Policies in Relational Contracts

How should a firm set policies– public decision plans that determine the role of its employees, divisions, and suppliers– to strengthen its relationships? We explore whether and how a principal might bias the decisions she makes to foster relational contracts with her agents. To this end, we examine a flexible dynamic game between a principal and several agents with unrestricted vertical transfers and symmetric information. We show that if relationships are bilateral– each agent observes only his own output and pay– then the principal may optimally make decisions in a systematically backwardlooking, history-dependent way in order to credibly reward agents who performed well in the past. We first show that these backward-looking policies are prevalent in a broad class of settings. Then we show by example how such policies might affect firm performance: for example, hiring might lag increases in demand or investment might be awarded in a biased tournament. In contrast to the game with bilateral relationships, we show that if monitoring is public, optimal policies never involve biased decisions.