Rules, Discretion, Corruption, and Efficient Sole-Sourcing in Procurement

I study a model of procurement with moral hazard and adverse selection. The procurer is either corrupt or honest and can choose between sole-sourcing and competitive tender. I compare two procurement regulations: no sole-sourcing is allowed (rigid regulation) and sole-sourcing is allowed in an emergency (flexible regulation). Sole-sourcing in an emergency is efficient. Whether there is an emergency that justifies sole-sourcing is the procurer's private information. A regulator may fire the procurer depending on his belief that the procurer is corrupt. I find the counterintuitive result that if the gain to corruption is big, a flexible procurement regulation may be better than a rigid procurement regulation. If the gain to corruption is sufficiently small, the flexible regulation may be worse or better than the rigid regulation. Interestingly, although the inefficient choice of sole-sourcing is not always penalized, there exists a perfect Bayesian Nash equilibrium in which a corrupt procurer could be given discretion and incentivized to use sole-sourcing efficiently but without an explicit monetary incentive contract. In this case, the flexible regulation is better than the rigid regulation. The results are driven by three sources of inefficiency that are discussed in the paper.