Sequential Contests with Incomplete Information: Theory and Experimental Evidence

Abstract

In this paper, we investigate behavior in two-player sequential-move contests with complete and incomplete information about the value of the prize, theoretically and experimentally. First, we describe a Bayesian equilibrium of a sequential contest in which both players have private prize values. Then, we test our predictions in the experimental laboratory. We analyze three settings: symmetric prize valuations with complete information, asymmetric prize valuations with complete information, and asymmetric prize valuations with incomplete information. In all settings, and regardless of the order of moves, we find that subjects' behavior is less consistent with theory and more in line with simple mental shortcuts. The decision of the first-mover is relatively more complex than that of the second-mover. Our data supports a simple investment heuristic: first-movers invest half of their own valuation. The decision environment is free of strategic uncertainty and ambiguity for second-movers, and hence, they can condition their investment directly on the investment of first-movers. Regardless of valuation, second-movers invest frequently in one of the following ways: drop out of the contest, i.e., bid sufficiently close to zero, or invest at or just above the first-movers' investment. We add to the growing literature by showing that experimental contest data can be better explained by simple heuristics, rather than equilibrium predictions.

Keywords: contest, sequential moves, heterogeneity, incomplete information, investment heuristic

JEL Classification: C72, C91, C99, D82, D91