Friends or foes? Optimal incentives for reciprocal agents

Widely used performance-based contracts put (positive or negative) externalities on co-workers. These externalities have been proven to shape an organization’s working climate especially when workers exhibit social preferences. However, it is a priori unclear whether a more friendly or a more competitive working environment should be encouraged. In this paper, we develop a theoretical model in which a self-interested principal has to motivate a set of agents to work hard. Agents are symmetric, potentially risk-averse, and exhibit reciprocity concerns towards each other. We show that the principal reduces the cost of achieving high effort provision by designing a psychological gift-exchange game, thereby replacing monetary with psychological incentives. We find that the optimal incentive scheme depends on the interplay between the agents’ attitudes towards risks and their preferences for reciprocity. In particular, the optimal scheme implements (i) a relative performance compensation scheme which induces an exchange of unkindness if agents are relatively little risk averse and (ii) a joint performance compensation scheme which induces an exchange of kindness if agents are sufficiently risk averse. Our findings can explain some puzzling empirical results.