Quantifying Homeless Populations

Quantifying homeless populations has never been more important. Accurate counts of populations in need are the fundamental data needed to decide policy and program recommendations. Historically, the Department of Housing and Urban Development has relied on a one-point-in-time count—that is, on a given day and time workers canvas a city or other jurisdiction to count the homeless people sleeping outside and those in shelters and they then add the two numbers to get the total number of homeless. Unfortunately, methods for this count are not standardized across jurisdictions. This leaves open the questions as to whether the counts are accurate in any given jurisdiction and whether these counts can be compared across jurisdictions.

In this issue of AJPH, Tsai and Alarcón (p. 633) propose novel methods to improve these counts. However, I believe there are additional methods that should be considered. While advocating for epidemiological surveys, the authors fall short of recommending key approaches to population size estimation. These include capture–recapture methods¹ and successive size estimation (built into respondent-driven sampling methods),² unique object multiplier, service multiplier,³ and multiple regression approaches.⁴ A useful primer to estimating the size of hard-to-reach populations is available from the United Nations Programme of HIV/AIDS and the World Health Organization.⁵ Finally, the definition of homelessness needs to be standardized. In my view, current definitions only encompass those sleeping rough on the streets, living in homeless encampments, or staying in shelters. These do not account for those who are homeless but may be staying with friends or family, although these individuals would undoubtably benefit from services.