An exact optimization approach to the principal-agent problem in infrastructure projects via PPPs

PPPs are a compelling mechanism for infrastructure procurement but are vulnerable to the principal-agent problem due to the divergence in the parties' utility functions. This research considers PPPs in infrastructure maintenance problems, in which the principal pursues high performance levels, whereas the agent pursues cost-efficiency. This paper proposes an optimization model that describes infrastructure performance as a result of deterioration and maintenance processes, and computes the associated benefits and costs for the principal and the agent, combining game theoretical notions with life-cycle analysis. Illustrative examples show how the organizational conditions surrounding PPPs shape the parties' decisions and outcomes by running the model under different scenarios. For instance, the model can anticipate the agent’s optimal reaction to specific contractual conditions; i.e., what maintenance plan can be expected from a contractor depending on whether certain performance-related constraints or incentives are in place. Finally, the model can find the optimal maintenance plan for one party conditioned on respecting the other party’s optimum within a specified tolerance. This allows to explore a variety of middle-ground solutions and choose an alternative in which one party’s gain exceeds the other’s loss, providing evidence for incentive design. This research provides a quantitative tool to support decisions and policy on PPPs by integrating physical, financial, and organizational aspects of infrastructure management.