Sung-Ha Hwang (Sogang University)
Abstract. Recent evolutionary models have represented bargaining over a surplus (as occurs between a land lord and a share-cropper, for example) by a perturbed Markov process. We study how the stochastically stable bargain depends on two distinctions governing how deviations from best response play occur. First, deviations may be directed (the sharecropper asks for more, not less) or undirected (the share cropper deviates from the current contract in either direction). Second, deviations may occur at some given rate, or they may be less likely to occur the more costly they are to the deviant. The resulting four stochastic processes select as the stochastically stable bargain four different outcomes (contracts), including the Nash, Kalai-Smorodinsky, egalitarian, and a novel bargaining solution that exhibits significant nonmonotonicities and demonstrates the interplay of two key drivers of evolutionary selection: (i) the ease of making errors; (ii) the ease of responding to errors. (with Jonathan Newton)