![Sample our Geography journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5937/TOC-Subject-Sample-2021-Geography.jpg"})
ABSTRACT
Stable housing is a fundamental platform for individual and collective well-being, and research indicates that a significant disruptive effect of severe environmental disasters is residential displacement. Despite extensive research on the intersection of disasters and housing, the effect of major disasters on evictions remains understudied. How do landlords and renters respond to the economic dislocation that accompanies disasters and to what extent do major disasters lead to evictions? To answer these questions, we adopt a mixed methods approach. Analyzing county-level data on evictions and disasters between 2000 and 2016, we find that disasters are associated with significant increases in evictions in the year of a disaster and the two years following a disaster and that increases in the housing cost burden are associated with higher eviction rates. We complement these quantitative findings with qualitative interviews and archival analysis from Panama City, Florida in the year after Hurricane Michael. The qualitative findings suggest that eviction dynamics may differ by landlord size and identify challenges for small landlords accessing federal assistance, particularly because of clouded titles from unrecorded property transfers. Together, the findings indicate that disasters increase evictions and lead to significant disruption for many low-income tenants for years after the disaster.
Acknowledgments
We thank all of the Panama City, Florida, residents who contributed their insights and experience to this research as well as the staff of Legal Services of North Florida, particularly Charlotte Waters, for their wisdom and assistance. We also thank Daniel Powers, Marisa Prasse, and Benjamin Walker for their research assistance. Finally, we thank the MIT Leventhal Center for Advanced Urbanism for its support for this research and the three anonymous reviewers for their very helpful comments.
Disclosure Statement
No potential conflict of interest was reported by the authors.
Notes
1. We recognize that the federal disaster declaration process considers multiple factors as set out in the federal regulations at 44 C.F.R. § 206.48 and is inherently political (Reeves, Citation2011; Salkowe & Chakraborty, Citation2009; Schmidtlein, Finch, & Cutter, Citation2008). Most disasters do not rise to the level of a federally declared disaster and do not qualify for federal assistance. Here we focus on the most severe 5% of federally declared disasters by value of property damage, which we believe includes those disasters most likely to have the largest effect on local housing markets and creates an unbiased sample of the most significant disasters over the study time period.
2. Much of the existing disaster relief and recovery infrastructure dates to the 1966 Disaster Relief Act (P.L. 89–769 (November 6, 1966)) and its subsequent amendments, particularly the Disaster Relief Act of 1974 (P.L. 93–288 (May 22, 1974)) and the Robert T. Stafford Disaster Relief and Emergency Assistance Act (P.L. 100–707 (November 23, 1988)). Recent revisions have included the Post-Katrina Emergency Management Reform Act of 2006 (PKEMRA, P.L. 109–295), the Sandy Recovery Improvement Act of 2013 (SRIA, Division B of P.L. 113–2), and the Disaster Recovery Reform Act of 2018 (DRRA) (P.L. 115–254 (October 5, 2018)).
3. The allocation of Community Development Block Grants for Disaster Recovery is authorized under Title I of the Housing and Community Development Act of 1974 (P.L. 93–383 (August 22, 1974)) and regulated by the federal regulations at 24 C.F.R. § 570.
4. This data set provides counts of eviction filings and completed evictions, and the filing and completed eviction rate. It also provides a count of rent-burdened households. We accessed the county-year file on September 19, 2018, and the most current version is available at https://evictionlab.org/, where an update entails a limited subset of validated records and the full set of records, the latter of which is the most current version of our data set. Modeling results are qualitatively parallel to ones we report when we substitute our data for the current, full set of eviction records. To account for outlying county-years (e.g., Los Angeles County), we replace the number of evictions in county-years in the top 0.01% with the number of evictions in the 99.99 quantile.
5. This data set provides panel data for disasters by year, county, and type (e.g., flood, fire). We accessed the file on April 18, 2019, which was available at https://cemhs.asu.edu/sheldus, where an updated version of the data is now available. Modeling results are qualitatively parallel to ones we report when we substitute our data for updated disaster records.
6. We fill in missing values for each county using the values from the subsequent census or survey year. We use values from subsequent censuses to populate missing values in the preceding one. We recognize that linear interpolation relies on the assumption that the measures change in a linear fashion between decades, which is not always true. But in the absence of annual data over the whole time period, we believe reliance on the common strategy of linear interpolation is the most appropriate approach.
7. This data set provides time-series data with county identifiers for population, share of non-Hispanic White and Black population, share of Hispanic population, share of households in poverty, share of population with bachelor’s degree or higher, share of multifamily, vacant, and rental housing, median rent, and median home value. The file is available at https://www.brown.edu/academics/spatial-structures-in-social-sciences/diversity-and-disparities; we accessed it on April 11, 2019.
8. This data set provides data on valid, inspected, approved renter households, as well as amounts of assistance by type to those households. We accessed the file on April 12, 2019, and the most current version is available at https://www.fema.gov/data-feeds. To account for outlying county-years (e.g., Queens County after Hurricane Sandy), we replace the value of assistance in county-years in the top 0.01% with the value of assistance in the 99.99 quantile.
9. In the immediate years after a disaster, the demographics of an area change (Schultz & Elliott, Citation2013). The reliance only on pretreatment confounders is a limitation that may reduce the accuracy of the estimates in years following the disaster and warrants future research on this dimension.
10. Available at https://baycoclerk.com/court-records/case-search/.
11. The state enacted the Hurricane Michael Recovery Loan Program to provide low-interest mortgages and down payment assistance to qualifying applicants whose incomes were less than 140% of the area median income (applicants did not need to be first-time homebuyers to qualify). This program was recently restarted by Governor DeSantis, noting that the storm’s damage called for “years and years of support” (Cassels, Citation2020). Additional relief efforts included a foreclosure counseling program and a recovery-specific $65 million expansion of the existing State Housing Initiatives Partnership program, which provides entitlement dollars to local governments that adopt a plan to produce affordable homeownership and rental housing for very low- to moderate-income Floridians (Florida Housing Finance Corporation, 2020). Although multifamily projects were eligible for these recovery funds, a minimum of 65% of the total funds were earmarked for homeownership projects (Ibid). Finally, the state approved $50 million in bridge loans to affordable housing developers in hurricane-affected counties and another $50 million in additional federal HOME fund allocations, the latter of which is expected to fund the construction of 200 affordable rental units (Ibid).
Additional information
Funding
Notes on contributors
Mark Brennan
Mark Brennan is a postdoctoral fellow in planning. He works on making essential goods and services more accessible to those who rely most on private markets and public programs.
Tanaya Srini
Tanaya Srini is a graduate of MIT’s Department of Urban Studies and Planning. She currently advises on a range of local-level technology policy issues as a fellow at the Ford Foundation and holds research affiliations with the National Housing Law Project and MIT.
Justin Steil
Justin Steil is an associate professor of law and urban planning. Broadly interested in spatial dimensions of inequality, his research examines the intersection of power, space, and the law in areas such as environmental justice and access to place based resources.
Miho Mazereeuw
Architect and landscape architect Miho Mazereeuw, is an associate professor of architecture and urbanism at MIT and is the director of the Urban Risk Lab. Working on a large, territorial scale with an interest in public spaces and the urban experience, Mazereeuw is known for her work in disaster resilience.
Larisa Ovalles
Larisa Ovalles is a Research Scientist at the MIT Urban Risk Lab. Larisa is currently working on a multi-year project to develop alternatives for FEMA post-disaster housing solutions across the US. Larisa holds a Bachelor of Architecture from Cornell University and a Master of Science in Architectural Studies in Urbanism from MIT.