Skip to main content
Blog
Are Voters Competent? An Interview with Neil Malhotra

election-2012In the latest installment of our Election 2012 series, political scientist and RSF Visiting Scholar Neil Malhotra discusses his research on retrospective voting and voter competence.

Q: In his book, Just How Stupid Are We?, the popular historian Rick Shenkman writes, "The consensus in the political science profession is that voters are rational." Before we go into the literature, I wanted to ask you to give your own assessment: How strong is the evidence that voters are rational? Would you agree with Shenkman’s conclusion?

A: I'll quote Vanderbilt political science professor Larry Bartels' response to Shenkman: "Well, no." If anything, the consensus in political science is that voters are uninformed and do not have well-structured preferences. Nonetheless, I think the question of whether voters are rational or irrational is not the right one. The important question is: Under what conditions does the American electorate collectively make decisions that benefit society and promote democratic accountability? That's a much tougher and more important question, I think.

Q: Let’s look at how political scientists have approached this question over the years. Let’s say I conducted a series of studies to find how much voters know about government, such as, "Who is the President?" or "What does the Federal Reserve do?" If I found that most people didn’t know the answers, could I conclude anything about voters' competence?

A: I don't think so. Skip Lupia of Michigan has rightly pointed out the question should not be "What do voters know or not know?" but rather "What do voters need to know?" Why is knowing the name of the Chief Justice an important or necessary job for voters? Indeed, the proponents of a research agenda called "retrospective voting" noted that voters actually need to perform fairly simple tasks: evaluate the health of the country and reward/punish the incumbent accordingly.

Q: So let me take another line of attack: Even if I don’t know the content of a politician’s policies, I can still judge the conditions on the ground. If my life has improved under a particular President and I vote to re-elect him or her, am I not being responsible? Is this not evidence of rationality on my part?

A: Perhaps, but it's a tricky question of what metrics voters should be using. Should they look at how their own life is doing, or should they think more collectively and altruistically? Should they judge the performance based on the current health of the economy or based on the change over four years? Or maybe the change over one year? Or maybe compared to the counterfactual of how his opponent might have performed over the last four years?

Q: Let me turn to your own research in this area. In “Random Events, Economic Losses, and Retrospective Voting,” you and Andrew Healy argue that it’s difficult to assess the impact of negative election outcomes on election results. Instead, you look at another area: the effects of tornadoes. Why is this a better way to assess retrospective voting? And how did voters who experienced tornado-related deaths and damage treat incumbents?

A: Using a natural experiment like tornadoes is helpful because unlike the health of the economy, we know when they begin and end, and can assess whether politicians' response to them was effective. We found that voters punish incumbents for tornado-related economic damages but not tornado-related deaths. At least in this domain, it appears that what voters are doing is "rational." They are not emotionally responding to tornado deaths, which are fairly random and not highly correlated to tornado damage. Rather, they are behaving like economic voters, punishing the incumbent based on the state of damage.

Q: Another study of yours I particularly enjoyed examined the effects of local college football games on election results. Tell us what you found.

A: Like tornadoes, seeing how college football outcomes is a good test case for rational retrospective voting, since incumbents cannot prevent, prepare for, or respond to college football wins and losses. They are totally irrelevant events. We find that on average an incumbent gets an extra percentage point of the vote if her local team wins as opposed to loses before Election Day. The results are not intended to mean that college football outcomes change election results, but rather that voters reward and punish politicians based on their mood. So we can interpret retrospective voting not simply as a voters acting like accountants, but also as voters expressing their emotions and frustrations.

Q: One of the implications of this study, you write, is that "events and information themselves may not be paramount in explaining election results. Rather, what may be most important is how campaigns use those events to affect voters’ perceptions of both their own well-being and the well-being others to whom they are socially connected…" Have you seen any examples of this strategy in this, or another recent, election?

A: Definitely. I think the right way to think about how the economy influences elections is not to view voters as closely reading the Wall Street Journal, but rather the economy as a narrative that the media and campaigns try to shape. For instance, in 1992, even though the economy was improving, President George H.W. Bush lost the war for the narrative; people still felt emotionally that the economy was in bad shape. Similarly, in this election, Romney is trying to emotionally tie Obama to the malaise felt with high unemployment rates and college graduates living with their parents. Conversely, Obama is trying to make voters feel better by reminding them that things are on the right track.

Q: A major issue this election is whom to blame – for the deficit, or the economy, or gridlock in Washington. President Obama, for example, has said he thinks voters will blame Republicans for the difficult debt ceiling negotiations last year, but I wanted to ask you: When things go wrong, or when government fails, what do we know about how voters decide to apportion blame?

A: Based on my own research, I have found that partisanship greatly colors how people decide which politicians to blame for bad circumstances, but partisanship is not over-powering. People have the ability to process actual information as well. One reason I think why Obama has exceeded expectations in this election is that people clearly do not blame him completely for the state of the economy given what was handed to him by the previous administration. Yes, Republicans are less likely than Democrats to do this, but a surprisingly high percentage of Republicans attribute the current economy to the Bush Administration.

Governance & Policies
Audited Financial Statements
Headquarters
Contact Us