Housing Recovery in Damaged Communities in New Jersey after Hurricane Sandy

Background: Hurricane Sandy landed in southern NJ on October 29, 2012, killed more than 100 people, left more than 8.5 million households without power, and damaged or destroyed more than 200,000 homes across the state's 21 counties (Neria & Shutz, 2012; FEMA, 2015).  

            Immediately after the disaster, public policy support was provided in nine coastal counties that constituted 98% of the major and severely damaged housing units in NJ overall. The programs include Individuals and Households Assistance Program, Small Business Loan Program, National Flood Insurance Program, and the FEMA Public Assistance Grant Program. In the most severely affected counties, additional funding was provided through Community Development Block Grant: Reconstruction, Rehabilitation, Elevation and Mitigation Program, and Homeowner Resettlement Program. How these policies individually and as a whole impacted housing recovery across communities with different socioeconomic and demographic profiles has not yet been studied, particularly in relation to damage levels.

            The concept of community vulnerability (Iversen 2008; Zakour & Gillespie 2013) has been widely used to explain many aspects of natural disasters. Nevertheless, how communities are vulnerable has been less fully examined. As housing recovery is a dynamic social and economic process that varies according to context (Quarantelli, 1995), this study examined the path from community vulnerability to housing recovery looking at both community characteristics and policy interventions.

Sample/Methods:  This retrospective longitudinal case study compared a literature-identified set of pre-disaster community characteristics and the effects of multiple public policy programs, while controlling for a set of time-variant community resilience indicators, including employment rate and income, on levels of housing recovery (operationally defined by percent of building permits to total occupied housing units/households [Census Bureau, 2014]) between 2010 and 2014, two years before and after Hurricane Sandy. The sample consisted of 51 municipalities in the nine counties in which between 4% and 58% of housing units suffered either major or severe damage.

            Structural equation modelling (SEM) was used to determine the path from vulnerability to housing recovery, via multiple public programs, separately and as a whole, and via the community resilience indicators. The spatial analytical tool ArcGIS 10.2 was used to show the spatial relations between housing recovery patterns and community vulnerability and resilience.

Findings: Holding damage levels constant, communities with higher proportions of Hispanic households had significantly lower levels of housing recovery while communities with households with an adult >age 65 had significantly higher levels of housing recovery. The contrast was partly due to the different levels of total public support the two types of community received. Further, while the public policy programs individually mediated the negative associations between African American and female-headed households and housing recovery, communities with larger proportions of African American, female-headed and Hispanic households were “vulnerable” to lower levels of housing recovery because they lacked sufficient public program support. Even so, higher employment rates and incomes buffered vulnerability to lower housing recovery. Because housing is the "wobbly pillar" of the welfare state, the housing needs of these particular groups should be more fully addressed by disaster policy.