Equity dynamics in bargaining without information exchange

Abstract

In this paper, completely uncoupled dynamics for n-player bargaining are proposed that mirror key behavioral elements of early bargaining and aspiration adjustment models (Zeuthen, 1930; Sauermann and Selten, 118:577–597 1962). Individual adjustment dynamics are based on directional learning adjustments, solely driven by histories of own realized payoffs. Bargaining this way, all possible splits have positive probability in the stationary distribution of the process, but players will split the pie almost equally most of the time. The expected waiting time for almost equal splits to be played is quadratic.