Homeless Children: Update on Research, Policy, Programs, and Opportunities. Section II. Research About Homeless Children


Beginning with the earliest published reports in 1987, the literature on homeless children now spans about 23 years. Viewing these studies in the aggregate, a set of “first generation” studies, many of which were reviewed 15 years ago by Rafferty and Shinn (1991), can be distinguished from a second stage of investigations published after the review. The first studies conducted on homeless children sounded a public health alarm (Alperstein, Rappaport, & Flanigan, 1987; Bassuk & Rubin, 1987; Miller & Lin, 1988; Rescorla, Parker, & Stolley, 1991; Wood, Valdez, Hayashi, & Shen, 1990). Findings indicated that homeless children had a range of health and mental health problems. Data for these investigations were collected in the mid-1980s, not long after the issue of homelessness for families became apparent in the United States.

A second generation of studies followed in the early 1990s to further clarify the impact of homelessness on children (Buckner, 2004). Some of these investigations were funded by the National Institute of Mental Health, while others were supported by foundations and other funding sources. Investigators who conducted these studies attempted to further an understanding of the effects of homelessness on children by involving larger study populations, including comparison groups of housed children from low-income families, administering a greater breadth and quality of assessment instruments, and applying more advanced statistical techniques with which to analyze the data (Bassuk, Weinreb, Dawson, Perloff, & Buckner, 1997; Buckner, Bassuk, Weinreb, & Brooks, 1999; Masten, Miliotis, Graham-Bermann, Ramirez, & Neemann, 1993; Rafferty, Shinn, & Weitzman, 2004; Rubin, Erickson, San Agustin, Cleary, Allen, & Cohen, 1996; Schteingart, Molnar, Klein, Lowe, & Hartmann, 1995; Shinn et al., 2008; Zeismer, Marcoux, & Marwell, 1994).

Almost all investigations of homeless children sample children in shelter in the midst of an episode of homelessness. As such, all homeless children in these studies fit the HUD definition of homelessness. Masten and colleagues (1993) described children as falling on a “continuum of risk” in which children who experience homelessness are worse off than other poor children, and both are worse off than middle class groups. That is, children who are homeless share all of the adversities of poverty and also experience additional risks associated with episodes of homelessness, which for most are temporary.

By and large, differences between children in homeless shelters and poor domiciled children are smaller in recent, second-generation, research than in earlier research (Buckner, 2008). Buckner suggests three principal reasons for this. First, as more and more families experience episodes of homelessness in tight housing markets, differences between poor families that do and do not become homeless are diminished. In a housing market in the early stages of dwindling supply, family difficulties (such as mental health or substance use problems in mothers), which can put children at risk, are prominent in explaining which families are most vulnerable to experiencing homelessness. Over time, as the housing supply worsens, families that become homeless will have characteristics increasingly similar to the broader group of low-income housed families. Thus, the family-level risks that children are exposed to will be more alike in each group. Second, as shelter conditions have improved in many jurisdictions and as legal changes and funding have reduced obstacles to stable schooling for children, the plight of sheltered homeless children may have become less severe. Third, homeless children and children from low-income families living in housed conditions are exposed to many of the same stressors, such as community and family violence. As such, the effects of homelessness per se, above and beyond the effects of poverty, can sometimes be difficult to discern.

Buckner (2005, 2008) conducted a comprehensive review of the published literature on homeless children from its inception in 1987 through 2005. The chief criteria for inclusion in this review were that the study be published in an academic journal and that it involve a comparison of homeless children to either normative data on children or a housed comparison group. This eliminated unpublished reports or a few studies that examined only homeless children. The dominant research question across these publications was ascertaining the impact of homelessness on children, an issue that requires comparative data. Some studies focused on a particular domain, such as mental health or academic achievement, while others looked across several domains. With studies that examined children of different age groups and/or across various domains, findings from the same overall investigation can sometimes be found dispersed across academic journals due to journals' space limitations and areas of focus (e.g., mental health, education, health). For instance, child-related findings from the Worcester Family Research Project, a comprehensive investigation of 436 homeless and low-income housed mothers and more than 600 of their children conducted in Worcester, Massachusetts, during the 1990s, can be found in various publications (Bassuk et al., 1997; Buckner et al., 1997, 1999, 2001, 2004; Garcia Coll et al., 1998; Weinreb et al., 1998, 2002). While these articles may seem to be separate studies, they report on data collected during the same time period, in the identical community context, and with the same overall study population.

This section of the paper begins with an overview of the demographic characteristics of homeless children and is followed by a review of studies, including those examined by Buckner (2005, 2008) and other studies published more recently. Table 2 (Parts A–D) distinguishes among studies focusing on the impact of homelessness in various domains: health status (Part A), developmental status (Part B), mental health and behavior (Part C), and education/academic achievement (Part D).  The table includes all published studies of the effects of homelessness on children, emphasizing studies that included housed comparison groups and that followed homeless families over time and used relatively comprehensive measurement. This section continues with a discussion of the small number of studies that follow children longitudinally and the finding that homeless children’s outcomes improve with the passage of time. Next, it describes the subgroup of children who become separated from their homeless families. Finally, it concludes with what is known about interventions for homeless families and children.

View full report


"index.pdf" (pdf, 419.29Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®