Farms produce value no one pays for
A regenerative acre produces about $477 a year in climate, water, biodiversity, soil, and resilience value. The farmer is paid about $30 of it — 6% — and only if they sell carbon credits.
A direction we are working toward
The Ecosystem Dividend
What if every acre paid its steward a recurring dividend for the clean water, carbon, and habitat it produces — funded by the people who benefit? This paper sets out the idea, the model behind it, and what the numbers say if it worked.
September 30, 2026 · 12 min read · Model v0.1 · Every assumption published below
- modeled public value of a regenerative acre, per year
$477
modeled public value of a regenerative acre, per year
- of that value is paid to the farmer today
6%
of that value is paid to the farmer today
- per acre per year: the dividend at prices already paid in the US
$42
per acre per year: the dividend at prices already paid in the US
- acres in stewardship by 2040 vs. today’s plateau
4.3×
acres in stewardship by 2040 vs. today’s plateau
The idea in 60 seconds
So charge the people who benefit
Water utilities, food companies, insurers, and public programmes each pay per unit of the outcome they use. The money is pooled, confidence-weighted, and paid to the steward as one annual per-acre dividend.
Recurring money moves acres
At prices already paid somewhere in the US, the dividend is $42/acre and pays back the switch in year 3. Calibrated to the census and to Maryland, it lifts stewardship from 19M to 84M acres by 2040.
01 · The value gap
Only the carbon sliver has a buyer.
Using TerraValue’s benchmark composite, an acre under regenerative management produces roughly $477 a year in public value. Carbon is the only part with a working market, and a US farmer selling soil-carbon credits clears about $30 an acre. The rest is a public good nobody invoices.
These are societal values — what the outcomes are worth to someone — not prices anyone has agreed to pay. The dividend below deliberately uses only prices that are already paid somewhere today, which is why it is far smaller than this bar.
Annual value of one regenerative acre, and what the farmer is paid for it
US$ per acre per year · mid estimates
- Climate (carbon)
- Water
- Biodiversity & pollination
- Soil health
- Yield resilience
02 · Why adoption stalls
Three censuses, one plateau.
Cover crops are the best-measured stewardship practice in the US. A logistic curve through the last three Censuses of Agriculture levels off at 19.3M acres — 5.1% of cropland. Adoption grew 50% from 2012 to 2017, then 17% to 2022. At today’s incentives it is close to finished.
US cover crop acres, and the curve they imply
Million acres
- USDA Census of Agriculture (reported)
- Logistic fit — dashed where extrapolated
27.3% vs 4.7%
cover-crop share of cropland: Maryland, which pays per acre every season, vs. the US (2022 Census)
USDA NASS, 2022 Census$6.34/t
average voluntary carbon credit price across all credit types, 2024
Ecosystem Marketplace, 2025| Dimension | Cost-share contract | Carbon credits | Ecosystem dividend |
|---|---|---|---|
| Pays for | Installing a practice | Tonnes of CO₂e | A basket of measured outcomes |
| Paid | For the life of a contract | Per verified issuance, after year 1 | Every year the outcomes persist |
| Funded by | Taxpayers — when funded (31% of EQIP applications were, FY17–22) | Carbon buyers | Each beneficiary, per unit of what it uses |
| Early adopters | Mostly pays for new practices | Excluded as non-additional | Paid against a regional benchmark |
| Share of an acre’s value addressed | One practice at a time | ~13% (climate only) | Carbon, water, habitat, resilience |
| Evidence | Practice was installed | Protocol sampling, pass or fail | Tiered — better evidence earns a larger share |
03 · How it works
Beneficiaries pay per unit. Stewards receive one dividend.
A dividend is what an owner receives for holding a productive asset. Farmland’s natural capital is that asset; its outcomes are the return. The mechanism connects the people who use those outcomes to the people who produce them — through a ledger both can audit.
01
Those who benefit pay
Water utilities · per lb of nitrate kept out
Des Moines runs a nitrate-removal plant at up to $16,000 a day.
Food & ag companies · per t CO₂e in their supply shed
Scope 3 targets sit on the same acres they source from.
Insurers & lenders · per resilient acre
Iowa and Illinois already discount premiums $5/acre for cover crops.
Public programmes · per habitat acre
CRP, EQIP, and EU eco-schemes already buy these outcomes piecemeal.
02
Dividend pool
- Buyers commit per-unit prices for 5+ years
- Each outcome is sold once — no double counting
- 10% risk reserve, released if outcomes hold
03
Outcome ledger
- Outcomes modeled against a regional benchmark
- Confidence tier sets the share paid (60–95%)
- Every number traceable to data and method
04
Stewards
- One annual payment per acre, for as long as outcomes last
- Early adopters paid too
- Owner–tenant split set in the lease
Pay for outcomes, not practices
The unit is a pound of nitrate kept out of the river or a tonne of carbon stored — whatever practice produced it.
Benchmark against the region, not your own past
A farmer who adopted cover crops a decade ago is still producing clean water. Paying against a regional baseline stops punishing the early movers.
Pay more for better evidence
A confidence factor κ pays 60% of a modeled outcome, 80% once sampled, 95% once measured. Measuring more earns more — so evidence improves itself.
Sell each outcome once
Reductions from stacked practices compound instead of adding, and a ledger assigns every unit to exactly one buyer.
04 · The model
One line of arithmetic.
For each service, multiply the verified quantity by the price its beneficiary pays. Then pay the steward the share that the evidence supports, after a risk reserve and the cost of measurement.
D = Σs Qs · Ps · κ · (1 − r − f)
- Qₛ
- Verified quantity of service s per acre: t CO₂e, lb nitrate-N, habitat acres, enrolled acres.
- Pₛ
- Price the beneficiary of s pays per unit. Scenarios below use prices already paid in the US.
- κ
- Confidence factor: 60% modeled, 80% modeled + sampled, 95% measured.
- r
- Risk reserve, 10% — held against shortfall, released if outcomes hold.
- f
- Measurement, verification, and administration, capped at 10%.
Worked example: the core bundle on one acre
Cover crops + No-till + Nitrogen right-sizing · Tile-drained corn–soybean field, US Corn Belt · base prices · modeled + sampled evidence · 2026.
| Service | Q per acre | P | Buyer pays | Steward receives |
|---|---|---|---|---|
| Carbon | 0.68 t CO₂e | $20 | $13.60 | $8.70 |
| Clean water | 9.47 lb nitrate-N | $5.00 | $47.38 | $30.32 |
| Habitat | 0.00 habitat acres | $145 | $0.00 | $0.00 |
| Resilience | 1.00 enrolled acres | $5.00 | $5.00 | $3.20 |
| Total | × 64% paid | $65.98 | $42.22 |
- Clean water carries the dividend — 72% of it. The bundle removes 38% of the field’s 25 lb baseline nitrate loss; reductions compound rather than add.
- Early adopters are paid too. The same acre, farmed this way for years, earns $38/acre. A carbon market would pay it nothing.
05 · Run it on a farm
Does it pay to switch?
Choose the practices, the price scenario, and how well the outcomes are evidenced. The cash-flow chart includes what the switch really costs: seed and passes, input savings, and the yield drag farmers report in the first years. At base prices, only the dividend ends ten years ahead ($140/acre); cost-share ends at −$42 once its contract lapses, carbon credits alone at −$209.
Practices on the field
Price scenario
Evidence tier
Dividend per acre, year 1
$42
10-yr average $42
Per farm, year 1
$19.5K
463 acres
Buyers pay per acre
$66
64% reaches the steward
Dividend breaks even
Year 3
Cost-share contract: not in 10 yrs · Carbon credits only: not in 10 yrs
Per acre per year this bundle stores 0.68 t CO₂e, keeps 9.5 lb nitrate-N (38% of the baseline loss) out of the water.
From what buyers pay to what the steward receives
US$ per acre, year 1
Cumulative net position per acre, by how the farmer is paid
US$ per acre · practice costs, input savings, and transition yield drag included
- Ecosystem dividend
- Carbon credits only
- Cost-share contract
- No payment
Year 1 = 2026. Cost-share pays for 3 years. Carbon credits start after the first verification and exclude early adopters. Prices follow the “Base” scenario; modeled + sampled evidence pays 80% of the modeled outcome.
06 · If it scaled
84M acres by 2040.
The projection joins two observations. The census fixes how fast adoption spreads and where it stops today; Maryland shows how far a recurring $50/acre payment moves it. Between them, an exponential response curve (scale λ ≈ $96/acre) maps any recurring payment to an adoption ceiling, capped at 60% of cropland.
In the base case, stewards receive $3.5B a year by 2040 — about 1.7× EQIP’s annual budget — while beneficiaries pay $5.5B for 57 Mt CO₂e and 794M lb of nitrate-N kept out of water each year.
Price scenario
Evidence tier
Enrolled acres, 2040
84M
22% of cropland · 4.3× status quo
Paid to stewards, 2040
$3.5B/yr
Buyers pay $5.5B · $42/acre
Carbon, 2040
57 Mt
CO₂e stored or avoided per year
Nitrate-N kept out of water, 2040
794M lb
per year, on enrolled acres
Acres enrolled in the core bundle
Million acres · 2026 starts at the census-fitted curve
- Ecosystem dividend
- Carbon credits only
- Status quo
Money flowing to stewards each year
US$ billions per year
- Buyers pay (dividend)
- Stewards receive (dividend)
- Carbon credits only
Acres diffuse at the rate fitted to the 2012–2022 census and rise toward a ceiling set by each policy’s recurring per-acre payment, calibrated so that $0 reproduces today’s plateau and $50 reproduces Maryland’s 27.3% share. Buyer demand at scenario prices is assumed, not modeled.
07 · What moves the number
Two unsettled prices, and one lever we control.
The nitrate price and the carbon price swing the dividend most — the two prices markets have not settled. The largest lever a programme controls is evidence: moving from modeled to measured outcomes adds $18/acre on the core bundle. That is why measurement is the product, not overhead.
What moves the dividend: each assumption swung low → high, all others at base
Net $/acre, 2030, all four practices, modeled + sampled · base $58
- Price of a pound of nitrate-N$40$114Price of a pound of nitrate-N: $2.50/lb gives $40; $13/lb gives $114.
- Carbon price$53$121Carbon price: $10/t gives $53; $146/t gives $121.
- Practice nitrate reduction$41$74Practice nitrate reduction: Low gives $41; High gives $74.
- Baseline nitrate loss$44$73Baseline nitrate loss: 15 lb gives $44; 36 lb gives $73.
- Evidence tier (confidence paid)$43$69Evidence tier (confidence paid): 60% gives $43; 95% gives $69.
- MRV fee + risk reserve$51$65MRV fee + risk reserve: 30% gives $51; 10% gives $65.
- Soil carbon accrual rate$53$62Soil carbon accrual rate: Low gives $53; High gives $62.
- Habitat price$53$61Habitat price: $72/ac gives $53; $191/ac gives $61.
- Insurer resilience discount$55$61Insurer resilience discount: $0/ac gives $55; $10/ac gives $61.
Price of a pound of nitrate-N: $2.50/lb → $40/acre; $13/lb → $114/acre. Swing $74.
- Below the base case
- Above the base case
Bundle: cover crops, no-till, nitrogen right-sizing, prairie strips. Hover or tab through a row for its input values.
08 · It has worked before
Beneficiary-pays is not new. Stacking it is.
Each piece of the dividend already exists somewhere. What does not yet exist is one mechanism that pays for all of them, on the same acre, from everyone who benefits.
New York City, USAsince 1997
Catskill/Delaware watershed
Paid by: City water ratepayers
~$2.5B vs $8–10B
spent protecting farms and forest upstream vs. the filtration plant it avoided
Instead of building a plant that would also cost roughly $1M a day to run, the city pays upstream landowners to keep water clean at the source — about $100M a year since 1997.
Beneficiaries pay when they can compare stewardship against an avoided cost.
National Academies, 2020Costa Ricasince 1997
Payments for Environmental Services
Paid by: 3.5% of national fuel-tax revenue
21% → 52%
national forest cover, 1987 to 2015
A dedicated, recurring revenue stream pays landowners roughly $64 per hectare a year to keep forest standing. More than a million hectares have been enrolled.
A permanent funding stream outlasts project-by-project money.
CBD case studyVosges, Francesince 1993
Vittel catchment
Paid by: Nestlé Waters (the bottler)
92%
of the sub-basin under protective farming by 2004
The bottler paid 26 farms about €200/ha a year through the transition, plus equipment, under 18- to 30-year contracts — €24.25M over 1993–2000.
One beneficiary can underwrite a whole catchment when the outcome it needs is specific.
IIED, Perrot-Maître 2006Iowa & Illinois, USAsince 2017
Cover crop premium discount
Paid by: State programmes via crop insurance
1.2M+ acres
enrolled in Iowa’s $5/acre discount to date
A $5/acre crop-insurance discount for cover-cropped acres. Illinois’ first 50,000-acre round filled in seven days.
Risk-bearers can pay for resilience, and a small, simple recurring payment gets taken up fast.
Iowa Dept. of AgricultureMaryland, USAsince 1990s
Cover crop grant
Paid by: State Chesapeake Bay restoration funds
27.3% vs 4.7%
of cropland in cover crops, Maryland vs. the US (2022 Census)
A per-acre grant paid every season — $50 base in 2023–24, up to $105 with add-ons — on nearly 450,000 acres.
Recurring per-acre payments move adoption; one-off contracts do not.
Maryland Dept. of AgricultureEuropean Unionsince 2023
CAP eco-schemes
Paid by: EU agricultural budget
€44.7B
ring-fenced for 158 eco-schemes over 2023–2027
At least a quarter of direct farm payments must now go to schemes that pay for environmental and climate practices.
Public farm budgets are already moving from income support toward paying for outcomes.
European Commission09 · What could break it
The failure modes, and the design answer to each.
| Risk | Why it matters | Design response |
|---|---|---|
| Double counting | The same pound of nitrate sold to a utility and a food brand inflates impact and destroys trust. | Multiplicative stacking in the model; one ledger entry per unit, one buyer per entry. |
| Over-crediting | Soil carbon and nitrate models can be wrong at field scale, and the error is not random. | Confidence tiers pay only a share of modeled outcomes; a 10% reserve absorbs shortfalls. |
| Reversal | Tillage can release stored carbon; a practice can simply stop. | Pay-as-you-go: the dividend stops when the outcome stops, so no one prepays for permanence. |
| Who gets paid | About 39% of US farmland is rented, and most landlords do not farm. | The dividend follows the acre; the split between owner and operator is written into the lease. |
| Buyer commitment | Prices for nitrate and carbon are thin and volatile; a dividend that swings wildly is not a dividend. | Multi-year price commitments per unit, with public programmes as anchor buyers. |
| Model risk (ours) | The adoption curve is fitted to three points; Maryland may not generalise; prices are scenarios. | Every assumption is published and tagged; a pilot exists to test exactly these numbers. |
10 · What we’re building toward
Measurement first. Then a pilot. Then the ledger.
The Ecosystem Dividend is a direction for TerraValue, not a product we sell today. Every step toward it starts with the same thing: outcomes measured well enough that someone will pay for them.
- 1Now
Measure the outcomes
The TerraValue dashboard already models carbon, water quality, water quantity, biodiversity, resilience, and economics for a field. Next: uncertainty ranges and evidence tiers on every output.
- 2Next
Pilot one watershed
One water utility, one supply-shed buyer, a cohort of farms. Pay on modeled outcomes, measure edge-of-field and soil alongside, and test whether the confidence ladder holds.
- 3Later
Build the ledger
A registry that sells each outcome once to the buyer who values it most, and pays stewards one annual dividend per acre.
11 · Methods & sources
Every assumption, tagged.
DATA a published figure · DERIVED arithmetic on published figures · DESIGN a parameter of the proposed mechanism, which a pilot would test.
Practices (per acre per year, mid and range)
| Practice | Carbon, t CO₂e | Nitrate cut | Cost − savings | Yield drag | Sources |
|---|---|---|---|---|---|
| Cover cropsA winter cereal rye cover between cash crops. | 0.47 (0.36–0.59) | 31% (15%–45%) | $37 | 2.25% | Poeplau & Don 2015 · Iowa Nutrient Reduction Strategy · SARE, cover crop cost baseline · Deines et al. 2023 |
| No-tillPlant directly into last season’s residue; no tillage passes. | 0.16 (0.00–0.31) | 0% (0%–5%) | −$15 | 5% → 3% → 1% → 0% | Powlson et al. 2014 · Pittelkow et al. 2015 |
| Nitrogen right-sizingTrim N to the economic optimum rate (about 15% less) with split timing. | 0.05 (0.03–0.08) | 10% (5%–15%) | −$11 | 0% | IPCC 2019 Refinement, Vol. 4 Ch. 11 · Iowa Nutrient Reduction Strategy |
| Prairie stripsConvert ~10% of the field to native prairie in contour strips. | 0.10 (0.05–0.16) | 10% (5%–20%) | $30 | 0% | Schulte et al. 2017 (PNAS) · STRIPS economics · Gebhart et al. 1994 |
Baseline: Tile-drained corn–soybean field, US Corn Belt; 25 lb nitrate-N lost per acre per year (15–36); gross crop revenue $546/acre. Yield drag is a share of that revenue by year; the last value persists.
Price scenarios
| Scenario | Carbon $/t | Nitrate $/lb N | Habitat $/ac | Resilience $/ac |
|---|---|---|---|---|
| ConservativeToday’s thin markets, frozen in place. | $10 | $2.50 | $72 | $0 |
| BasePrices already paid somewhere in the US today. | $20 | $5.00 | $145 | $5 |
| AmbitiousRemoval-grade carbon; utilities pay at their marginal cost. | $20 → $146 by 2030 | $13.00 | $191 | $10 |
Carbon: 2024 land-use credit average; BloombergNEF 2030 high-quality ($20) and removals-only ($146) scenarios. Nitrate: PENNVEST auction, Virginia trading, Maryland Conowingo fee. Habitat: CRP averages (all, continuous, CREP), per acre of habitat. Resilience: Iowa/Illinois premium discount.
Mechanism and national anchors
| Confidence factor κ | 60% / 80% / 95% | DESIGN |
|---|---|---|
| Risk reserve r | 10% | DESIGN |
| MRV & admin fee f | 10% | DESIGN |
| Early-adopter carbon rate | 50% | DESIGN |
| Carbon-only farmer share | 60% | DATA |
| Cost-share contract | 3 years | DESIGN |
| US cropland, 2022 | 382M acres | DATA |
| Cover crops, 2012 / 2017 / 2022 | 10.3M / 15.4M / 18M | DATA |
| Logistic ceiling · rate | 19.35M · 0.246/yr | DERIVED |
| Maryland anchor | $50/ac → 27.3% | DATA |
| Technical potential | 60% | DESIGN |
Limitations we know about
- The logistic fit uses three census points for three parameters; it matches them exactly and cannot test itself.
- Maryland’s adoption also reflects Chesapeake Bay regulation and a long winter; it may overstate the payment response elsewhere.
- The national projection assumes buyers exist at scenario prices. It sizes the pool that would be needed; it does not prove demand.
- Cover-crop adoption stands in for the whole bundle, because it is the only practice with a clean census time series.
- The value-gap bar uses societal values (the carbon slice at $51/t, the 2021 interagency interim value); the dividend itself uses market prices only.
- Outcomes are field-average estimates for tile-drained corn–soybean in the Corn Belt; other systems will differ.
Model outputs are estimates, not guarantees or offers of payment. The model and its tests live alongside the site code, and every chart on this page is computed from it.
Pressure-test this with us
Tell us where it breaks.
We are looking for a watershed to pilot in, buyers who would commit a per-unit price, and people who think an assumption above is wrong. All three help.
← All perspectives