Evolutionary dynamics in the voting game

Voter participation is immense but theoretically doubtful because there exists cost of voting and the probability of casting the deciding ballot is low. Game theoretic models (Palfrey and Rosenthal, 1985) confirm this paradox of voting. Individual optimization in the voting game is problematic with respect to the rationality and information requirements of traditional game theory. Therefore in this paper a non-optimizing but learning individual is considered. By individual learning the adjustment processes and equilibria of voter turnout are determined. Voters are able to learn to participate and substantial voter turnout is possible.