Social Norms and Voter Turnout

Ben Klemens
Ben Klemens Senior Statistician - Office of Tax Analysts, U.S. Department of Treasury

January 26, 2004

As much as we would like to believe otherwise, people do not turn out to vote solely to have a part in deciding public policy. After all, the probability that one person’s vote will swing the election is almost nil, while the effort of taking an hour or two out of one’s day to cast a vote is decidedly not nil. There are other considerations which induce people to take time out of their day to vote, and to the extent that these considerations are related to voters’ true opinions, they will affect the outcome of the elections.

Voting is affected by social norms

The focus of my own research, and that of the Center on Social and Economic Dynamics at Brookings, is on the effects of peers on a person’s actions. For example, a man hired in to a firm of men who wear ties will feel uncomfortable without a tie, but if the same man transferred to a company where no one wears ties, he would feel uncomfortable with a tie. There are a dozen plausible explanations for why people feel the need to emulate those around them, such as a desire to establish an identity as part of the team, fear of rocking the boat, or taking others’ behavior as evidence of what is right. As an economist, I can make no guesses at why people wear ties, but I can say what the data on tie-wearing look like: some offices are filled with tie-wearers, some have none, and very few have an even mix of tie-wearers and bare-necked men.

As it turns out, data on voter turnout look the same. My work on the U.S. National Election Studies data shows that even after controlling for all possible individual characteristics, some groups show very high rates of turnout, some show very low rates, and too few show moderate rates.

The princial investigators of the British Election Study (Whiteley, Clark, Sanders, and Stewart) took the direct approach and asked people if they felt that most of their friends and family think that voting is a waste of time. They found that, controlling for all else, if a person believes that his or her peers think voting is a waste of time, then that person is less likely to vote.

In short, my findings and the findings from the British National Election Study both show evidence that one person’s decision to turn out to vote is strongly influenced by the customs of the groups of which that person is a member.

Multiple norms can affect the outcome of an election

Tie-wearing, like all norm-influenced behavior, is `clumpy’, with a concentration of ties in some situations and a concentration of bare necks in others. In the same way, the data show clumps of people who turn out to vote and clumps of people who do not turn out.

If we want our elections to be representative of the population, then clumpy turnout is a disaster. Our first ideal is that everybody who can vote does. Barring this, the next ideal is that the opinons of the people who turn out are in proportion to the opinons of the electorate at large. But if people turn out partly based on the norms of the community, we fall short of even this limited goal, because there is a strong correlation between a community’s norms to turn out and its political leanings.

There is a reason for this. Social norms are often products of a group’s history, and changing past norms is difficult. If a group recently did not have the right to vote, or was discouraged from voting, then without some sort of internal or external pressure, it will maintain that norm of not turning out. So it is no surprise that Blacks, who were pushed away from the polls forty years ago, and immigrants, who did not have the right to vote before they were naturalized, are among the groups with the least voter turnout today.

These are groups that traditionally vote Democrat, which makes voter turnout a political issue. This is a shame. We all benefit from elections which are a truer picture of what the public feels.

Social norms distort election results more than bad ballots

There is little or no systematic relation between bad polling equipment and the opinions of the people who use it, which makes reforming antiquated equipment an easy issue to discuss. After all, the famous butterfly ballot which cost Gore votes could just as easily have been printed in a Bush-supporting area and cost Bush votes. Neither party stands to suffer immensely from better equipment, and everyone is better off when the election procedure gives a definite answer. Although it is also an important issue, debate over better voting equipment is a distraction from issues such as differential turnout, which distort our record of public opinon in a much more significant way.

Because establishing consistent social norms regarding turnout would generally favor Democrats, Republicans have a strong incentive to block bipartisan efforts to change the status quo. If the bipartisan groups fail to implement efforts to establish these norms, then Democrats need to bear the cost of improving this aspect of polling themselves.

As much as we would like people to turn out to vote for pure interest in public policy, there is abundant data that show that this is not the case. If our goal is to get an honest reflection of public opinion, we must do this by influencing the social, non-policy aspects of voting—and we know how to do this. For example, the social norms regarding smoking or the wearing of seat belts have changed significantly in the past decade, partly due to efforts by the federal government. A consistent social norm that voting is a part of an individual’s civic duty is also important to our society, and one we can and should work to achieve.