Community-Level Determinants of Homelessness

In 2013, a startling image entered the public’s consciousness representing the scale of homelessness in New York City: Yankee Stadium. According to numbers from the Coalition for the Homeless, the 50,287 seats that make up the South Bronx giant would not be enough to contain the Big Apple’s homeless population. The sheer scale of homelessness poses a problem to policymakers and researchers grasping to understand the issue in this country; the ways in which counts of the homeless population are gathered and evaluated using samples and estimates is the underbelly of those efforts.

In an article recently published in the Journal of Urban Affairs, Thomas Byrne, Ellen A. Munley, Jamison D. Fargo, Ann E. Montgomery, and Dennis P. Culhane use 2009 US Department of Housing and Urban Development (HUD) data to identify community-level determinants of homelessness. Building upon the research of Barrett Lee, Townsand Price-Spratlen, and James Kanan in 2003, the article “New Perspectives on Community-Level Determinants of Homelessness” points to poverty, rent levels, the aging baby-boomer population, and the growing number of Hispanics in the country as points of consideration for policymakers working to craft solutions to a problem that is plaguing metropolitan areas nationwide.

Although homelessness is not a new concern to urban policymakers, new tools have been made available to researchers in just the past decade. As a result, Byrne et al. consider past research on homelessness to be limited because reliance on informant interviews or proxies may have lead to underestimations of the actual homeless population. In what they consider a more robust approach, the authors differ from previous research in their utilization of point-in-time (PIT) counts—a tally of the population on a specific night—of the homeless population as well as an expanded geographic scope that includes the suburban and rural populations surrounding cities.

Using data available for 338 metropolitan and 76 non-metropolitan areas in the United States that were collected on a January 2009 night, Byrne et al. stratified the samples into metropolitan and non-metropolitan areas and conduct their statistical analysis separately for each subgroup. In doing so, the authors use the estimates of the number of homeless persons in those communities as the dependent variable and six characteristics of those areas (housing market factors, economic conditions, demographic composition, safety net, transience, and climate) as the independent variables.

Ultimately, Byrne et al. find that several variables are positively associated with homelessness in metropolitan areas: rent level, homeownership rate, the size of Hispanic and baby-boomer populations, the proportion of single-person households, and the proportion of recently moved households. Further, these variables are found to be strongly significant among those living below the poverty line in cities as well as in areas surrounding urban hubs. Notably, the authors cite a 15 percent (metro areas) and 39 percent (nearby suburbs and rural areas) increase in homelessness per $100 increase in median rent for the examined area.

Giving these findings, Byrne et al. direct policymakers toward affordable housing considerations as well as additional safety net supports for baby-boomers, the Hispanic population, and recently moved households. The authors also indicate migration patterns and estimates of sub-population, such as families, the mentally ill, and veterans, as areas for further research.

With 633,782 people reported as homeless nationwide on a single January 2012 night, homelessness has been described as a problem both visible on American streets as well as strikingly hidden in countless shelters. Consequently, Byrne et al.’s upstream analysis of the issue presents itself as a useful resource. Nevertheless, the work done here, as the authors themselves recognize, is only a first step. Diving deeper into the numbers is likely to yield even more insights and considerations for policy-level interventions.

Article Source: Thomas Byrne, Ellen A. Munley, Jamison D. Fargo, Ann E. Montgomery, and Dennis P. Culhane, “New Perspectives on Community-Level Determinants of Homelessness,” Journal of Urban Affairs 35, No. 5 (December 2013): 607-625.

Feature Photo: cc/(N. Feans)