In This “Research Insights” Edition

This edition of the Research Insights series –which brings together related studies from NNERPP members to discover connections across research and advance our collective understanding of these topics– focuses on students experiencing homelessness [1]. Three research-practice partnerships (RPPs) within our network have examined this topic recently in an effort to better understand this vulnerable population of students and support more equitable outcomes:

Research Alliance for New York City Schools, an RPP between New York University and the New York City Department of Education.

Houston Education Research Consortium, an RPP between Rice University and a number of surrounding school districts. The study examined here focuses on the Houston Independent School District (HISD) and was conducted by Houston Education Research Consortium external researchers at Southern Methodist University.

John W. Gardner Center for Youth and Their Communities, which has several partnerships with California school districts and communities. The study examined here focuses on the San Francisco Unified School District (SFUSD). 

As we have examined in previous editions of this series, not all studies on similar research topics and around similar research questions lend themselves to direct comparison, due to differences in sample constructions, outcome variables chosen for study, definitions of outcome variables, and so on. The three studies we examine here do ask similar questions around similar outcomes for similar groups of students, but for the reasons outlined previously, we do not attempt to directly compare findings across studies/districts. Rather, we highlight the contributions of each individual study and make broader observations about similarities and differences in the themes of the study findings. As you read with your own context in mind, we encourage you to think about how your own partnership or education agency might study student homelessness and which takeaways may be applicable to your context.

Why This Article

Amid the current pandemic, growing concerns about students who experience homelessness and an already increasing awareness of the challenges they face are amplified and especially top of mind for many education leaders. With this group of students facing greater obstacles in the best of times, a disruption of this magnitude raises many concerns about their trauma and learning loss in the wake of COVID-19. Even reaching these students now and once schools reopen in the fall is a considerable challenge. In this article, we hope to offer a starting point for addressing such concerns by examining what we already know through RPP-conducted research about students experiencing homelessness.

Table 1. List of RPPs + Artifacts Included in This Article 

Research Questions 

We first share the individual research questions that were addressed in each report:

New York:

1. Who experiences homelessness in New York City’s elementary aged population?
2. What are the different ways in which students experience homelessness?
3. How does homelessness disrupt students’ educational experiences?
4. How are schools supporting homeless students?

Houston:

1. Who are HISD’s homeless students? 
2. What are the educational outcomes of HISD’s homeless students, in terms of attendance, discipline, achievement, and attainment? 
3. How do the educational outcomes of HISD’s homeless students depend on their family context? 
4. How do the educational outcomes of HISD’s homeless students depend on their residential context?

San Francisco:

1. What is the size and distribution of homeless and highly mobile (HHM) students in SFUSD across different grades, schools, racial ethnicities, and language backgrounds? 
2. How heterogeneous is this student population in terms of chronicity of HHM status, instability of living arrangement, placement in foster care, and number of siblings? 
3. What promotes resilience and positive outcomes for HHM students? 

a) What are the implications of being HHM for students’ school attendance, graduation rates, and academic achievement as indexed by GPA and standardized tests? 

b) What student assets help explain variability in HHM students’ academic outcomes? 

c) Do HHM students fare better in some schools than in others? Are there any shared attributes among these schools?

Research Methods

Here we share a high-level overview of each report’s research methods for some more context around the studies and how they are similar and different from each other. We encourage readers to explore each individual research artifact for more details.

The New York study focuses on homelessness and homelessness supports among elementary school students, given the foundational role the early grades play for long-term academic success. It follows the cohort of students who began kindergarten in the fall of 2012 through the end of the 2016-17 school year, when they should have been completing 4th grade. In addition to the system-wide administrative data used for this part of the analysis, the study also included interviews and focus groups with 18 school staff members, including principals, social workers, guidance counselors, and teachers, in five elementary schools with unusually high proportions of students living in shelters, and where those students had outcomes that were similar to those of housed students across the City. Researchers also interviewed four district officials in charge of supporting students experiencing homelessness to gain additional insight into how schools are supporting these students and the challenges they face in doing so.

The Houston study examines homelessness in HISD in all grade levels (Kindergarten through 12th grade) using district data from the 2012-13 school year to the 2016-17 school year to compare students who experienced homelessness during that time to all non-homeless HISD students and to a matched sample of non-homeless students with otherwise similar characteristics in terms of grade level, school year, race/ethnicity, gender, economic disadvantage, at-risk status, Limited-English Proficiency, and Special Education Enrollment, and school mobility.

Finally, the San Francisco study examines homelessness in all grade levels (Kindergarten through 12th grade) in SFUSD for the 2013-14 through the 2015-16 school years using SFUSD administrative data as well as students’ self-reports on social and emotional learning survey items. Where appropriate, the study compared students experiencing homelessness to the whole SFUSD student population and/or to the population of students eligible for free lunch.

What Does the Research Show?

All three research artifacts provide descriptions of the students experiencing homelessness and all observe differences in the homelessness experience in terms of type and duration. All of the studies additionally examine similar outcome variables, including school mobility, attendance, suspensions, and academic achievement. The New York study takes a closer look at promising school supports as well as the challenges schools face in providing these, and the San Francisco study uses insights from social emotional learning questionnaires to examine what factors help students experiencing homelessness better overcome the obstacles they are facing. Here, we highlight just some of the major findings along these dimensions.

I. Describing Students Experiencing Homelessness  

A. Definitions and Measures of Students Experiencing Homelessness

To define students experiencing homelessness, all three studies use the federal McKinney-Vento Homeless Education Assistance Act, which considers as homeless all students who live doubled up with other family, friends, or acquaintances, who live in emergency or transitional shelters, those awaiting foster care placement, those living in hotels/motels, and those living in temporary housing solutions such as trailer parks, campgrounds, cars, or public places (categorized as “unsheltered” in the Houston and San Francisco reports). The San Francisco study uses the term “homeless or highly mobile (HHM)” to refer to this population of students. The New York and San Francisco studies point out that the data might not capture all students experiencing homelessness, since a student’s housing status is generally self-reported, and that it does not capture all aspects of homelessness, such as the exact length of homelessness within a given school year or multiple forms of homelessness during the same school year. 

In New York, almost 13% of the examined cohort of students (those who began kindergarten in the fall of 2012 and were then followed through the end of the 2016-17 school year) experienced homelessness at some point during that time. In Houston, approximately 2.5 to 3.5% of HISD students were homeless each year over the course of the study period (2012-13 to 2016-17) and 7.5% of HISD students were ever homeless over the entire study period. In San Francisco, approximately 4% of SFUSD students were reported homeless or highly mobile in the 2013-14, 2014-15, and 2015-16 school year.

B. Characteristics of Students Experiencing Homelessness 

In examining ethnicity/race of students experiencing homelessness in the respective samples the studies considered, all three studies found that Black students were overrepresented. In New York, Latino students were also overrepresented, and students experiencing homelessness disproportionately qualified for free or reduced price lunch every year between kindergarten and 4th grade, and were disproportionately more likely to be identified as English Language Learners or special education students. The study also found that students experiencing homelessness were heavily concentrated in particular neighborhoods and schools. In Houston, White and Asian/Pacific Islander students were underrepresented among the district’s homeless students and Hispanic students were slightly underrepresented. In San Francisco, Hispanic students were overrepresented and Asian/Pacific Islander and White students were underrepresented. This study also examined gender among students experiencing homelessness (distributed fairly evenly) and home language among students experiencing homelessness (English and Spanish were the most common home languages).

C. Duration of Homelessness

In the New York study, which focused on elementary-aged learners, the students in the cohort who experienced homelessness did so on average for just under three years. Almost 70% experienced homelessness for more than one year, and over 25% experienced homelessness for all five years that were examined (2012-13 through 2016- 2017). 

In the Houston study, which examined students across all grade levels from the 2012-13 through the 2016-17 school year, about 89% of students who experienced homelessness did so only for one year and only 0.3% of students who experienced homelessness did so for all five years of the study period. 

Among students experiencing homelessness in San Francisco examined in the Gardner Center study (reminder: these included students across all grade levels in the 2013-14 through the 2015-16 school years), more than half were reported homeless in all three years, about 40% were reported homeless during one of the three years, and less than 30% were reported homeless for two consecutive years.

D. Type of Homelessness

Living doubled up and living in a shelter emerged as the two most common forms of homelessness in New York (58% and 30% of students who experienced homelessness, respectively – these percentages refer to students who were only doubled up or only in shelter). Students who were doubled up were disproportionately Asian or Latino, while students in shelters were disproportionately Black. Additionally, results showed that 90% of homeless students in the cohort experienced the same form of homelessness from year to year. Given these findings, the study then divided all students who experienced homelessness into the following four groups to better understand the different experiences and their implications: 1) Doubled Up – Fewer Than Three Years (28% of homeless students), 2) Doubled Up – Three Years or More (30% of homeless students), 3) Shelter – Fewer Than Three Years (24% of homeless students), and 4) Shelter – Three Years or More (19% of homeless students). The qualitative data from the interviews with school staff showed that students in shelters were often described as the most vulnerable group.

Similarly, the two biggest groups of students experiencing homelessness in Houston were those living doubled up with family and friends (82% of homeless students in 2012-13; 81.9% in 2016-17) and those living in shelters (13.3% of homeless students in 2012-13; 10.2% in 2016-17). This study additionally examined the family context of students experiencing homelessness, as defined by students living unaccompanied (not living in the physical custody of a parent guardian) versus accompanied (living in the physical custody of a parent guardian). The share of students experiencing homelessness that were unaccompanied increased from 9.2% in 2012-13 to 11.5% in 2016-17, and unaccompanied homeless students were more than twice as likely to live in shelters (20.2% versus 8.9%) or to be unsheltered (8.0% versus 3.8%) as accompanied homeless students.

In San Francisco, 60% of students experiencing homelessness lived doubled up and 28% lived in shelters, where these two types of homelessness emerged again as the two most common types. This varied by race/ethnicity: For example, living doubled up was far more common among Hispanic homeless  students than other groups. Additionally, Hispanic and Asian students were more likely than White and Black students to experience multiple years of homelessness.

II. Outcomes of Students Experiencing Homelessness  

A. School Mobility

Not surprisingly, students experiencing homelessness were more likely than students not experiencing homelessness to move schools in all three studies (for example, 49.0% versus 18.7% between school years and 20.4% vs. 7.7% within school years in Houston and 12% of homeless students in SFUSD changing schools at least once during the school year versus 2% of free lunch students). In New York, homeless students in shelters for three or more years changed schools most often (with the average student changing schools at least once between kindergarten and 4th grade, and some changing schools seven times). In the interviews, school staff highlighted the challenges of teaching highly mobile students. 

B. Attendance and Attainment

In examining attendance, the three studies found that students experiencing homelessness did have lower attendance and higher rates of chronic absenteeism than their non-homeless peers. Homeless students in the cohort examined in the New York study had almost double the levels of chronic absenteeism (defined as missing 20 days or more in a given school year) between kindergarten and 4th grade of never homeless students (almost 59% for homeless students and just above 32% for never homeless students), with students living in shelters for three years or more having especially high levels of chronic absenteeism (over 80% were chronically absent). These attendance issues were driven at least in part by transportation challenges in getting to school and back, as was discussed by school staff in the study’s qualitative interviews. 

Similarly, homelessness was associated with lower levels of attendance across all family and residential contexts in the Houston study, compared to all students who were never homeless and to the matched group of non-homeless students who had similar characteristics otherwise. For example, homeless students attended 5.5 fewer days of school than all non-homeless students and 3.3 fewer days of school than matched non-homeless students. The Houston study additionally examined dropout and on-time graduation, finding that students experiencing homelessness were substantially more likely to drop out and substantially less likely to graduate on time than students not experiencing homelessness, even when compared to the matched set of non-homeless students. 

The San Francisco study found that 25% of students experiencing homelessness were chronically absent, compared to 11% of free lunch students. Homeless students attended school 92% of the time, compared to free lunch students’ average attendance rate of 95%. Additionally, homeless 12th graders were less likely to complete the A-G course portfolio required for graduation in California (32% of homeless 12th graders versus 52% of free lunch 12th graders completed the A-G course portfolio in 2015-16), and less likely to graduate than their free lunch, reduced lunch, and non-homeless and non free or reduced price lunch peers (71% of homeless 12th graders versus 88% of free lunch 12th graders graduated in 2015-16).

C. Discipline

Two of the studies examined this dimension: The Houston study examined disciplinary infractions (defined as every disciplinary action that results in removal of a student from any part of their program, including administrative interventions, suspensions, expulsions, or removals to Disciplinary Alternative Education Programs), and the San Francisco study looked at suspensions. The Houston study found that homeless students received more disciplinary infractions per year than their non-homeless peers in general (.23 more infractions), but when compared to their matched group of non-homeless students received slightly fewer infractions (0.1 fewer infractions). Homeless students’ disciplinary infractions varied significantly depending on their family and residential contexts. The San Francisco study found that students experiencing homelessness were 2.5 times more likely to be suspended out of school than free lunch students (4.4.% of homeless students were suspended in 2015/16 versus 1.7% of free lunch students).

D. Academic Achievement

In terms of academic achievement, the New York study examined students’ proficiency on state math and English tests in 4th grade, the Houston study examined rates of participation and pass rates on the  State of Texas Assessment of Academic Readiness (STAAR) exams in reading and math for students in 3rd through 8th grade, and the San Francisco study examined students’ GPAs and standardized test scores in English language arts, math, and science. Among the students in the New York study, only about a quarter of students who experienced homelessness attained proficiency on state math and English tests in 4th grade, compared to about 48% of non-homeless students in this cohort. Homeless students who lived doubled up were more likely than homeless students in shelters to achieve proficiency. Students who were in shelters for three years or more had the lowest scores (less than 20% attained proficiency on either test). Similarly, homelessness was associated with lower academic achievement in Houston: Students experiencing homelessness in the Houston study had substantially lower pass rates on the STAAR exams compared to non-homeless students; this varied by residential and family context. Notably, when compared to the matched group of non-homeless students, homeless students achieved slightly higher pass rates on the STAAR exams. Homelessness was also associated with lower rates of participation on the STAAR exams across all family and residential contexts. The San Francisco study similarly found that students experiencing homelessness scored lower on standardized tests in ELA, math, and science than their peers. In math, being homeless was associated with a 0.05 point reduction in standardized scores even when controlling for gender, race/ethnicity, grade level, and prior achievement. When examining students GPAs, SFUSD homeless students in middle and high school had GPAs that were on average half a point lower compared to the GPAs of free lunch students (2.4 vs. 2.9), and almost a full point below the GPA of their reduced price lunch peers (3.3).

III. School Supports and Factors Promoting Positive Outcomes 

A. School Supports

The qualitative interviews with school staff undertaken as part of the New York study highlighted several challenges schools face in supporting students experiencing homelessness, as well as promising practices. Generally, school staff reported that there were not enough staff to meet the needs of these students, especially given the fact that helpful interventions and supports required significant personalized and time-intensive attention and care. For example, school staff found themselves needing to advocate for homeless students and their families with other agencies (such as social services) or to coordinate with other, similarly overburdened services (such as shelter-based services). Additionally, identifying students experiencing homelessness was a challenge, as was addressing barriers to school attendance. Finally, schools had limited funds that often were not sufficient to create the kinds of supportive programming and services needed as well as to hire designated staff for homeless students and other groups of vulnerable students. Deciding how to allocate those limited funds was a significant challenge. Practices that school staff identified as contributing to the improvement of homeless students’ experiences were building trusting relationships with students and their families, which helped identify and support them, setting aside time to analyze trends in attendance data (including aspects like differentiating between morning and afternoon attendance), having additional non-instructional staff who could focus specifically on supporting homeless students, and partnering with community-based organizations. 

B. Factors Promoting Resilience and Positive Outcomes for Students Experiencing Homelessness

In examining the variability in homeless students’ outcomes and student assets that might explain why some homeless students fare better than others, the San Francisco study finds that many homeless students are able to thrive despite their overall lower outcomes when compared to non-homeless students. For example, three-quarters of SFUSD students who experienced homelessness had an attendance rate of 90% or above in 2015-16 and 56% had an attendance rate of 95% or above. 52% of students experiencing homelessness had a GPA of 2.5 or higher in 2015-16, 32% had a GPA of 3.0 or higher, and 15% had a GPA of 3.5 or higher. Additionally, one-third of students experiencing homelessness scored at or above the district mean on standardized state tests. The study tested multiple domains of students’ self-reported social and emotional learning (including self-management, social awareness, growth mindset, and self-efficacy) to better understand which factors promote such resilience and positive outcomes, finding that student self-management and growth mindset positively predicted ELA and math achievement, controlling for race/ethnicity and attendance.

Implications for Policy and Practice

As these findings demonstrate, student homelessness is a diverse and varied experience, with type and length of homelessness and students’ social and emotional learning skills being just some of the factors shaping students’ outcomes. Students experiencing homelessness do face much adversity, but many are also able to overcome significant challenges and still thrive in school. Likewise, although schools may face myriad challenges in trying to support homeless students, the New York study highlights school staff’s willingness to go above and beyond to help. Overall, the studies put forth the following implications for policy and practice for better understanding student homelessness and supporting students experiencing homelessness:

• Improvements in identifying homeless students and capturing more information in their administrative records are needed.
• More insight into effective school-level strategies for supporting students experiencing homelessness is needed, while being attentive to the unique and varied needs of different student groups by residential and family context. Specific areas for support that emerged were school mobility, attendance, attainment, test participation, and discipline.
• Non-instructional staff and partnerships with community-based organizations seem critical in supporting students experiencing homelessness.
• Targeting social and emotional skills seems promising for promoting better outcomes for students experiencing homelessness.

Current Efforts

As we conclude, we want to highlight ongoing current efforts that will be important additions to the work examined in this article: In addition to the study featured above, the Gardner Center is also currently in the midst of another, larger study on housing instability among San Mateo County students, with the first phase completed (we invite you to explore the associated research brief in greater detail here). This first phase examined the size, distribution, and heterogeneity of housing instability among youth enrolled in Sequoia area public schools and investigated the relationships between each type of housing instability and several academic outcomes. Key findings were that African American, Latinx, and English Learner students were overrepresented among students experiencing housing instability, that the type of housing instability experienced by students varied greatly across all districts, and that experiencing housing instability was associated with lower academic achievement and school attendance. Academic achievement varied greatly by type of housing instability students experienced. Next, the Gardner Center will conduct similar analyses for the remaining public school districts in the county and conclude the project with a comprehensive, county-wide report on housing instability among students across all participating districts that will also include a comprehensive connection to the prevailing literature in the field. 

Finally, we hope the findings and implications featured here provide a helpful overview of what we know about students experiencing homelessness and the policy and practice needs in this area during regular times – and serve as a starting point for determining homeless students’ needs and supports in these most unprecedented times, as well.

[1] “Students experiencing homelessness” is the preferred term and used whenever possible throughout this article. The term “homeless students” is used at times for the sake of brevity.

Nina Spitzley is Marketing Specialist at the National Network of Education Research-Practice Partnerships (NNERPP). 

Suggested citation: Spitzley, N. (2020). When Students Experience Homelessness: Learning From 3 RPPs’ Work. NNERPP Extra, 2(2), 2-9.