Haute Lumière

Commerce · VII.06 · MMXXVI · daylight

La Bourse  /  Volume VII  /  Nº VII.06  /  Ten concept briefs

TEN CONCEPT BRIEFS · Chapter VII.06 — Watershed as the Unit of Account

One page each. A reader who reads only these ten pages has the chapter.

---

BRIEF 1 — The Boundary That Closes

The idea. A watershed is the area of land from which water drains to a single point. That definition has an unusual consequence: the accounts must balance.

        P  =  ET  +  Q  +  ΔS  +  ε

  P    precipitation onto the basin         the only credit
  ET   evapotranspiration                   the loss to the sky
  Q    streamflow past the gauge            the transfer downstream
  ΔS   change in storage                    soil, snow, aquifer, reservoir
  ε    the residual                         what you have not found yet

Every drop that falls inside the line leaves past the gauge, leaves as vapour, or is still inside. There is no fourth destination. Water cannot be offshored, restated, or booked to a subsidiary in another jurisdiction.

Worked example. New York City's Catskill and Delaware watershed is 1,000,000 acres — 404,686 hectares, 4,047 square kilometres. The city draws about 900 million gallons a day from it. Spread across the basin, that draw is a depth of 307 millimetres a year, and it is 58 percent of the runoff the forested basin produces. You can state both of those because the boundary closed.

Why it matters. Almost no economic boundary has this property. A firm, a country and a sector are all boundaries you can move water across without anyone noticing. A catchment is not. It is the only economic frontier drawn by gravity, and gravity does not lobby.

You already know this because you have reconciled a bank statement. The reason that exercise is worth doing is not the arithmetic — it is that the two sides have to agree, so a difference is information rather than opinion.

---

BRIEF 2 — The Residual Is the Instrument

The idea. In a set of books that must close, the interesting number is the one that does not fit.

ε — precipitation minus evapotranspiration minus outflow minus the change in storage — is the basin's suspense account. If it is near zero, the account is consistent. If it is not, then one of exactly three things is true: somebody is abstracting water nobody has recorded; the gauge is wrong; or the storage estimate is wrong. All three are findings and all three are actionable.

Worked example. A basin's annual account comes back with a residual of 40 millimetres — about 3 percent of rainfall. That is too large to be measurement noise on a well-run gauge. Somebody drives up the valley and comes back with an unmetered abstraction that has not been on a register since the 1970s. Nothing in the report said "there is an unmetered abstraction." The arithmetic simply left no other place for the water to be.

Why it matters. Conventional environmental reporting has no residual, because nothing has to sum. A number you cannot fail to reconcile is a number nobody can learn from. Publish the residual, every year, including the years it is embarrassing. A basin whose residual is always exactly zero is running a press release.

You already know this because every good auditor you have met went straight to the reconciliation, and every bad one went straight to the summary.

---

BRIEF 3 — Shares, Not Volumes

The idea. Never apportion a fixed quantity of a variable flow. Apportion a percentage of the measured flow.

Worked example. The Colorado River Compact of 1922 gave 7.5 million acre-feet a year to each of the Upper and Lower Basins; the 1944 treaty added 1.5 for Mexico. 16.5 million acre-feet a year, obligated. What the negotiators were looking at is now checkable. In the Meko tree-ring reconstruction of the Upper Colorado, the wettest twenty-five-year window in the entire 1,244-year record begins in 1905, at 16.83 maf a year — and Reclamation's own gauged natural flow for water years 1906 to 1922 averages 18.03. They picked the wet period, and the measured record in front of them was wetter still.

Now rewrite the identical deal as shares — 45.4545 percent to each basin, 9.0909 to Mexico. At 12.40 maf, each basin receives 5.636 instead of 7.5. The 25 percent haircut arrives by arithmetic, in the year the water does not, and nobody has to be sued.

Why it matters. A volumetric entitlement makes shortage a legal event. A share makes shortage an arithmetic event. Australia writes entitlements as shares of an announced allocation, which is why the Murray–Darling's crisis was a crisis of volume and not additionally a crisis of law.

You already know this because you would rather hold equity in a company whose revenue might fall than a fixed claim against it. One of those instruments survives a bad year. The other one goes to court.

---

BRIEF 4 — Avoided Cost Is the Downstream Price

The idea. The most defensible value of a watershed service is not what it is worth. It is what not having it would cost you to replace.

Avoided cost is a price a finance function already knows how to handle. It needs no contingent valuation, no willingness-to-pay survey, no shadow price — only a capital plan and an engineering estimate.

Worked example. At New York's Catskill/Delaware basin the replacement is a filtration plant. A $9.0 billion plant at a 4 percent real discount rate over a 40-year life carries a capital recovery factor of 0.050523 and a capital charge of $454.7 million a year. Add the city's own stated operating figure of at least $100 million and the avoided cost is $554.7 million a year — against 328.5 billion gallons delivered, $1.69 per thousand gallons.

Why it matters. It puts the watershed and the plant in the same table, in the same units, competing for the same money on the same test. That is the only comparison that ever changes a capital decision.

The trap. Avoided cost is only as good as the alternative's price, and alternatives get cheaper. Membrane filtration costs have fallen for thirty years. Re-run the arithmetic every five years and expect it to erode.

You already know this because when you priced your own insurance you did not ask what your house was worth to you. You asked what rebuilding it would cost.

---

BRIEF 5 — Check the Estimate Against a Receipt

The idea. An avoided cost drawn from an estimate is a forecast. An avoided cost drawn from a project somebody actually completed is a measurement. Always look for the receipt.

Worked example. The famous Catskills ratio entered the literature through Chichilnisky and Heal in Nature in 1998 and has circulated for nearly thirty years as roughly $1–1.5 billion of watershed protection against $6–8 billion for a plant. New York did build one plant: the Croton Water Filtration Plant, in service 2015, final cost $3.2 billion against an original projection of $800 million — a 4.0× overrun — for 320 million gallons a day.

  realised capital per mgd     $3.2bn / 320 mgd   =  $10.0 million per mgd
  Catskill/Delaware at 900 mgd                    =  $9.0 billion
  Catskill/Delaware at 1,200 mgd design           =  $12 billion

The city's own later figure is "more than $10 billion", reached by a different route. The famous ratio was not inflated. Checked against a receipt, it was low.

Why it matters. The reflexive move on meeting a famous number is to suspect it upward or downward on instinct. Neither is analysis. Find the nearest thing somebody actually paid for and scale from that.

You already know this because you have watched a builder's quote and a builder's invoice differ, and you know which one you would plan against next time.

---

BRIEF 6 — The Detection Threshold

The idea. At basin scale, hydrological attribution from upstream land use to downstream water is weak — and the literature says how weak, in a number.

Bosch and Hewlett's 1982 review of catchment experiments established the threshold the field still uses: a change in forest cover below about 20 percent of catchment area cannot normally be detected in streamflow at all. Andréassian (2004) and Bruijnzeel (2004) reach the same place: attribution weakens as basins get larger, and at basin scale it is mostly gone.

Worked example. New York has protected 130,000 of 1,000,000 acres — 13 percent. The threshold is 200,000 acres. The most successful watershed payment programme on earth has protected less land than would be required to measure its own effect.

Why it matters. Most watershed payment schemes cannot demonstrate the service they buy. Pattanayak, Wunder and Ferraro (2010) went looking for schemes with credible counterfactual evaluation and found very few; Naeem and colleagues, in Science in 2015 found most lacked an adequate scientific basis for the service claimed. This is the strongest objection in the field and it is correct.

What follows is not despair. It is a design instruction: stop trying to attribute, and move the settlement point to somewhere measurement already happens. That is Brief 10.

You already know this because you have seen an advertising budget defended by a correlation, and you know the difference between a number that is true and a number that is attributable.

---

BRIEF 7 — Trees Are a Quality Strategy with a Quantity Price

The idea. Forests do not make water. On net they consume it. A watershed programme that buys trees in a basin short of water is paying to deepen the scarcity it was convened to relieve.

The evidence. Farley, Jobbágy and Jackson (2005) analysed 26 catchment data sets and 504 observations. Annual runoff fell by 44 percent (±3) where grassland was afforested and 31 percent (±2) where shrubland was — and by 75 percent (±10) with eucalypts against 40 percent (±3) with pines.

Worked example. Run the Zhang, Dawes and Walker (2001) relation over the Catskill/Delaware basin at 1,200 mm of rainfall and 700 mm of potential evapotranspiration, with the published coefficients of 2.0 for forest and 0.5 for grass:

The forested basin yields 154 millimetres a year less. Across 4,047 square kilometres that is 451.9 million gallons a day that does not arrive, against a city draw of 900. Keeping the Catskills forested costs New York about half as much water as New York drinks.

Why it matters. New York is buying quality with quantity and it is right to, because it has quantity to spare. A basin that does not have quantity to spare must not copy the purchase without copying the arithmetic. Only a closed account shows you the trade.

You already know this because you have met a cost saving that turned out to be a cost transfer, and you found it by looking at the other department's budget.

---

BRIEF 8 — The Third Term

The idea. A payment for watershed services has three terms. Published business cases print two: what upstream gives up, and what downstream avoids. The one almost nobody prints is what it costs to make the trade happen at all — and it is the term that kills most schemes.

        A*  =  t / (b − c)

  t   fixed transaction cost per contract per year
  b   downstream benefit per hectare per year
  c   upstream opportunity cost per hectare per year
  A*  the parcel size below which the deal cannot pay for itself

Ferraro (2008) named the mechanism: transaction costs and asymmetric information are largely fixed per contract, so cost per hectare falls as contract size rises.

Worked example, twice. In the Catskills, with 15 percent of the basin at genuine risk of conversion, b is $9,138 per hectare per year. Against a forgone rent of $250 and $1,200 a contract, A\* is 0.135 hectares. Anything above a seventh of a hectare pays for its own paperwork. The third term is invisible — which is exactly why the famous case taught the world a lesson the world could not use.

Now a basin shaped like most basins: 100,000 people at 400 litres a head a day, avoiding $0.30 per thousand gallons of treatment, across 10,000 at-risk hectares. b is $115.71 per hectare per year. At a forgone rent of $250, b − c is negative and no parcel size clears. At $80, b − c is $35.71 and A\* is 33.6 hectares — and where the median holding is 2 hectares, transaction cost runs at $600 per hectare against a net benefit of $35.71. Short by 17×.

Why it matters. It fails on paperwork, not on ecology, and nobody said so on the page. Print all three terms or you have not priced the scheme.

You already know this because you have abandoned a genuinely worthwhile small transaction once you saw the legal fee.

---

BRIEF 9 — Aggregate the Counterparty

The idea. If transaction cost is fixed per contract, the design lever is the number of contracts. Contract with one intermediary that holds membership, not with every landholder.

Worked example. Take the failing basin from Brief 8 — net benefit $35.71 per hectare a year, transaction cost $1,200 a contract, median holding 2 hectares.

  1,500 separate contracts   $600.00 per hectare per year    short by 17x
  one 3,000-hectare contract   $0.40 per hectare per year    clears by 89x

It replaces 1,500 contracts with one — and it replaces 1,124,250 pairwise relationships, n(n−1)/2, with one.

Why it matters. This is the cheapest structural change available in the whole field and it requires no new science, no new valuation and no new law. New York's Watershed Agricultural Council is exactly this: the city does not contract with four hundred farms, it contracts with one farmer-led organisation that contracts with four hundred farms.

Two conditions. The intermediary must be governed by the people it contracts for — that is why farmers sign — and its administration margin must be disclosed as a line, because the third term only stays managed while it stays visible.

You already know this because you have watched a union, a buying group or a trade association negotiate something no individual member could have afforded to negotiate alone.

---

BRIEF 10 — Buy the Deferral, Not the Service

The idea. The design that does not depend on attribution. Stop purchasing a hydrological outcome from a parcel. Purchase a reduced probability of having to build a plant, and settle it where measurement already happens.

Three moves, and each removes a dependency:

1. Settle downstream at the intake, against the regulatory threshold the utility already monitors under compulsion. Zero marginal measurement cost, an existing chain of custody.
2. Contract upstream on practice — the septic system replaced, the buffer standing, the plan followed. Practice is observable and inspectable. Outcomes at the parcel are not.
3. Do attribution once, at basin scale, against the threshold — never per hectare, because per hectare it cannot be done.

Worked example. Deferring a $9.0 billion plant by ten years at 4 percent real is worth $2.92 billion in present value. The programme that buys the deferral costs $55.6 million a year, a present value of $450.6 million.

  net present value                    $2.47 billion
  the number that decides it              6.5 : 1
  the walk-away price             $360 million a year

That last figure is the honest ceiling on what upstream can ask and the honest floor under what downstream should pay.

Why it matters. It converts an unfalsifiable ecological claim into a capital decision with a date and a price. And it protects the scheme from the audit that arrives in year four asking which hectare produced which milligram — because the instrument said, in writing, on day one, that it was never buying that.

You already know this because you have paid for an option before. You were not buying the asset. You were buying the right to decide later, and you knew exactly what that right was worth.