Thursday, October 20, 2005

Why we vote

Henry excerpts the elegant mathematical example from Andrew Gelman et. al's paper on why voting is rational. The passage that jumped out at me was the following about why voting can't be all about self-interest, because the elections most clearly affecting self-interest, where self-interest should drive high turn-out, have the lowest turn-out:
Voter turnout tends to be higher in large elections—in the United States, highest for presidential elections, then congressional and state, then finally local elections tend to have the lowest turnout. Theories of voting that focus on instrumental benefits (e.g., the theory that says that voters are instrumental utility-maximizers who happen to overestimate small probabilities) would tend to predict higher turnouts in small elections. In contrast, the social-benefit theory predicts a slight increase in turnout for national elections, if the issues at stake are perceived as more important, on a per-voter level, than in local elections.
The proposition that local elections have the most impact on your self-interest is disputable. After all, federal taxes are over twice as large as state and local taxes combined (your mileage may vary by specific locality). So the self-interest rationale is somewhat consistent with higher turn-out in federal elections. Still, given all the other issues that local and state governments get involved in that the federal government doesn't (education, zoning, transportation...), the story Gelman et. al. tell is a compelling illustration of the importance of motives other than self-interest in decisions to vote.

Though there is something confusing in using notions of expected utility to explain social voting: in national elections all the issues at stake that raise the expected value of voting (and why it is rendered socially valuable to vote) would raise the importance of federal elections beyond state and local. So I think there is a hidden problem there, though this may be an artifact of fatigue and not an actual problem.


Post a Comment

<< Home