Is There Really a Turnout Paradox?
This paper revisits the turnout paradox and shows that the correct computations of endogenous pivot probabilities provide plausible upper bounds for the cost parameter in high-turnout equilibria. Thus, the widespread view that said equilibria are fragile is overstated. Typically, these equilibria have excessively narrow margins of victory and mixed pure strategy profiles. But with a continuous version of the outcome function and differential costs of voting it is possible to generate realistic equilibria in a voting game with costly voting.