- The Housing-Health Connection
- Volume 20, Number 2
- Managing Editor: Mark D. Shroder
- Associate Editor: Michelle P. Matuga
The Geography of Vacant Housing and Neighborhood Health Disparities After the U.S. Foreclosure Crisis
Georgia Institute of Technology
Georgia State University
- Objectives: We examined the impact of long-term (6 months or more) vacant housing and various durations of vacancy on a variety of health outcomes at the neighborhood level across three types of U.S. metropolitan areas (metros): (1) those that have experienced consistently strong growth, (2) those that have undergone weak growth, and (3) those hit hardest by the foreclosure crisis.
- Methods: We used hierarchical linear modeling with long-term vacant housing data derived from the U.S. Postal Service as well as data for health outcomes obtained from the Centers for Disease Control and Prevention to examine the health effects of residents who resided in 19,243 neighborhoods (census tracts) in the 50 largest metropolitan areas during the housing recovery.
- Results: Neighborhood long-term vacancy is significantly associated with neighborhood health problems in adults, but the association between vacant housing and neighborhood health outcomes varies based on the growth trajectory of the metropolitan area. For most health outcome measures, long-term vacancies are more strongly associated with poor outcomes in strong-growth and hard-hit metros than in weak-growth metros, but the reverse is true for asthma and mental health. Our findings also suggest that very long-term (more than 3 years) vacant housing increased significantly after the housing crisis and was significantly associated with health problems in all three types of metros.
- Conclusions: The differences in the relationship between neighborhood-level longterm housing vacancy and health outcomes across the three types of metros should be considered when addressing community development strategies for decreasing vacancy rates aimed at improving health outcomes.
Previous Article | Next Article