Coate and Conlin: A group rule-utilitarian approach to voter turnout

Coate and Conlin. A group rule-utilitarian approach to voter turnout: Theory and evidence.

In Brief

Coate and Conlin discuss a rule-utilitarian theory for voting where individuals vote to maximize the utility of their side, and use data from local Texas liquor referendums to show that the model fares well. Similar to the Feddersen and Sandroni model, but with an emphasis on group benefits.

Theory

The authors conceive of the decision to vote as a game not between individual citizens but between groups. Each group wants to maximize the chance that they are successful in getting their measure passed or candidate elected, but they would also prefer to do this with the least effort possible, to maximize the expected utility of the group. This tension can help explain the variety in turnout, because duty is seen as conditional. Their rule-utilitarian approach to voting is that an individual votes if their voting increases the expected utility to their 'group', which contrasts with other models like Feddersen and Sandroni where the aggregate benefit to 'society' (i.e. to everybody, not just to the group) is what matters.

Hypotheses

• H1 (Rule utilitarianism): Voter turnout by each side (supporters and opponents) should depend on (1) the relative size of each group and (2) the willingness of each side to bear the costs of voting.
• H2: The five parameters determined through analysis of the referendums should model the decision to vote well

Independent Variables

• Proportion of Baptists
• alcohol-related auto fatalities
• policy compared to neighboring policies
• Whether the election was held on a weekend, in the summer, or on a bad weather day
• Age, sex, race, marital status
• metropolitan area

Dependent Variable

• Turnout in liquor referenda (broken into both supporters and opponents)

Findings

Highlights include increasing population decreased turnout for both sides and that the demographic variables pointed in the expected direction for almost all the variables; however the static analysis does not stand as a good test of the rule-utilitarian model because other predictions might posit similar directions

Including the five parameters into the model improved its predictive power.

Conclusion

The authors also compare the findings with what would be expected under an intensity hypothesis, like the expressive account of Brennan and Lomasky. They find that this model does a decent job predicting this outcome but that it can be rejected at a 5% significance level by a comparison with their rule-utilitarian approach, which they conclude does better. They also say that it would be hard to distinguish their findings from those predicted by other rule-utilitarian models like Feddersen and Sandroni's, and can also be compared to the pivotal voter model.

Comments and Criticism

The ecological fallacy could be a significant threat because they took the aggregate measures for the various independent variables and assumed they were evenly distributed among the population, which is likely untrue; the aggregate characteristics need not represent the characteristics of individuals within the study.