Website CV
Fields
Urban & Regional, Theory, Environmental & Energy
Job market paper
Dynamic spatial equilibrium models: theoretical and numerical foundations
We study the existence and uniqueness of dynamic equilibria in forward-looking quantitative spatial models in continuous time and space. We formulate the dynamic equilibrium problem as a mean-field game coupled with a static model that determines agents' flow utility as a function of the spatial distribution of the population. We establish existence for a broad class of models featuring monopolistic competition and, with discrete locations, perfect competition. We prove uniqueness under the Lasry–Lions monotonicity condition. We illustrate our results in discrete time and space with the canonical model of Caliendo et al. (2019). We develop an algorithm to compute the dynamic trajectory of solutions, using numerical methods from mean-field game theory. We illustrate the efficiency of our method through a large-scale example taken from the literature: a forward-looking version of the model of climate migration in sub-Saharan Africa of Conte (2026), with 949 locations over 60 years, whose full transition path we compute in less than 30 minutes. To the best of our knowledge, this is the first time that the full transition path of a forward-looking spatial model of this size has been computed.