Résumé
Over the long term, sometimes on the scale of several centuries, transport investments will shapethe fundamental character of the country, the quality of its environment and its ability to meet futurechallenges. This underscores the importance of making decisions in the most informed mannerpossible, making the best evaluation of the benefits they will provide and the costs they willengender, especially since their funding is largely based on public funds, a scarce resource ingeneral and even more so today.Project cost-benefit analysis (CBA) often makes use of parameters related to the public financesystem, which are designed to take into account imperfections in the tax structure and the shortfallof public resources. This paper discusses how to define, estimate and use for CBA the “opportunitycost of public funds”, which measures the inefficiency in the structure of the tax system. Itdiscusses also the methods for prioritising projects in times of limited budgets when, for a given taxstructure, tax revenue does not provide sufficient resources for public spending on all projects thatmerit it. To take into account that constraint, it suggests the introduction of a "scarcity cost ofpublic funds" coefficient, which applies to public spending, and which should be used together withthe opportunity cost of public funds when public funds available are insufficient to carry out allworthwhile projects.The opportunity cost depends on the tax structure, regardless of its level: as the distortive effect ofthese taxes increases (reducing the incentive for economic activity), so does the opportunity cost.The scarcity cost depends on the level of public funding for investment: as the difference betweenthe funds for investment and the volume of worthwhile investment increases, so does the scarcitycost. The opportunity cost applies to the project's expenditures and revenues; the scarcity cost onlyapplies to public expenditures.Methodology :Refering to Beaud(2013) we analyse the main sources of public revenues and their distortiveeffects, deriving analytical formulas first. Then, using statistical data for France and introducinginformed assumptions, we issues numerical estimations of distortion coefficients for each mainsource public revenue in France, and for each main type of public entity (each of these having itsown sources of revenues), among those the French state.The scarcity cost depends more on the balance between a level of budget constraint and the list andcosts of potential transport projects: an illustrative simulation is made so as to estimate an order ofmagnitude for the scarcity cost in France.The limits and extensions of these notions and methodologies are then discussed. Analyses on thepractical use of OCPF and SCPF does lead to cross political economy considerations withnormative economics, and compare methodologies used in several countries.Results :Numerical results indicate that the OCPF for the main sources of revenues range between 1,01 to2,2, the combined OCFP for the French state being close to 1,2. The scarcity cost is estimated to be0,05 when systemic risks are fully taken into account in project CBA, but comes out higher whenrisks are not fully taken into account in net present value estimations.The discussion drives to question other types of similar distortions, which do not necessarily relateto public funds, some of them leading to imperfect competition issues for instance. Practical advicefor the use of OCPF and SCPF at the project level and for programming, and their limits, are given.