Selfish Voters are Not the Problem that Ails Democracy

Both right and left have standard stories explaining how selfish voters often prevent the government from adopting good policies. The right-wing story – made famous by Mitt Romney’s much-derided “47 percent” speech last year – is that voters become dependent on government transfer programs and then support those programs despite the harm they cause to the economy. The left-wing counterpart is the theory that middle and upper class voters selfishly refuse to support programs that benefit the poor or provide useful public goods, because doing so would require them to pay more taxes.

In reality, however, the data suggest that there is only a weak correlation between narrow self-interest and public opinion on most issues. For example, the young support Medicare and Social Security almost as much as the elderly do. During the Vietnam War, men eligible for conscription actually supported the war at higher rates than the general public. Although there are some exceptions, in general voters’ decisions are based on their perceptions of the public interest far more than on narrow self-interest. In this recent article, economist Dwight R. Lee has a good discussion of the reasons why voters act this way:

People face very different incentives in markets than they do in the voting booth. In the marketplace the shopper’s choice decisively determines what he receives and pays for. In the voting booth, the voter’s choice is almost never decisive in determining what he receives and pays for. At best, voters receive and pay for what the majority of voters choose, whether they vote for it or not. Your vote is as decisive as a market purchase only if the election would have been a tie without your vote….

The costs and benefits of what we are thinking about buying in the marketplace are far more important to us personally than the costs and benefits of what we are thinking about voting for. Assuming the probability of casting a decisive vote to be 1 in 500,000, we weight the importance of costs and benefits of a marketplace purchase 499,999 times more than we weight the same cost and benefits of a choice we are considering in the voting booth. For example, if you can buy a high-performance sports car that is worth $25,000 more to you than its price, you would obviously buy it, and make a quick trip to the automobile dealership to do so. In sharp contrast, if a poor person is considering a vote for a government transfer program that, in present value terms, would pay him $25,000 more than it cost him in taxes, his financial motivation to vote for the program is 5 cents. The probability may be high that the program will receive a majority vote, in which case he will be $25,000 better off. But since the probability that he will receive the extra $25,000 from the program only because he voted for it is 1 in 500,000, his expected value of voting for the program is 1/500,000 times $25,000, or a nickel. That is not nearly enough to cover the cost of driving to the nearest polling location to vote.

Thus, few people are likely to take the time and effort to vote if their sole motivation is narrow self-interest. As Lee explains, most of those who vote do so because they care at least somewhat about the interests of others in society. Even if you care a lot less about your fellow citizens than yourself, it may still be rational to take the time to vote, given that the cost of doing so is low. While most people probably don’t do precise, detailed calculations about the costs and benefits of voting, they do intuitively understand the idea that it may be worth investing a small amount of time and effort to get a small chance at creating a huge benefit for the rest of society if your vote is decisive in helping the “right” candidate win.

The problem, however, is that moderate altruism combined with a low probability of decisiveness is not enough to incentivize most voters to acquire more than a minimal amount of knowledge about they’re voting on. Thus, we get an equilibrium where lots of people vote out of a sincere desire to promote the public interest, but most of them have very little understanding of which policies are actually likely to do that. This problem is the focus of my new book Democracy and Political Ignorance. As I will explain in an upcoming post, an electorate of narrowly self-interested but knowledgeable voters might actually produce better policies than one where most voters are well-meaning, but largely ignorant.