HUD credited its policies for a historic homelessness drop. Two cited actions came after the count.

The Department of Housing and Urban Development announced on November 6 that 698,425 people were counted as homeless on one night in January 2026, down 6.3% from the previous year. HUD called it the largest decrease on record and said President Donald Trump’s policies drove the result.

The headline decline is real in HUD’s aggregate: the January 2025 count was 745,652, so the new figure is 47,227 lower, a 6.33% decrease. But HUD released no 2026 national report, state tables or breakdowns for sheltered, unsheltered, chronic, veteran, youth or family homelessness with the announcement.

Independent work supports the scale of the decline. An analysis commissioned by the National Alliance to End Homelessness estimated about 693,000 people, down 6.7%. Community Solutions estimated about 699,700, down 6.2%, from 182 reporting communities that represented 74% of the 2025 national count. Across reporting communities it found declines in both sheltered and unsheltered homelessness; 103 of the 182 communities recorded an overall decrease.

HUD used the drop to credit border enforcement, a Supreme Court ruling that allowed local camping bans, a July 2025 executive order and its shift from Housing First toward Treatment First. Secretary Scott Turner said Trump’s actions drove the decrease and described the count as proof that the administration’s approach works.

Two measures HUD highlighted cannot explain the January 2026 count because they came later. The department announced its revised $4.04 billion homelessness funding competition on June 1, 2026. It released the Treatment First toolkit on August 12. Both occurred months after communities conducted the count.

Trump’s July 24, 2025 executive order did precede the count. It directed agencies to favor treatment, civil commitment and enforcement against public camping. The aggregate total alone does not show whether those policies, border changes, local housing programs, shelter capacity, migration or counting practices produced the decline.

HUD’s own submission guidance shows what the missing national record should contain. Communities report whether they counted sheltered people only or both sheltered and unsheltered people, along with household, demographic, chronic-homelessness, veteran, youth and methodology information. Those fields are necessary to identify who experienced improvement and whether the comparison is consistent.

Local results show why composition matters. Utah reported declines in unsheltered and chronic homelessness, while homelessness among people 65 and older increased and youth trends worsened in some areas. Salt Lake County reported a 13.6% increase in housing inventory since 2024 alongside fewer unsheltered and more sheltered chronically homeless people.

The decrease follows a very high baseline. HUD’s May report counted 745,652 people homeless in January 2025, including 266,320 people living unsheltered. Even its November release said homelessness remains near a historic high.

The independent estimates are important contrary evidence against dismissing the new total as mere spin. They make a major nationwide decline—and likely the largest since comparable national Point-in-Time reporting began in 2007—plausible. But the reporting communities were not a random sample, early-reporting bias could overstate the decrease, and both analyses preceded complete public federal tables. Neither establishes HUD’s account of why the change occurred.

The accountability question is not whether a decline should be welcomed. It is whether HUD will publish the evidence needed to distinguish fewer people living outside from changes in shelter use, geography or counting—and whether it will separate policies operating before January from programs launched after the measurement.

The next test is straightforward: release the full 2026 Point-in-Time and Housing Inventory Count tables, document methodology changes, identify the communities and populations driving the decline and show how each claimed policy could have affected the count before it occurred. Until then, HUD has announced a consequential improvement without providing the record needed to audit its explanation.