05/29/2026
Vermont Is Already Paying for Homelessness. The Question Is Whether We Want to Keep Paying for Failure.
A proposal for transitional community housing, public safety, and fiscal responsibility
Vermont is already spending a great deal of public money on homelessness, mental health crises, addiction, incarceration, emergency medicine, motel rooms, shelters, police response, and psychiatric hospitalization. The question is not whether Vermont will pay. Vermont is already paying.
The real question is whether we want to keep paying for homelessness in the most expensive, chaotic, and least effective ways possible.
We can continue paying for emergency rooms, ambulance calls, police responses, court processing, jail beds, motel vouchers, psychiatric hospital beds, encampment sweeps, crisis services, and human deterioration. Or we can build a structured, humane, accountable, medically supported transitional housing system that reduces the number of people living outside and reduces the public cost of unmanaged crisis.
This is not a proposal for “free housing” as a reward for addiction, criminal behavior, or refusal to work. That criticism deserves to be answered directly because many taxpayers are asking it in good faith. A person who works, pays rent, follows the law, and struggles to afford groceries, property taxes, health care, and transportation has a right to ask why the state would fund housing, food, showers, medical care, case management, and security for someone living outside.
The answer is simple: because the current system is already costing us more than most people realize.
When someone is sleeping outside, untreated, medically unstable, addicted, psychotic, traumatized, or cycling between the street and jail, the cost does not disappear. It moves. It moves to police departments. It moves to sheriffs. It moves to emergency rooms. It moves to hospitals. It moves to Medicaid. It moves to the Department of Corrections. It moves to the state psychiatric system. It moves to municipalities, downtown businesses, libraries, parks, first responders, shelters, and exhausted nonprofit workers.
Refusing to house people does not make homelessness free. It often makes it more expensive.
The proposed model
Vermont should consider piloting a medically supported transitional community housing campus for unsheltered single adults.
Each community would house approximately 60 individuals in 60 small, private, three-room studio cottages or pods, each no larger than 300 square feet. Each unit would include a toilet and sink room, a bedroom, and a small common room. The units would be prefab wooden cottages, winterized for Vermont, with running water, electricity, heat, and Wi-Fi.
The campus would include a 30,000-square-foot main building with offices for residential staff, case managers, medical staff, security, administrative staff, and other support personnel. The main building would also include a community health center, a full cafeteria with kitchen staff, staffed private restrooms, showers, laundry capacity, and practical support spaces.
A secondary 2,000-square-foot community center would support structured programming, peer support, skills training, group activities, visiting providers, and community-building.
The campus would be gated, with unarmed security staffing the entrance and patrolling the grounds. Security would not be there to criminalize residents. It would be there to protect residents, staff, neighbors, and the integrity of the program. Violence, trafficking, dealing, weapons, predatory behavior, and intimidation would not be tolerated.
The staffing model would include nurses, case managers, residential staff, security staff, kitchen staff, cleaning staff, and per-diem coverage. Residents would be assigned case managers who help them move toward one of two outcomes: transition into a higher level of care when clinically necessary, or transition into permanent subsidized housing when they are stable enough to live more independently.
This is not meant to become a permanent warehouse for poor people. It should be designed as a bridge: from the street to stability, from crisis to care, from unmanaged public cost to measurable public benefit.
What would it cost?
A realistic planning estimate for one 60-person site is approximately $49 million in startup capital, with a likely range of $35 million to $71 million depending on land, construction costs, utilities, permitting, environmental work, site conditions, and inflation.
Annual operating costs would likely fall between $6 million and $11 million per year, with a base-case estimate around $8.2 million per year before debt service.
If construction is financed over time, debt service could add approximately $3.2 million per year in the base case. That means the full annual cost, including operations and debt service, could be approximately $11.4 million per year for one 60-person site.
That equals roughly $137,000 per resident per year for operations only, or about $190,000 per resident per year including construction financing.
At first glance, that sounds extremely expensive. But that number only becomes meaningful when compared to what Vermont already pays for unmanaged homelessness, incarceration, and psychiatric hospitalization.
What does Vermont already spend?
Vermont’s 2025 Point-in-Time count found 3,386 unhoused Vermonters on a single January night, including 270 people counted as unsheltered. That unsheltered number is almost certainly an undercount because Point-in-Time counts are snapshots, not complete population registries.
Vermont’s homelessness-response budget discussions have involved figures around $82.6 million for a continuum of emergency housing, shelter, case management, rental assistance, supportive housing, cold-weather shelter, and related services.
When that kind of spending is divided only by the officially counted unsheltered population, the number appears shockingly high: roughly $300,000 per unsheltered person per year. That figure must be interpreted carefully because homelessness funds do not serve only unsheltered people. They also serve people in shelters, hotels, temporary housing, supportive housing, prevention programs, and rental-assistance programs.
Still, the comparison matters. Vermont is already spending large sums responding to homelessness. The problem is that much of the money is being spent reactively.
The same is true in corrections. Vermont incarceration has been estimated at just under $300 per incarcerated person per day, or more than $100,000 per person per year. Other comparative analyses place Vermont’s incarceration cost even higher, around $134,000 per incarcerated person per year.
Psychiatric hospitalization is far more expensive. Vermont Psychiatric Care Hospital’s FY2024 financial template listed operating expenses of approximately $34.2 million for a 25-bed state hospital. Divided across 25 beds, that equals roughly $1.37 million per bed per year. Even if one uses older and lower daily-rate comparisons, high-acuity inpatient psychiatric care still falls in the hundreds of thousands of dollars per person per year.
So the fiscal reality is this: the proposed housing campus is expensive, but it is not obviously more expensive than the status quo for the highest-cost individuals. It is likely cheaper than long-term psychiatric hospitalization. It is in the same broad range as Vermont’s highest public-cost systems when accounting for homelessness response, incarceration, crisis care, and emergency medicine.
Could Vermont save money?
Yes, but only under specific conditions.
A 60-person site could plausibly save Vermont approximately $4 million to $11 million per year if it is targeted to the highest-cost unsheltered adults and if it actually reduces spending in other systems.
A reasonable base-case estimate is that one site could avoid approximately $18.4 million per year in public costs, while costing about $8.2 million to operate. After debt service, the net savings could be about $7 million per year per 60-person community.
But this is not guaranteed. It depends on targeting and ex*****on.
The program saves money only if it reduces motel use, shelter overflow, police calls, ambulance transports, emergency department visits, jail days, psychiatric hospitalization, detox readmissions, and crisis-service utilization. If Vermont simply adds this program on top of the existing system, without reducing existing emergency spending, then there is no savings. There is only another expensive program.
This point matters politically and ethically. Vermont cannot afford symbolic compassion. It needs measurable compassion. It needs humane programs that work.
A response to conservative critics
Many people on the political right ask: Why should addicts, criminals, and drug dealers get free housing, food, and medical care?
The answer should not be defensive. It should be direct.
They should not receive housing because addiction, criminal conduct, or drug dealing is being rewarded. They should receive structured housing because untreated homelessness is dangerous, expensive, and socially destabilizing. The public has a rational interest in reducing the harm caused by unmanaged addiction, untreated mental illness, survival crime, public disorder, and repeated crisis-system use.
This is not about giving people something for nothing. It is about replacing a disorderly and expensive system with a structured and accountable one.
A person sleeping outside with untreated addiction does not become cheaper to the public because we refuse to house them. A person with untreated psychosis does not become safer because we leave them under a bridge. A person cycling between the street, the emergency room, jail, detox, and motel rooms does not become more accountable because the system refuses to stabilize them.
But conservatives are right about one thing: housing cannot mean lawlessness.
A housing campus must have rules. It must have expectations. It must prohibit violence, dealing, trafficking, exploitation, threats, weapons, and predatory behavior. It must protect staff, neighbors, and residents. It must include consequences for serious misconduct. It must work with law enforcement when public safety requires it. It must also distinguish between people who need housing support, people who need addiction treatment, people who need psychiatric care, people who need medical respite, people who need disability services, and people who pose a serious danger to others.
A serious system does not pretend everyone has the same needs. It sorts people accurately and responds proportionately.
The conservative case for this program is not that it is soft-hearted. The conservative case is that it could reduce disorder, reduce emergency spending, reduce public nuisance, reduce pressure on police, reduce jail cycling, reduce hospital burden, and create a more accountable pathway out of street homelessness.
A warning to the political left
Progressives also need to hear a hard truth: moral urgency does not repeal arithmetic.
Vermont is a small state with a limited tax base, high housing costs, high construction costs, rural geography, harsh winters, property-tax pressure, health care workforce shortages, and a population that already feels financially strained. If advocates propose large programs without credible cost controls, outcome measures, staffing plans, and tax fairness, the public will eventually turn against the programs.
The left should not argue as if every compassionate program automatically deserves indefinite funding. That is not policy. That is moral expression without fiscal discipline.
A real program needs a real budget. It needs a sustainable tax structure. It needs economies of scale. It needs shared administration. It needs Medicaid billing where legally permissible. It needs federal funds. It needs philanthropic capital. It needs hospital partnerships. It needs opioid settlement funding where substance-use services are involved. It needs measurable offsets from corrections, emergency housing, psychiatric hospitalization, emergency departments, and crisis services.
Most importantly, it needs public trust.
Vermonters should not be asked to fund an open-ended system with vague promises. They should be shown what the program costs, what it replaces, how many people it serves, how many people exit into permanent housing, how many return to homelessness, how emergency service use changes, how jail days change, how psychiatric bed days change, how staff safety is protected, and whether the public is getting a better outcome for the money spent.
Compassion without accountability will fail. Accountability without compassion will also fail.
The problem with dehumanization
The phrase “addicts, criminals, and drug dealers” is politically powerful because it reduces a complicated population to its most frightening examples.
Some unsheltered people do have substance use disorders. Some have criminal histories. Some engage in harmful or illegal behavior. Some are dangerous. Pretending otherwise insults the public’s intelligence.
But many unsheltered people are disabled, elderly, medically fragile, traumatized, cognitively impaired, neurodivergent, mentally ill, fleeing violence, priced out of housing, discharged from institutions, or unable to compete in a rental market that has no realistic place for them.
A humane society does not erase people’s harmful behavior. But it also does not erase their humanity.
When we reduce people to labels, we stop thinking clearly. “Addict” replaces a person’s developmental history, trauma, neurobiology, family losses, economic exclusion, and medical condition. “Criminal” replaces the question of whether a person needs accountability, treatment, supervision, housing, or all of the above. “Homeless” replaces the fact that this is a human being whose life has collapsed in public view.
Dehumanization is not only morally corrosive. It is bad policy analysis.
If we define people only by the worst thing they have done, we will build a system designed only for punishment. If we define people only as victims, we will build a system with no accountability. Vermont needs neither cruelty nor naivete. Vermont needs disciplined humanism.
Who should be at the table?
This proposal would affect far more than housing agencies.
The primary audience should include the Vermont Legislature, the Governor’s Office, the Agency of Human Services, the Department for Children and Families, the Department of Mental Health, the Department of Vermont Health Access, the Department of Corrections, the Vermont State Housing Authority, the Vermont Housing and Conservation Board, and municipal governments.
Secondary partners should include designated mental health agencies, preferred providers, local hospitals, the University of Vermont Health Network, Brattleboro Retreat, Rutland Regional Medical Center, Central Vermont Medical Center, federally qualified health centers, substance use disorder treatment providers, recovery centers, community justice centers, police departments, sheriffs, Vermont State Police, emergency medical services, fire departments, public health officials, housing nonprofits, shelters, domestic violence agencies, disability rights organizations, legal aid, peer-support organizations, veterans’ services, food security organizations, transportation providers, and neighborhood/community representatives.
This cannot be treated as only a housing problem. It is also a health care problem, a corrections problem, a municipal problem, a disability problem, a labor problem, a Medicaid problem, a public safety problem, and a moral problem.
How to fund it
A serious funding model should not rely on one source.
Capital costs could be supported through state bonding, federal housing funds, philanthropic contributions, hospital community-benefit investment, municipal participation, federal grants, climate-resilient building funds where applicable, and public-private partnerships.
Operating costs could be supported through a combination of state appropriations, Medicaid-billable services, federal homelessness funds, opioid settlement dollars for qualifying substance-use services, hospital partnerships, corrections reinvestment, mental health system savings, and local contributions where municipalities benefit from reduced emergency response.
The state should also consider a pay-for-performance structure. If the program reduces emergency room visits, jail days, psychiatric hospitalizations, motel use, and unsheltered homelessness among enrolled residents, future funding should expand. If it fails, it should be redesigned.
The funding logic should be replacement, not accumulation. Vermont should not simply create a new expensive program while preserving every inefficient emergency expenditure that the program was supposed to reduce.
What should be measured?
The program should publish a public dashboard.
At minimum, Vermont should measure the cost per resident per month; average length of stay; exits to permanent housing; exits to higher levels of care; returns to homelessness; emergency department use before and after admission; ambulance calls before and after admission; arrests and jail days before and after admission; psychiatric hospitalization before and after admission; overdose reversals; deaths; resident grievances; neighborhood safety incidents; staff injuries; staff turnover; employment or benefits stabilization; Medicaid enrollment; primary care connection; behavioral health connection; substance use treatment engagement; and resident satisfaction.
If the program cannot measure outcomes, it should not receive long-term funding.
The moral argument may open the door. The data must keep the door open.
The policy case
Vermont should pilot this model because the current system is too expensive, too fragmented, too reactive, and too inhumane.
A structured transitional housing campus would not solve homelessness by itself. It would not replace permanent affordable housing. It would not eliminate the need for psychiatric beds, addiction treatment, shelters, rental subsidies, or corrections reform. It would not be appropriate for every unsheltered person.
But it could serve a specific population extremely well: high-need unsheltered adults who are currently cycling through the most expensive parts of the public system and who need stabilization before they can move into permanent housing or higher-level care.
The program should be compassionate, but not permissive. Structured, but not punitive. Clinically informed, but not medically coercive. Fiscally disciplined, but not cruel. Transitional, but not disposable. Accountable to residents, staff, taxpayers, neighbors, and policymakers.
The public should not accept a false choice between punishment and permissiveness. There is a third path: structured care with public accountability.
The bottom line
Vermont is already paying for homelessness.
We are paying through emergency rooms, police calls, corrections, psychiatric hospitalization, motel rooms, shelters, crisis services, municipal disruption, and preventable human suffering.
The question is whether we want to keep paying for failure, or whether we want to pay for a system that has a chance to produce stability.
A 60-person community housing campus would be expensive. It might cost around $49 million to build and around $8.2 million per year to operate, with a full annual cost around $11.4 million if construction debt is included.
But if it is targeted to the highest-cost unsheltered adults, if it reduces motel use, jail days, hospitalizations, emergency department use, and crisis-system cycling, it could plausibly save Vermont several million dollars per year while producing a more humane and orderly public system.
That is the standard Vermont should use.
Not charity without limits.
Not punishment without results.
Not ideology without math.
The standard should be: Does it reduce human suffering? Does it improve public safety? Does it reduce emergency-system costs? Does it move people toward permanent housing or appropriate care? Does it protect taxpayers? Does it work?
If the answer is yes, then Vermont should build it.
Key source notes for hyperlinking or endnotes:
Vermont’s 2025 Point-in-Time count reported 3,386 unhoused people statewide, while VTDigger reported that 270 were counted as unsheltered in January 2025. ([Vermont Housing Alliance][1])
The Vermont Legislature’s H.938 fiscal comparison showed a homelessness-response total of about $82.6 million. ([Vermont General Assembly][2])
Vermont Psychiatric Care Hospital’s FY2024 financial template listed $34.2 million in total operating expense; the state hospital is described as a 25-bed facility in state budget materials. ([Vermont Department of Health][3])
WCAX reported a Vermont corrections estimate of just under $300 per incarcerated person per day, and a 2025 comparative analysis reported Vermont at about $134,000 per incarcerated person annually. ([wcax.com][4])
[1]: https://helpingtohousevt.org/wp-content/uploads/2025/07/FINAL-State-of-Homelessness-2025.pdf?utm_source=chatgpt.com "2025 State of Homelessness in Vermont Report"
[2]: https://legislature.vermont.gov/Documents/2026/Workgroups/Senate%20Health%20and%20Welfare/Bills/H.938/Drafts%2C%20Amendments%2C%20Legal%20Documents/H.938~Nolan%20Langweil~Fiscal%20Side%20by%20Side%20Comparison%20%E2%80%93%20H.938%20vs.%20Gov%20Rec~4-9-2026.pdf?utm_source=chatgpt.com "H.938"
[3]: https://www.healthvermont.gov/sites/default/files/document/VPCH%20financial%20template%202024.pdf?utm_source=chatgpt.com "Vermont Psychiatric Care Hospital"
[4]: https://www.wcax.com/2025/11/20/vermont-prison-population-reaches-highest-level-since-2019/?utm_source=chatgpt.com "Vermont prison population reaches highest level since 2019"
During the pandemic, the state worked to release as many inmates as possible. Today, the state’s six prisons are up to 1,632 people.
