Neighborhood Housing Resilience: Examining Changes in Foreclosed Homes During the U.S. Housing Recovery

ABSTRACT

The surge in foreclosures in the United States that began in 2007 reached a peak in mid-2011, and since then, the rate of foreclosures has been decreasing, providing evidence of the housing market recovery. This study examines factors that affected changes in ZIP code-level foreclosure rates in more than 300 U.S. metropolitan areas during the national housing recovery. Using multivariate analyses of the long- and short-term effects of foreclosures simultaneously, this finding shows that certain characteristics of the mortgage and housing markets led to more rapid neighborhood recovery. Results also indicate, however, that most urban-form variables led to neighborhood resilience over the long term, that high shares of mixed land use were strongly associated with fewer foreclosures, and that high shares of auto dependency were associated with high foreclosure rates. Finally, findings suggest that low- and moderate-income neighborhoods, particularly in cities, were more vulnerable and less resilient to economic shock, and the accumulated effects of foreclosures worsened over the long term. However, low- and moderate-income neighborhoods surrounded by suburban affluent neighborhoods recovered more rapidly than those in cities did. Understanding such resilience to economic crises will provide policymakers with insights that they can leverage to establish housing policies for sustainable neighborhoods.

KEYWORDS:

• housing
• neighborhood
• resilience
• recovery
• foreclosure
• housing crisis

Acknowledgments

I would like to thank the Community and Economic Development Department of the Federal Reserve Bank of Atlanta for supporting this work through its Co-op Graduate Researcher program. All opinions are those of the author, and any errors in this research are her sole responsibility.

Disclosure Statement

No potential conflict of interest was reported by the author.

Notes

1. These threshold effects were inspired by the tipping point theory of Schelling (1971). The racial tipping point is a threshold at which white residents tend to move out of their neighborhoods, when black residents reach a threshold percentage value. Black residency remains endogenously stable as long as exogenous shocks occur below its tipping point. Galster, Cutsinger, and Lim (2007) claimed that the threshold effect enables researchers to analyze how neighborhoods endogenously respond to transient and exogenous shocks or changes.

2. The housing policies by state governments such as an established preforeclosure period (from a foreclosure notice to a foreclosure auction) and a postforeclosure period (from a foreclosure auction to a redemption period) grant time and opportunity for borrowers to pay off loans and related fees. Although the national average preforeclosure period ranges from 38 to 312 days,18 states have mandated postforeclosure periods that range from 10 to 180 days (Immergluck, 2010b).

3. Some large MSA are divided into MD, smaller groupings of counties within the MSA that include Chicago, Illinois; Dallas, Texas; Detroit, Michigan; Los Angeles, California; Miami, Florida; New York, New York; San Francisco, California; Philadelphia, Pennsylvania; Seattle, Washington; Washington, DC; and Boston, Massachusetts. As CoreLogic and LPS have their housing-related data in MD and because MSA are too large to capture metropolitan housing market characteristics, MD are more appropriate representatives of housing markets.

4. For the long term, the initial time may need to be set to an earlier year such as 1990 instead of 2000. However, as shown in the bottom of Figure 3 and the first 2006 map of Figure 4, foreclosure rates prior to 2006 exhibited constantly lower levels (less than 1%). Thus, because of limited data availability and lower foreclosure rates before 2000, this study sets 2000 as the initial year for the long term.

5. Based on the classification from the Community Reinvestment Act, “low income” denotes a family with a medium income of less than 50% of that of a metropolitan family, “moderate income” between 50% and 80%, “middle income” between 80% and 120%,” and “upper income” 120% or more (U.S. Government Publishing Office, 2017).

6. The proportion of low-cost loans is identified in the spread of the loan rate in the HMDA. High-cost loans are reported when the annual percentage rate of a loan is 3 percentage points higher than the comparable security rate of the U.S. Treasury. Lower priced loans are aggregated at the census tract level by excluding these high-cost loans and then converting them at the ZIP code level using the HUD ZIP code crosswalk file.

7. This study identifies factors associated with neighborhood resilience (declining foreclosure rates), so it uses low-cost loans (instead of high-cost loans).

8. Based on previous studies, auto-dependent neighborhoods may be less resilient (Dong & Hansz, 2016; Hartell, 2016; Hepp, 2013).

9. Because of a lack of LPS data, the numbers of ZIP codes in the short and long term differ: The short-term variables matched between 2011 and 2014, and the long-term variables matched between 2000 and 2014.

Additional information

Notes on contributors

Kyungsoon Wang

Kyungsoon Wang is a researcher interested in housing and real estate market analysis and community and economic development. She received her PhD from Georgia Tech.