LA’s Declining Homelessness Numbers: Real or Illusory?

Earlier this year, Los Angeles Mayor Karen Bass issued a statement titled “Lasting Change: Annual Homelessness Count Down Two Years in a Row for the First Time Ever in Los Angeles.” LA officials crowed about the numbers, but the count leaves room for reasonable doubt.

The annual “point-in-time” (PIT) count conducted in February (released in July) reported that homelessness in the City of Los Angeles decreased by 2.2 percent in 2024 and 3.4 percent in 2025. The report also stated that homelessness across the County of Los Angeles decreased by 0.27 percent in 2024 and 4 percent in 2025.

Those small recorded improvements translate into 1,553 fewer homeless people citywide and 3,004 fewer people countywide—not exactly eye-popping reductions. More than 72,000 people remain homeless in Los Angeles County, equivalent to a moderate-sized town. But even those small improvements could be illusory.

Researchers at the University of Southern California, who help design and analyze LA’s PIT count, report that the sampling margin of error is about 1,300 people, plus or minus, for the Los Angeles County Continuum of Care and about 920 people, plus or minus, for the city. In other words, much of the so-called “lasting change” could plausibly be the result of mere error and not because of a true reduction in the number of homeless people. Without changes in the sample design, there is reason to believe that errors will increase.

After the 2024 U.S. Supreme Court decision in City of Grants Pass, OR v. Johnson, many local governments adopted tougher anti-camping laws that resulted in more sweeps of encampments, collection of personal belongings, arrests, and disruption of lives. That created stronger incentives for homeless people to hide, making it harder for PIT counters to find everyone and skewing results toward undercounting. In LA, underground homeless dwellers have been found in tunnels and vaults. It’s unknowable how many homeless people live undetected, but the number is likely increasing.

In addition to sampling errors and hiding, homelessness numbers drop due to overdose deaths, suicides, homicides, vehicle-related deaths, and fire deaths. (A third of all fires in the City of Los Angeles involve unhoused people.) Those factors reduce homelessness, but in tragic ways.

Looking at the most recent data on deaths of homeless residents from 2023, the LA County Department of Public Health reported 1,128 deaths due to alcohol or drug overdoses—70 percent of which involved fentanyl—and 2,508 total homeless deaths from all causes. Thus, it is plausible that sampling errors plus homeless deaths accounts for the much-celebrated declines in homelessness.

Undercounting might also occur for purely political reasons. Like most workers, PIT bureaucrats want to please their bosses. California Gov. Gavin Newsom and Mayor Bass have made it clear that they want homelessness numbers to fall—and fast.

“It is time to take back the streets. It’s time to take back the sidewalks,” Newsom bluntly stated in May. “No more excuses. Time to deliver.” On her first day in office, Mayor Bass declared a state of local emergency on homelessness, but there’s more to it than rhetoric. If additional funding flows to jurisdictions with falling homelessness numbers, then stronger incentives exist to cook the books. Taxpayers should take notice of curious happenings.

In July, before the public release of the final numbers, 475 sheltered people were suddenly removed from the LA City homeless count without notifying elected officials. Separately, a former information officer for the LA Homeless Services Authority (LAHSA), the organization that conducts the official count, testified that her supervisor instructed her “to do whatever we can to make the mayor look good.”

In a 2025 report, the Santa Monica-based RAND Corporation counted significantly more homeless people and dwellings in three LA neighborhoods than the official count revealed. In Venice, for example, RAND counted 554 people and dwellings. The official count was 173. RAND found a 32 percent systematic undercount by LAHSA in those three neighborhoods alone.

Data discrepancies, premature deaths, and sampling errors provide strong evidence that LA’s celebrated drop in homelessness is illusory. LA officials should stop gloating and work to achieve real results: Get homeless people off the streets and back on their feet as independent, self-sufficient individuals.