Association for Public Policy Analysis & Management

Natural disasters cause substantial social losses and they have particularly challenged the capacity of those economically and socially disadvantaged populations (e.g., low-income families, the elderly and persons with disabilities) to respond to and recovery from disaster shocks. Due to their destructive nature, disasters may impede the provision of public housing in the affected communities. They may increase the local demand for public housing and result in a temporary housing shortage and prolonged waiting time. These follow-on impacts may in turn affect the residential location decisions as well as the community’s long-term recovery following the disaster shock. However, little is known about the post-disaster dynamics in public housing, particularly considering the fact that the stock of public housing is in decline in the United States and there is a long-waiting list for housing assistance.

In this paper we address two related research questions. First, how do extreme weather shocks affect the availability of public housing? Second, how do post-disaster assistance programs alleviate the disaster shock to public housing? For the latter question we specifically focus on the effects of the Public Assistance (PA) program spending of the Federal Emergency Management Agency (FEMA) and the Small Business Administration (SBA) disaster loans. We expect these programs to facilitate the post-disaster recovery of public housing through at least two channels: (i) the PA projects that fund repair and rebuilding of public buildings would ensure public housing is functional soon in the aftermath; (ii) the PA infrastructure rehabilitation projects would help private landlords and households to return quickly and rebuild; 3) The SBA low-interest loans may help private landlords to repair and make rental units available to needy residents soon after a disaster shock.

To answer these questions, we employ a panel data set of all U.S. counties over the period from 1993 to 2015, using a variety of variables indicating the number of subsidized housing units, rents, occupancy rate and the waiting time, merged with disaster aid data (PA expenditure and SBA disaster loans). To address the potential endogeneity of the public disaster programs, we use the state representation in overseeing congressional committees as an instrumental variable. Our preliminary analysis, based on the instrumental variable model, shows that weather incidents have a significant and negative impact on the public housing units, and they also increase the occupancy rates. We also find that the PA program expenditure can significantly reduce the negative disaster shock to public housing. Overall, our research sheds light on the challenges confronting the socially vulnerable populations relying on public housing when a disaster strikes. Our findings are also beneficial for policymakers in making disaster aid allocation decisions.