Chapelle, G., Ubeda M (2025). Competing for opportunity : Transport infrastructures and localized unemployment (revisions requested at Economic Journal)
Slides / Paper
Unemployment rates vary significantly across neighborhoods, yet the role of transport infrastructures in explaining these disparities remains relatively unexplored. Building on evidence that highlights persistent disparities despite significant reductions in transport times and the relative proximity of unemployed to jobs, we propose a novel quantitative urban model with frictional unemployment and heterogeneous workers. In this framework, better connections between neighborhoods might exacerbate unemployment disparities due to competition among workers with different productivity levels. We document this phenomenon using a difference-in-differences panel approach to estimate the impact of the creation of the Paris Regional Express Rail (RER). We find that the project increased the unemployment rate of low-skilled workers. Finally, we decompose the employment gap between skill groups. We find that, in Paris, differences in job market access actually reduce unemployment inequalities between college graduates and the rest of the population.
Chapelle, G. and Eyméoud, J.-B. and Wolf, Clara (2023). Land use regulation and the housing supply elasticity: evidence from France.
Slides / Paper
This study gathered original data on French metropolitan statistical areas to estimate and decompose their inverse housing supply elasticity, describing how housing prices react to demand shocks. Our findings confirm that French cities are highly inelastic, with an estimated average supply elasticity of 0.5. Furthermore, leveraging a nationwide regulation protecting historical monuments as an instrument, we found that land-use regulations controlled by local authorities appear to be mainly responsible for this low supply elasticity.
Chapelle, G. (2017). Does social housing crowd out private construction?
This paper provides new evidence on the impact of subsidized housing programs when access to developable land is strictly regulated. I study the impact of social housing in France, which represents about 1% of the French GDP in yearly subsidy. Specifically, I use the fact that subsidized programs are twice less dense than private projects in similar areas to test whether access to developable land explains the crowding out effect of private starts by public starts. Exploiting a a national law forcing some municipalities to build more social dwellings as a natural experiment in a Regression-in-Discontinuity framework , I find that one social dwelling crowds out about two private units which is proportional to the relative density of both sectors. This effect is robust with alternate instrumental variables and disappears when the share of land developed is low. This result sheds a new light on the way to think about place based policies when the land supply is inelastic which is the case in most of the major cities where these programs are used as Paris, London or New York.
Chapelle, G., Urvoy,C. and J-P. Venturine (2026). Mayors strike back, Sciences Po mimeo
Slides / Paper
In this paper, we study the impact of a national-level reform that abolished locally set floor area ratios (FAR) in France. FAR is a piece of regulation that local policy makers use to prevent urban densification. The goal of the reform was to foster the supply of housing projects in cities while reducing the consumption of agricultural land by densifying urban places. To identify its impact on the Paris metropolitan area, we use a difference-in-difference (DID) framework comparing housing blocks where a maximum FAR existed before the reform with the closest blocks where no FAR existed. We find that treated areas in which FAR were removed didn't experiment any change in their density and no particular dynamic in their price. This suggests that the suppression of FAR imposed by the central government didn't have any positive impact on the housing supply.
Chapelle, G and A, Trannoy, (2026). Estimating the demand and supply price-elasticity of housing floor space using a quasi experimental VAT cut
Slides / Paper
Using French cadastral and transaction data, we study the effects of a VAT cut from 20% to 5.5% on new residential construction in and around about 600 deprived neighborhoods. The reduced rate raised the housing stock by 2% in floorspace terms, equivalent to roughly two years of residential investment over a decade, and generated a gain in total surplus of about €13.6 billion. We estimate the price elasticity of floor-space demand at −0.24, with a 60% pass-through to buyers; landowners and developers each capture roughly half of the remainder. We also confirm that housing supply in France is price-inelastic (elasticity of 0.35).