Compromise or Alternation? Experimental Evidence on Coordination Under Payoff Asymmetry

While the previous literature identifies the compromise option as a potential coordination device, it remains unclear how varying degrees of payoff asymmetry affect its adoption. This study experimentally examines coordination behavior in a repeated battle-of-the-sexes (BOS) game with a compromise option across three distinct levels of payoff asymmetry. We implement three between-subjects treatments that vary the degree of payoff asymmetry in the original BOS game while fixing the payoff for the compromise option. Under a hybrid matching protocol, we find that when the payoff asymmetry is higher, more groups coordinate on the compromise option. While payoff asymmetry initially reduces the coordination rate, repeated interaction mitigates this effect through learning. The alternation strategy is shown to be more efficient than the compromise one, though both enhance fairness. Our results reveal how the degree of payoff asymmetry influences subjects' strategy adoption between compromise and alternation.