Quantal response equilibrium in a citizen-candidate experiment

Abstract

In this thesis I address the next question: can the QRE theory, with more realistic assumptions, explain better the behavior than Nash Equilibrium (NE) in an experiment based in Citizen-Candidate model? The contribution of this thesis is to answer this question with the analysis of data from the election experiment assuming that agents are not rational, but they are strategic. In this sense, this work is located at the intersection of two perspectives: first, Behavioral Game Theory (Camerer, 2003; Gächter, 2004), where QRE can be considered; and second, Political Economy (Besley, 2007), where the citizen-candidate model is an important instance. I conclude that, considering stochastic and non-perfect maximizing decision rule, it is possible to better describe candidates’ decisions in the experiment.