Defense against an opportunistic challenger: Theory and experiments

This paper considers a contest setting in which a challenger chooses between one of two contests to enter after observing the level of defense at each. Despite the challenger’s chance of success being determined by a proportional contest success function, the defenders effectively find themselves in an all-pay auction that largely dissipates the value of the defended resources because the challenger will target the weaker defender. However, if the defenders form a protective alliance then their expected profits increase despite the fact that a successful challenge is theoretically more likely, given the overall reduction in defense. Controlled laboratory experiments designed to test the model’s predictions are also reported. Observed behavior is generally consistent with the comparative static predictions although challengers exhibit the familiar overbidding pattern. Defenders appear to anticipate this reaction and adjust their behavior accordingly.