The White House recently took an important step toward rethinking America’s approach to homelessness. An executive order signed by President Trump last year directs federal housing funds to support a broader set of interventions beyond the long-dominant “Housing First” model. This shift is long overdue — and it should be reinforced and codified into law by Congress. New nationwide data now confirms what many local leaders have learned through experience: homelessness is driven by more than housing affordability alone, and policy that treats it otherwise is destined to fail.
That evidence comes from a landmark new analysis by the Common Sense Institute (CSI), which examined homelessness data across all 50 states to test a core assumption made by many — that homelessness is primarily a function of the cost of housing. If that premise were true, housing costs would show the strongest statistical relationship with homelessness. CSI’s analysis found otherwise.
Before turning to the data, it’s worth understanding how federal homelessness policy arrived at this point. Housing First rose to prominence with a simple and appealing idea: provide permanent housing — without preconditions such as sobriety, employment, or participation in treatment — and stability will follow. Over time, this approach became the default — and often the exclusive — model funded by the federal government. Billions of dollars have flowed into permanent supportive housing and rental subsidies, while alternative approaches were marginalized or dismissed as outdated or punitive.
Meanwhile, homelessness has continued to increase dramatically in many states and cities, including those that invested most heavily in Housing First. According to CSI’s analysis, Colorado ranks among the top 10 in the nation for its rates per capita of homelessness, chronic homelessness, unsheltered homelessness, homelessness with severe mental health issues, and homelessness with chronic substance abuse. The other states most consistently represented in each of those categories are California, Hawaii, Oregon, Washington, and Vermont — each of which have invested heavily in Housing First response.
This contradiction has often been explained away by rising rents, leading to calls for even more housing-only funding. CSI’s research finally puts that explanation to the test.
Using consistent, nationwide data, CSI evaluated the relationship between homelessness and a wide range of variables, including housing prices, rent levels, illicit drug use, crime rates, and policing levels. The findings are striking. While housing affordability does correlate with homelessness, it is not the strongest link. The variables most closely associated with higher homelessness rates are illicit drug use, crime, and lower levels of policing.
CSI established this by finding what’s called the “correlation coefficient” — simply put, the strength of one set of numbers’ relationship with another. The correlation coefficient between state-level homelessness rates and rent affordability is 0.35, which is a weak to moderate relationship Other correlations, however, are stronger, e.g., the state homelessness rates and state rates of illicit substance use. There the correlation coefficient is 0.57, which is stronger than the relationship between homelessness and rent affordability.
In most of the homelessness subcategories, items like police per population, illicit substance use, and NIBRS combined crime rate had stronger positive or negative correlations to state homelessness rates than did rent.
In other words, states with higher drug abuse rates and crime -and fewer law enforcement officers per capita — consistently experience higher levels of homelessness, regardless of housing costs. These relationships are stronger and more statistically significant than housing prices alone.
This matters because it exposes a fundamental mismatch between the drivers of homelessness and the way public dollars are allocated to address it. Federal, state, and local funding has focused overwhelmingly on housing supply and rental assistance, while comparatively little has been invested in treatment capacity, public safety, or systems that promote accountability and reintegration into work.
It is worth reviewing Colorado’s recently introduced House Bill 26-1202. The legislation focuses on housing-centric strategy by requiring a statewide plan through the Department of Local Affairs, creating multi-jurisdictional homelessness authorities empowered to levy new sales taxes, and redirecting real estate documentary fees toward affordable housing for the homeless. This bill likely promotes a Housing First model over work-first alternatives, potentially fostering dependency by prioritizing unconditional housing without tying it to employment incentives. It could risk expanding government administration, ignoring local variations in homelessness drivers like those identified in CSI’s data.
Housing First advocates often respond with a predictable counterargument: people cannot address addiction, mental illness, or employment instability until they are housed. Housing, they argue, is the necessary foundation upon which all other progress depends. Some point to individual success stories or short-term housing retention rates as proof that the model works.
But this argument confuses necessary with sufficient — and ignores the broader evidence. Housing may be part of the solution, but the data shows it is rarely the decisive factor. Providing housing without addressing substance abuse, mental health, or chronic disengagement from work does not resolve homelessness at scale; it often relocates it indoors or concentrates it in ways that destabilize neighborhoods and overwhelm service providers. Bills like HB26-1202 should be closely examined, with their emphasis on housing resources and regional navigation without prioritizing work-first pathways.
CSI’s data and findings align with the real-world experience of cities that have pursued more comprehensive approaches. San Antonio, for example, has paired housing with mandatory treatment, employment pathways, and coordination with law enforcement — achieving sustained reductions in homelessness
In Aurora, where I served on the City Council, we sought to pursue a similar philosophy: meeting people where they are, but not pretending that housing alone would solve addiction, mental illness, or criminal activity. We believed this approach will deliver better outcomes than the housing-only strategies adopted elsewhere along the Front Range, including those advanced in HB26-1202.
Critics often characterize these models as “treatment first” or “work first” in a pejorative sense, suggesting they impose unreasonable barriers. In reality, they reflect a more honest understanding of homelessness as a complex social failure, not merely a market failure. Addiction, crime, and untreated behavioral health conditions are not side issues; they are central drivers, as CSI’s analysis makes clear.
None of this means housing affordability should be ignored. In high-cost markets, rent pressures can push vulnerable individuals closer to the edge. CSI’s analysis confirms that housing costs do matter. But they are part of a constellation of factors — not the dominant force.
The White House’s executive order acknowledges this reality by allowing greater flexibility in how federal homelessness funds can be used. That flexibility should not be temporary or discretionary. Congress should codify it, ensuring that federal policy supports a full spectrum of interventions: housing paired with treatment, employment pathways, and public safety measures that protect both unhoused individuals and the broader community.
Compassion does not require ignoring evidence. And accountability is not cruelty. True compassion means confronting the conditions that keep people homeless — even when doing so is politically uncomfortable.
For years, policymakers have followed ideology instead of outcomes. CSI’s nationwide data analysis removes any remaining excuse for doing so. The data is clear: homelessness cannot be solved by housing alone. The sooner federal policy reflects that reality — the sooner real progress can begin.
