and here is GROK's answer (I posted the Colonel's question exactly:
The question treats "people living on the street" as one group that produces no value and should be handled as a fiscal problem. That framing is too crude to produce a useful policy. The useful answer starts by splitting the population, then applying different tools.
They are not one group
U.S. point-in-time counts for January 2025 put total homelessness around 746,000, with about 266,000 unsheltered.
Those numbers mix very different situations:
Families and people in short-term housing crises. Family homelessness fell in the latest count; most families are sheltered. This group is closer to a poverty/housing-cost problem.
Workers living in cars or motels after a job loss, eviction, or medical bill.
Veterans. Veteran homelessness has fallen substantially over the last 15 years. That is one of the few clear policy successes.
Chronic unsheltered homelessness: long-term street living plus a disability. This is the group the question is mostly about. It hit a record high of about 156,000 individuals in 2025 and is up on the order of 80% since 2013. About six in ten of that group are unsheltered.
Serious mental illness and substance use are concentrated in that last group, not evenly across "the homeless." Estimates vary by method, but chronic street homelessness is where psychosis, addiction, and medical collapse cluster. A JAMA review found very high rates of mental disorders among unhoused people generally; SAMHSA-type figures put serious mental illness much higher among the chronically homeless than in the general adult population.
So "providing no value" is a slogan. Some people on the street are temporarily poor. Some are sick enough that they cannot currently function as workers, parents, or neighbors. Some are both. Policy that pretends they are all the same fails.
The "drain" part is real — and so is the cost of doing nothing
Street homelessness is expensive even if you ignore compassion. Cities already spend large sums on encampment cleanup, EMS, emergency rooms, jails, and policing. Those costs do not buy stability; they buy management of collapse. Studies of encampment response put city spending in the thousands of dollars per unsheltered person per year just for clearance, sanitation, and outreach. Sweeps without an exit often move the same people a few blocks.
Housing First / permanent supportive housing does a specific thing well: it keeps high-need people housed better than "get sober first, then maybe a bed." Multiple trials and reviews show large gains in housing stability versus treatment-first or usual care. Cost-offset studies are mixed: pre-post studies often claim savings; better-controlled studies often find partial offsets, not a free lunch. Typical U.S. program costs are on the order of mid-teens of thousands of dollars per person per year, sometimes more for the highest-need cases. Benefits show up as fewer shelter nights, some ER/jail reduction, and longer life — not as a sudden conversion into net taxpayers.
That is the honest fiscal picture: leaving people on the street is costly and ugly; housing them is also costly; the question is which mix reduces death, disorder, and long-run public spending.
Housing First as national doctrine is weaker than Housing First as a tool. Since HUD leaned hard that way after 2013, taxpayer-funded beds and Continuum of Care spending rose a great deal, while chronic homelessness still climbed. Housing stability is not the same as recovery, employment, or public order. Reviews consistently find weaker or mixed effects on mental illness, addiction, and work.
What actually moves the numbers
Places that reduced street homelessness did more than one thing.
Houston cut area homelessness sharply by coordinating agencies, using a shared list of people, housing them quickly, and clearing encampments after there was a place to send people. Supply of relatively cheap apartments mattered. "Housing first, then services" worked there as an operating system, not as a slogan that banned treatment or enforcement.
Finland treated a permanent apartment as the starting point, converted shelters into flats, built social housing, and kept doing it for years with cross-party continuity. Long-term homelessness fell a lot. That model assumes a welfare state and a housing pipeline most U.S. cities do not have. Recent welfare cuts there have tested the model.
Japan has very little unsheltered homelessness by U.S. standards. The mix is different: more housing production (national rules that make it easier to build), a comprehensive public-assistance system that can actually pay rent quickly, stronger family fallback, no U.S.-scale opioid market, and a much higher rate of institutional care for severe mental illness. You cannot copy the culture. You can copy "build housing" and "aid that arrives before someone is living under a bridge."
The common ingredients are not mysterious: enough cheap housing, a real off-ramp from the street, treatment capacity, and a rule that sidewalks and parks are not housing.
A policy that does not require pretending
If the goal is fewer people dying and camping in public, plus less drain on hospitals and police, do this:
1. Stop using one program for every case.
Economic homelessness needs cheaper housing, eviction prevention, and wages that clear local rents. Chronic street homelessness needs housing plus psychiatry, addiction treatment, and, for a subset, long-term supported or locked care. Mixing those populations in the same slogan is how systems fail both.
2. Build and legalize housing.
No service model beats a city that forbids enough homes. Tokyo and parts of Houston are not magic; they allow more housing than coastal U.S. cities that then act shocked when tents appear.
3. Offer an immediate alternative to the street, then enforce the street.
A real offer is a bed or a room tonight, not a waitlist and a pamphlet. Once that exists, camping, open-air drug markets, and blocking sidewalks are code-enforcement and public-health problems, not a housing philosophy. Clearing people with nowhere to go is theater. Leaving encampments indefinitely is also theater, just with different voters.
4. Treat incapacity as incapacity.
A person with untreated psychosis or catastrophic addiction is not exercising a lifestyle preference in any meaningful sense. Voluntary Housing First should remain the default for people who can use it. For people who are gravely disabled and repeatedly dying in public, civil commitment, conservatorship, and locked treatment are more honest than "housing with optional services" that they cannot use. Early LA County pilots aimed at gravely disabled unsheltered residents show high rates of getting people indoors when the legal and clinical machinery actually exists. Scale and due process matter; pretending this group is just "underserved renters" does not.
5. Measure exits, not spending.
Track: people permanently off the street, overdose and mortality, return-to-homelessness, employment or disability stabilization, and public-order complaints. Beds funded and dollars obligated are inputs.
6. Do not outsource family and work to the state, then act surprised.
A large share of adult street homelessness sits downstream of family collapse, foster-care exits, prison release with no plan, and untreated illness. Prevention is boring: keep people housed after jail and hospital discharge, treat first-episode psychosis, and stop discharging foster youth into nothing.
On "no value to society"
Net fiscal contribution is not the same thing as moral status. A society that only houses the currently productive will still pay for the rest — in ERs, morgues, and parks. A society that treats public space as a right to camp will lose the public space.
The adult position is both:
People on the street are not a single disposable class.
The public is not required to accept open-air disorder as the price of compassion.
What to do: house the housable, treat the treatable, confine the small group that cannot safely live in ordinary housing, build enough homes that the first two groups have somewhere to go, and stop confusing unmanaged street living with a social good.
