Free Access
Issue
E.J.E.S.S.
Volume 15, Number 3, 2001
Evolution and Learning in Markets
Page(s) 171 - 184
DOI https://doi.org/10.1051/ejess:2001103
DOI: 10.1051/ejess:2001103


European Journal of Economic and Social Systems 15 N°3 (2001) 171-184

An evolutionary model of voting

Juan D. Montoro-Pons and Miguel Puchades-Navarro

Departamento de Economía Aplicada, Universitat de València, Spain.

Abstract
Collective allocation of resources that takes place in political markets is characterized by the complex exchange that emerges among the individuals involved. Traditional Public Choice models depart from individual rational choice in a setup in which many of its strict requirements need not hold. This paper introduces a model of social interaction among agents in a simple political market which departs from bounded rationality and evolutionary dynamics as the key mechanisms that drive individual behavior. Learning plays a significant role as it allows to establish an individual link between decisions and collective outcomes. The model is that of a representative democracy with two parties in which individuals are restricted to a one dimensional policy space. The main findings from computational experiments allow us to revise the results of traditional models, specially those related to the voting paradox. We find that turnout levels may be higher than expected in a population composed of fully rational agents, and that there is a rationale for abstention that stresses the role of limited information, the discounting of the future, and the extent of the redistributive policies.


Key words: voting, vote motive, collective action, evolution, bounded rationality, learning, classifier systems.


© EDP Sciences 2001

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.