Homelessness is not one problem.

Each session brings proposals for a solution to homelessness, stated in the singular, as though a single condition awaited a single remedy. The evidence does not support that framing. What we call homelessness comprises at least four distinct populations with distinct causes, and policy that treats them as one will succeed with some and fail the rest.

Consider what the data establish.

The 2026 Point-in-Time Count for Oʻahu, conducted in January and released in May by Partners In Care,¹ identified 4,539 people experiencing homelessness on a single night, approximately 1% above the 4,487 counted in 2024.² That stability conceals two substantial and opposing movements. Unsheltered homelessness declined by roughly 20%, from 2,756 to 2,208. Sheltered homelessness rose by roughly 35%, from 1,731 to 2,331.²

This represents movement indoors rather than movement into housing.³ It is a genuine achievement, and it should not be mistaken for the resolution of the underlying condition.

Two findings warrant particular attention. People in households with adults and children increased 19% in two years, reaching 1,334,² and family homelessness on this island has risen 43% since 2022,⁴ a period corresponding to the withdrawal of pandemic-era federal support. This year’s count included 802 children and youth under eighteen.⁵ Separately, Native Hawaiians and Pacific Islanders constituted 43% of the counted population while representing approximately 10% of Oʻahu’s residents.²

On the neighbor islands, Bridging the Gap counted 1,863 people, a 2% overall decline accompanied by an increase in unsheltered homelessness.⁶

Earlier statewide surveys found that approximately 62% of those counted reported at least one disabling condition and roughly 29% reported a mental illness.⁷ The corollary deserves equal emphasis: for most people in these counts, the precipitating condition is the cost of housing.

Four populations, four obligations

Kānaka Maoli: a question of land, not of services

Native Hawaiians and Pacific Islanders account for 43% of the count. Among unsheltered Native Hawaiians surveyed, approximately half reported eligibility for Hawaiian home lands,³ while the Department of Hawaiian Home Lands waitlist stands at 29,365 applicants,⁸ some of whom have waited more than fifty years and many of whom die before receiving an award.⁹

We are describing a population dispossessed of land in its own country, holding a legal entitlement under the Hawaiian Homes Commission Act that the State has administered slowly for a century. To respond with shelter beds and case management is to mischaracterize the obligation. This is not a gap in social services. It is an unmet legal and moral debt.

And the debt is not owed only by the State, nor only in the past tense.

A great many Hawaiian families hold live claims against private parties. Following the Māhele of 1848 and the Kuleana Act of 1850, land granted to native tenants passed down through generations, often without wills, deeds, or probate, until an ancestral parcel was held by dozens of descendants in undocumented fractional shares.¹⁰ Plantations, ranches, and land speculators used that condition against them. Dispossession was accomplished at scale through adverse possession claims brought in quiet title actions — a practice the Native Hawaiian Legal Corporation dates as peaking in volume in the 1970s and 1980s, and which has not stopped.¹¹

The mechanism is cost, not merit. Quiet title and partition suits run for months and often years, and require surveys and appraisals and expert testimony that ordinary families cannot finance.¹¹ Partition by sale then forces the parcel to auction, severing a family’s connection to ancestral land — the specific harm identified by the Center for Excellence in Native Hawaiian Law.¹² Our own Legislature has said as much: a 2021 Senate committee report found that the existing framework leaves kuleana owners vulnerable to dispossession through processes complex and costly enough to foreclose them from mounting a defense at all.¹³ Families do not lose these cases. They lose the ability to appear in them.

Then there is the government’s own hand. Kalama Valley in 1971 — residents resisting eviction after their land was condemned for residential and commercial development — is conventionally remembered as the beginning of the modern Hawaiian movement.¹⁴ It is less often remembered as what it was procedurally: a taking. And the case that made Hawaiʻi famous in American property law, Hawaii Housing Authority v. Midkiff, upheld a condemnation scheme that took title from lessors and transferred it to lessees in the name of breaking up a land oligopoly.¹⁵ That decision is usually taught as a progressive victory. Legal scholarship has since argued the opposite reading — that land reform predictably moved Indigenous land to relatively wealthy tenants, and that Midkiff belongs in a long history of taking Native land to accomplish progressive ends.¹⁶ Whichever reading persuades you, the point for our purposes is narrow: public power has been used to move Hawaiian land into other hands, and no process exists to look back at that.

So I want to name a fourth intervention alongside the three above, and I recognize it is the largest.

We need a land-justice process. Not a study. A standing body with the authority to examine claims — against private parties who quieted title against Hawaiian families, and against the State and counties that condemned Hawaiian land — and to move parcels back into Hawaiian hands where the record supports it.

The components are not exotic:

• A funded defense. Kuleana defendants should have counsel, surveys, and title research paid for, in the same way we fund indigent defense. Most of these losses are defaults, and a default is not an adjudication.
• Reform of the actions themselves. Restrict who may bring a quiet title action against kuleana land, restrict partition by sale where an ancestral interest is at stake, and toll the clock where heirs were never located.
• A trustee for unclaimed interests, so that a fractional share whose owner cannot be found is held for the ʻohana rather than swept to the plaintiff. The Legislature has considered making the Office of Hawaiian Affairs that trustee before.¹⁷
• A claims commission with a remedy in land. Money damages have their place, but the injury here is the loss of a specific place. Where the State holds suitable land, restitution should be measured in acres.
• A reacquisition fund, so that when an ancestral parcel comes to market, a public or community entity can buy it and convey it back rather than watch it go to a bidder from elsewhere.

I am aware how this sounds beside a discussion of shelter beds. But the connection is direct and it is the reason I put it here rather than in a separate piece. A family sleeping in a vehicle in Kalihi may hold an undivided interest in a parcel their great-grandmother was awarded, lost in a proceeding no one in the family could afford to attend. We are currently offering that family a case manager. The honest response is a lawyer, a surveyor, and the land.

The appropriate interventions are accordingly land-based: accelerated awards accompanied by the infrastructure that makes lots habitable, rather than leases on parcels lacking water, sewer, and roads;⁹ conveyance of suitable state lands; kauhale sited on Hawaiian land under beneficiary governance; and community land trusts that hold land outside the speculative market in perpetuity.

Women and children leaving violence: independence requires a lease and resources

Families are the fastest-growing group in the count, and a substantial share of women’s homelessness originates in flight from an abusive household. The prevailing response is congregate shelter subject to a time limit, after which a survivor is expected to have secured housing in one of the most expensive markets in the country.

Genuine independence requires options that do not depend on an abuser’s cooperation or an administrative timeline: rental assistance for which survivor status establishes eligibility, without a prior shelter-residency requirement; the legal right to terminate a lease or remove an abuser from one without damage to credit; address confidentiality; and flexible cash assistance, since the immediate obstacle to leaving is most often a deposit or a vehicle repair. Other jurisdictions have established survivor-specific housing programs and flexible funding mechanisms with demonstrated results.¹⁸ Hawaiʻi has not.

The organizations conducting this work here — the Domestic Violence Action Center, the Hawaiʻi State Coalition Against Domestic Violence, Child & Family Service — operate on grant cycles that expire. Intermittent funding is not a policy.

Severe mental illness and substance dependency: a question of appropriated resources

This population is the most visible in public discussion and the least adequately funded. Outreach alone cannot address untreated psychosis or opioid dependency. What is required is crisis stabilization capacity, detoxification beds, permanent supportive housing with clinical staffing, and a behavioral health workforce the State has not built and does not presently compensate at competitive rates.

The cost is substantial: tens of millions of dollars annually, on a recurring basis. The relevant comparison is not to zero. We currently absorb these costs through emergency medical response, hospital utilization, correctional facilities, and law enforcement time. We are paying already, through the least efficient mechanism available to us.

Households priced out of the market

This is the largest group: people living in vehicles, doubled up with relatives, or one financial disruption away from displacement. The remedies here are structural rather than clinical — housing supply, eviction prevention, and rents that bear some relationship to local earnings.

The international evidence

Finland is the most frequently cited example, and it merits that attention, provided the full record is presented.

Beginning in 2008, Finland’s Housing First model provided permanent apartments without sobriety or treatment preconditions. Long-term homelessness fell by approximately 68% through 2022, with roughly 80% of participants housed through the programme.¹⁹ The mechanism was not an improved shelter system but the conversion of shelters into permanent housing: Helsinki moved from nearly 600 hostel and shelter beds in 2008 to a single emergency service center with 52 beds by 2019,¹⁹ and reliance on temporary accommodation fell 76% between 2008 and 2017.²⁰ The State financed this at over €270 million between 2008 and 2019, cost-shared with municipalities.¹⁹ Finland became the only country in the European Union recording sustained reductions in homelessness.²¹

The subsequent record is equally instructive. Finnish homelessness rose approximately 20% in 2025, to roughly 4,579 people, following reductions in housing benefits and social supports.²² The Y-Foundation, the nonprofit housing provider central to the model, has stated directly that changes in the operating environment can reverse long-accumulated progress,²² and its chief executive has written candidly about both the model’s strength and the pressures now bearing on it.²³

The lesson is therefore more demanding than the slogan suggests. Sustained reductions require permanent housing, a substantial public housing stock, and a benefit system maintained through fiscal pressure. Vienna reflects the same principle from a different direction: a city in which a large share of residents occupy social or subsidized housing is a city whose private market cannot dictate terms to its people.

If Finland, with universal health coverage and a functioning income floor, can lose ground in a single year of retrenchment, we should not expect a pilot program to hold here.

An accounting of our own program

The kauhale initiative has produced real results, and those operating individual sites are doing difficult work under difficult conditions. The Hoʻokahi Leo kauhale on Middle Street recently opened its second phase, adding thirty units in this district,²⁴ and Alana Ola Pono is at full occupancy with a waitlist of approximately 400 people.²⁵ Demand is not the constraint on this program.

The State’s Act 309 report to the 2026 Legislature records 918 active beds across 24 sites as of November 2025, 721 residents, 2,128 people served since 2023, and 888 recorded exits: 40% to permanent housing, 32% to hospitals, treatment programs, or family, and 28% to literal homelessness.²⁶ The administration reports investment exceeding $100 million since 2023.²⁷

The State has not published a cost per outcome. The calculation below is therefore my own, and I present the method alongside the figures.

What the program costs

Measured against 359 permanent housing exits, $100 million yields approximately $278,000 per permanent exit. Excluding construction and measuring annual operating expenditure against the annual rate of permanent exits yields approximately $210,000. Counting all positive exits — permanent housing together with family, treatment, and medical destinations — yields approximately $157,000.²⁸

A range of $157,000 to $278,000 is not a matter of rounding. It reflects the absence of a defined success metric for a program of this scale.

I would advance a different figure as the more accurate one. The quarterly expenditure filing for October through December 2025 records approximately $8.1 million in reimbursements — $6.5 million operating and $1.6 million construction²⁵ — which annualizes to roughly $32 million and corresponds closely to the $33 million appropriated for kauhale statewide.²⁹ Distributed across 721 residents, this represents approximately $45,000 per resident-year, or roughly $123 per night, and approximately $98 per night for operations alone.²⁸

I offer that framing in preference to the more severe one, including to those who share my broader concerns. Act 309 defines a kauhale as affordable housing charging no more than 30% of area median income, with no predetermined length of stay.²⁶ A resident who remains for three years reflects the program operating as designed. Cost-per-exit analysis records that resident as expenditure without result, and thereby measures the program against an objective it did not adopt.

The 28% of exits returning to literal homelessness is nonetheless the program’s weakest reported outcome and warrants a substantive response. For comparison, the ʻOhana Zones programs, which rely on rental subsidies and scattered-site units rather than purpose-built villages, report 196 permanent housing exits out of 298 total, a rate of 66%.²⁶ The served populations differ materially, and I do not present these figures as equivalent. But a difference of that magnitude between two state-administered approaches is a question the Legislature should resolve before authorizing the next phase of development.

Why the calculation cannot presently be verified

Three structural features of the reporting prevent independent confirmation, and each can be corrected.

The annual performance report contains no expenditure data. It records beds, occupancy, length of stay, and exit destinations, and no dollar figures.²⁶ Expenditures appear in separate quarterly filings.²⁵ No requirement obliges anyone to reconcile the two documents.

The quarterly filings record payments rather than accrued costs. Reimbursements are booked in the quarter an invoice is approved and may cover service periods of widely varying length. The quarter examined here includes a $3 million predevelopment payment covering twenty-one months and a $1 million shelter payment covering eleven.²⁵ Adjusting for this back-billing suggests in-period expenditure closer to $2.2 million.²⁸ My annualized figure may therefore overstate the true rate — or understate it, once one accounts for projects presently carried by county, Medicaid, and philanthropic funds.²⁵

Kauhale and ʻOhana Zones expenditures are reported together, and several sites are co-funded. Aʻala Medical Respite is currently sustained by City funds.²⁵ State contract expenditure is consequently not equivalent to program cost.

I do not conclude from this that the kauhale initiative has failed. I conclude that after an investment exceeding $100 million, its cost per outcome should not require reconstruction from two documents never intended to be read together. Members appropriating these funds should be able to see cost per occupied bed-night and cost per permanent exit, by project, in a single place.

Legislative priorities

A dedicated behavioral health trust fund with an identified revenue source, appropriated in multi-year increments, supporting crisis stabilization, detoxification, mobile crisis response, and clinical staffing within supportive housing — structured to survive the annual appropriations cycle.

Housing First established in statute, prohibiting sobriety or treatment compliance as a condition of publicly funded housing, with a scheduled conversion of emergency shelter appropriations into permanent supportive housing. This is the mechanism that produced Finland’s results, as distinct from the principle alone.

A survivor housing measure providing rental assistance and flexible cash support to those fleeing domestic and sexual violence without a shelter-residency precondition, together with lease bifurcation, penalty-free early termination, and confidentiality protections.

Land justice as housing policy: accelerated DHHL awards against a published delivery schedule, conveyance of state lands, beneficiary-governed kauhale, and a right of first refusal for community land trusts — together with the land-justice process set out above: funded defense in quiet title actions, reform of quiet title and partition as applied to kuleana land, a trustee for unclaimed interests, a claims commission empowered to remedy in land, and a fund to reacquire ancestral parcels when they reach the market.

Reporting sufficient to permit oversight: a statewide by-name list and public dashboard; and a statutory reporting mandate requiring total expenditure by project with all funding sources identified, cost per occupied bed-night, cost per permanent exit with the denominator defined in statute, returns to homelessness at six, twelve, and twenty-four months, and expenditures reported on an accrual basis matched to service period rather than payment date. The final requirement is the one that makes the others verifiable.

Kalihi carries more of this than most of this island, and it also holds a considerable share of those doing the work — Kōkua Kalihi Valley, the churches, and the neighbors who feed people without waiting for a program to authorize it. What they lack is funding and legal authority. Supplying both is the responsibility of this office.