Hurricane Katrina and the Long-Term Outcomes of Subsidized Housing Recipients in New Orleans

• In Event: Climate Shocks and the Reallocation of Households and Housing

Abstract

Hurricane Katrina shocked New Orleans in 2005, destroying nearly 60% of the rental stock and displacing more than 1 million people (Vigdor, 2008).  Considerable work has documented the impact of this shock on migration (e.g. Sastry and Gregory, 2014), individual financial outcomes (e.g. Deryugina, 2018) and house prices, both in New Orleans and in the receiving neighborhoods of migrants (Daepp, 2023).  However, very little is known about one of the more vulnerable populations in New Orleans: the roughly 20,000 households that were receiving housing subsidies through the Department of Housing and Urban Development (HUD).In this paper I study the impact of Hurricane Katrina on several outcomes for Housing Choice Voucher holders and public housing residents. Specifically, the project will investigate (i) the impact of Katrina on neighborhood characteristics and individual outcomes (e.g. income, housing stability) of Housing Choice Voucher (HCV) recipients living in New Orleans at the time of the storm, (ii)  how these impacts differed between HCV recipients who remained in New Orleans and those who left and (iii) how these impacts differed for residents of public housing in New Orleans prior to Katrina. Combining geospatial data on water inundation after the storm with restricted-access HUD administrative data I employ a difference-in-difference regression framework where I track outcomes for voucher recipients and public housing residents that resided in New Orleans the year prior to the storm, 2004, in areas flooded during Katrina relative to areas with little or no water. This identification strategy relies on the fact that, controlling for the salience of flood risk before the hurricane, any variation in surge heights (the storm’s impact) is conditionally random in the context of very fine-grained locational (e.g., census tract) and temporal controls.  Results from this paper have important policy implications.  First, it documents how a shock like a natural disaster can impact the well-being of the households HUD serves.  Second, it explores the differential impacts of policy-levers at HUD's disposal. In particular, New Orleans' replacement of its original public housing after Katrina with new, mixed-income developments offers an opportunity to compare longer-term outcomes for those that were permanently displaced and those that returned to the new developments, potentially generating insights for HUD's approach to re-imagining public housing moving forward. Results from this project also have the potential to contribute to the debate between the use of place-based and tenant-based subsidies.

Author

• Kate W H Harwood, University of North Carolina, Chapel Hill
  Presenting Author