Research

 

Working Papers

 

Does Building New Housing Cause Displacement?: The Supply and Demand Effects of Construction in San Francisco

San Francisco is gentrifying rapidly as an influx of high-income newcomers drives up housing prices and displaces lower-income longtime residents. In theory, increasing the supply of housing should mitigate increases in rents. However, new construction could also increase demand for nearby housing by improving neighborhood quality. The net impact on nearby rents depends on the relative sizes of these supply and demand effects. This paper identifies the causal impact of new construction on nearby rents, displacement, and gentrification by exploiting random variation in the location of new construction induced by serious building fires. I combine parcel-level data on new construction with an original dataset of historic Craigslist rents and panel data on individual migration history to test the impact of proximity to new construction. I find that rents fall by 2% for parcels within 100m of new construction. Renters' risk of being displaced to a lower-income neighborhood falls by 17%. Both effects decay linearly to zero within 1.5km. Next, I show evidence of a hyperlocal demand effect, with building renovations and business turnover spiking and then returning to zero after 100m. I find that gentrification follows the pattern of this demand effect: parcels within 100m of new construction are 2.5 percentage points (29.5%) more likely to experience a net increase in richer residents. Affordable housing and endogenous construction do not affect displacement or gentrification. These findings suggest that increasing the supply of market rate housing has beneficial spillover effects for incumbent residents, reducing rents and displacement pressures while improving neighborhood quality.

Works in progress

Evicted from the Land of Opportunity: Evidence on Displacement from California Rent Control

With Brian Asquith and Charly Porcher

We investigate the socioeconomic impacts of plausibly exogenous eviction from rent-controlled housing in California. Under the Ellis Act, landlords are allowed to evict all tenants from a building and withdraw it from the rental market. We assemble panel data on address histories, employment, income, and neighborhood characteristics for all Ellis-evictees in San Francisco and a control group of non-evictees in the same block. We confirm that those large-building evictions appear orthogonal to evictees’ individual characteristics after controlling for observable household and neighborhood effects. Comparing those two groups with a difference-in-difference approach, we find that evicted tenants between 1998 and 2012 not only exhibit a higher propensity to exit the city but also endure a reduction in nominal income that reaches 20 percent eight years after the eviction, the end of our analysis window. Those income losses are also experienced, in the same magnitude but more gradually, by evictees who remain in San Francisco. Those large income losses contrast with the income improvement we observe for non-eviction-related relocations. The negative impact extends to their residential destinations post-eviction, which tend to be neighborhoods with lower job density, higher unemployment rates, and diminished school quality, particularly for tenants with lower pre-eviction income. Specifically, children from evicted households face significant setbacks, evidenced by diminished earnings in early adulthood, suggesting a persistent intergenerational impact.