The externalities from social housing, evidence from housing prices,” mimeo

I investigate the impact of social housing on the sales price of neighboring flats in Paris. I construct a unique dataset including flat sales and social housing projects at the building level. To account for endogenous placement of social housing projects, I use a difference-in-differences strategy that includes fine geographical controls and trending unobservables. In my preferred specifications which control for building fixed effects, a particular spatial pattern emerges: a 10 percentage points increase in the social housing share implies a 1.2 % increase in housing value within a radius of 50 meters. However, private properties located farther away from the social projects within a 350 to 500 meter belt experience price decrease by 5.5%. The positive effects appear more important for small dwellings and for properties located in poor neighborhoods while negative impacts dominate in high income neighborhoods and for family dwellings. Further estimates exploit the unexpected win of a left-wing mayor in Paris, which was followed by a sharp increase in social housing units driven by the direct conversion of private rental flats into social units without any accompanying rehabilitation. This natural experiment allows to identify the impact of the inflow into the neighborhood of low income tenants, separately from the effects of social housing on the quality of the existing housing stock. I do not find evidence of a positive impact of the conversion projects on housing prices