We perform an experimental test of the canonical mechanism for Nash implementation, using 3 subjects in non-repeated groups, as well as 3 out-comes, states of nature, and integer choices. We nd that this mechanism successfully implements the desired outcome a large majority of the time, while an embedded comprehension test indicates that subjects were gener-ally able to comprehend their decision tasks. The performance can also be improved by imposing a ¿ne on a dissident. However, the strategy prole corresponding to the pure Nash equilibrium is seldom played. Thus, we are left with the puzzling fact that implementation of the desired result occurs by means of strategy pro les that are not pure Nash equilibria. We offer an explanation to this fact based on the existence of a mixed-strategy equilibrium when preferences show risk aversion and love for truth.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados