Working Paper: CEPR ID: DP3126
Authors: Micael Castanheira
Abstract: Why do people vote? This question received a lot of attention for more than thirty years, and yet remains unanswered. In this Paper, we take stock of existing empirical regularities and argue that we can use them to improve the model of instrumental voting. Once this is done, we show that purely rational/instrumental factors actually explain a large fraction of turnout variations. To perform our analysis, we use Myerson?s (1997, 2000) advances on Poisson Games and generalize the Riker and Ordeshook (1968) seminal model of instrumental voting. Applying our results to US data, we show how our model can explain several stylized facts, like the secular fall in turnout rates in the US.
Keywords: Paradox of Voting; Poisson Games; Rational Voter Hypothesis
JEL Codes: C72; D72; D81
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| traditional models of voting (D72) | failure to predict turnout accurately (D79) |
| variations in voting costs (K16) | variations in turnout (D72) |
| incorporating uncertainty (D80) | enhanced explanatory power of the model regarding turnout (D79) |
| post-electoral bargaining (D72) | moderating party positions (D72) |
| moderating party positions (D72) | affecting turnout (K16) |
| relaxing assumptions (C29) | explaining the decline in U.S. turnout rates since 1960 (K16) |
| instrumental motivations (C26) | substantial role in voting behavior (D72) |