Should we punish non-voters or peer-pressure them?

Presidential election years have a way of focusing attention on voter turnout.  Each season there seems to be renewed focus on voting participation rates in America and recognition that they tend to be quite a bit lower than in other democracies (e.g., America ~50-60%; Australia ~80%, Malta 94%, Iraq 58%).

What if everyone voted?  Would election outcomes be different?  Research is somewhat divided on this.  Some say no, some say yes. Peter Orszag argues that regardless of the effect on outcomes, voting should be mandatory and punishable.

Peter Orszag rightly points out the “paradox of compulsory voting” is that it could only be enacted if it could be shown to have a null effect; in other words, both parties would have to support it and if one party was seen to have an advantage, then the reform would fail to be enacted.  Orszag suggests that compulsory voting is a sensible reform that would “make democracy work better, in the sense of being more reflective of the population at large.”

Orszag’s argument correctly assumes that elections, or the outcomes they produce, are a public good—no individual has any incentive to contribute to them (via voting) because an individual’s action can make no difference; yet, collectively we are all better off if everyone participates.  Almost by definition, democracy can only work if its citizens participate in selecting representatives.  If no one participated, democracy would cease to exist.

One of the classic tenets of political science is that public goods result in collective action problems, and that we can traditionally use institutions to solve these dilemmas. For example, we can prevent car crashes at intersections using traffic lights; we can create safe drinking water by regulating polluters; we can create safe places for children to play and learn by using tax dollars to build parks and schools; we can create safe communities and perhaps prevent military actions against us by having police and national defense systems.  When individuals defect from participating in these public goods (by, for example, refusing to pay taxes), we punish them (by, for example, charging fines or sending them to jail). Our laws and system of sanctions act as institutions that solve the collective action problems created by the desire for public goods—this is what we teach in political science 101.

Orszag’s argument is in direct line with this type of approach to governing. Create an institution (mandatory voting) to solve the collective action problem (non-voting) that results from the desire for the public good (elections). But Orszag himself notes that it might be “unpopular” and difficult to enact. In light of this problem this blog is a good place to point to recent research on the causes and consequences of citizens’ participation in politics (especially voting behavior), which suggests that other tactics might be more effective, easier to enact, or perhaps more welcome by a citizenry increasingly suspect of government.

For example, Meredith Rolfe shows that voting is a social exercise. People tend to vote when they are active in large social circles where others vote. Likewise, in a large-scale experiment using Facebook, James Fowler and collaborators found that when an individual believes his/her friends have voted, the individual is more likely to vote. Also, Betsy Sinclair shows that many types of political behavior, including voting tendencies, are influenced by one’s peer network and their perceptions of their friends’ inclinations toward voting. Simple positive social messaging, such as “people like you have voted today, will you?” have a meaningful positive impact on voter turnout.

This is consistent with the mission of new organizations like Votizen, which aims to reduce the role of money in campaigning by activating individual’s social networks to help encourage their friends to vote.

So, mandatory voting laws would likely increase participation in elections in America, but it is unclear whether such reforms are feasible, and whether the benefits of increased participation would outweigh the costs of enforcing the law.  Rather, social media, social networks, and peer influence may be inexpensive and perhaps more effective ways of increasing political participation. Political parties and candidates have already been figuring this out. The use of social media in modern political campaigns has quickly become a standard tactic, rather than a novelty. Perhaps the government in general, or those who champion higher turnout rates, should take this research to heart and deploy a more systematic system of encouraging voter participation through social networks and their media.