Fresno, Madera Homeless Count Drop May Be Misleading — Methodology Change Makes Comparison Unreliable

Fresno and Madera counties reported what appeared to be California’s steepest drop in homelessness this year — a 34 percent decline — but officials acknowledged the figure may not reflect reality. This year’s volunteers counted only people who confirmed they were homeless and agreed to be surveyed, a shift from 2025’s broader method that counted anyone visibly living on the street. Officials told the Fresno Bee the change makes year-over-year comparisons unreliable.

The admission comes as 28 California counties released preliminary results from January’s federally mandated point-in-time count. More than half showed decreases, with San Joaquin County’s overall homeless population falling nearly a third and its unsheltered count cut nearly in half. San Francisco’s count dipped about 4 percent. Eight counties didn’t conduct a count this year; seven haven’t released results.

Local officials credit new shelter beds and affordable housing units funded by recent state spending — the Homekey program, the state’s Homeless Housing, Assistance and Prevention fund, and targeted dollars for homeless residents with mental illness or substance use disorders. San Joaquin County added 304 shelter beds and 1,078 permanent housing units between 2024 and 2026, boosting the number in permanent housing by 86 percent.

Advertisement

But homeless advocates say the numbers may be obscured by intensified encampment sweeps and camping citations following the Supreme Court’s 2024 ruling allowing cities broader authority to clear camps. “It’s a lot harder to find one person with a blanket somewhere than it is to see a camp,” said Niki Jones, executive director of the Sacramento Regional Coalition to End Homelessness, who called the reported progress “baloney.”

A RAND Corporation study found that in Los Angeles’ Skid Row, Hollywood, and Venice neighborhoods, overall unsheltered homelessness stayed roughly flat between December 2024 and January 2026 — but tent dwellers fell 23 percent while people sleeping without any shelter rose 20 percent.

County officials warn that even real progress may not hold. Federal cuts to Medi-Cal and other social services threaten recent momentum, and state homelessness funding typically arrives as one-time grants, leaving counties uncertain year to year. “The reality is that it’s going to be hard to hold this line if we don’t start having more stable funding sources,” said Monterey County homeless services director Roxanne Wilson.

The Breakdown

• Fresno and Madera counties’ 34% homeless count drop is undermined by a methodology change — this year’s count only included people who confirmed they were homeless and agreed to be surveyed, versus 2025’s broader method of counting anyone visibly living on the street
• State programs funding the reported progress include Homekey (converting hotels to homeless housing), the Homeless Housing, Assistance and Prevention fund, and targeted funds for homeless residents with mental illness or substance use disorders
• Point-in-time counts are federally mandated tallies conducted on a single January night by volunteers and outreach workers — widely acknowledged to undercount the true homeless population but remain the primary nationwide tracking tool
• The Supreme Court’s 2024 ruling gave cities broader authority to clear homeless encampments, leading to increased sweeps and citations that advocates say scatter unhoused residents and make them harder to count

What This Means for You

The state legislature and Governor allocated billions in homelessness spending over the past four years — Homekey, housing assistance funds, mental health and substance use disorder programs — and these county-level results are how voters assess whether that spending is working. Fresno and Madera’s methodology change means you can’t trust this year’s 34% decline as evidence of actual progress.

The acknowledged unreliability of the count also raises the question of whether county officials are presenting the most favorable number available rather than the most accurate one.

What to Watch

The core tension here is uncertainty about whether the reported decline is real progress or an artifact of how the count was conducted. Fresno and Madera officials admitted the methodology shift makes comparisons unreliable, which means the 34% drop could be entirely a counting artifact rather than actual movement of people into housing.

The enforcement angle compounds the problem. Post-2024 Supreme Court ruling, cities have intensified encampment clearings and camping citations. That scatters people, making them harder to count — which shows up as a decline even when no one actually found housing.

The RAND study in LA found tent dwellers down 23% but people sleeping rough up 20%, suggesting dispersal rather than housing.

There’s also a sustainability question. State homelessness funding arrives as one-time grants, not recurring line items, leaving counties uncertain year to year about what support they’ll have.

Federal cuts to Medi-Cal and other social services add another layer of risk — gains funded by one level of government may be erased by cuts at another.

The accountability question: why did Fresno and Madera change the counting methodology in a way that makes year-over-year comparison unreliable, and why report a 34% decline when officials acknowledged the comparison doesn’t work? A methodology that only counts people who confirm they’re homeless and agree to be surveyed is narrower than one that counts anyone visibly living on the street.