Locked Twice

The proposition, stated first

Most of what the Commission will hear this summer is an argument about how to divide $54.5 million. This paper makes that argument too, because the division is genuinely inequitable — fewer than one acre in five draws more than half the money.

But the Commission should know at the outset that the argument has a ceiling, and the ceiling is low. The Department of Revenue's own estimate puts the entire aggregate shortfall at $4.5–5.5 million. Every dollar could be appropriated tomorrow and the disparity would remain, because a proportional increase applied to a disproportionate base preserves the disproportion exactly.

So this paper's principal recommendation is not about the formula at all.

Massachusetts holds roughly 461,000 acres of PILOT-eligible land, much of it in towns with almost no other path to tax base growth. Some fraction of the perimeter of that land is already-disturbed ground fronting existing public roads — old field edges, former landings, roadside strips, parcels acquired incidentally as part of larger tracts. None of it is ecological core.

Four things make this work, and all four already exist:

• Proposition 2½ makes it permanent. New growth is added to the levy limit and stays there. A PILOT dollar is an appropriation subject to next year's politics; a converted acre is capacity the town owns outright and compounds on.
• BioMap already draws the line. The Natural Heritage program's Core Habitat and Critical Natural Landscape layers, with connectivity modeling, exist and are currently linked to no disposal pathway at all.
• The conservation literature already sets the buffer. The Environmental Law Institute's synthesis of more than 1,400 studies recommends 230–300 meters from habitat edges.
• Other states already run the swap. New York has amended its constitution roughly sixteen times to execute equal-or-greater-value Forest Preserve exchanges. Every one ran net-additive to conservation.

No state has assembled these into a single pathway. The pieces sit in New York, Maine, New Jersey, Florida, and Massachusetts, unconnected. The Commission has the opportunity to recommend a first-in-nation framework in which edge release funds core consolidation — peripheral acres out, higher-value core acres in, under the existing Article 97 two-thirds vote and the Public Lands Preservation Act's no-net-loss test.

The obstacle is not cost, and it is not ecology. It is that Article 97 currently does double duty against the very communities this Commission was convened to examine — holding their land off the tax rolls, and simultaneously holding it out of the Commonwealth's flagship land-for-housing program.

A note on the scope of this paper

The comparative section rests on nine programs verified against current primary sources — New York, Minnesota, Michigan, Wisconsin, Pennsylvania, Connecticut, Vermont, Maine, and the federal PILT program — not on fifty.

Approximately 28 states could not be re-verified. The most complete existing fifty-state inventory is a 1996 New York Department of Taxation and Finance survey built on 1990 and 1994 data, and it is no longer safe to republish: for at least one state whose mechanism it recorded, no current authorizing statute could be located at all, suggesting repeal in the intervening thirty years.

We have therefore published depth rather than breadth. A commissioner who checks their own state will find it either accurate as of 2026 or honestly labeled as unconfirmed.

The formula compensates value. The cost scales with acres.

A municipality's PILOT payment is its share of statewide eligible land value multiplied by the annual appropriation. Using the Commission's own worked example for Athol — $10,922,000 in eligible land value against $5,132,068,700 statewide, times a $53,000,000 General Appropriations Act line — the program pays a uniform $10.33 per $1,000 of assessed land value to every municipality in the Commonwealth.

The program is already tax-rate-like in its internal logic. It is not a grant and not discretionary aid; it is a flat statewide levy-equivalent, set below what any actual municipality charges, applied to a land valuation the municipality did not perform and cannot appeal.

What the formula does not contain is anything about cost. The Commission's own materials state it plainly: the distribution does not take into account local property tax rates, municipal budgets, population, municipal service costs, or ecological or recreational value.

But the obligations state-owned land imposes on a host community scale with area and frontage, not value: road miles maintained to reach and border it, road-stream crossings and culverts carrying water off it, emergency medical and fire response into it, search and rescue on it, wildfire and invasive-species management at its edges. In a county where the Commonwealth holds 135,417 acres of low-value forest, cost and compensation run in opposite directions.

This is the largest unquantified term in the entire policy debate, and it is unquantified because no one has been asked to measure it. The Commission should direct that a municipal service-cost study for host communities be conducted before any formula change is recommended. This paper publishes no service-cost estimate and will not.

One rate is applied to two entirely different kinds of land

Fewer than one acre in five draws more than half the money.

Stating this as a land-class problem rather than a regional one matters for two reasons. It is more accurate — Plymouth and Worcester sit on the large-tract side of the line, and Dukes and Nantucket on the high-value side, so this is not a map of east against west. And it identifies the design defect rather than the political geography: the Commonwealth applies a single value-proportional rate to a portfolio containing both urban parcels and twenty-thousand-acre forests.

Value-proportionality is defensible for the first. For the second it measures the wrong thing entirely, because a forest's burden on its host town — road frontage, drainage structures, emergency response, wildfire exposure — has almost no relationship to its assessed value.

Every comparison state has reached the same conclusion and acted on it. Minnesota sets a flat $3.087 per acre for tax-forfeited and other DNR land and the greater of $5.281 per acre or 0.75% of appraised value for acquired natural resources land. Wisconsin pays the greater of a local-tax-rate calculation or $3.50 per acre. Michigan pays $4.00 per acre indexed. Pennsylvania pays $7.20 per acre.

Massachusetts is the outlier not in what it pays but in refusing to classify what it owns.

Share erosion is automatic, directional, and permanent

Because the appropriation is fixed before the distribution runs, the program is zero-sum by construction. One municipality's gain in share is another's loss. A town whose eligible land value rises 4% in a year when statewide value rises 6% receives less money than the year before, despite its own land having appreciated. Nothing in that outcome reflects a judgment that the town needs less. It is arithmetic.

The framework re-indexes every municipality's base value to its Equalized Valuation on a two-year cycle. Land values in the eastern half of the Commonwealth have grown faster than in the west for the entire life of the current system. Faster eastern appreciation therefore transfers share eastward at every reset. The transfer requires no vote, appears in no line item, and is invisible in year-over-year payment figures because the appropriation has grown fast enough to mask it.

The Commission's appendix demonstrates the mechanic. Between FY2017 and FY2026, Pittsfield's payment rose 61.9% and Greenfield's 70.5%, while Holyoke's rose 176.0%. All three gained dollars. Only one gained share. In a fixed-pool program, only share is real.

Article 97 does double duty

Article 97 requires a two-thirds roll-call vote of both legislative branches to dispose of or change the use of conservation land, reinforced by the 2022 Public Lands Preservation Act's no-net-loss requirement (M.G.L. c.3 §5A). It is a strong and, in this paper's view, worth-keeping protection.

It also produces a consequence nobody designed. Article 97 land is untaxable, so the host town depends on PILOT. Article 97 land is also excluded from State Land for Homes, so the host town cannot convert any of it to taxable housing. The town gets the fiscal burden of hosting and access to neither remedy.

The exclusion is not incidental. The §122 surplus-land guidance states that all state-owned land not subject to Article 97 is eligible to be declared surplus, and the June 2025 inventory release confirms the review exempted protected open space, wetlands, and flood plains outright.

The result is a program that distributes housing-development opportunity in inverse proportion to conservation-land hosting. This is worth stating precisely because it is not a criticism of State Land for Homes, which is doing what it was designed to do. It is an observation that the design leaves out exactly the communities this Commission was convened to examine.

What the money actually looks like

A note on the Commission's published statistic

The listening-session materials report per-acre reimbursement in a column labeled Average of Actual PILOT Per-Acre, characterizing the range as $47/acre in Berkshire County to $4,885 in Suffolk.

Those figures cannot be aggregate ratios. Dividing each county's own reported dollars by its own reported acres — both from the same table — produces materially different values, shown above. The Essex divergence is tenfold: $3,105 as published, $316 in aggregate.

We should be plain about the limits of this observation. What is certain is the arithmetic, and it is reproducible from the Commission's own table. What is inferred is the explanation. We read the published column as an unweighted mean across municipalities because that construction would produce exactly this divergence pattern. We have not seen documentation of the calculation. If the Division of Local Services computes the figure by some other method, our characterization is wrong and we will correct it — the underlying disparity finding would be unaffected, because it rests on the aggregates rather than on the published column.

What Massachusetts is not

Massachusetts is not underfunding the program in aggregate. DOR's own local-rate estimate implies a shortfall of $4.5–5.5 million, or 8–10%.

Massachusetts is not ungenerous by national standards — but the right comparison is like-for-like. The blended $117/acre is not a meaningful cross-state number, because it averages land classes the comparison states price separately. Measured against comparable land, Berkshire's $34/acre still exceeds Minnesota's $5.281 acquired-natural-resources floor by roughly six times.

Track A: Fixing the formula

Track A is worth doing. It is also bounded, zero-sum, and — on the cross-state evidence — the track most likely to produce a formula that looks corrected without being funded.

A1. Publish the aggregate statistic. (No cost. No legislation.)

Report per-acre reimbursement as county and municipal aggregates, alongside share-of-acres against share-of-dollars, and retire the unweighted mean. This is the highest-leverage item available to the Commission because every subsequent argument runs through the number.

A2. Fund to the Department of Revenue's own local-rate estimate

Cost: $4.5–5.5 million above FY2026. Closes the aggregate gap entirely. Objection, which we think is correct: this does nothing about distribution.

A3. Classify the portfolio before adjusting the rate

The single change with the largest structural effect is also the one no Massachusetts reform has attempted: stop applying one value-proportional rate to every acre the Commonwealth owns.

The "greater of" construction used by Minnesota and Wisconsin does exactly what the Massachusetts formula cannot: it preserves value-proportionality where value is the right measure, and installs a per-acre floor where it isn't. A developed urban parcel clears the floor on value alone. A twenty-thousand-acre forest does not, and receives the floor instead.

Objection: a per-acre floor is not, strictly, a payment in lieu of taxes — it compensates burden rather than foregone revenue. That objection is correct on the merits and should be met by naming the component honestly rather than folding it into the value formula.

A4. Adopt a concentration adjustment — but not at Minnesota's rate

Minnesota's 2023 reform (H.F. 825 / 2023 Ch. 64) added per-acre supplements keyed to the share of a county's total area in PILT-eligible ownership: +$0.185/acre where that share is 25% or more, +$0.082/acre where 10–25%. Minnesota also codified an express statutory purpose at Minn. Stat. §477A.10.

The principle is exactly right. The magnitude is not. Those supplements total roughly $1.36 million against a $48.8 million program — 2.8%. Applied at Minnesota's rate to Berkshire County's 135,417 acres, the supplement would yield $25,052 per year. Minnesota identified that six counties hold 62% of its PILT acres while receiving 39% of payments, then corrected it with a payment that moves under three percent of the program.

The Commission should take Minnesota's principle and reject its scale.

A5. Put the hold-harmless in statute

The Commonwealth has been appropriating to avoid year-over-year decreases without being required to. Municipalities cannot budget against a practice. Vermont's PILOT statute (32 V.S.A. §3708) contains a floor; Massachusetts's own Watershed PILOT already carries a statutory hold-harmless.

A6. Rationalize eligibility

Approximately 245,000 acres — 34.7% of the Commonwealth's 706,000 state-owned acres — generate no payment. The Commission's own materials conclude that current eligibility "reflects historical decisions rather than intentional policy design."

A7. The lesson Track A must not ignore

Two states prove that formula design without funding architecture is decorative.

Connecticut enacted a genuinely sophisticated tiered PILOT in 2021 (P.A. 21-3), keyed to Equalized Net Grand List Per Capita, with a Tier 1 trigger for municipalities where the state owns more than 50% of the property. Full implementation was scored at roughly $137 million per year in additional cost. It has never been funded to the statutory formula; when the appropriation is insufficient, grants are prorated.

Michigan wrote "payment in full" into statute in 2012 (PA 603/604). House Fiscal Agency materials confirm appropriations have repeatedly been prorated below statutory levels, and that prorated payments satisfy the obligation.

For scale: Connecticut's unfunded formula gap alone is about 2.5 times the entire Massachusetts SOL PILOT program.

Track B: Growing the base

Track A argues about how to divide $54.5 million. Track B asks a different question: how does a town with 30% of its land area in state ownership ever grow its own tax base?

The mechanism, and why it is structurally superior

Under Proposition 2½, new growth is added permanently to a municipality's levy limit. It is capacity the town may levy without an override, on top of the annual 2.5% and any override, and it compounds into the base as units come online.

A PILOT dollar is an annual appropriation subject to the politics of the next budget cycle. A new-growth dollar is permanent taxing capacity the town owns outright.

One acre converted from untaxed state land to taxable housing is worth structurally more to a host municipality than the equivalent PILOT increase, forever — and it takes nothing from any other community.

It also breaks the circular trap that afflicts rural Massachusetts: infrastructure cannot be funded without new growth, and new growth cannot happen without infrastructure. Edge parcels adjacent to existing roads break the circle at exactly the point where the infrastructure already exists.

The proposal: screened edge release

Edge release should fund core consolidation.

This is not a proposal to trade conservation for housing. It is a proposal to trade low-value perimeter for high-value core, using mechanisms other states have already built:

• New York's equal-or-greater-value land swap model. Because Article XIV forbids sale, lease, or exchange of Forest Preserve outright, New York has amended its constitution roughly sixteen times via parcel-specific, voter-ratified swaps — Perkins Clearing (~8,500 acres), Keene (12 acres out, 144 in), Raquette Lake (1 out, 12 in). Every one ran net-additive to conservation.
• Maine's same-region replacement mandate (Art. IX §23), pairing a two-thirds vote with a same-county, same-purpose replacement-land requirement.
• New Jersey's Green Acres no-net-loss diversion regime, requiring State House Commission approval, an ecological-impact test, and replacement land of equal or greater value.
• Massachusetts's own Public Lands Preservation Act (M.G.L. c.3 §5A), which already codifies no-net-loss for Article 97 dispositions.

The screening framework is likewise already built and currently unused for this purpose. BioMap — the Natural Heritage and Endangered Species Program's joint product with The Nature Conservancy, updated 2022 — maps Core Habitat (~1.24 million acres) and Critical Natural Landscape with connectivity modeling. It is non-regulatory and is not linked to any disposal pathway.

For edge-depth thresholds, the Environmental Law Institute's Conservation Thresholds for Land Use Planners (2003), synthesizing more than 1,400 studies, recommends buffer zones of 230 to 300 meters from habitat edges, maintaining habitat patches larger than 55 hectares, and conserving 20–50% of the total landscape.

A defensible framework therefore looks like: protect all BioMap Core Habitat and Critical Natural Landscape, plus a 230–300 m buffer; consider for release only already-disturbed parcels outside that envelope, fronting existing public roads; screen out wetlands, priority habitat, steep slope, and connectivity corridors; route every disposition through the existing Article 97 two-thirds vote and PLPA no-net-loss test; and direct proceeds and swaps to core consolidation.

The objection the Commission will hear, and the answer

The Department of Conservation and Recreation and the conservation members of this Commission will say — correctly — that intact forest is the cheapest stormwater and climate infrastructure a municipality has, that edge development is where fragmentation harm concentrates, and that a release program creates a precedent that will be widened by future legislatures.

Three responses, offered honestly:

First, the screening framework above is designed so that no core tract is touched, and the buffer thresholds come from the conservation literature, not from a development proponent.

Second, the swap structure means the program is net-additive to conservation on the New York model. If it cannot be run net-additive, it should not be run.

Third — and this is the argument the Commission should weigh most carefully — the Commonwealth's conservation and biodiversity targets are currently being financed disproportionately by a small number of rural tax bases that were never asked. A program that permanently removes parcels from both the tax base and the development base, and compensates at $34 an acre, is not a stable long-term basis for public support of conservation in the communities that host it. Edge release, done right, is a conservation-durability measure.

We state plainly that this is a contested judgment, not a finding. The Commission should hear DCR's version of it directly.

What other states actually do

Full local taxability. New York taxes designated state land "for all purposes" under RPTL §532, assessed by the local assessor as if privately owned, with the state holding only the appeal rights of any taxpayer. Roughly 3.9 million acres are taxable. The state's all-in liability was about $250 million in FY2016 — up from roughly $100 million twenty years earlier.

Per-acre payment. Minnesota, Michigan, Pennsylvania, Wisconsin, Vermont. Rates are low in absolute terms but they are entitlements, not shares of a pool.

"Greater of" hybrids. Minnesota pays the greater of $5.281/acre or 0.75% of appraised value on acquired natural resources land. Wisconsin §70.114 is the closest analogue to what Massachusetts municipalities have asked for: for land acquired after 1991, the payment is the greater of the parcel's estimated value times the aggregate local net general property tax rate, or $3.50/acre.

The lesson that matters most

New York's executive branch attempted three separate times — under Governors Pataki, Paterson, and Cuomo — to convert full taxability into a capped PILOT. The best-documented attempt is Cuomo's 2018–19 Executive Budget, Article VII Part F, which would have frozen Forest Preserve assessments, replaced locally-assessed taxation with PILOTs at a rate set by the State Comptroller, and capped growth by a state-determined factor.

A coalition of local governments, school districts, and conservation organizations defeated it. Their arguments: it broke the 1885–86 bargain; central rate-setting would erode payments in real terms and shift burden onto private landowners in Forest Preserve towns; and the Budget Division had not analyzed local impacts.

All three attempts failed. Full taxability survives. In every case the fight was over who assesses and whether payments can decline — not over the rate. Massachusetts currently sits on the losing side of both questions.

Migration in the other direction

New York's Putnam County is the one documented case of a jurisdiction moving from a formula PILOT to full taxability. Putnam had the highest percentage of land in exempt state ownership (8.4%) and little taxable state land; the state instituted a formula PILOT ($400,000 in 1994, raised to $600,000 in 1995). In 2004, Chapter 364 amended RPTL §532(d) to make Putnam state land directly taxable for all purposes.

New York also operates concentration triggers Massachusetts lacks entirely. RPTL §534(3) makes certain reforestation lands taxable only where they exceed 3,000 acres and comprise at least 9.5% of the assessing unit's acreage — a dual absolute-and-proportional test directly applicable to the Massachusetts problem.

The analysis that has not been run

Every claim above about edge acreage is conceptual. Parcenomics has published that limitation before, and it remains true. We recommend the Commission direct that this analysis be conducted, and we specify it here so that it can be.

Inputs. MassGIS DCR and Division of Fisheries and Wildlife ownership layers; MassGIS road centerline file; BioMap Core Habitat and Critical Natural Landscape; MassDEP wetlands; NHESP Priority Habitat; slope derived from statewide elevation data; municipal assessor parcel boundaries.

Method. Intersect state ownership against road centerlines. Buffer to a defensible edge depth from the road, tested at multiple depths. Subtract the full BioMap Core Habitat and Critical Natural Landscape envelope plus a 230–300 m ecological buffer. Subtract wetlands, Priority Habitat, and slope above a buildability threshold. Classify the residual by existing disturbance. Aggregate to the municipality.

Outputs. Developable edge acreage by municipality; estimated unit yield at stated density assumptions; and permanent new-growth levy capacity at stated assessed-value assumptions, presented as a band with every modeling assumption labeled.

Verification discipline. Measured quantities — acreage, slope, distance — reported as measured. Modeled quantities — unit yield, assessed value, absorption pace — reported as labeled scenario bands, never as point estimates.

Why this specification is credible

This is the same structure Parcenomics has already published in municipal work. In the Great Barrington fiscal levers report, we measured the town's impervious cover directly from the MassGIS 2016 Land Cover / Land Use data: 44.45 million ft² total (~1,020 acres), 15.2 million ft² of road right-of-way excluded, 29.2 million ft² billable = 8,883 Equivalent Residential Units, with the tax-exempt share of the billable base at 16.9%. The revenue requirement was published as a band ($200K–$620K), not a point estimate.

Edge inventory is the same operation against different layers: ownership and roads in place of land cover, ecological screens in place of right-of-way exclusions, and the same measured-quantity-plus-banded-scenario discipline throughout.

Answering the Commission's questions

Does the current PILOT program work as intended? It works exactly as designed. The design compensates land value and nothing else. Whether that was the intent is the question the Commission should put to the record.

How does the program affect municipalities across the Commonwealth? Asymmetrically and directionally. 55.4% of eligible acres sit in four western counties receiving 18.9% of the dollars, and the gap widens on its own at every valuation cycle.

Are there opportunities to improve how the program operates? Yes, and the least expensive costs nothing: publish the aggregate per-acre and share-based statistics so the distributional question is visible in the Commission's own materials.

What tradeoffs should be considered? The central one is not fiscal. Closing the aggregate gap costs $4.5–5.5 million. The real trade is between the Commonwealth's budget flexibility and municipalities' ability to plan — the same trade already made, in the other direction, for Watershed PILOT.

Does the program appropriately balance conservation, biodiversity, and climate goals? The program is currently structured so that the more successfully the Commonwealth conserves land in a low-value county, the less that county is compensated per acre for hosting it — while Article 97 simultaneously excludes it from the state's principal land-for-housing program. Whatever balance that represents, it was not chosen.

Open issues

This paper is deliberately bounded. The following bear directly on the Commission's mandate, are not answered here, and are not answerable from the public record as it currently stands. Each is a defined piece of work rather than an unknown.

1. Municipal service costs attributable to hosting state-owned land. The formula excludes service costs by design, and no one has measured them. Until this is quantified for a representative sample of host communities, any claim that the current payment is adequate or inadequate is an assertion rather than a finding.
2. Developable edge acreage. The specification above is complete and runnable, but the analysis has not been performed, and this paper publishes no acreage, unit, or revenue figure derived from it.
3. Current program status in approximately 28 states. The 1996 baseline has decayed past the point of safe use.
4. The construction of the Commission's published per-acre statistic. Disclosed as an inference rather than a finding. If DLS supplies the actual construction, we will update this paper to reflect it.
5. The composition of the town-level payment series. Whether the Marion, Greenfield, Holyoke, and Pittsfield series are SOL PILOT only or composite is not stated. The share-erosion mechanic does not depend on them.
6. The cost of eligibility rationalization. Bringing 245,000 acres into the base without a corresponding appropriation increase would dilute every current recipient's share. The cost of doing it properly has not been priced.
7. Whether a Massachusetts concentration supplement can be keyed to municipal land-area share. Minnesota's trigger is the percentage of county area; Massachusetts distributes to municipalities.

Parcenomics prepared this paper independently and at its own cost as a contribution to the Commission's public record. The open issues above are stated as they are so that anyone — the Commission, a regional planning agency, or an individual municipality — can commission the work from any competent analyst. Parcenomics is available to scope or perform any of it.