This story is from UCLA Today, a discontinued print and web publication.

In defense of the undecided

|
Lynn Vavreck is a UCLA associate professor of political science and communication studies and co-author, with John Sides, of “The Gamble: Choice and Chance in the 2012 Presidential Election.” She's also co-principal investigator of the 2012 Cooperative Campaign Analysis Project. Her op-ed appeared Nov. 1, 2012, in the New York Times.
 
The caricatures flourish. People who can’t agree on anything political find common ground on one point: undecided voters are vapid, disengaged and a little bit frightening.
 
And yet there they are, still part of the electorate. Over the past few weeks, I have been using data from the Cooperative Campaign Analysis Project to explore the thoughts and feelings of voters who were undecided last December.  The takeaway has been that they are typically less interested in politics; they pay less attention to news in general; and they are more likely to have moderate ideas on contemporary political questions (but 30 percent report that they have no political ideas at all).  So far, the caricature pretty much holds.
 
They are also more likely to be women (63 percent), more likely to have less than a high school education; less likely to hold a graduate degree; and more likely to have family incomes of less than $30,000 a year.
 
The data, however, also reveal that almost half of the undecided consider themselves Democrats or Republicans. But even this cue is not as strong for them – only 65 percent eventually choose their party’s candidate (as opposed to 95 percent of early-deciding partisans).
 
It is possible that political signals like party mean less to late-deciders because politics is like a foreign country to them. But another possibility is that some of their ideas conflict with the positions their putative party takes.
 
Here’s an example. On the question of whether to raise taxes on families earning more than $200,000 a year, most Democrats say yes – roughly 80 percent of early-deciding and undecided voters favor this policy. But among Republicans, 60 percent of early deciders oppose this policy while only a quarter of undecided voters oppose it.  Even at the highest levels of income (more than $100,000 a year), the difference among Republican early and late deciders on this issue is more than 30 points.
 
Similar differences can be found on whether undocumented immigrants who have been living in the United States should have a legal path to citizenship. Compared to early-deciding Democrats, undecided Democrats are more than 30 points less likely to favor the pathway; undecided Republicans are 20 points less likely to oppose it compared to their co-partisans.
 
But these out-of-step late-deciding semi-partisans do not always abandon their party at the ballot box. Conflicted, undecided Democrats are more likely (75 percent to 25) to break for Romney over Obama if they oppose raising taxes on the rich, but no more likely to break for Romney if they oppose the pathway to citizenship. The reverse is true for conflicted, undecided Republicans – more likely to choose Obama if they disagree with Republicans on immigration, but no more likely to choose Obama if they disagree with Romney on taxing the rich.
 
Looked at in this light, the undecided actually exemplify a type of political flexibility we often claim to admire, but often denigrate in practice. A healthy portion of undecided voters seem to understand when they are out of step with their party and this sometimes drives them to the opposing candidate. They may not be as interested in news or politics as you are, but they consider their preferences relative to party positions when making up their minds. Adjust those caricatures.
 
Let’s briefly move away from party to another canonical driver of voter choice in American elections: attitudes about race. Despite the subject’s prevalence in 2008, when the economic collapse portended a strong Democratic victory, the discussion of race has been conspicuously absent in 2012 or at least much less of a focus. The problem with this is that the 2012 election is going to be much closer. Attitudes about race could be pivotal in a way they almost surely could not have been in 2008, when we couldn’t stop talking about it.
 
I’ve modeled votes for Obama or Romney separately for early deciders and initially undecided voters using only party identification, ideology, retrospective evaluations of the nation’s economy, and a measure called racial resentment.
 
Racial resentment is one of a set of regularly used political science measures of attitudes about race.  It is born from the concept of symbolic racism, which has its share of critics. Essentially, it is a scale of four survey questions asking people to agree or disagree with questions about whether “generations of slavery” have made it hard for blacks to work their way up the economic ladder – or whether blacks would be as well off as whites if they only “tried harder.”
 
Racial resentment is related to voter choice for every presidential election in modern history.  Interestingly, even within the 2008 Democratic primary, these attitudes robustly predicted a voter’s choice between Obama and Hillary Clinton. Evidence from 2008 also suggests that racial resentment plays a larger role when Obama is running relative to others who have run for president before.
 
Vavreck.chart.Here is what the relationship looks like in 2012 plotting across deciles of resentment in the overall sample. After controlling for party, ideology and economic judgments, increasing levels of racial resentment (moving from left to right on the horizontal axis) decrease the likelihood of voting for Obama – not a shock.  But here’s where it gets interesting. For voters who were able to make up their minds early, moving from the lowest levels of resentment to the highest drops the chance of their voting for Obama by more than 70 points. The comparable drop among initially undecided voters is only slightly more than 10 points.
 
As we saw with party, attitudes about race among the undecided are related to their choices, but the relationship is weaker than it is for those who decide early. In other words, the racial attitudes of undecided voters do not affect their vote for or against Obama as dramatically as those same attitudes affect otherwise-similar early deciders.
 
On the one hand, this could be interpreted as more good news — another blow at the caricature. Perhaps undecided voters are truly post-racial. If race mattered to them as much as it does early deciders, they’d have already made up their minds, as the more partisan do. Maybe these voters are the ones who have moved “beyond” race, at least in terms of their candidate selection.
 
On the other hand, I’m already catching sight of Seth Meyers over at “Saturday Night Live “ working on the next skit about undecided voters with the too-good-to-pass-up punch line: “Wait, what do you mean the president’s black?”
 
Media Contact