摘要：In homelessness research and policymaking, it seems to be axiomatic that single adults experience 3 temporally based types of homelessness: chronic, episodic, and transitional. We discuss problems with the theorization of this typology and with the research design, data analysis, and time-aggregated conceptualization and measurement of temporality in the empirical work supporting the typology. To address the latter, we suggest a time-patterned approach to temporality and report a 10-group typology that differs significantly from the more familiar 3-group typology. We argue that which approach is used—and how typologies are developed more generally—should be based on theory and the uses to which typologies are put rather than on claims to being more true. IN RESEARCH AND POLICY -making on homelessness, adults unaccompanied by children axiomatically experience 3 temporal kinds of homelessness: chronic, episodic, and transitional. Kuhn and Culhane extracted this typology from the research literature and provided initial empirical support for it. 1 In research, this typology has subsequently been employed by Kertesz et al. to show that episodic or chronic homelessness is linked to worse mental health than is transitional homelessness, 2 by Goering et al. to suggest that the problems faced by first-time homeless people are similar to those faced by people who are chronically homeless, 3 and by Caton et al. to identify risk factors for chronic homelessness. 4 Culhane et al. also used this typology to analyze family homelessness and found the same 3 categories among homeless families. 5 In policymaking, this 3-category typology has been employed by the federal government to focus on chronic homelessness and especially to encourage localities to develop 10-year plans to end chronic homelessness, 6 – 8 by the National Alliance to End Homelessness to urge a “housing first” approach to chronic homelessness, 9 and by the Urban Institute to specify steps for preventing and ending homelessness in general. 10 Although other typologies of homelessness among single adults combine temporal and nontemporal information 11 , 12 or employ only temporal information (as do Kuhn and Culhane), 13 these other typologies have not been used as extensively in research and policymaking as has the one analyzed by Kuhn and Culhane. Jahiel and Babor provided a more general review of typologizing homelessness. 14 Despite its extensive research and policy influence, the initial theorization of and supporting research for this typology have flaws that undermine its utility. This work has been a useful starting point for better incorporating the temporal character of homeless people's lives into research and policymaking, but it should not be the end point. In this article, we describe problems with the logic of how this typology was theorized as well as with the design, analysis, and analytic approach of the initial empirical research supporting that theorization. As part of this critique, we outline an alternative approach to categorizing temporality and cite evidence that uses that approach to identify a typology structurally and substantively different from the prevailing 3-category typology. Based on this evidence, we conclude that other temporal and nontemporal typologies may be useful both for constructing theories of homelessness and for policymaking. We also contend that no single correct typology exists 15 ; rather, we argue for constructing typologies based on theorizing, policymaking, and other uses to which they are put.
