Reciprocal benevolence and rent sharing in the employment contract: from experimental results to econometric applications.

Summary "The existence of non-compensatory wage differences is a stylized fact of labor economics. One explanation can be found in the gift exchange hypothesis initiated by employers. Employers adapt the wage to the standard considered fair by employees, inciting them to maximize their effort in return. The definition of the "fair wage" is firm-specific (Akerlof & Yellen), justifying the persistence of wage differences between firms. However, this interpretation contradicts an elementary behavioral assumption of economic analysis: egoi͏̈sm. Theoretically, shared giving stumbles on a lack of incentive for the employee to produce costly effort. One must then reconsider the scope of the gift exchange hypothesis as a determinant of wage individualization. Experiments show that individuals spend a share of their earnings to punish a selfish partner or reward a benevolent action. Reciprocity produces stable behaviors that must be taken into account in the case of the labor contract. We therefore incorporate these considerations into the agents' utility functions. The goal is to determine their impact on the degree of cooperation. Despite a structure that encourages opportunism, cooperation can occur if the players are reciprocal. We simulate equilibria to reveal the trade-off between the value of gains and incentives for reciprocity. The predictions of the model are tested on individual data from the REPONSE 98 survey. We test the determinants of punishment behavior and present a direct estimate of the gift exchange game. We evaluate the probability of obtaining productivity gains following the introduction of profit sharing in the firm. We highlight the reciprocal nature of the choices to enter conflict, engage in profit-sharing, or increase productivity, corroborating the Arkerlof and Yellen hypotheses."