New Federal Data on Student Homelessness: Six Major Trends from the 2024–2025 School Year
For the second year in a row, public schools identified more than 1.5 million students experiencing homelessness.
- See the six major trends from new federal data.
- See interactive table of student homelessness by state, 2024-2025
Recently released federal data show that public schools identified 1,553,989 children and youth experiencing homelessness in the 2024-2025 school year, the highest number since national reporting began in 2004-2005. It slightly exceeded the previous record, set in 2023-2024, even as overall public school enrollment continued to decline. Even so, these numbers likely understate the scale of child and youth homelessness: research shows that many K-12 students experiencing homelessness are not identified by their schools, and infants, toddlers, and other young children are largely missing from these data.1
The data show six major trends:
-
- Public schools identified more than 1.5 million students experiencing homelessness for the second year in a row, the highest number on record and 12% above pre-pandemic levels.
- Despite a steady national total, most states identified fewer students in 2024-2025, possibly reflecting the loss of pandemic-era education funding.
- As overall public school enrollment declined, students identified as homeless became a larger share of all students: about 1 in 32, the highest share on record. The share rose in 45 of 53 states and jurisdictions.
- Chronic absenteeism declined and graduation rates rose, but students experiencing homelessness still lag behind other low-income students.
- The number of students in shelters or unsheltered situations rose for the fourth year in a row, even though the January 2025 Point-in-Time count by the U.S. Department of Housing and Urban Development (HUD) found fewer homeless families. Most students experiencing homelessness were staying temporarily with other people because they had nowhere else to go, a situation HUD’s count leaves out.
- A record number of young children, from birth to age 5, were identified and served through Education for Homeless Children and Youth subgrants.
The 2024-2025 school year was also the year that American Rescue Plan Homeless Children and Youth (ARP-HCY) funds expired. During the three years those funds were available, schools in nearly every state identified more students experiencing homelessness. In 2024-2025, the national total stayed approximately the same as the previous year, but most states identified fewer students. Because children and youth must be identified before they can receive help, that decline is a serious concern. In 2025-2026, the first full year without ARP-HCY funds, many schools no longer had the staff and resources that helped them find these students.
Meanwhile, the pressures that push families and youth into homelessness have not let up: housing cost burdens for renters reached a record high in 2024, child poverty has climbed above its pre-pandemic level, and hurricanes and wildfires have displaced thousands of families. The cost of housing and other basics remains high, and many families now face new limits on SNAP and Medicaid.
Taken together, these findings show both the scale of the challenge and the importance of sustained investment and targeted support to help students experiencing homelessness enroll, attend, and succeed in school. Dedicated support matters not only for immediate educational stability, but also for longer-term homelessness prevention: lack of a high school degree or GED is the single greatest risk factor for homelessness as a young adult. This makes targeted resources, policies, and effective practices critical for preventing future homelessness and for closing academic achievement gaps. ARP-HCY funds ran out in September 2024; in many school districts, the staff and services they paid for are gone. Today only about one in five school districts receives Education for Homeless Children and Youth (EHCY) funding.
We urge Congress to increase EHCY funding so schools can rebuild the capacity to identify and serve students experiencing homelessness. We also urge state lawmakers to increase their investments in these students.
Six Major Trends from the 2024–2025 School Year
1. Public schools identified more than 1.5 million children and youth experiencing homelessness in 2024-2025, the highest number on record and 12% above pre-pandemic levels.
Children and youth must first be identified as experiencing homelessness in order to receive the protections and services they need to enroll, attend, and succeed in school – and to decrease their risk of experiencing homelessness as adults. When identification rises, whether because schools have more capacity to find students or because more families have lost their housing, more children and youth are connected to essential support.

The number of identified homeless students was 12.0% higher than in the 2018-2019 school year, before the pandemic – and despite the fact that public schools now enroll about 1.3 million fewer students (see Trend 3). It has increased 28.9% since 2021-2022 (the first full school year of ARP-HCY funding) and 0.4% over the previous school year.
The increases were widespread. Forty-nine of 53 states and jurisdictions identified more homeless students in 2024-2025 than in 2021-2022, and 43 identified more than in 2018-2019. Some of the largest increases since 2018-2019 were in West Virginia (43%), Pennsylvania (43%), South Carolina (37%), Arkansas (30%), Montana (29%), Wyoming (29%), and Tennessee (26%).
The increases since 2021-2022 coincided with three notable conditions: significantly increased federal funding for homeless education, growing economic pressures on families, and major natural disasters.
Federal funding
The American Rescue Plan Homeless Children and Youth program, created by a bipartisan and unanimous amendment, provided $800 million for identifying and supporting homeless students over a three-year time period (for most states, between June 2021 and September 2024). This one-time investment more than doubled the number of school districts receiving dedicated resources to identify students experiencing homelessness, respond to urgent needs, and strengthen school stability.
Housing costs and child poverty
Rents rose faster than inflation in both 2023 and 2024, and by 2024 a record 22.7 million renter households – nearly half of all renters – were spending more than 30% of their income on housing. The child poverty rate under the Supplemental Poverty Measure was 13.4% in 2025, above its pre-pandemic level of 12.4%, after falling to a record low of 5.1% in 2021 when temporary pandemic relief was in place.
Natural disasters
Wildfires, floods, and hurricanes likely contributed to localized increases in student homelessness in certain states, such as in Hawaii, Florida, Texas, and Vermont in 2023-2024. In 2024-2025, Hurricane Helene led to more than 2,500 North Carolina students being identified as homeless, and the January 2025 Los Angeles wildfires displaced more than 100,000 people, including 1,100 Pasadena Unified students whose families lost their homes. In Hawaii, the number of identified homeless students fell in 2024-2025 but remained nearly 50% above its level before the 2023 Maui wildfires.
2. Despite a steady national total, most states identified fewer students in 2024-2025, the school year ARP-HCY funds expired.
Fewer students identified does not necessarily mean less homelessness. A decrease also can reflect a lack of resources, outreach, training, and proactive identification.
Thirty-four of 53 states and jurisdictions identified fewer students experiencing homelessness in 2024-2025 than in 2023-2024. Most of those decreases were under 5%, but 15 states and jurisdictions identified at least 5% fewer students. Increases in California and New York, which together account for nearly a third of all identified homeless students, kept the national total steady. Without those two states, the national total would have fallen by 1.3%.
Declining overall K-12 enrollment does not explain these decreases. In the 34 states and jurisdictions that identified fewer students experiencing homelessness, the number fell by a median of 4.5%, while total enrollment in those states fell by a median of less than half a percent. In 29 of the 34, students experiencing homelessness also made up a smaller share of enrollment than the year before.
For most states, ARP-HCY funds were available only through September 2024. Students are counted if they are identified at any point during the school year, and many are identified when they enroll at the start of the year, so ARP-funded staff likely still contributed to 2024-2025 numbers in many districts. The 2025-2026 school year was the first full year without these funds. Federal data for that year will not be available until 2027.
3. As public school enrollment declined, students identified as homeless became a larger share of all students: about 1 in 32, the highest share on record.
Public school enrollment has declined since the pandemic, while the number of students identified as experiencing homelessness has grown. As a result, schools are serving more students experiencing homelessness, often with less funding to support them.
In 2024-2025, students experiencing homelessness made up about 3.1% of public school students, or 1 in 32, up from 2.7%, or 1 in 37, in 2018-2019. Over that period, public school enrollment fell by about 1.3 million students, while the number of students identified as experiencing homelessness rose 12%. That share is the highest since national reporting began in 2004-2005, matched only by 2023-2024.
The share rose in 45 of 53 states and jurisdictions. Some of the largest increases were in states where overall public school enrollment fell while homeless identification rose, including West Virginia (from 3.9% to 6.2%), Maine (1.4% to 2.7%), Vermont (1.2% to 2.1%), Hawaii (2.0% to 3.1%), Montana (2.8% to 3.7%), and Arkansas (2.8% to 3.7%). In West Virginia, about 1 in 16 public school students were identified as experiencing homelessness in 2024-2025.

The drop in enrollment raises two concerns for students experiencing homelessness: tighter school budgets, and children who may not appear in school records at all.
School budgets
Many state funding formulas are tied to enrollment, so districts losing students face tighter budgets, even as a larger share of the students they serve are experiencing homelessness. For many districts, this is happening as federal pandemic funds, including ARP-HCY, have ended.
Children missing from school records
Some of the decline in public school enrollment may reflect children who are not in school at all. According to a 2025 Brookings analysis, the share of school-age children who are either in private school or do not appear in any public school record rose from 9.7% before the pandemic to 12.6% in 2023-2024, while private school enrollment held steady. That growing group includes homeschooled children, but also children who are not enrolled anywhere. Children and youth experiencing homelessness, who are chronically absent at higher rates and less likely to graduate than other low-income students, are especially at risk of falling out of school records. Declining enrollment may therefore hide additional children and youth experiencing homelessness who do not appear in these data at all.
4. As schools identified more students, chronic absenteeism declined and graduation rates rose, although students experiencing homelessness still lag behind other low-income students.
Improvements in attendance and high school graduation came while public schools were serving more students experiencing homelessness than ever before, and during the years when ARP-HCY funds gave more than twice as many school districts dedicated resources to support those students. A U.S. Department of Education study found that districts used ARP-HCY funds for transportation, tutoring, summer programs, temporary motel stays, and more staff support, with a focus on finding more students and reducing chronic absence. A companion analysis of the nearly 12,000 districts that received ARP-HCY funds found that by 2022-2023, the graduation rate for students experiencing homelessness in those districts (72.4%) was nearly three percentage points above the national rate for these students, and their reading and math proficiency rates were also higher, though many districts did not report proficiency data.
Chronic absenteeism
In the 2024-2025 school year, 47.3% of students experiencing homelessness were chronically absent, down from 49.6% in 2023-2024.2

The reduction in chronic absence, while modest, occurred during a period when public schools were identifying many more students experiencing homelessness and had more resources to support their attendance. Some of those newly identified students may already have missed substantial school before schools knew they were homeless. That means the recent improvement in chronic absenteeism may actually understate progress.
National data on chronic absence rates for economically disadvantaged3 students are not yet available for 2024-2025. However, an Attendance Works and SchoolHouse Connection analysis of 2024-2025 state data found that, on average, students experiencing homelessness were chronically absent at 1.5 times the rate of economically disadvantaged students.4 Because some states may count students experiencing homelessness within the economically disadvantaged subgroup, this comparison likely understates both the size of the gap and the distinct effect of homelessness on chronic absenteeism. The same state data show that chronic absence among students experiencing homelessness fell in 24 states between 2023-2024 and 2024-2025.
High school graduation
The 2024-2025 national average graduation rate for homeless students was 71.8%, an increase of 1.5 percentage points over the previous year and 3.8 percentage points since 2020-2021.
The gap between students experiencing homelessness and other students persists. In 2023-2024, the most recent year with comparable data, the homeless student graduation rate (70.3%) was 12.2 percentage points below other low-income students and 16.7 percentage points below all students.

5. The number of students in shelters or unsheltered situations rose for the fourth year in a row, but most students experiencing homelessness were staying temporarily with other people when schools identified them.
HUD’s latest Point-in-Time count is being cited as evidence that homelessness is declining. But schools identified more students in shelters and unsheltered situations in 2024-2025, the fourth year in a row of increases. Most students experiencing homelessness were not in shelters at all, but rather staying temporarily with other people, where they are less visible and not eligible for HUD homeless assistance.
Shelters and unsheltered situations
Public schools identified 200,299 students in shelters or transitional housing in 2024-2025, up from 196,812 the year before and the most since 2008-2009. Schools identified 70,359 unsheltered students, up from 67,536. Both numbers have risen every year since 2020-2021 and are higher than before the pandemic (up 19% and 27% from 2018-2019). The number of students staying temporarily with other people or in motels is also higher than before the pandemic (up 9% and 33%).

Shelters and unsheltered locations are the living situations that most closely match HUD’s definition of homelessness. Yet HUD’s January 2025 Point-in-Time count, taken during the same school year, found 11% fewer people in homeless families than the year before. A smaller count is not the same as fewer homeless families. In HUD’s own report, local communities attributed some of the decline, including in Massachusetts, to fewer shelter beds and stricter eligibility rules. When shelter is harder to get, families end up staying with others or in motels, where HUD’s count doesn’t reach them. HUD has not yet released 2026 numbers for families with children. Read our analysis of HUD’s 2026 count.
Staying with others temporarily and in motels
Most students experiencing homelessness were staying temporarily with other people, often referred to as “doubled-up” (74.2%, down from 76.3% in 2018-2019), while others were in shelters or transitional housing (12.9%), hotels or motels (8.4%), or unsheltered situations (4.5%), all larger shares than in 2018-2019. Homeless families and youth who stay temporarily with others, or in self-paid motels, due to lack of adequate alternatives are not eligible for homeless assistance from HUD and are not included in its Point-in-Time count. But these categories are not fixed. Schools record where students were staying when they were first identified, and families and youth often move among these situations during the school year.

The increase in motel stays is especially notable, rising from about 7.0% of identified homeless students in 2018-2019 to 8.4% in 2024-2025, an increase of 32,426 students. This may reflect both rising homelessness and the temporary availability of ARP-HCY funds, which for the first time could be used for emergency motel stays through September 2024. Many schools and communities regarded that flexibility as highly stabilizing for children and youth, especially in rural areas. In 2024-2025, the school year those funds expired, the number of students identified in motels declined slightly, from 132,010 to 130,066. In August 2026, at the direction of Congress, the U.S. Department of Education issued guidance clarifying that EHCY funds may be used for short-term emergency motel stays on a case-by-case basis if there are no other resources available and if tied to school attendance.
6. The number of young children ages 0 to 5 identified and served through Education for Homeless Children and Youth (EHCY) subgrants reached a record high in 2024-2025.
Homelessness in early childhood is associated with delays in language, literacy, and social-emotional development, and one recent study linked it to lower achievement in third grade. Early childhood programs can help mitigate these harms, which is why the McKinney-Vento Act requires school district liaisons to help young children experiencing homelessness and their families access Head Start, early intervention, and preschool programs.
Local educational agencies (LEAs) receiving EHCY subgrants reported identifying and serving 90,748 children from birth to age 5 who were not yet enrolled in kindergarten in the 2024-2025 school year. This was a 6.9% increase over the previous year and a 21.2% increase over the prior two years (since 2022-2023), the highest number reported to date.
The largest increases occurred during the period when ARP-HCY funds were available. LEAs receiving regular EHCY subgrants were especially well positioned to benefit from ARP-HCY, because ARP Homeless I could be used to supplement existing EHCY subgrants and ARP Homeless II was distributed broadly by formula to reach the vast majority of LEAs. That likely gave many EHCY subgrantee districts substantially greater capacity to identify and serve young children experiencing homelessness during this period.
These figures likely understate the number of young children experiencing homelessness who were served by school districts in some way, because this dataset includes only children served through EHCY subgrants, and only 20% of all LEAs receive McKinney-Vento EHCY subgrants due to the national funding amount for the EHCY program.
All LEAs, regardless of whether or not they receive an EHCY subgrant, are required to collect and report data on the number of children ages 3-5 identified as experiencing homelessness who are enrolled in an LEA-administered preschool program, but who are not yet in kindergarten. These LEA numbers are not yet available.
About the Data
This analysis draws primarily on the National Center for Homeless Education’s (NCHE) newly released national data summary for the 2024-2025 school year, as well as prior NCHE reports used for pre-pandemic and state-by-state comparisons. Because these reports aggregate data reported by schools, LEAs, and state educational agencies (SEAs), additional analyses will be possible once the full SEA and LEA data files are released.5 The national and state totals in this analysis include children ages 3 to 5 enrolled in public preschool programs, students in kindergarten through grade 12, and ungraded students. They do not include the count of young children served through EHCY subgrants described in Trend 6, which is reported separately. Shares of enrollment are based on NCES public school enrollment counts (Digest of Education Statistics 2025, Table 203.20), including Puerto Rico and the Bureau of Indian Education to match the coverage of the homeless student totals.
SEA and LEA files will also allow SchoolHouse Connection and U-M Poverty Solutions to update our interactive homeless student data profiles, which let users explore data at the local level, including within congressional and state legislative districts. The data profiles allow users to explore important data on student identification and under-identification, graduation rates, chronic absenteeism, and trends over time.
Student Homelessness by State, 2024-2025
Search for your state, or select a column heading to sort.
Note: Students identified includes all students public schools identified as experiencing homelessness at any point during the school year (preK, K–12, and ungraded). “1 in X students” shows the share of all public school students identified as experiencing homelessness; a smaller number means a larger share. Shares are SchoolHouse Connection calculations based on National Center for Education Statistics public school enrollment data. % chronically absent is the share of students experiencing homelessness who missed at least 10% of the school days in which they were enrolled. Graduation rates are four-year rates for 2023–24, the most recent year with state data, comparing students experiencing homelessness with low-income (economically disadvantaged) students. National totals include the District of Columbia, Puerto Rico, and the Bureau of Indian Education.
Source: U.S. Department of Education, EDFacts, as reported in National Center for Homeless Education, Education of Children and Youth Experiencing Homelessness in the United States: Data Summary for School Years 2022–23 to 2024–25, and earlier NCHE reports; SchoolHouse Connection analysis.
Related Resources
Footnotes
[1] Shephard, D. D., Hall, C. C., & Lamberton, C. (2021). Increasing identification of homeless students: An experimental evaluation of increased communication incorporating behavioral insights. Educational Researcher, 50(4), 239–248. https://doi.org/10.3102/0013189X20981067
Moore, H., Astor, R. A., & Capp, G. (2023). Districts’ and schools’ role in identifying and providing services for homeless students: Nested ecological case studies in school districts with high- and low-socioeconomic status. Journal of Community Psychology, 51(3), 1124–1148. https://doi.org/10.1002/jcop.22925
[2] Beginning with 2023-2024, NCHE changed how it calculates this rate, so these figures are not directly comparable to earlier years.
[3] In this brief, we use “low-income” and “economically disadvantaged” interchangeably for ease of reading. “Economically disadvantaged” is the term used in federal education data and refers to students who meet the definition established by their state.
[4] Attendance Works & SchoolHouse Connection. (2026). Connecting the dots: How states can integrate addressing student homelessness and chronic absence. https://e1.nmcdn.io/assets/schoolhouse/wp-content/uploads/2026/09/Connecting-the-Dots-SEA-Brief.pdf
Analysis of publicly available state-level chronic absence data for 2024-2025. States include: AZ, CA, CO, CT, DE, DC, GA, HI, ID, IL, IA, KS, KY, LA, ME, MD, MI, MO, MT, NJ, NM, ND, OH, OR, RI, SC, SD, UT, VT, VA, WA, WV.
[5] Please note that national figures are not subject to the same suppression standards as SEA and LEA files and may reflect different de-duplication methods. As a result, there may be small differences between national totals and SEA- or LEA-level data files.