This article proposes a method to determine the optimal program for a set of projects, taking into account their links (incompatibilities, competition, or complementarity), funding sources (public-private partnership), and possible budget constraints. We analyze the method in formal terms. By applying the Kuhn-Tucker theorem, we identify the direction of the dual variables associated with the constraints and link them to classical optimization methods. We use simulations to test the robustness of the programming criteria commonly used in practice, the scope for decentralized decision-making, and the use of private funding and public-private partnerships to carry out infrastructure projects.