The Design of Incentives in the Presence of Lobbying Activity

The effects of unproductive lobbying have so far mainly been addressed by the public choice literature on rent-seeking and by Milgrom/Roberts' (1988,1990) work on influence activities in organizations. Our paper makes an attempt to incorporate lobbying into the simple principal-agent framework that was proposed by Holmstr'm/Milgrom (1991). This framework has been used by Itoh (1992) to analyze the costs and benefits of relative performance evaluation (RPE), a method of pay that can reduce agency costs by inducing competition between agents. One of Itoh's main findings is that the principal may prefer the agents to cooperate if there exist productive interactions that cause sabotage under a system of RPE. In our model, each of the two agents can affect the output signal that determines his pay both by productive effort and by unproductive lobbying that makes his performance appear better. We show first that the presence of lobbying leads to lower-powered incentives in the optimal contract. Secondly, complementing Itoh's findings, we conclude that even if there are no productive interactions between agents, RPE may be non-optimal, because its use would provide the agents with an additional incentive to engage in lobbying. In this case, the principal will instead create a positive interdependence between agents' income (a kind of implicit teamwork) to reduce the attractiveness of lobbying.