This paper aims at explaining the casual evidence that monetary disinflations produce recessions, using the evolutionary game theory framework. This is an appealing framework for analyzing this issue because it combines two ingredients commonly blamed for the costs of credible disinflations: limited rationality and coordination failures. The study of a disinflation episode is the analysis of a transition between two steady state equilibria. The dynamics between those two equilibria is generated by the replicator dynamics, which is based on an appealing strategy selection rule: the use of strategies that have a worse than average performance is reduced through time. The disinflation costs in our model depend on how fast passive agents - who continue to use the strategy that was optimal in the inflationary equilibrium - are converted into active agents - who adopt maximizing strategies during the transition.
