§ 03

Homelessness Fell in 2025 for the First Time Since 2016

The federal government counts homeless people once a year, on a single night in January. In January 2025, volunteers and outreach workers fanned out to shelters, streets, and encampments and counted 745,652 people without a place to live, according to HUD's 2025 Annual Homelessness Assessment Report. That is roughly 25,000 fewer than the record set in 2024, a decline of about 3%, and the first year-over-year decrease since 2016. It is still about 100,000 more people than in 2007, when HUD began tracking homelessness.

The gap across states is enormous. New York now leads the nation at 72.8 people per 10,000 residents, followed by Oregon at 63.7 and Vermont at 52.5. Hawaii, which topped the ranking a year earlier, dropped to fourth at 47.7. At the other end, Mississippi reports 4.2 per 10,000, a rate roughly 17 times lower than New York's. The average across the 50 states sits at 18.9 per 10,000, but that number is misleading: most states fall well below it. A handful of states at the top, led by New York and Oregon, pull the average far above what the typical state actually experiences.

The concentration is even starker in raw numbers. Just five states, California (181,900), New York (145,600), Washington (31,700), Massachusetts (28,300), and Texas (28,000), account for 56% of every homeless person counted across the states and DC. The remaining 45 states and DC split the other 44%.

Where Homelessness Fell and Rose Between 2024 and 2025

Homeless rate per 10,000 residents. Illinois and Hawaii posted the steepest drops while Oregon and North Carolina climbed.

Homeless Rate 2024 → 2025 0 10 20 30 40 50 60 70 80 Hawaii Oregon Illinois New York Colorado North Carolina Montana Maine Utah Delaware Maryland Missouri Massachusetts Minnesota Florida Kentucky California New Hampshire South Dakota New Jersey Alabama Oklahoma West Virginia Connecticut Vermont Mississippi Nevada Rhode Island Arizona Tennessee Pennsylvania Wyoming Iowa Indiana Kansas Louisiana South Carolina Nebraska Michigan Idaho New Mexico Wisconsin Arkansas Ohio Alaska Georgia North Dakota Texas Virginia Washington 32.8 +10.2 8.8 6.7 3.6 +3.4 +2.1 2.1 +2 +1.9 +1.7 +1.6 1.5 1.5 1.5 +1.2 1.2 +1.1 +1 +1 0.9 0.8 +0.8 +0.8 0.8 +0.7 0.7 0.7 +0.6 +0.6 +0.6 +0.6 +0.6 +0.5 0.5 +0.5 0.5 0.4 +0.4 0.4 +0.3 +0.3 0.3 +0.3 0.3 0.1 0.1 0.1 +0.1 0
§ 04

The Poorest States Have the Fewest Homeless People

The most counterintuitive finding in the data is what does not predict homelessness. Across all 50 states, the poverty rate shows essentially no correlation with the homeless rate. The relationship is so weak that knowing a state's poverty level tells you almost nothing about how many people are sleeping on its streets or in its shelters.

Louisiana has the highest poverty rate in the country at 18.7%. Its homeless rate is 8.0 per 10,000, third-lowest in the nation. Mississippi has the second-highest poverty rate at 17.8% and the lowest homeless rate of all, 4.2 per 10,000. Meanwhile, Hawaii has a poverty rate of just 10%, well below the national figure, and still sits fourth in the country for homelessness even after a steep fall in 2025.

The pattern holds at a regional level. The South averages the highest poverty rate of any region at 13.96% but the lowest average homeless rate at 10.3 per 10,000. The West averages a poverty rate of 10.93%, lower than the South, yet its average homeless rate of 29.9 per 10,000 is nearly three times higher.

What does track with homelessness is rent. The National Alliance to End Homelessness identifies housing cost burden, not poverty, as the primary driver of homelessness. The data supports that finding. States where the median gross rent exceeds $1,800 a month, including California ($2,104), Hawaii ($1,942), Massachusetts ($1,848), and Washington ($1,824), dominate the top of the homeless rankings. States where rent falls below $1,000 a month, such as Mississippi ($990), West Virginia ($883), and Arkansas ($982), cluster at the bottom.

Rent Prices Track With Homelessness Rates Across All 50 States

States with the highest median rents consistently report the highest rates of homelessness per 10,000 residents, while the cheapest rental markets barely register.

20 40 60 $1K $1.2K $1.4K $1.6K $1.8K $2K Homeless Rate Gross Rent $ California Hawaii Colorado Nevada Virginia New Hampshire Alaska Vermont Minnesota New Mexico South Dakota West Virginia
§ 05

What Changed in Illinois, Hawaii, and New York

The 2025 count did not just end a run of record highs. It unwound much of the surge that reshuffled the rankings a year earlier, and the states that had risen fastest fell fastest.

Illinois posted the steepest drop in the country: its homeless rate fell from 20.3 to 11.5 per 10,000, a 43% decline that moved it from 13th to 30th in the rankings. The 2024 spike had been driven by migrants and asylum seekers arriving in Chicago from the southern border. HUD reports that Illinois and New York, which recorded the largest increases between 2023 and 2024, recorded the largest decreases between 2024 and 2025, by roughly 11,300 and 12,500 people respectively, and that some local agencies in those states attributed the declines in part to changes in federal immigration policy.

Hawaii fell from 80.5 to 47.7 per 10,000, a 41% decline that dropped it from first to fourth. Its 2024 number had been inflated by the Maui wildfires of August 2023, which destroyed the town of Lahaina and displaced thousands of residents. Even after the fall, Hawaii's rate remains more than double the 50-state average.

New York eased from 79.5 to 72.8 per 10,000, an 8% decline, yet it now tops the ranking because Hawaii fell further. New York is unique among states: a 1981 consent decree known as Callahan v. Carey requires the city to shelter anyone who asks, so its count captures people that other states never record.

Not every state fell. Oregon rose from 53.5 to 63.7 per 10,000, a 19% increase that lifted it to second in the nation. North Carolina climbed from 10.5 to 13.9, up 32%, and Mississippi rose from 3.5 to 4.2, though it remains the lowest in the country. Nationally, HUD counted 22 states and DC with decreases, with the rest flat or higher.

§ 06

Nearly Two Out of Three Homeless Californians Sleep Outside

The national homeless count treats all homeless people equally, but the experience of homelessness varies enormously depending on where a person lives. In some states, being homeless means sleeping in a shelter bed. In others, it means sleeping under a highway overpass.

California reports an unsheltered rate of 63.5%, meaning nearly two out of every three homeless Californians are living on streets, in vehicles, or in places not meant for habitation. Oregon (60.7%), Hawaii (59.2%), Arizona (55.7%), and Georgia (54.2%) show similar patterns. Nationally, HUD counted about 36% of homeless people in unsheltered locations in 2025.

The Northeast looks like a different country. New York reports an unsheltered rate of just 4.0%. Massachusetts is 6.2%. Vermont is 8.0%, up from 1.6% in 2022, when pandemic-era hotel and motel voucher programs kept nearly everyone indoors. Maine moved the same way, from 3.7% to 11.6%. New York's right-to-shelter mandate explains its number directly: the legal obligation to house anyone who presents at a shelter means almost no one goes uncounted as unsheltered.

The regional divide is sharp. The West averages an unsheltered rate of 43.3%. The Northeast averages 15.5%. That gap means the same homeless rate in two different states can describe completely different realities: one person in a shelter bed with access to services, and another sleeping in a tent with none.

Mississippi, the state with the lowest overall homeless rate, still reports an unsheltered rate of 41.7%, down from 63.6% in 2022. Only about 1,200 people in Mississippi were counted as homeless, so even a modest lack of shelter capacity translates to a high unsheltered percentage.

§ 07

A Single Night in January

Every number in this ranking comes from the same source: HUD's Annual Homelessness Assessment Report, built on a nationwide Point-in-Time count. On a single night during the last 10 days of January, local agencies send volunteers to count every person they can find sleeping in shelters, on park benches, under bridges, and in encampments.

The count catches people in emergency shelters and transitional housing. It catches people sleeping visibly outdoors. What it misses is everyone else: families doubled up on a relative's couch, individuals staying in motels they cannot afford past the week, people cycling through hospitals or jails who have no home to return to. The Department of Education, which uses a broader definition under the McKinney-Vento Act, consistently identifies far more homeless children and youth than HUD's count captures.

Timing matters too. The count happens in January, when freezing temperatures in the Midwest and Northeast drive people indoors or into hiding. A homeless person in Chicago who spends one night with a friend to escape the cold will not appear in the data. A homeless person in Phoenix sleeping in a public park likely will. That asymmetry means warm-climate states may appear to have higher unsheltered rates partly because their unsheltered populations are easier to find in January.

None of this means the data is useless. The PIT count is the only consistent, nationwide measurement of homelessness that exists. Its year-over-year trends are meaningful even if the absolute number is almost certainly an undercount. But readers should understand that the 745,652 figure is a floor, not a ceiling.