Working paper

Learning to Disagree in a Game of Experimentation

Alessandro Bonatti, and Johannes Hörner

Abstract

We analyze strategic experimentation in which information arrives through fully revealing, publicly observable “breakdowns.” When actions are hidden, there exists a unique symmetric equilibrium that involves randomization over stopping times. With two players, this is the unique equilibrium. Randomization leads to dispersion in actions and to belief disagreement on the equilibrium path. The resulting lack of coordination has significant welfare consequences. In contrast, when actions are observable, the equilibrium is pure and welfare improves.

Keywords

Experimentation; free-riding; mixed strategies; monitoring; delay;

JEL codes

  • C73: Stochastic and Dynamic Games • Evolutionary Games • Repeated Games
  • D83: Search • Learning • Information and Knowledge • Communication • Belief
  • O33: Technological Change: Choices and Consequences • Diffusion Processes

Replaced by

Alessandro Bonatti, and Johannes Hörner, Learning to Disagree in a Game of Experimentation, Journal of Economic Theory, vol. 169, May 2017, pp. 234–269.

Reference

Alessandro Bonatti, and Johannes Hörner, Learning to Disagree in a Game of Experimentation, TSE Working Paper, n. 17-791, March 2017.

See also

Published in

TSE Working Paper, n. 17-791, March 2017