Learning Benevolent Leadership in a Heterogenous Agents Economy

This paper studies the potential commitment value of cheap talk<br>inflation announcements in an agent-based dynamic extension of the<br>Kydland-Prescott model. In every period, the policy maker makes<br>a non-binding inflation announcement before setting the actual<br>inflation rate. It updates its decisions using individual evolutionary<br>learning. The private agents can choose between two different<br>forecasting strategies: They can either set their forecast equal to<br>the announcement or compute it, at a cost, using an adaptive learning<br>scheme. They switch between these two strategies as a function of<br>information about the associated payoffs they obtain through<br>word-of-mouth, choosing always the currently most favorable one.We<br>show that the policy maker is able to sustain a situation with a<br>positive but fluctuating fraction of believers. This equilibrium is<br>Pareto superior to the outcome predicted by standard theory. The<br>influence of changes in key parameters and the impact of transmission<br>of information among nonbelievers on the dynamics are studied