Informed Voters Should See Psychiatrists

October 24, 2012, 8:09 PM GMT+0

In deciding between Barack Obama and Mitt Romney, we hope that voters are choosing between candidates based in part on candidates’ policy stances, such as their positions on taxes and war. In a book I have coming out this month, Follow the Leader (press release), I find surprisingly little evidence that policy stances matter to voters. Specifically, I find that voters’ prior policy views rarely lead them to change votes or even approval of candidates during campaigns. Instead, voters usually choose candidates for some other reason, such as the election-year economy, and then follow their preferred candidate on policy issues, adopting that candidate’s stances as their own. Rather than leading politicians on policy, voters mostly follow.

Are voters choosing between Obama and Romney on policy issues such as healthcare, taxes, and the war in Afghanistan? Or are they merely following, adopting their preferred candidates' views? Determining whether citizens are leading or following on policy is tricky—it's a “chicken and egg” problem. In the book, I use repeated interviews with the same individuals. But I can illustrate this tendency to follow with a surprising pattern that shows up in a single survey.

In a recent YouGov poll, I asked participants about their views on abortion policy and what position they thought Obama, Romney, the Democratic Party, and the Republican Party took on abortion. Only about 60% of respondents knew that Obama and the Democrats supported more pro-choice policies than Romney and Republicans. Given that the parties have had clear and long-standing positions on this issue, it's astonishing that 40% of Americans don't know this basic fact (other surveys find even higher levels of ignorance).

If voters tend to adopt their preferred politicians’ stances on policy issues, we should expect to find the following pattern: among those who know the candidates'/parties' positions on abortion, their attitudes about abortion should predict their views on other issues, even when those issues have nothing to do with abortion.

In the YouGov survey, I find exactly this pattern. Take foreign policy. Among respondents who said that abortion should always be legal (pro-choice respondents), 74% approved of Obama's handling of the war in Afghanistan. In contrast, among those who said abortion should never be legal (pro-life respondents), only 14% approved. This 74-14 difference is large, especially since these issues are unrelated. It can only come about if voters are following either Obama or Romney on one or both issues—bending their policy opinions to match their votes. They could be voting for Obama, for instance, because he's pro-choice and then adopting the belief that Obama has handled the war in Afghanistan well. They could be voting for Romney because he's pro-life, and then adopting the belief that Obama has handled the war poorly. Although it seems less likely, respondents could also be voting for the candidates because of the war, and then bending their abortion views accordingly. Most likely, however, many respondents are choosing between Obama and Romney on other grounds—such as the candidates’ parties, the economy, or their appearances—and then bending their opinions on both issues to match their votes.

The same tendency to follow shows up with all the issues I examined. Given Romney’s statement about the 47%, his concern for the middle-income and low-income individuals is an issue in this election. The YouGov survey asked how well “cares about the poor” described Romney. Again, we find that abortion attitudes almost perfectly predicted perceptions. Among pro-choice respondents, only 12% said that the statement described Romney well. Among pro-life respondents, almost 90% said so. Since this issue has nothing to do with abortion, citizens must be bending their opinions to match their prior views.

Another controversy for Romney in this election, whether he flip flops, exhibits a similar pattern. Almost 80% of pro-life respondents took the hard-to-square-with-reality view that Romney “takes positions on issues and sticks by them,” while only 18% of pro-choice respondents agreed. Importantly, these patterns only show up among people who know the candidates'/parties' positions on abortion. They are absent among the 40% who don't. (For similar evidence in the 2004 American National Election study, see this plot.)

The most important issue in this election is the economy. Are people bending their perceptions here as well? They are. Almost 80% of pro-choice respondents approved of Obama's handling of the economy, compared with only 11% of pro-life respondents. How about the trend in the economy? More than 60% of pro-choice respondents said it’s getting better, while only 9% of pro-life respondents concurred.

New York Mayor Ed Koch allegedly said, “Pick a dozen issues. If you agree with me on eight out of twelve, you should vote for me. If you agree with me on 12 out of 12, see a psychiatrist.” Based on these findings, informed voters should be keeping psychiatrists busy.

For democracy, these patterns suggest that voters often fail to lead politicians on policy. Instead, they follow. In Follow the Leader, I find that voters exhibit a similar tendency to follow in US, British, Canadian, and Dutch elections, that voters follow across a wide variety of policy issues, and that they often do so blindly, even when they don't know their preferred candidates' or parties' broader ideological orientations.