Why Some Cities Have Far More Homelessness: It's the Housing Market

By Common Ladder · June 25, 2026 · 8 min read

Spend any time in a debate about homelessness and you will hear two stories. One says homelessness is about people — addiction, untreated mental illness, bad choices, broken families. The other says it is about housing — rents that outran wages, a market with no room left at the bottom. The two camps tend to talk past each other, and the disagreement shapes everything downstream, from what cities fund to how neighbors react to a shelter on their block.

The useful thing is that this is partly an empirical question, and the data are clearer than the shouting suggests. The honest answer is that both stories are true, but they answer two different questions — and confusing the two is where most of the argument goes wrong.

The puzzle: same problems, very different outcomes

Start with a fact that should bother anyone confident in the "it's just addiction" story. Rates of poverty, mental illness, and drug use are not dramatically higher in San Francisco, Seattle, or Boston than in Detroit, Cleveland, or rural West Virginia. In several cases they are lower. Yet per-capita homelessness in the expensive coastal metros runs several times higher than in poorer places where the same personal struggles are at least as common.¹

If individual vulnerabilities were the driver, the map of homelessness would track the map of addiction and poverty. It doesn't. That mismatch is the puzzle that the research set out to explain.

What the cross-city evidence found

The most thorough attempt to settle it is the 2022 book Homelessness Is a Housing Problem, by University of Washington researcher Gregg Colburn and data scientist Clayton Page Aldern. They lined up the leading explanations — poverty rates, mental illness, drug use, climate, the generosity of public benefits, even the politics of a city — against actual regional homelessness data, and tested which ones predicted the differences between metro areas.¹

Most of the usual suspects fell away. The variables that held up were about housing: the absolute level of rents and the rental vacancy rate. Places where rent was high and empty units were scarce had more homelessness, full stop. One finding captures how counterintuitive this is: regions with higher poverty often had lower homelessness, because in those places housing was cheap enough that a low income still reached it. Poverty plus expensive housing produces homelessness; poverty plus cheap housing, much less so.

Two other lines of research point the same way. Analysts at Zillow, working with academics at the University of New Hampshire, Boston University, and the University of Pennsylvania, found a threshold effect: once a community's median rent climbs past roughly 32% of median income, homelessness starts rising at an accelerating pace rather than a gentle one.² And in 2020 the U.S. Government Accountability Office ran its own econometric analysis of Point-in-Time data and found that a $100 increase in median rent was associated with about a 9% increase in an area's homelessness — a relationship that held even after controlling for wages, unemployment, and poverty.³ Three different teams, three different methods, the same direction.

Who versus how many: the distinction that ends the argument

None of this means addiction and mental illness are irrelevant — and it is worth being precise here, because the housing finding is easy to overstate in the other direction. The cleanest way to hold both truths is a distinction between who and how many.

Picture a game of musical chairs with ten players and seven chairs. When the music stops, three people are left standing — and you can predict, fairly well, who they will be: the slowest, the most distracted, the one who tripped. Individual disadvantage decides who loses. But notice that no amount of personal improvement changes the fact that three people end up with no chair. That number is set by the gap between players and chairs. Add three more chairs and the same "slow" players sit down comfortably.¹

Homelessness works much the same way. Addiction, mental illness, a lost job, a fled abuser, a medical bill — these are what determine who, among the many people under strain, ends up without housing. But how many people end up homeless in a given city is set mostly by how many affordable units exist relative to the people who need them. In a loose, cheap market, the same personal crisis ends in a friend's spare room or a modest apartment. In a tight, expensive one, it ends on the street. The person didn't change; the market did.

The national numbers followed the rent

The recent national trend is consistent with all of this. HUD's January 2024 Point-in-Time count recorded roughly 771,000 people experiencing homelessness on a single night — the highest figure since the count began, and an 18% jump in a single year.⁴ HUD attributed the rise to the worsening shortage of affordable housing, wages that lagged housing costs, and the expiration of pandemic-era rental aid and eviction protections that had briefly held the line. When the protections ended and rents stayed high, the count climbed.

What it looks like in Phoenix

Metro Phoenix is close to a textbook case. Rents here rose roughly 72% over the past decade, and Arizona is estimated to be about 270,000 homes short of what its population needs.⁵ That is the tight, expensive market the research describes. The 2025 Point-in-Time count found 9,734 people experiencing homelessness in Maricopa County, and while the overall number ticked up only slightly, unsheltered homelessness rose 28% in a single year — driven in large part by the expiration of federal funding that had supported more than a thousand shelter beds.⁵ Fewer beds and unaffordable rents are exactly the squeeze the housing-market model predicts.

Why this matters for what we do

Getting the diagnosis right changes the response. If homelessness in a city were purely a matter of individual pathology, the fixes would be entirely individual — treatment, sobriety, services — and a city could expand all of them and still watch its count rise, because the number of chairs never moved. The evidence says services are essential for the people who need them, but they cannot, on their own, lower the total in a market that has run out of affordable units.

That is why the strongest local responses do both at once: help the individual in front of you, and add to the supply of housing they can actually afford. It is also why Common Ladder spends as much effort on eviction prevention, rent and utility assistance, and getting people into units as it does on connecting them to treatment. The personal crises are real and deserve care. But if you want a city to have less homelessness, not just different homeless people, the lever with the most leverage is the housing market itself.

Have a correction or an updated figure? Email corrections.commonladder@gmail.com.