Abstract
This article addresses school and community-level issues associated with the expanding crisis of student homelessness in the United States. We note that while an increased geographic dispersion of homeless and highly mobile (HHM) families is largely attributed to the widespread effects of the economic recession, it is also furthered by shifting federal policy on the engagement of homelessness. Highlighting the greater Madison, Wisconsin, area, we use geospatial analysis to document the area’s expanding “geographic centers” of HHM families. We then consider how, in these new geographic centers, schools may lack readiness to serve HHM students and HHM families may not have access to vital community-based resources. We conclude by suggesting that research and practice should more purposefully consider school effects and community effects in contexts of homelessness.
A range of post–recession era data indicates the historic economic and social difficulties facing the United States. In 2010, nearly 50 million residents were without health insurance and 46.2 million were living in poverty—the largest numbers since the Census began tracking poverty in 1959 (United States Census Bureau, 2010). Half of all workers in the United States have been affected by job losses, reduced hours, or pay cuts (Pew Research Center, 2010) and for the 18 million very-low-income renters in the United States, only 11.6 million affordable and adequate units were available—a housing supply gap of 6.4 million (Joint Center for Housing Studies, 2011). These indicators of widespread hardship were associated with a 20% increase in family and youth homelessness and, in schools, an 18% increase in student homelessness. Thousands of additional students and families go unidentified during periods of residential instability. Given that student homelessness may be associated with a host of problems—including lowered grades (Obradovic et al., 2009), achievement test scores (Masten, Sesma, & Si-Asar, 1997), attendance rates (Kennedy, 2007), and parental engagement (Haig Friedman, 2000) and increased school mobility (National Center on Family Homelessness, 2009) and social isolation (Anooshian, 2003)—attention to the issue is warranted in both research and practice.
Particularly noteworthy amid this emergent crisis is the expanding geographic dispersion of homelessness. Although states with some of the largest urban centers identified high numbers of homeless students during the 2010–2011 school year (220,738 in California, 90,506 in New York, 85,155 in Texas, and 55,953 in Florida), the numbers have also swelled in states such as Idaho (76% increase during 3 years), Arkansas (52%), Kansas (34%), and South Carolina (21%). In fact, the National Center for Homeless Education (NCHE, 2012)—which compiles federal data on student homelessness—noted that 44 states reported increases in homeless student identification between the 2008–2009 and 2010–2011 school years (refer to Table 1). Local-level data further illustrate the widespread, expanding nature of the crisis during this period. For example, 70% of U.S. school districts have identified significantly more cases of student homelessness—including those in midsized communities like Vista, California; Adrian, Michigan; and Boulder, Colorado—and, according to the U.S. Interagency Council on Homelessness (2011), there was a 57% increase in the proportion of people using homeless programs in suburban and rural areas. Perhaps more than ever, then, student homelessness is not just a “big city” issue.
Growth in Homeless Student Identification by State Between 2008–2009 and 2010–2011 (NCHE, 2012)
Note. NCHE = National Center for Homeless Education; SY = school year.
Although the spreading of homelessness across the United States is fundamentally tied to widespread recession effects, federal policy also plays a key role in shaping the residential patterns of homeless and highly mobile (HHM) families. Both the Homelessness Prevention and Rapid Re-Housing Program (HPRP)—which was implemented as part of the American Recovery and Reinvestment Act of 2009—and the similarly focused but longer-term Homeless Emergency Assistance and Rapid Transition to Housing (HEARTH) Act of 2011 are designed to help families who are homeless or on the verge of homelessness maintain or transition into independent housing. The programs clearly signal a pivot away from well-established “continuum of care” models of engaging homelessness that focus mostly on emergency shelter and transitional housing interventions. Although the independent living opportunities toward which HPRP and HEARTH aim are posited to provide families with stability and, in turn, more fertile life improvement opportunities, the policies’ prioritization of rental assistance and housing relocation/stabilization services across scattered sites furthers the community-level decentralization of the HHM population. That is, rather than multiple families being clustered in common shelters and transitional housing settings, HPRP and HEARTH families will be more geographically dispersed.
We suggest that the crisis of student homelessness must be examined at its intersection with this shift in federal homeless engagement policy. Focusing on the greater Madison, Wisconsin, area, we use geospatial, school, and community data to (a) compare the dispersion of sheltered and HPRP families across local school districts; (b) discuss these districts’ readiness to serve HPRP students; and (c) examine HPRP families’ access to out-of-school resources. We then describe some key implications for future research and practice.
Community-Level Dispersion
The greater Madison metropolitan area is a midsize community (population 488,000) in the Midwestern United States. Although Madison’s rates of unemployment (mostly between 3% and 6% since 2007) and poverty (3%–4% lower than national averages) indicate that it has fared the recession better than many other towns, it is like most others in that it has experienced an upsurge in homelessness in recent years. In 2010, area shelters served 16% more households and 19% more school-age children than they did in 2007, and the Madison Metropolitan School District (MMSD) identified almost 50% more homeless students during the 2009–2010 school year than it did only 3 years earlier. We include eight bordering communities—DeForest, McFarland, Middleton, Monona, Oregon, Sun Prairie, Waunakee, and Verona (each of which has its own school district)—in the discussion, for although they have traditionally been less affected by issues of poverty and homelessness than the city proper, they are widely recognized as part of the larger Madison metropolitan area. Our intent is not to suggest that the particularities of Madison’s homeless crisis are generalizable to all other communities, but to illustrate some family, school, and community-level factors that bear considering as homelessness increasingly reaches beyond cities. In this way, the Madison case has heuristic value.
Notwithstanding the fluid and often under-the-radar nature of student and family homelessness in Madison—which inhibits our attempts to comprehensively gauge the population—geospatial analysis can give us a general sense for which parts of town are most affected by residential instability. Including only families with school-age children, we used the directional distribution tool in ArcGIS software to compare the geographic dispersion of sheltered and HPRP households in the Madison area. Specifically, we used deidentified addresses from the local Homeless Management Information System (HMIS) to develop ellipses that represent the “geographic centers” of three groups: (a) families who used shelter services in 2007 (pre-recession); (b) families who used shelter services in 2010 (post-recession); and (c) families who used HPRP services between 2009 and 2011 (since the program’s inception). 1 Each ellipse represents the mean center of all the address points as well as both the geographic distribution and orientation of each population group. The 2010 shelter ellipse expanded, showing increased dispersion since 2007. The ellipse representing the standard distance between households that experienced sheltered homelessness in 2007 covered 8.5 square miles encompassing 540 sheltered school-age children within the Madison city limits. The 2010 ellipse grew to 18.9 square miles and 757 sheltered school-age children—though all still within city and MMSD boundaries (refer to Figure 1). These numbers indicate that families from an increasingly diffuse geography resorted to using shelters.
The geographic centers of homeless and highly mobile families in relation to Madison-area school districts
The region’s HPRP population is even more spread out. The HPRP standard deviational ellipse encompassed 156.6 square miles, 1,497 school-age children, and 89 schools—all significant increases from the 2010 sheltered-family ellipse. Generally put, when we move from a shelter-centric vision of homelessness to a more comprehensive one of HHM families, 740 more school-age HHM children, 80 more schools, and eight more school districts (as opposed to just MMSD) can be seen to be “in the center” of the region’s homeless crisis. This expansion is relevant given what appears to be these districts’ varying capacities to serve HHM students and families.
School Readiness to Serve HHM Students
Although the Madison data indicate that homelessness and residential instability increasingly reach beyond city limits, it is evident that many MMSD schools—the “city schools”—may be better positioned to respond to HHM student needs than those in adjacent districts. This readiness is partially a matter of infrastructure. MMSD sponsors a districtwide “Transition Education Program” that strategically works toward the provision of equitable opportunities for HHM students. The program, which shares some features with those in the Indianapolis Public Schools, the Minneapolis Public Schools, and a number of other urban districts throughout the United States, employs four full-time teachers and social workers who direct HHM student transportation, connect families with school and community services, assist with academic screening, and facilitate ongoing professional development throughout the district. Every MMSD school also has a point of contact who works in tandem with the Transition Education Program on site-specific issues that emerge relative to HHM students. MMSD’s program is undoubtedly among the most focused and advanced models in the state and was even recognized in 2004 by the National Association for the Education of Homeless Children and Youth as the “Outstanding School-Based Educational Program Providing Services to Students in Homeless Situations.”
MMSD’s program is informed—and partially funded—by the federal McKinney-Vento Homeless Assistance Act (reauthorized in 2002 as Title X of the No Child Left Behind Act of 2001), which delineates students’ rights and directs school action amid conditions of homelessness. Through McKinney-Vento’s Education of Homeless Children and Youth program, states and, in turn, districts are awarded funds to serve HHM students. State data indicate that “McKinney-Vento districts”—those that receive this funding and explicitly attempt to carry out its stipulations regarding student enrollment, transportation, and support—are better equipped to identify HHM students and respond to their needs. For instance, the 3% of school districts in Wisconsin that received McKinney-Vento subgrants during the 2009–2010 school year identified more than 70% of the total number of students identified statewide. This trend parallels a national one, where the 19% of districts receiving McKinney-Vento funds identify 80% of the overall U.S. total. 2 Although many McKinney-Vento districts are not as progressive as MMSD in their service models, most of them provide HHM students with transportation, school supplies, clothing, and instructional support. They also collaborate with community-based organizations to ensure that students’ and families’ outside-of-school needs are addressed (NCHE, 2011; Shields & Warke, 2010).
MMSD is one of only 12 districts in Wisconsin that received McKinney-Vento funding in 2009–2010 and, among the eight bordering communities that we identified as also being in the area’s HPRP “center of homelessness,” only the Middleton-Cross Plains District was funded. The other seven districts nominally designate district-level homeless liaisons (per federal law, which applies to all districts, not just those that are funded), but these titles are merely add-ons to other full-time roles. For example, in Monona Grove, Verona, and McFarland, the directors of pupil services concurrently serve as district homeless liaisons. In DeForest, the role is held by the director of business and auxiliary affairs. Regardless of how well intentioned these personnel are, research suggests that those who hold multiple roles simply do not have much time to devote to homeless-specific issues such as transportation, school mobility, and social isolation. The NCHE’s annual report (2011), for example, mentioned that 2,019 such district liaisons across the United States failed to submit yearly homeless counts to their local districts and Hernandez and Israel’s (2006) study found that in many districts, designated liaisons had little knowledge of federal policy and were not even aware that they held the liaison role. In comparison with MMSD and Wisconsin’s other McKinney-Vento districts like Racine, which has a “families in transition coordinator,” and Milwaukee, which has a “Homeless Education Program,” these communities lack school-based infrastructure for the direct engagement of homeless-related issues.
More broadly, poverty and homelessness-responsive practices have not traditionally been central to the “core work” (Spillane, 2006) of MMSD’s bordering districts. This is probably because before the 2009–2010 school year, such issues had an increasing, but still largely peripheral, presence in the surrounding communities. For instance, whereas about half of MMSD’s students qualified for free or reduced-price lunch in 2009–2010, only 18% did in its eight bordering districts. Furthermore, these districts identified, on average, about 28 homeless students per year between 2007 and 2010—less than the totals of 15 individual MMSD schools and only 2.8% of MMSD’s districtwide total in 2009–2010 (refer to Table 2). By no means do districts’ relative paucities of experience with poverty and homelessness connote their lack of capacity to address them. Some characteristics of suburban districts, such as lower school-level student mobility rates and larger district pools of financial resources can, in fact, be of great benefit to HHM students. Most of these districts, however, do not have systems of practice that are explicitly geared toward serving students in these situations. In the Madison area, for example, only MMSD and the Middleton-Cross Plains District—the two McKinney-Vento funded districts—develop, evaluate, and submit specific plans and data related to 10 federal and state standards for serving HHM students. These plans delineate specific homeless-related responsibilities for social workers, teachers, administrators, and front office staff—and fundamentally inform their everyday practice. Devoid of similarly focused systems of practice, educators’ readiness to serve HPRP students—some of whom are not classified as homeless per McKinney-Vento Policy, but all of whom are highly mobile and/or residentially unstable—likely lessens. 3
Poverty and Homelessness in the Madison-Area School Districts
Note. SY = school year; MMSD = Madison Metropolitan School District
Access to Education-Related Resources
Beyond school readiness issues, the expansive dispersion of HPRP families also raises concerns about their mobilization of broader out-of-school resources. These concerns are tied to both the presence of resources and families’ access to them. Small (2009) notes that neighborhoods with the densest concentrations of poverty—most of which are in inner cities—are typically more replete with social services than socioeconomically diverse neighborhoods. Accordingly, as poverty, homelessness, and residential instability spread to suburban and rural communities like those that border Madison, families will likely be further removed from afterschool programs, community centers, and the like. This removal is relevant because, amid wider empirical uncertainties about how and to what extent homelessness affects specific academic outcomes, we do know that HHM students and families are often socially and institutionally disconnected and, as a result, especially dependent on social services (Miller, 2011a, 2011b). HPRP families may be especially at risk in this regard, for, notwithstanding the perceived benefits of independent living arrangements, these settings are both physically removed from useful organizations and they lack the education-oriented resources (tutors, school supplies, computer access, etc.) that are embedded in many shelters.
In the Madison area, more than 77% of the libraries, community centers, Head Start programs, and Boys & Girls Clubs—the region’s most prominent nonschool organizations that provide education-related services—are located within MMSD boundaries. Although each of the bordering communities has a public library, none of them have Boys & Girls Clubs or full-service community centers and only Sun Prairie has a Head Start program. Only 86 of the 381 Madison-area HPRP families (22.6%) live within a half mile of more than one community organization and only 25 families (6.6%) live within a quarter mile of multiple resources (refer to Figure 2).
Homeless Prevention and Rapid Re-housing Program families’ proximities to out-of-school education resources
HPRP families’ distances from key resources are all the more problematic when we consider their limited transportation options on the periphery of town. Neither walking nor bussing—the primary modes of getting around town for most HMM families (Miller, 2011b)—are easy in communities like Monona, McFarland, and the other six bordering communities we examined. For instance, a Walk Score® analysis—which considers matters of both resource proximity and walking safety—reveals 1% (4 of 384) of HPRP families’ neighborhoods to be “walkers’ paradises,” 58% of them to be “somewhat walkable,” and 40% to be “car dependent.” These are considerably less walker-friendly settings than those marked by the 2007 and 2010 shelter ellipses (refer to Table 3). As such, the communities’ sparse public transportation infrastructures, where 55% of HPRP households are at least a quarter mile from the nearest bus stop and 27% are at least a half mile away (refer to Figure 3), are concerning. Simply put, in comparison with those who live in the city, HPRP families who live in suburban and rural areas tend to be farther away from vital out-of-school resources and to have few means for traversing these distances.
Walkability in the Geograhic Centers of Homelessness
Note. HPRP = Homelessness Prevention and Rapid Re-Housing Program.
Almost all errands require a car. bSome amenities within walking distance. cMost errands can be accomplished on foot. dDaily errands do not require a car.
Homeless Prevention and Rapid Re-housing Program (HPRP) families’ proximities to bus stops relative to the HPRP geographic center
In sum, it is evident that (a) the Madison area’s crisis of homelessness is marked by not only the highest numbers of HHM families and students in recent memory but also by significant recession and policy-affected geographic dispersion; (b) many schools new to the “center” of the crisis have little experience and/or infrastructure for addressing matters of poverty and homelessness; and (c) HPRP students and families may lack access to important community-based resources. Taking a wider view, trends and dilemmas such as these facing HPRP families are likely to continue with similar momentum in Madison and other places that have witnessed similar “HHM dispersion,” such as the wider Detroit, Pittsburgh, and Indianapolis communities. Middle and working classes face glum job markets and inadequate affordable housing options, leaving many on the brink of homelessness. Although some of those who ultimately lose housing will use shelter-based services in inner-city communities, the looming federal implementation of the HEARTH Act—which, like HPRP, decentralizes the engagement of homelessness—and ever-increasing populations of unaccompanied youth and doubled-up families suggest that many others will seek more institutionally detached options in suburban and rural areas. All told, matters of student homelessness and residential instability in increasingly diverse areas warrant strategic attention from the educational research field.
Future Directions for Research and Practice
Moving forward, our understandings of HHM students should be developed in intersection with wider school, community, and policy contexts, for although the impacts of homelessness on school matters are clearly mediated by certain student and family-level variables (students’ ages/developmental stages, families’ reasons for being homeless, the durations for which they experience homelessness, etc.), they are also affected by specific school and neighborhood ones. As the effects of the economic recession and changes in federal policy further expand the dispersion of homelessness, schools and communities might, in fact, surface as particularly important units of analysis. That is, in addition to focusing upon how homelessness affects school outcomes (i.e., student achievement scores, attendance, graduation), recent trends indicate that we need to know more about how schools and neighborhoods affect students’ and families’ broader experiences of homelessness.
Analyses of schools should move beyond traditional district-level programs addressing homelessness and, instead, focus upon regional and/or inter-organizational efforts. The Wayne (MI) Regional Education Service Agency (RESA), for example, centralizes McKinney-Vento-related action in the areas surrounding Detroit—including identification, enrollment, transportation, and professional development—in an attempt to mitigate district-level resource shortages and operational discrepancies. The RESA model and comparable efforts in other states (e.g., those operating through California’s County Offices of Education, Wisconsin’s “Cooperative Educational Service Agencies” and Pennsylvania’s “Intermediate Units”) appear to be promising ways of coordinating HHM student services across diverse districts, 4 but further research is needed to parse out the particular tools and routines that animate such initiatives and to determine whether they lead to better student outcomes amid shifting public policy contexts.
In addition, there is a dearth of research examining how and to what extents schools’ brokerage practices—those that connect families with relevant people, institutions, and/or resources (Small, 2009)—affect otherwise institutionally detached doubled-up and HPRP families in suburban and rural settings. Interorganizational information-sharing agreements, data management systems, and service integration appear to be both feasible and promising elements of systemic brokerage practice in several urban contexts (refer, e.g., to the Pittsburgh-area Homeless Education Network and the Philadelphia-area Kids Integrated Data System), but less is known about how schools can develop similar capacities in suburban and rural communities—especially those with expanding populations of institutionally detached HHM families.
Finally, a range of neighborhood effects questions await. Can resource-rich neighborhoods mitigate academic losses and social isolation during periods of homelessness? Are there within- district differences in HPRP and doubled-up students’ academic, social, and/or professional outcomes that are associated with particular residential and spatial variables? How do transportation infrastructures—including bus/subway options and neighborhood walkability—shape HHM students’ educational resource mobilization? By addressing such questions, we can not only cultivate more nuanced understandings of HHM students’ and families’ experiences but, as the geographies of homelessness expand, move toward the development of appropriately responsive schools and communities.