Who Wins in November?

It's a standard survey question:  who ya gonna vote for?

Less standard:  who ya think is gonna win?

A new WPost/ABC News poll has both questions.  The one getting attention is the tie, at 47 percent apiece, on who people say they're going to vote for in the presidential election.  In fact most people are settled on their candidate (4 percent of Obama supporters say there's a chance they'll change their mind, 8 percent of Romney supporters say so).  Obama's also ahead in the "enthusiasm" scores.

Less attention is going to the question asking people to predict the outcome.  Six-in-10 say Obama wins come November, and only a third predict a Romney victory.  In the scholarly biz, we talk about wishful thinking, which means we tend to predict our preferred candidate will win.  Yes, I've written about this topic on my own blog, in at least one newspaper column, and in my own academic research. 

It's fascinating that the two candidates are in a tie but there's a significant difference in predictions of who will actually win, leaning heavily toward Obama.  Lemme break this question down a bit more for you.

• Party ID:  87 percent of Dems predict an Obama victory, while only 60 percent of GOPers predict a Romney victory.  For you budding scholars out there, this supports the wishful thinking hypothesis.  It also demonstrates the softness of Romney's support among Republicans.  Oh, among "independents," 51 percent predict an Obama win and 38 percent predict a Romney victory.

• Ideology:  Looks like party identification above, with 52 percent of conservatives expecting Romney to win and 86 percent of liberals predicting an Obama victory.

• Education:  By a small amount, the greater the education, the more likely you are to predict an Obama victory (56 percent predicting Obama at the lowest level of education, 62 percent at the highest level).

• Region:  Respondents in the Northeast were the most likely to predict an Obama victory (68 percent), while in the South 54 percent predicted an Obama win.  It's interesting that even in solid GOP South, over half expect Obama to win.

• Income:  No real effect here, ranging from 58 to 62 percent predicting an Obama victory.  I find this fascinating in that it suggests the 1 percent and the 99 percent are about the same, at least in predicting an election outcome.

• Sex:  No real difference between men and women.

• Religion:  This gets a bit messy.  Among those with no religious convictions, 77 percent expect Obama to win.  White evangelical protestants are the least likely to say Obama wins (31 percent).  No real surprise if you read this in terms of the results above on party ID.

• Previous Voting:  Not surprisingly, 86 percent of the people who voted for Obama in 2008 expect him to do it again in 2012.  Among those who say they voted for McCain last time around, 66 percent expect Romney to win.

What's this tell us?   First, it says I need to find another way to spend my summer afternoon.  Second, it tells us that Romney's support is indeed soft, or softer than he'd like.  Third, it tells us that people really do expect outcomes based on preferences.  There's a danger here, or at least theoretically a danger, when you expect one kind of outcome and you get another.  That could possibly lead to less trust in democracy and voting, at least it's a plausible argument, but I've tinkered with this question and have never found this to actually be the case, even in the tortured 2000 U.S. prez election. 

Wishful Thinking

I've been wrestling with a wishful thinking manuscript, and of course -- as always seem to be the case -- the data are getting in the way of a good theory.  Wishful thinking is when preference leads to expectation.  In other words, if I'm for Candidate X, I'm probably also likely to predict Candidate X will win the election.  How often does it happen?  See below. This graphic I created from the ANES cumulative data file and goes only to 2004, but I'll tell ya now the wishful thinking in 2008 was just as high. The predictive accuracy is just that, what percentage accurately predicted the election. As can be seen, in close elections the accuracy tends to be low, in runaway elections, it tends to be higher.

I may write up a version for Like the Dew or one of my other favorite places. The paper I'm working on takes a simple idea -- that those who expect to win, but don't, are likely to have lower scores in government trust or satisfaction with democracy. And, from the media angle, exposure to the news should moderate this effect.

But do the data show that? Not as well as I'd like, so I may push this idea aside.

Who Will Win

Predictions of who is going to win a political campaign are fascinating stuff. Many people project their own preferences onto an outcome. In the literature that's called wishful thinking. The origins of wishful thinking research is out of studies of sport fans who think their team is going to win even when all the evidence suggests otherwise. Then the notion got moved over into politics. A few studies have been done. Here is some my own data.

Respondents were asked who is going to win. I then combined that with who they supported. Looking at presidential elections from 1968 to 2004, you see how much preference guides prediction.


Republicans: Out of 6,960 surveyed, 81 percent predicted their preferred candidate would win.
Democrats: Out of 6,664 surveyed, 64 percent predicted their preferred candidate would win.


The amount of wishful thinking from year to year is fairly stable, but predictive accuracy (getting it right) ranges wildly from election to election. This is important. When elections are close (like 2000), personal preference and wishful thinking gets in the way and people are less accurate in predicting the outcome.


When it's a runaway election, like 1972, then accuracy improves. You'd think media exposure would improve accuracy, if for no other reason than reading or watching the news would expose people to public opinion polls and the viewpoints of others, but it does damn little to dampen the preference -> prediction bias.