HUD’s 2025 Annual Homelessness Assessment Report recorded 745,652 people experiencing homelessness on a single night in January, a 3.3 percent decline and the first national drop since 2016. The figure comes from the annual point-in-time (PIT) count, a snapshot that policymakers use to gauge progress.
National count shows decline but tells a mixed story
The PIT count tallies individuals inside shelters and, before dawn, volunteers search streets, cars and encampments. Because it captures only one night, the method can miss people who cycle in and out of shelter or who are hidden from view.
Family homelessness fell by 11 percent, while the number of people without children rose. Both individual homelessness and chronic homelessness, defined as lasting at least a year or recurring frequently, reached record levels.
The report listed 515,286 individuals and 155,750 chronically homeless persons. About seven-in-ten of the individuals were men in 2024, a demographic that family-focused programs rarely reach.
The National Alliance to End Homelessness notes that new entrants to homelessness are outpacing the capacity of assistance programs.
HUD Secretary Scott Turner interpreted the data as evidence that the Housing First model has failed.
State-level shifts drive most of the drop
Only 22 states and the District of Columbia recorded any decline, and two states, New York and Illinois, accounted for roughly 92 percent of the national reduction of 25,828 people.
While the headline suggests progress, the underlying churn remains high. In a study of 1,462 homeless families in Salt Lake County, nearly half returned to shelters within a year, often after short-term Rapid Rehousing vouchers expired.
In Utah, the PIT count rose 18.5 percent in the same year, with 95 percent of the increase coming from sheltered counts after winter shelter beds were expanded. More beds make more people visible, inflating the apparent rise.
Hidden populations and counting gaps
Local data show that roughly ninety-seven percent of nearly 190,000 recent arrivals exited shelters without relocation assistance.
Although the Philadelphia count is not broken out in the national summary, it serves as a micro-example of the same methodological constraints, single-night timing, reliance on volunteer outreach, and the possibility that people who move in and out of shelters between counts go unrecorded. Local officials use that count to calibrate shelter funding and to identify gaps in outreach, reinforcing how each jurisdiction’s effort feeds into the broader picture, and inform policy adjustments at the state level.
Policy makers rely on the nightly tally to determine the size of federal and local homeless assistance allocations.
