Résumé
Following the example of pioneering countries (UK, USA in particular), since 2010, in France, local managers have been required to justify the efficiency of hydraulic infrastructure projects (dyke, dam, etc.) for flood prevention in order to claim subsidies granted by the State. The efficiency assessment must be carried out according to a cost-benefit analysis (CBA) respecting the State's recommendations. Among these recommendations, it is required to estimate the impact of the project on the monetised damage of floods (difference in mean annualised damage between the reference situation and the situation with the project). This recommendation has led to many advances in French flood damage modelling practice.
The French government also requires that a so-called "constant land use" assumption be made. In concrete terms, this means estimating damage by considering that the land use remains "frozen" in the state it is in at the time of the study. This assumption greatly facilitates the implementation of the method, as it makes it unnecessary to make projections on the evolution of land use, either in the reference situation or in the project situation. The main reason for this recommendation is not that, but that the State considers that anticipating changes in land use as a result of the project would be contrary to the policy of regulating land use in flood-prone areas that it has been pursuing since the early 1980s. At the same time, the French government is pursuing a policy to encourage territories not to rely entirely on hydraulic infrastructures, by promoting so-called integrated flood management projects (combining crisis management, control of land use and adaptation of existing stakes). In practice, the "constant land use" assumption is not verified, even less so on the time scale of the economic evaluation. It is very common that the installation of a hydraulic infrastructure is associated with an intensification of urbanisation in the newly protected area, less so with a disuse of the area.
The objective of our work is to show how relaxing the "constant land use" assumption in the economic evaluation based on the estimation of flood damages can, contrary to the fears expressed by the French State, allow to better clarify the irrelevance of focusing exclusively on structural solutions, without questioning the interest of controlling the use of flood-prone areas. To this end, we show how to relax this assumption in practice within the framework of a CBA. Then, based on a database of about 200 CBAs carried out to obtain subsidies over the last 10 years, we analyse the influence of the "constant land use" assumption on the estimation of the efficiency of projects.
In perspective, we argue that relaxing this assumption could also allow to enter a virtuous circle of knowledge production by encouraging the consolidation of the understanding of the co-evolution of urbanisation and hydraulic infrastructures.