Relational contracts with and between agents

We study a dynamic multi-agent model with a verifiable team performance measure and non-verifiable individual measures. The optimal contract can be interpreted as an explicit contract that specifies a minimum bonus pool as a function of the verifiable measure and an implicit contract that gives the principal discretion to increase the size of the pool and to allocate it among the agents. To mitigate the threat of collusion, the optimal contract often converts any exogenous productive interdependence into strategic payoff independence for the agents. Under productive complements, an unconditional bonus pool (pay without performance) can be less costly than one conditioned on the verifiable team measure.