The shortage in numbers

Homes short. The state's 2023 housing study found that Maine needs about 38,500 homes to make up for past underbuilding and 37,900 to 45,800 more to keep up with growth through 2030: 8,500 to 9,300 a year, when Maine was permitting about 4,800.¹

Homes being built. Census data show 6,979 units permitted in 2025, up from 4,760 in 2019; per resident, that is the most in New England and above the national rate.^2,3 The state's town survey, which also counts accessory apartments (ADUs), found 7,499, including 516 ADUs.⁴ By the Census count Maine is still 1,500 to 2,300 homes a year short of the need, and permits are not completions: Portland approved a record 1,420 units in 2025, but only 187 were completed.⁵

Prices and rents. The median single-family sale price rose 80 percent, from $225,000 in 2019 to $405,000 in 2025, and reached $424,550 in August 2026.^6,7,8 Median household income rose 30 percent (to $76,442) and median gross rent 39 percent (to $1,210) from 2019 to 2024.^9,10 HUD's two-bedroom fair market rent rose about 70 percent from 2019 to 2027, to $2,353 in the Portland area and $1,722 in the Bangor area.¹¹ In 2024 the median household earned only 61 percent of the income needed to buy the median-priced home.¹² Only 3.6 percent of rental units were vacant and for rent.^9,10

Who is squeezed. Of Maine's roughly 164,000 renter households, 50 percent paid 30 percent or more of their income for rent in 2024 (44 percent in 2019), and 23 percent paid half or more (our calculation from Census tables).^9,10 The January 2026 count found 2,126 people homeless, 361 of them outdoors, a record, and 663 chronically homeless, up from 226 in 2019.¹³ Landlords filed 4,385 eviction cases in 2025.¹⁴

Why owners and builders aren't adding enough homes

Land and zoning. The state's consultant found that Maine towns have “traditionally had highly restrictive development approval processes” and that the 2022 zoning law has “limited enforcement mechanisms outside of litigation.”¹⁵ LD 2003 (2022) required towns to allow two to four homes per residential lot and an ADU; LD 1829 (2025) allowed three per lot (four in growth areas), capped lot sizes where there is sewer, and barred planning board review for four or fewer units.^16,17 In April 2026, LD 2173 paused LD 1829 and the 2025 parking cap until July 1, 2027.^18,19,20 Permits for units in two- to four-unit buildings rose only from 225 in 2019 to 341 in 2025.^2,3

Construction cost. A new apartment now costs roughly $340,000 to $375,000 per unit. MaineHousing's average cost per affordable unit rose from $153,415 in 2019 to $348,145 in 2025,^21,12 and a 204-unit market-rate project in Westbrook will cost close to $70 million, about $343,000 per unit.^22,23 The 2021 energy code, in force since April 2025, adds about $2,000 to $3,400 per apartment compared with the 2009 code, by a federal laboratory's estimate, against energy savings of about $978 a year for a house.^24,25

Financing. At today's interest rates, rents do not cover those costs without help. Portland's consultant found that construction costs rose from $248 a square foot in 2021 to $392 in 2025, and that under the city's 2020 inclusionary rule only 8.6 percent of approved units were completed within 62 months, against 17.8 percent under its 2015 policy.^26,27 Our example in proposal 3 finds a gap of about $69,000 per unit in Greater Portland and $122,000 in a Bangor-type market (our estimate).

Why subsidized homes cost more. Public money comes with rules, and the rules cost money. In California, prevailing wage requirements added about 13 percent to the cost of tax credit projects, and each additional source of funding added about $6,400 a unit.²⁸ Rhode Island's public expenditure council found subsidized projects cost 49.5 percent more per unit than a small sample of new private homes for sale, and blamed goals unrelated to housing, such as green certifications, labor requirements and community benefit agreements.²⁹ Bangor's Yellen Pines shows where the money goes. It cost about $11.2 million for 30 studio apartments of about 400 square feet, or $373,000 each. Its construction ran about $394 a square foot, close to Portland's market average, but the apartments fill only about 63 percent of the building, and everything other than construction, from land and site work to financing, reserves and developer fees, took a third of the budget (our arithmetic from the city's figures).^{30,31,32,33,26}

Property taxes. Maine's own tax code treats 15 percent of a tenant's rent as property tax.³⁴ Statewide, property tax levies rose 7.2 percent in 2023 and 7.8 percent in 2024, the fastest increases since at least 2013, and from 2021 to 2024 they rose 16 to 34 percent in Augusta, Bangor, Lewiston, Portland, South Portland and Westbrook.^35,36 Revaluations after years of stale values add sharp jumps for individual owners (proposal 8).

The risk of a placement that fails. A national study of low-income buildings estimates that filing an eviction costs a landlord the equivalent of 2 to 3 months of rent; evicted tenants owed about 3.7 months of rent at move-out, and units then sat empty about 2.4 months.³⁷ At Maine's median rent that is roughly $10,000 to $11,000 before damage (our estimate: $1,210 times 8.1 to 9.1 months). When a voucher tenancy fails, HUD's contract bars the housing authority from paying the landlord's other claims.³⁸

Rent control uncertainty. Portland's ordinance, adopted by voters in 2020, covers new buildings once the first rent is set and limits annual increases to a percentage tied to inflation (2.6 percent for 2027), with a 10 percent ceiling.^39,40 A 2025 bill titled “An Act Enabling Municipalities to Protect Tenants and Stabilize Rents” (LD 1534) died, but the question remains open in every other town.^41,42 Investors price that risk in.

1. Let owners and builders add homes by right

How it would work. A small project that meets clear, published standards should get a staff permit on a fixed timeline, without a discretionary hearing. The bill would:

1. Keep July 1, 2027 as the firm start date for LD 1829 and the parking cap.
2. Define “by right”: up to four homes on a residential lot, an ADU, or conversion of an existing building to up to four homes, meeting objective standards, gets a staff decision within 45 days, or the permit is deemed approved.
3. Extend by-right approval to buildings of 5 to 12 units in designated growth areas with public water and sewer.
4. Let the Maine Office of Community Affairs (MOCA) review local ordinances and refer violations to the Attorney General, and let an applicant who wins an appeal of an unlawful denial recover attorney fees. California gave its housing department this power over ADU rules in 2024.⁴³
5. Adopt two code changes the state's consultant recommended: sprinklers only in newly added units when a two-unit building gains a third, and single-stair buildings up to six stories with suitable fire protection. Tie state housing and infrastructure grants to compliance, as it also recommended.¹⁵

The numbers. MOCA already offers towns technical help and grants.⁴⁴ An enforcement unit of four staff would cost about $500,000 a year (our estimate: four positions at about $125,000 each with benefits). After California made ADUs a right, its ADU permits rose from 1,336 in 2016 to 26,924 in 2023, more than 21 percent of homes permitted there.⁴³ Maine's 516 ADUs were 7 percent of its 2025 permits.⁴ Raising that share to 14 to 21 percent would add about 500 to 1,050 homes a year (our estimate: 7 to 14 percentage points times 7,499). With duplexes, triplexes and small apartment buildings, 1,000 to 2,000 more homes a year within three to five years is a reasonable target. Auckland, New Zealand, which upzoned three quarters of its residential land in 2016, permitted about 21,800 more homes over five years, about 4 percent of its housing stock; narrower reforms elsewhere showed little short-run effect.⁴⁵

Legal basis and risks. Maine towns may exercise powers “not denied either expressly or by clear implication” (30-A M.R.S. §3001), so the Legislature can set these rules by statute, as it did in LD 2003 and LD 1829.⁴⁶ Neighbors lose a hearing on small projects, and water, sewer and school capacity must keep up; rural lots without sewer still face septic lot-size rules.

2. A statewide ban on local rent control

How it would work. Amend Title 30-A to deny municipalities the power to limit the amount of rent for residential units. Limit the ban to rent amounts, so local rules on notice, deposits and discrimination stay in place. Apply it at once to new buildings and vacant units, and phase out existing limits (in practice, Portland's) over two to three years. Keep the state's 75-day notice for increases of 10 percent or more.⁴⁷ Leave manufactured housing lot rents to the 2025 state law, which requires 90 days' notice and mediation for increases above inflation plus 1 percent, because owners of those homes cannot easily move them.⁴⁸

The evidence. When the University of Chicago's panel of leading economists was asked in 2012 whether local rent limits had improved the amount and quality of affordable rental housing, only one agreed.⁴⁹ In San Francisco, a 1994 expansion of rent control helped covered tenants stay in their homes, a real benefit, but owners of covered buildings cut the rental housing they offered by 15 percent; the authors' working paper estimated that this raised market rents citywide by 5.1 percent.^50,51 After St. Paul, Minnesota adopted a strict 3 percent cap in 2021, its permits for new homes fell 82 percent in the next six months from a year earlier while Minneapolis's rose, and the city exempted new construction the next year. Permits in both cities fell later as interest rates rose; Minneapolis Federal Reserve researchers note they lack hard data on the cause, though developers blamed the ordinance.^52,53,54 Economists studying St. Paul found the cap cut property values 4.0 to 5.5 percent and that higher-income renters gained more than lower-income renters.⁵⁵ No independent study has measured Portland's ordinance; a Realtor-funded study estimated it shrank the taxable base 3.2 to 5.4 percent, which a Rutgers economist called correlation, not causation.⁵⁶

Other states. Industry counts put the number of states that bar local rent control at 33 to 36; the Urban Institute counted 31 as of 2020.^57,58,59 Washington and Oregon bar local ordinances but cap rents statewide, and both exempt newer buildings (for 12 and 15 years).^60,61,62,63 Maine's last attempt at a ban, LD 522 in 2019, died when the House accepted an “ought not to pass” report, 81 to 50.⁶⁴

Costs, legal basis and risks. A statute is enough; no constitutional amendment is needed.⁴⁶ The ban costs the state nothing and would not add homes overnight, but it removes a risk priced into every rental project. It also overrides Portland voters, who adopted rent control in 2020 and in 2023 rejected an exemption for small landlords, 66 to 34 percent, and the tenants of about 11,000 covered units would face market increases as limits end.^65,66 The phase-out and the Property Tax Fairness Credit, which renters can claim and which rises to $1,500 for people under 65 from tax year 2026, soften that but do not remove it.⁶⁷

3. Tax increment financing that makes rental projects pencil out

What Maine law already allows. A TIF lets a town dedicate some or all of the new property taxes a project generates to the project's costs. Maine's affordable housing TIF (30-A M.R.S. §§5245 to 5250-G) is approved by MaineHousing; at least one-third of the district's units must serve households at or below 120 percent of area median income, rental units must stay affordable 30 years, and the increment may pay operating costs, expressly including “real estate taxes,” as well as support services and child care. A 2025 amendment lets the program run up to 30 years after the certificate of occupancy.^68,69 Since 2021 ordinary municipal TIFs may also pay “costs associated with the development and operation of housing,” market-rate or not.⁷⁰ Towns usually return the increment through a credit enhancement agreement (CEA). Portland allows up to an average of 75 percent for up to 30 years for affordable housing and 65 percent for up to 20 years otherwise, requires proof that help is needed, and adds its own conditions, including state prevailing wages and its green building code.⁷¹ Its 2023 agreement for 54 affordable units at 25 Casco Street returns 75 percent.⁷² Bangor approved one in August 2026 for Yellen Pines: 30 supportive efficiency apartments costing about $11.2 million, or $373,000 each, more than the roughly $343,000 per apartment of a 204-unit luxury complex under construction in Westbrook, whose apartments average 1,000 square feet.^30,31,22,23 A TIF should reward building for less, not cover the cost of building for more.

A key feature is often overlooked: captured value in a housing TIF is left out of the town's state valuation, which sets its school aid, revenue sharing and county tax.⁷³ Without that shelter, new value raises a town's required local school share by 5.645 mills in 2026-27, and a town that receives state aid loses that much aid.⁷⁴ Bangor gets about 61 percent of its school formula from the state; Portland about 12 percent.⁷⁵ Because state aid is a fixed pot, the shelter moves some aid away from other towns.⁷⁶

The proposal. (1) MaineHousing and DECD publish a model housing TIF and CEA; a town that adopts it for a project meeting the by-right standards in proposal 1 gets state approval within 30 days. (2) Default terms: up to 75 percent of new taxes for up to 30 years where at least a third of units are income-restricted, and up to 65 percent for up to 20 years otherwise, with an independent review of the financing gap for any agreement above 50 percent. (3) The state template adds no cost mandates; a town that adds them publishes their cost per unit. (4) The town captures the whole increment, returns 75 percent and puts 25 percent in a local housing fund, which shelters all of the new value.

A worked example. A new 48-unit building of two-bedroom apartments, in two markets. Inputs are assumptions drawn from the Maine data above; results are our estimates.

Arithmetic (Portland): $2,450 times 12 is $29,400; less 5 percent vacancy and $6,500 of costs leaves $21,430. Dividing by the cap rate plus the tax rate (7.758 percent) gives a value of about $276,000 and a tax of $3,475; after-tax income of $17,955 at 6.5 percent gives the same value, $69,000 below cost. Seventy-five percent of the tax is $2,606 a year, worth $35,900 over 30 years at 6 percent (annuity factor 13.76). A TIF on tax-credit buildings is worth less because they are assessed lower: about $18,000 per unit at Portland's 25 Casco Street, by our calculation from the city's figures.⁷²

So a standard TIF closes about half the gap in Greater Portland and under a third in a lower-rent market. Stacked with the workforce credit (proposal 4) and the savings from by-right approval, it can carry a marginal project.

What it costs a town, and why the town still gains. The Portland-type building pays about $166,800 a year in new tax. The town returns $125,100, keeps $41,700 for its housing fund, and, because all the new value is sheltered, loses no school aid; without the shelter an aid-receiving town could lose up to about $74,800 a year ($13.3 million of value times 5.645 per $1,000; about $53,500 in the Bangor case). If the building would not be built without the TIF, the town gains $41,700 a year and 48 homes, though new residents also use services, including schools. If it would have been built anyway, the town is about $50,000 a year worse off: without the TIF it would net about $92,000 ($166,800 less $74,800 of lost aid), against the $41,700 it keeps. That is why the gap review matters.

Risks. A TIF is a subsidy, binds future councils for decades, and shifts some school aid from other towns. The gap review, the public hearing the statute requires, and annual reports certifying that units remain affordable are the safeguards.^68,69

4. Tax credits for low-income and workforce rentals

What Maine has now. The Maine Affordable Housing Tax Credit (36 M.R.S. §5219-WW) is paired with the federal 4 percent Low-Income Housing Tax Credit; where at least 60 percent of credit units serve households at or below 50 percent of median income, the state credit can equal the federal one. It is refundable and claimed once the building is finished, capped at $10 million a year (up to $15 million with carryforward), with 45-year affordability. In April 2026 the Legislature extended it to 2036 without raising the cap.^77,78,79

The Legislature's watchdog office, OPEGA, reported in February 2026 that the credit had supported 429 completed units at about $72,000 of state credit per unit, in projects costing about $375,000 per unit, and 2026 awards averaged about $100,000 per unit. MaineHousing says subsidy, not bonds or federal credits, is its binding constraint. OPEGA's full evaluation is due in 2027; Maine Revenue Services puts the revenue loss at $14.25 million a year.^80,81,82 A 2025 federal law halved the bond financing a project needs to get the 4 percent credit, so more projects can qualify if gap money exists.⁸³

The proposal. Raise the cap to $25 million a year for 2027 through 2036, with a ceiling of $100,000 of state credit per unit, indexed to construction costs, and require OPEGA's evaluation to report cost per unit and whether projects would have been built anyway. The added $15 million a year would support about 150 to 210 more affordable apartments a year at $72,000 to $100,000 per unit (our estimate). The risk is rising per-unit costs, which the ceiling and review address.

A workforce credit: a pilot, not a broad program. Maine's rural rental program reaches 80 percent of median income at about $171,500 of subsidy per unit,⁸⁴ and the Middle-income Housing Support Program created in April 2026 got a $500 base allocation and no appropriation.^67,85 The numbers support only a narrow credit. In the Portland area the maximum two-bedroom rent is $2,504 at 80 percent of median income and about $3,130 and $3,756 at 100 and 120 percent, while new suburban apartments rent for $2,600 to $2,800.^22,23 A cap at 100 or 120 percent there would buy no affordability. In rural counties market rents are often below the 80 percent limit, so the problem is that rents do not cover construction costs at any income.

We propose a five-year, $10 million-a-year pilot: up to $100,000 per restricted unit at or below 80 percent of median income and up to $50,000 at 81 to 100 percent; at least 20 percent of a building restricted for 30 years; competitive allocation with per-unit cost caps; stackable with the housing TIF; and an OPEGA review after five years. That supports about 100 to 200 restricted units a year inside larger mixed-income buildings (our estimate: $10 million divided by $50,000 to $100,000). MassHousing's workforce loans (up to $100,000 per unit) had financed 1,308 workforce units by 2021, and Colorado's new middle-income credit awarded $4.8 million in 2025 for 220 units; neither has been rigorously evaluated.^86,87,88

What Maine can learn from Vienna

Vienna is often held up as the city that solved housing by partnering with builders. The record is more useful than the legend.

What Vienna does. About 42 percent of Vienna's households rent from the city (20 percent) or from limited-profit housing associations (22 percent).⁸⁹ Most new subsidized homes are built by those associations, which federal law holds to a return of 3.5 percent on equity, 5 percent at most, with rents set to cover costs and surpluses kept in housing.⁹⁰ Commercial developers can compete too, if they accept cost-based rents while the public money is in.^91,92 The city banks land and offers it in competitions where the land price is fixed in advance and juries pick the project on cost, quality, ecology and social mix.⁹¹ The winners get a public loan at 1 percent for up to 40 years, covering roughly a third of construction.⁹³

What it gets. For buildings built since 2011, for-profit rents in Vienna run 91 percent above rents in limited-profit buildings.⁹⁴ The difference comes mainly from capped returns, capped land prices and cheap public money, not from building more cheaply: subsidized projects cost about €1,930 a square meter to build in 2019, excluding land, and costs have climbed since.^95,94

What Maine shouldn't copy. Vienna pays for its loans partly with a payroll tax, raised this year to 1.5 percent of wages, split between employer and worker, and no longer reserved for housing.^96,94 Since 2019 it has required two-thirds of the floor space on large newly zoned sites to be subsidized, and capped what that land may cost.⁹⁷ It regulates rents on older private buildings, and since January a federal law has limited rent increases on most leases.^98,99 Most households qualify, and incomes aren't rechecked after move-in.^100,94 And it hasn't solved housing: the OECD finds that prices and rents have risen sharply over two decades, “particularly in Vienna.”⁹⁴ Even there, commercial builders put up 61 percent of new homes from 2018 to 2021, and about 70 percent of new building was unsubsidized.¹⁰¹

What Maine can borrow. One idea travels well: make builders compete for public help. When the state or a town offers land, a TIF or a tax credit, it should fix the terms in advance (the rent, the years, the return) and award the help to whoever delivers the most homes for the money, with a capped profit for as long as the public money is in. MaineHousing already caps developer fees on tax credit projects at 15 percent of development costs;¹⁰² cost per home should sit at the top of its scoring, and every TIF should face the gap review in proposal 3. Cheap public money is the other lever. Montgomery County, Maryland, citing Vienna, runs a $100 million revolving fund that lends low-cost construction financing to mixed-income projects with at least 30 percent of units income-restricted.^103,104 Its advantage is cheaper money, not cheaper building: its first project cost about $451,000 a unit, including retail and parking.¹⁰⁵ And subsidy can crowd out private building: careful U.S. studies find subsidized rental projects largely replace homes the market would have built, especially where demand is strong.^106,107,108 The best use of public help is the project that wouldn't happen without it.

5. A state guarantee for landlords who rent to program tenants

The problem. When a voucher tenancy fails, the landlord bears the loss: HUD's contract says the housing authority “shall not pay any other claim by the owner against the family.”³⁸ Maine bars refusing a tenant “primarily because of” receiving housing assistance, but the Law Court held in 2014 that a landlord may decline because of the voucher program's requirements, and bills to change that died in 2021 and 2026.^{109,110,111,112,113,114} In 2023, 30 percent of vouchers MaineHousing issued in towns without their own housing authority expired unused (45 percent in Waterville).¹¹⁵ MaineHousing's landlord incentive, including up to $1,500 for damage, ended on December 31, 2025,^116,117,118 and a 2024 guarantee bill (LD 1710) became a study and died.^119,120 The risk is smaller than many landlords fear: in a national study, voucher tenants paid 96.7 percent of rent due, against 86.2 percent for other tenants, and the authors found direct financial incentives more likely than procedural delays to prevent evictions durably.³⁷

How it would work. A Maine Rental Guarantee Fund at MaineHousing, modeled on Washington's program, would cover private-market tenancies of households with a Housing Choice Voucher or state rental assistance, including people leaving homelessness, for as long as the assistance lasts. It would pay the tenant's share of unpaid rent for up to three months, damage beyond normal wear after the deposit is applied, and up to $1,000 of legal costs, to a total of $5,000 per tenancy, for claims of at least $500. Washington caps claims at $5,000 per tenancy; Oregon now allows up to $20,000.^121,122 Landlords would need a signed move-in report with photos, would accept the hearing track in proposal 6, and, as in Washington, could not pursue the tenant for losses the fund pays. Claims would be paid within 60 days; Washington's underfunded fund now warns payments may take “multiple months,” which defeats the purpose.^123,124,125 Tenants could contest any claim, and the fund would not sue them or report them to credit bureaus.

The cost. Washington's core program has paid 9,726 claims totaling $26.2 million since 2018, an average of $2,694, and paid 1,502 claims in fiscal 2024, about 3 to 4 per 100 voucher households a year.^123,124,125 Oregon's program for high-barrier tenants drew requests on 13 of 226 guarantees (5.8 percent), and its main voucher guarantee paid $895,409 to 202 landlords in 2023-25.^126,127 Minnesota's Beyond Backgrounds paid 10 claims for more than 498 renters.¹²⁸ Maine has 9,216 tenant-based vouchers in use.¹²⁹ Adding about 1,800 other assisted tenancies (our assumption) gives about 11,000 covered tenancies:

Arithmetic: 11,000 tenancies times the claim rate times the average claim, plus 15 percent. We propose a standing appropriation from the HOME Fund, which paid for the incentive program, or the General Fund, plus a one-time $3 million reserve so claims are never delayed. A shelter bed costs about $37,300 a year,¹³⁰ so the central cost equals about 37 beds; helping 40 more households a year move from shelter to an apartment they keep would cover it.

Risks. Guarantees can invite padded claims (Washington flagged 5 of more than 1,600 early claims as possible fraud) and looser screening; the move-in report, minimum claim, hearing review and cap limit both.^123,124,125 A guarantee does not fix payment standards below market rents, and Maine's voucher programs were spending 106.4 percent of their federal budget authority in early 2025, so in the near term it mainly speeds lease-ups and keeps landlords in the program.¹¹⁸ Washington and Oregon paired their guarantees with laws barring landlords from refusing voucher holders; this plan relies on the incentive alone.^{109,110,111,112,113,114}

6. A fast, fair hearing track for program tenancies

What happens now. Maine's eviction law is faster on paper than many landlords assume. A landlord may serve a 7-day notice once rent is 7 days late; paying in full voids it. The summons must be served at least 14 days before the hearing; if the landlord wins, the writ issues 7 days after judgment, and the tenant has 48 hours after it is served. A tenant can still stop a nonpayment eviction by paying all arrears, current rent and the landlord's filing and service costs before the writ issues, and a General Assistance payment or written promise counts as cash. Either side may appeal to the Superior Court, including for a new jury trial.^47,131 Our legal research puts the earliest lawful removal at about 31 days after the notice (5 to 8 weeks is typical when uncontested, our estimate); for voucher tenancies, the federal CARES Act's 30-day notice, which most courts treat as still in force, pushes the minimum to about 54 days.^132,133

Other problems matter more than the day count. An eviction case cannot include a claim for back rent, so the landlord must sue again.^47,131 In a review of 2,300 Maine files, landlords had lawyers in 81 percent of cases and tenants in 20 percent, and represented tenants were far more likely to avoid an eviction judgment.¹³⁴ The filing fee is only $100; the real costs are lost rent and turnover.¹³⁵

Can tenants agree in advance to an arbitrator or hearing officer? Not safely under current law. It is an unfair trade practice for a landlord to require a lease term that waives a tenant's eviction-law rights, and evictions “without resort to” that law are illegal; the statute allowing advance agreement on an arbitrator covers only commercial leases. On July 28, 2026, the Law Court expressly left open whether a residential lease's arbitration clause is enforceable. For voucher tenancies, federal rules require that an owner evict only “by instituting a court action.”^{136,137,138,139} Washington bans naming an arbitrator in a rental agreement.¹⁴⁰ A federal court in California has held that the Federal Arbitration Act overrides a state ban on such clauses,^{136,137,138,139} so a Maine lease clause would invite years of litigation either way.

How it would work. The goal, a quick hearing before a neutral officer with the guarantee covering the landlord's losses, can be met inside the courts, where it will hold up. A new section of the eviction chapter would create a program tenancy docket for tenancies covered by the guarantee:

1. Guaranteed scheduling on the first hearing date, with no delay for docket congestion.
2. Cure first: the guarantee fund may pay a tenant's arrears once in 12 months, with the same legal effect as a General Assistance payment, and the tenant repays the fund on an affordable plan.
3. A court magistrate (a new position like Maine's family law magistrates) or a referee the parties choose after the dispute arises, which court rules already allow, hears the case, and a judge reviews objections within days.^131,141,142
4. One hearing for possession and money: the magistrate also decides unpaid rent and damages, as commercial eviction cases may, and the fund pays the landlord within 30 days, up to its caps.^47,131
5. Due process built in: written notice of the grounds and amount; access to documents; a lawyer or trained advocate paid by the program; a hearing with evidence and cross-examination; all existing defenses (retaliation, habitability, disability accommodation, domestic violence protections); a written decision; and the existing appeal, with rent paid into escrow rather than a bond.^{143,144,145,146}
6. Federal rules for voucher tenancies continue to apply.^132,133

The numbers. If 3 to 5 percent of about 11,000 covered tenancies reach a filing each year, somewhat above the 2.7 percent of all Maine renter households who faced one in 2025 (4,385 filings for about 164,000 households), the docket would see about 330 to 550 cases a year (our estimate).^14,9,10 At about $400 per case for the magistrate or referee and $1,000 for tenant counsel, the cost is about $460,000 to $770,000 a year (our estimate: cases times $1,400). Payments to landlords are counted in proposal 5.

What the evidence says. New Hampshire's court diversion program reached agreement in 71 percent of mediations, and 63 percent of those agreements let the tenant stay. In Philadelphia's 2021 program, 7 percent of cases that received rental assistance had an eviction filing within six months, against 42 percent without, but 33 percent of assisted cases had one within 21 months, so assistance sometimes delayed eviction rather than prevented it. Cleveland's housing court uses magistrates and, with a right to counsel, reports that 83 percent of clients seeking to avoid eviction or an involuntary move since 2020 succeeded. North Carolina's magistrates hear evictions with a new trial before a judge on appeal, the same two-step structure Maine's courts have approved.^{147,148,149,150,151} Money to cure arrears does more than procedure to keep tenancies alive, but a one-time payment does not fix a rent the household cannot afford.

Risks. Faster hearings raise the cost of mistakes, which is why counsel and the cure payment are part of the design. Magistrates require legislation and money. Cure repayment must be affordable and never reported to credit bureaus while the tenant stays housed.

7. Put families, workers and real treatment first

What Maine prioritizes now. MaineHousing, which serves more than 3,800 voucher households a month, sets aside 60 percent of its undedicated voucher funding for homeless households referred through coordinated entry; its waiting-list preferences put elderly, disabled and two-or-more-person households, including parents with children, in the same group.^152,153 Coordinated entry scores people mainly on how long they have been homeless, with added points for sleeping outside or fleeing violence and a few for barriers such as a disabling condition; it gives no points for a diagnosis. Scores of 12 to 16 (of 16) go to permanent supportive housing, which federal rules limit to households with a disability; lower scores go to rapid rehousing, MaineHousing's STEP program or vouchers.^154,155 DHHS's supportive housing vouchers go to people with a serious mental illness, a substance use disorder or HIV, chronically homeless people first,¹⁵⁶ and the state's Home for Good program builds apartments with around-the-clock voluntary services for people homeless a year or more with a disabling condition.¹⁵⁷ In 2026 the Maine Continuum of Care began requiring participation in supportive services in the projects it funds, within federal limits.¹⁵⁸

Federal policy has shifted. HUD withdrew its “chronically homeless first” guidance on November 21, 2025,^159,160 and its coordinated entry rules already let communities weight families with young children and pregnant women higher, an option Maine's tool does not use.¹⁶¹ A July 2025 executive order directs agencies to end support for “housing first” policies that “deprioritize accountability and fail to promote treatment, recovery, and self-sufficiency.”¹⁶² In January 2025, about 34 percent of homeless adults reported a serious mental illness and 21 percent a substance use disorder; at least 45 percent reported neither.¹⁶³

What the evidence says. In Canada's five-city randomized trial, Housing First participants with serious mental illness were stably housed 71 percent of the time, against 29 percent with usual services, but gains in quality of life faded by the second year.^164,165 Assertive Community Treatment (ACT) teams reduced homelessness 37 percent more than standard case management; overdose deaths run about five times higher among people out of methadone treatment than in it; contingency management, small rewards for negative drug tests, was one of the few approaches that clearly reduced drug use among homeless adults in a 2024 review; and supported employment was about 2.5 times as effective as pre-job training.^{166,167,168,169} For homeless families, a long-term rent subsidy reduced homelessness and food insecurity at a cost about 9 percent above usual care, while short-term help and transitional housing had little effect.¹⁷⁰

The proposal. Organize state-funded housing help into three tracks, each with access to the landlord guarantee and hearing track:

1. Families with children: weight families with young children and pregnant women higher in coordinated entry, as HUD allows, and give homeless families and families facing eviction first priority for state rental assistance and the guarantee, with a long-term voucher where needed.
2. Work and housing: priority for adults who are working or in a job program or supported employment, with up to 24 months of rent help on the STEP model and a path to a regular voucher.¹⁷¹
3. Treatment and housing: fund ACT teams, addiction medication, contingency management and certified recovery residences alongside the apartment, and keep a Housing First option for people who decline treatment.

The numbers. Reordering priorities costs little. Treatment costs money: 300 more ACT slots at about $14,000 a year each (Washington State's 2013 cost benchmark, likely higher now) would cost about $4.2 million; ACT is a MaineCare service, and at the 60.62 percent federal match for October 2026 to September 2027, the state share is about $1.7 million (our estimate).^172,173

Legal limits and risks. A voucher preference for working families must extend to elderly and disabled applicants, and preferences for a specific disability are barred.¹⁷⁴ Projects funded by HUD's Continuum of Care program may not require mental health treatment as a condition of housing, except projects whose purpose is substance use treatment.¹⁷⁵ So treatment should be required only in housing built around treatment or recovery and offered everywhere else. A preference for families with children is permitted. Moving the longest-homeless people down the list could leave the sickest outside longer, which is why the treatment track must be funded and results published for each track.

8. Predictable property taxes for rental owners

The problem. Our companion paper on property taxes finds that in Maine, budgets, not values, set the total levy, while revaluations decide who pays it. Statewide levies rose 24.6 percent from 2019 to 2024, a little faster than inflation (22.7 percent), including 7.2 and 7.8 percent in 2023 and 2024.^35,36 Maine's municipal levy limit (LD 1) has been repealed,¹⁷⁶ and the sharpest jumps for individual owners come from revaluations after years of stale values. Portland's 2025 revaluation raised residential values an average of 43 percent and commercial values 19 percent. Augusta's 2026 revaluation, its first since 2006, followed years in which homes were assessed near 43 percent of market value and commercial property near 77 percent. Lewiston, in its first full revaluation since 1988, expects residential assessments to rise about 180 percent when new values take effect for 2027-28, with the rate falling from $32.78 to about $16.90 per $1,000.^{177,178,179,180} Owners of older rental buildings in these cities face the same jumps as homeowners. How much reaches tenants is debated: a study of New York data from 1975 to 1994 found owners passed on about 14 percent of tax increases, while recent studies found 50 to 89 cents per dollar passed to new tenants in Berkeley and 83 percent over three to four years in Germany.^181,182,183 The fair reading is that owners pass on a substantial share over time, especially in tight markets. In Portland, rent control allowed an automatic tax pass-through until voters removed it in 2022.¹⁸⁴

What will not work. Maine's Constitution requires property taxes to be “apportioned and assessed equally according to the just value thereof.”¹⁴⁶ A cap or freeze for rental property would need a constitutional amendment, and caps shift taxes rather than cut them. Owner-occupied homes paid about $1.59 billion of the $3.17 billion levied in 2023; if a cap held their assessments 20 percent below market and budgets did not change, the tax rate on everything else, rental housing included, would rise about 11 percent (our arithmetic).^185,35 Maine's senior property tax freeze, created in 2022, was ended after the tax year beginning April 1, 2023.¹⁸⁶

How it would work. Four steps within the Constitution:

1. Keep values current: require assessment updates at least every five years, with annual market adjustments between, and fund state help for towns that fall behind. Stale values turn ordinary market changes into 40 to 180 percent jumps.
2. Spread the shock, not the assessment: when a revaluation raises the tax on a rental property or home by more than 10 percent in a year, let the owner pay the excess over three years, with interest at the town's cost of funds and a lien until paid. Values stay at just value, and the town is made whole. Lewiston's assessor rejected a three-year phase-in of values as confusing and costly;^177,178,179 a payment plan avoids both problems.
3. Multi-year agreements for new rental buildings through the housing TIF in proposal 3, so an owner can count on the same share of taxes coming back for up to 30 years.^68,69
4. No shifting onto rentals: avoid homestead-only caps without replacement revenue, and publish each new exemption's effect on rental property tax rates.

The numbers. The payment plan costs the state little if towns charge interest. Revaluation help is modest: Lewiston's citywide revaluation contract was $574,000,¹⁸⁰ so $2 million to $3 million a year could fund several city revaluations or many small-town updates (our estimate). At Maine's 15 percent convention, a 20 percent jump in a building's tax implies about a 3 percent rent increase; spread over three years, about 1 percent a year (our arithmetic).³⁴ The risk is delayed town revenue if many owners use the plan, which is why it carries interest.

What it costs and what it builds

All cost and output figures below are our estimates, built from the assumptions shown in each section.

At central estimates the added state cost is about $32 million a year, mostly the two tax credits, plus the one-time guarantee reserve. By our estimates, the zoning and credit proposals could add about 1,250 to 2,400 homes a year within five years, plus what the TIF makes feasible, roughly the size of the 2025 shortfall against the state study's target (1,500 to 2,300 homes), but only if the Legislature stops delaying the zoning reforms it has already passed. For scale, General Assistance, the town-run aid of last resort, cost $26.2 million in fiscal year 2025.¹⁸⁷

What we do not recommend. The evidence does not support several popular versions of these ideas:

• Arbitration clauses in residential leases: they risk liability under Maine law and conflict with federal rules for voucher tenancies; the fast track belongs in the courts (proposal 6).
• A workforce credit reaching 120 percent of median income in Greater Portland: rent limits at that level sit above new-building rents, so the state would pay for affordability it does not get (proposal 4).
• A cap or freeze on assessments for rental buildings: it needs a constitutional amendment and shifts taxes to other owners (proposal 8).
• Making HUD-funded supportive housing conditional on mental health treatment: federal rules forbid it outside treatment programs; fund and offer treatment instead (proposal 7).
• Relying on the rent control ban or the fast track to add homes: both reduce risk, but neither builds housing by itself.