Housing as a Foundation, Not a Finish Line: Why Evidence Demands We Refine, Not Discard, What Works for the Most Vulnerable

By Gina Williams, M.N.M., M.B.A

Founder and CEO, GirlWrites LLC

“I felt compelled to write this after having to advise a client that, under the new policy shift, I cannot recommend that their amazing program invest resources in me writing this grant for them”

Their model is grounded in deep harm reduction and Housing First principles that have produced real stability for people others had given up on. Yet the FY 2026 HUD CoC NOFO's direction made a competitive application a poor investment of their limited resources—and my effort.

Imagine a woman in her 40s, diagnosed with schizophrenia and struggling with methamphetamine use after years on the streets. She has cycled through shelters, jail holds for petty theft to survive, and emergency rooms for frostbite and infections.

Then she receives a key to her own apartment through a scattered-site permanent supportive housing program grounded in Housing First principles: no prerequisites for sobriety or treatment compliance to get housed. Intensive case management follows through voluntary, persistent outreach.

Over 18 months, she stabilizes enough to engage in psychiatric care. She reconnects with a sibling. She begins part-time work. Her emergency room visits drop dramatically. She is still housed three years later.

This is not a hypothetical.

It reflects documented outcomes from well-implemented Housing First programs serving people with severe mental illness and co-occurring substance use disorders.

Systematic reviews of dozens of studies show Housing First consistently achieves housing stability rates of 75–98% in the first year for chronically homeless individuals, far outperforming "treatment first" or linear continuum-of-care models that require sobriety or program compliance before housing. One major analysis found an 88% reduction in homelessness and a 41% improvement in housing stability compared to treatment-first approaches.

Cost-offset studies in places like Denver show participants spending hundreds more days housed, with meaningful drops in jail bookings, arrests, and emergency service use.

These are real people—mothers, veterans, young adults fleeing domestic violence or aging out of foster care, individuals with profound trauma and brain disorders.

For many with serious mental illness, stable housing is the prerequisite for any recovery. Without it, treatment engagement is nearly impossible.

The evidence here is robust: Housing First, when delivered with high fidelity and paired with assertive community treatment or intensive case management, delivers on its core promise for this population.

The FY 2026 HUD Critique

Yet the FY 2026 HUD CoC NOFO levels a blunt critique.

Since the 2013 emphasis on Housing First, national literal homelessness has risen 27%, unsheltered homelessness 36%, and chronic homelessness 81%, even as permanent housing beds increased 151% and CoC spending rose 111%.

The document calls the approach a "profound failure" by its own metrics of ending homelessness and optimizing self-sufficiency.

This is where the baby risks being thrown out with the bathwater.

The Data Does Not Support a Simple Narrative

National Point-in-Time counts and state-level data reveal concentration—not uniform failure.

California and Oregon consistently report among the highest rates of unsheltered homelessness. Portland's tri-county area saw sharp recent increases. Visible encampments, public drug use, and street disorder in parts of Los Angeles, San Francisco, and Portland fueled public backlash and political reversals, including Oregon's partial rollback of Measure 110, the 2020 drug decriminalization initiative.

Measure 110 coincided with the fentanyl wave that devastated the West Coast—and much of the country.

Overdose deaths were already rising sharply before the measure. Studies on its direct causal impact on fatal overdoses are mixed, with some finding no clear additional spike attributable to decriminalization alone.

However, the policy environment—reduced criminal consequences for possession alongside visible public use, encampments, and slower-than-hoped treatment infrastructure ramp-up—contributed to perceptions (and, in many neighborhoods, realities) of disorder.

California's Proposition 47 similarly reduced penalties for certain theft and drug offenses and has been linked by critics to retail theft cycles and strained local responses.

These are real governance and policy failures in specific jurisdictions.

High housing costs driven by restrictive land-use rules, under-building, and NIMBYism have made supply chronically inadequate in high-cost coastal metros. Migration to areas with services, post-COVID behavioral health surges, and the synthetic opioid crisis amplified inflows.

Attributing the entire crisis to Housing First ignores these structural and supply-side drivers.

It also ignores places where coordinated Housing First approaches delivered measurable reductions.

Houston's region cut homelessness by roughly 63% since 2011 through data-driven prioritization, landlord partnerships, and rapid re-housing alongside permanent supportive housing, without massive new local spending.

Earlier Utah efforts showed large drops in chronic homelessness when focused on the highest-need individuals with robust support.

Where the Critique Has Merit—and Where It Overreaches

Pure "housing only" implementations that provide apartments with minimal services often struggle with the most complex cases: people with severe untreated psychosis or stimulant use disorders who do not spontaneously engage in recovery.

Substance use outcomes in many randomized trials of Housing First are mixed.

Housing stability improves, but reductions in use or related harms are not consistently superior to other models.

Some studies show no worsening of substance use, which itself is a harm-reduction win. Others find limited impact on employment or broader self-sufficiency without intentional, well-resourced treatment and workforce pathways layered on top.

The NOFO's new requirements—including greater emphasis on treatment integration, employment income outcomes, 20-plus hours of weekly engagement in services and activities for certain transitional models, and restrictions on specific harm-reduction practices like safe consumption sites—respond to these gaps.

They aim to prevent models that house people while visible chaos persists on surrounding streets.

That impulse is understandable after years of public frustration in hard-hit cities.

But broad application risks punishing programs that are already doing the hard work well.

The woman in the opening example did not need coercion into treatment to accept housing.

She needed safety and relationship-based support to eventually choose engagement.

For people with serious mental illness, forced "treatment first" or high-barrier models have historically failed at scale, producing higher returns to homelessness.

The new project quality thresholds and scoring factors, while well-intentioned, could disadvantage providers serving the hardest-to-house if CoCs interpret them rigidly or if narratives cannot authentically demonstrate rapid employment gains among populations with profound barriers.

The administration is correct that some policy environments—decriminalization without treatment capacity, tolerance of open-air markets and encampments, and underinvestment in recovery infrastructure—have made certain cities magnets for visible suffering while eroding public support for funding.

Those problems are real.

They demand accountability, public order measures, and supply-side housing reform.

They are not, however, synonymous with every Housing First or harm-reduction-informed program operating in those same states.

What Actually Serves Those Who Need Help Most

The profoundly moving truth is this:

Housing is necessary but rarely sufficient for the people carrying the heaviest burdens—severe mental illness, complex trauma, long-term substance use disorders, and repeated system failures.

The programs that change trajectories combine immediate, low-barrier housing with persistent, voluntary, high-quality outreach and treatment access.

They track real outcomes on stability, health, and eventually self-sufficiency.

They partner across systems—behavioral health, workforce, and criminal justice reentry—without pretending one tool solves everything.

Houston's coordinated model, Denver's supportive housing social impact bond results, and rigorous evaluations of assertive community treatment paired with housing all point in the same direction:

Fidelity matters.

So does refusing false choices between "Housing First" and "treatment matters."

The new NOFO's critique contains uncomfortable data that cannot be wished away.

But a swing that broadly devalues proven housing-stability tools—or pressures providers into inauthentic narratives—will not fix the governance and supply failures driving visible crises in specific regions.

It risks destabilizing the very people who have nowhere else to turn.

Those who need the most help deserve housing as a foundation for dignity and recovery, paired with the treatment, structure, and accountability that evidence shows can turn stability into a life worth living.

We can demand better implementation and honest outcome measurement without discarding what works for the most vulnerable among us.

The alternative is more cycling, more visible suffering, and more lives lost to the streets.

That is not a policy any administration should accept.