Election Day Approaches
Why Some People Shouldn't Vote
By N. Gregory Mankiw
As election day gets close, get ready to hear the usual exhortations about voting. Whether Bush or Gore is the better choice is debatable, but responsible people all agree that everyone should be encouraged to vote. It's a national disgrace, the hand-wringers say, that millions of eligible voters fail to turn out in presidential elections. Voting is a civic responsibility, they tell us, because democracy works best when everyone participates.
The problem is, this isn't true. Sometimes the most responsible thing a person can do on election day is stay at home.
Before we're too hard on people who don't vote, we should ask why they don't. It is tempting for us voters to view abstainers as lazy free riders who aren't holding up their end of the democratic bargain. But is it really this simple? Not according to economists Timothy Feddersen and Wolfgang Pesendorfer. They wrote a 1996 article in the American Economic Review that is now on the cutting edge of explaining why people don't vote.
Feddersen and Pesendorfer suggest that nonvoting has to be understood together with a related phenomenon--the decision of voters to skip some items listed on the ballot. This behavior, which political scientists call roll off, is common. Feddersen and Pesendorfer give an example of a 1994 Illinois gubernatorial race in which about 3 million citizens showed up at the polls, but only 2 million voted on a proposed amendment to the state constitution.
Anyone who has ever entered a polling booth can easily see why roll off occurs. You come ready to vote for your favorite candidate in some race you've been following closely, but then you face a whole list of races and ballot questions, most of which you know little or nothing about. What do you do? You could quickly make a decision based on your scant knowledge. But what if the contest is very close? Do you really want the outcome based on your almost random vote?
So you choose another course: You skip the item. In practice, this means that you are relying on your fellow citizens to make the right choice. But this can be perfectly rational. If you really don't know enough to cast an intelligent vote, you should be eager to let your more informed neighbors make the decision.
Feddersen and Pesendorfer suggest that not showing up to vote is motivated by the same reasoning as roll off. Eligible voters who are less informed about the candidates than their fellow citizens choose to stay at home, knowing the outcome will be more reliable without their participation. By not voting, they are doing themselves and everyone else a favor. If the ill-informed were all induced to vote, they would merely add random noise to the outcome.
What's the evidence that this theory is right, that nonvoters are less informed than voters? Studies of voter turnout have found that education is the single best predictor of who votes: The highly educated turn out more often than less educated. A classic argument for why democracies need widespread public education is that education makes people better voters. If this is true, then the less educated should show up at the polls less often. They are rationally delegating the decision to their better educated neighbors.
So the next time a friend of yours tells you he's not voting, don't try to change his mind. It's a good bet that if he's not voting, he's not been following the election closely anyway. Maybe he watched a baseball game instead of the debates. Maybe he is bored silly with all the talk of targeted tax cuts, privatized social security, and campaign finance reform. Maybe he's as ignorant about public policy as those focus groups of undecided voters that are the media's latest darling.
So rather than pushing your friend to the polls, perhaps you should thank him for staying at home. He's making your vote count just a little bit more.