Bundling Tasks and Contracts The Case of Public-Private Partnerships

Recent contributions to the literature on Public-private Partnerships identify conditions that support task bundling vis-à-vis a conventional model of procurement in which the State delegates investments and operations into di¤erent …rms. When asymmetric information between agents is a concern, the literature …nds that risk allocation, long term contracting, and commitment matter for sector performance. Naturally, ownership has an important say in the debate but does not matter if complete contracts can be written. When non-contractible situations do arise, the owner of the asset can bene…t from a renegotiation capturing pro…ts otherwise held by the agent that operates the asset and thus reducing investments in asset-speci…c human capital or in other speci…c investments. In this article, we look for conditions that are beyond ownership to analyze task bundling. We …nd that a PPP better internalizes the externalities between activities thus limiting the need of transferring costly informational rents. This advantage disappears if the government cannot separate incentives at the service provision stage according to the builder’s e¢ ciency or when it implements a conventional model. Conversely, the latter may perform better than a PPP if the government can obtain some advisory services from the service provider about the builder’s construction cost. I am grateful to Marisela Montoliu Muñoz and Antonio Estache for their useful comments. [email protected]