44,000 respondents.  Then each week from January until Election Day, YouGov reinterviewed a different subset of 1,000 respondents from among these 44,000.  In other words, we have a two-wave panel, where the date of the second wave varies.

Lynn’s suggested approach was: take the people who were undecided in December 2011 but decided in a later survey, model their choices, and then see what this predicts for the remaining undecided voters (as in my first post, these are the 195 undecided respondents in the 3 YouGov polls since the first debate).  The assumption here is that the remaining undecideds will make up their minds the way the earlier undecideds have—which is more plausible than assuming, as I was, that undecideds make up their minds like the decideds do.  (Of course, it’s still an assumption.)

In this case, the answer is much the same: the remaining undecided voters are predicted to split fairly evenly—about 55% for Obama and 45% for Romney.  Given the uncertainty in these estimates, I don’t think this implies a definitive Obama “win” among these voters.  Ultimately, these models do not suggest either candidate will get a significant advantage from late-breaking undecided voters.  The same caveats noted in my earlier post apply here as well.


10 Responses to A Bit More on How Undecided Voters Might Break

  1. Ed October 31, 2012 at 12:42 am #

    You might also look at the Rand poll as a basis for estimating how undecideds might break.

  2. Jacopo October 31, 2012 at 9:55 am #

    How the hell is a 10-point advantage for Pres. Obama “split fairly evenly”? If Romney had this advantage, I feel like we’d be seeing headlines saying: “Last-minute undecideds break Romney’s way.”

    • leonard October 31, 2012 at 10:00 am #

      that was a great comment !!!!!!!

      JACOPO get a journalism job NOW whoever you are we need you!!!!!!!!

      • John Sides October 31, 2012 at 10:05 am #

        Read the first post and you’ll understand. The caveats come about because my model of how the undecideds will vote is very simple, and because it’s possible that the undecideds who do vote will not be as favorable to Obama, and because small advantages in the small number undecideds don’t translate into much of a margin.

      • Jacopo October 31, 2012 at 10:23 am #

        I do a little. Here is a piece I co-wrote at the Huffington Post, if you’re interested: http://www.huffingtonpost.com/ray-errol-fox/emancipation-proclamation-150th-anniversary_b_1905792.html

  3. Jacopo October 31, 2012 at 11:12 am #

    Re John: I read your previous post, and I think the 50.1%-49.9% split you initially described requires a whole different vocabulary than a 55% – 45% split. One scenario is Missouri in the 2008 presidential election, and the other is South Carolina that same year: It’s only close if you zoom out to the point that such infinitesimals no longer matter. I’d say they do matter, and perhaps enough to decide who our next President will be considering the closeness of the 2012 campaign.

    As for whether such last-minute voters can decide an election, exit polls from 2008 show that they definitely can. A total of 14% of Missouri voters decided their vote in the last week of the 2008 election. Those who made up their mind exactly one week prior chose McCain over Obama 57% – 40%, and those who made their decision on Election Day voted 49% to 42% for McCain. Had these numbers been closer, say, within .2% of each other, Obama would have probably carried Missouri and the state would have retained its bellwether status.

    Naturally, we won’t know how these last-minute undecideds factor in the 2012 election until next week, but I’m guessing that if the Romney campaign was told they’d be winning undecideds 55%-45% instead of 50.1%-49.9% in the aftermath of Hurricane Sandy and a backfiring Jeep ad, I think they’d hold a press conference about it.

    Source: http://www.cnn.com/ELECTION/2008/results/polls/#MOP00p1

    • Matt Jarvis October 31, 2012 at 1:42 pm #

      One very important clarification. Those who CLAIM to have decided their vote at the last minute split that way. People are really quite terrible at explaining why they did what they did.

  4. Frankly Curious October 31, 2012 at 12:49 pm #

    I did my own calculation with Ohio voters using the “undecideds will break as the decideds” assumption. (I still think this is the best assumption.) I did make one important change: I filtered out the Democratic and Republican leaning independents to give Romney the best chance possible. The result: Romney still barely wins the undecided vote 53% to 47%.

    Details at my site:

    Undecideds Unlikely to Affect Presidential Race

    My feeling is that your 55-45 split is probably about right. To a first aproximation it is always going to be 50-50. Regardless, contrary to the claims of Jay Cost and others, the undecideds are not going to break hard for Romney. If Romney is going to win, it will be for more fundamental reasons.

  5. Arlo November 1, 2012 at 2:47 am #

    I spoke with an undecided voter from NH today. He’s 64. Here’s how the conversation goes after introductions and ascertaining he’s still undecided, and that he might be accessible to some very quick persuasion before I move on to GOTV:

    Me: “What’s going to help you make up your mind about who to vote for?”
    He: “Who’s on the ballot.” (The intonation affirms that this is an answer, not a question. In short, he will make up his mind based on the names that appear on the ballot.)
    Me: “Are you concerned about your retirement?”
    He: “I’m 64!” (This was preceded by a snort that is the mature equivalent of a teenage “Well, duh”.)
    Me: “Mr. Romney and Mr. Ryan have some plans that might not be to your liking. Mr. Obama on the other hand… ” (I won’t give you my whole pitch.) “What do you think about that? Is that something that might persuade you to vote for Mr. Obama?”
    He: “I don’t like any of them.”
    Me: “Will you be voting for another party? Libertarian?”
    He: “I don’t know.”

    I tick the undecided box and move on, because this is GOTV.

    I don’t envy you trying to read minds like his.

  6. white label surveys November 3, 2012 at 6:50 am #

    I agree that your 55-45 split is probably about right. To a first aproximation it is always going to be 50-50.