An Experimental Evaluation of Strategic Preemption

This paper reports the results of a series of two-stage, two-person non-cooperative games where one player can strategically preempt the other. In one of our designs, the subgame perfect equilibrium entails complete preemption; in the other, it entails partial preemption. Logit analysis indicates that players tend to completely preempt when it is optimal. However, when partial preemption is optimal, a non-trivial fraction of players persist in choosing the non-preemptive structure. This may result because of occasional irrational behavior following preemptive play, which induces some dominant agents to play less aggressively.