La Bourse / Volume VI / Nº VI.05 / Ten concept briefs

A woman seated on the floor of a long white room reading, a cup and a teapot beside her, candles lit. — Plate VI.05 · Ten concept briefsSix Years of Paper.Every collapse in this chapter was measured before it happened, written down, and published. The interval between the writing and the falling is the only number here that generalises.

TEN CONCEPT BRIEFS · Chapter VI.05 — The Commons That Fell

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

BRIEF 1 — The Lag

The idea. Every commons that fell was measured before it fell. The load-bearing quantity is not the collapse. It is the interval between the first published, quantified warning and the end.

    L  =  t(collapse)  −  t(first public, quantified warning)

The warning has to be a number in public — not a worry, not an anecdote, not a conversation in a corridor. That standard is strict on purpose. It is what makes L comparable across a fishery, an aquifer and a website.

Worked example. Northern cod. In 1986 a stock assessment found the biomass far smaller than the standing figure and advised cutting the quota to roughly half. The moratorium came on 2 July 1992. L = 6 years. For the High Plains aquifer, a federal appeals court allowed irrigators to deduct groundwater depletion as the exhaustion of a capital asset in 1965 — the tax code acknowledged the overdraft — and there is still no end date. L ≥ 50 years, and unbounded.

Why it matters. A collapse looks like a shock and behaves like a schedule. Once you have L you are no longer asking whether a thing is at risk. You are asking whether your institution can move inside it.

You already know this because you have been in an organisation where the problem that eventually broke something had been written up, circulated, and agreed to be real, for years before anyone changed a rule. The paper was never the issue.

BRIEF 2 — The Three Geometries of Delay

The idea. What a lag costs depends entirely on the kind of stock it is drawn against. There are three shapes and they are not interchangeable.

Geometry	Stock	Cost of one more year of delay
Exponential	A living population under excess mortality	Compounds. Year five costs more than years one to four together.
Linear	A mined stock with negligible recharge	Constant. A foot a year is a foot a year.
Step	A rule-governed commons under someone else's rules	Zero, then total, on a date you do not choose.

Worked example. Fish: survival is e^(−(F+M)t), so six years of fishing at F = 0.50 instead of F = 0.20 leaves e^(0.30 × 6) = 6.05 times less. Water: the High Plains lost 273.2 million acre-feet over sixty-five years, and the sixty-sixth year will look like the sixty-fifth. Platform: Reddit's third-party ecosystem cost nothing for eight years and then ended over eighteen days.

Why it matters. People apply fishery intuitions to aquifers and aquifer intuitions to platforms, and both errors are expensive in opposite directions. The exponential case punishes patience. The step case punishes the assumption that you will see it coming.

You already know this because you have watched a slow budget overrun and a sudden contract termination, and you know one of them could be caught in the fourth quarter and the other could not.

BRIEF 3 — Fishing Mortality, F0.1, and the Cost of Delay

The idea. Fishing mortality F is an instantaneous rate, not a tonnage. Natural mortality M sits alongside it, and a cohort's survival over t years is e^(−(F+M)t). F0.1 is a reference point — the fishing mortality at which the marginal yield per recruit has fallen to a tenth of its value at the origin — and it is used as a ceiling, not a target.

Worked example. Northern cod, late 1980s: realised F ≈ 0.50/yr against a reference F0.1 ≈ 0.20/yr, with M ≈ 0.20/yr. Over the six-year lag:

  survival at realised F        e^(−0.70 × 6)  =  0.0150
  survival at F0.1              e^(−0.40 × 6)  =  0.0907
  ratio                         e^(ΔF × L)     =  6.05 ×

Why it matters. It converts a delay into a number a finance committee understands. Six years of not acting cost roughly five sixths of what was left. And note the form: the cost of a lag is e^(ΔF · L) — exponential in the delay and only linear in the overshoot. Halving the excess buys you less than halving the delay does.

The honest limit. This is a single-cohort survival model. It carries mortality only: no recruitment, no individual growth, no depensation. It therefore understates the real cost, which is the safe direction for an argument to be wrong in, and it should be labelled every time it is used.

You already know this because you have seen interest compound and you know that the last year of a delayed payment hurts more than the first.

BRIEF 4 — The Failure Order

The idea. "Governance failed" is not a finding. Ostrom's eight design principles fail in sequence, and the sequence is the diagnosis. Date each one: working, degraded, gone.

Worked example — northern cod.

#	Principle	What happened	When
1	Clear boundaries	Trap berths drawn by lot stopped defining anything once an offshore fleet arrived	1950s
2	Congruence with local conditions	A rule fitted to a small boat replaced by a quota fitted to a stern trawler	1960s–70s
4	Monitoring	Catch-per-unit-effort rose while the stock fell, because the fleet's electronics improved	1980s
3	Collective choice	Inshore fishermen reporting empty traps had no standing; the 1986 advice went to a federal department	1986–91

Four principles, four dates, one decade. Note that the order is not 1-2-3-4: here monitoring failed before collective choice, and that ordering is the useful part.

Why it matters. A failure with dates is a failure you can build against. A failure without them produces a lesson about human nature, which is unactionable by construction.

You already know this because when a project fails you do not accept "people didn't communicate." You want to know which decision, on which date, made the next one worse.

BRIEF 5 — The Instrument That Lied

The idea. In every collapsed commons, the number people watched moved the wrong way relative to the number that mattered. Find it before you need it.

The question to ask of every governing metric: under what circumstance does this rise while the thing it represents falls?

Worked examples.

Cod: catch per unit effort. A modernising fleet with better sonar and bigger nets keeps its catch rate up on a shrinking, aggregating stock. CPUE is most reassuring immediately before it is useless.
The High Plains: the area-weighted average. 273.2 million acre-feet over 174,000 square miles divides into a comforting aggregate; the Texas Panhandle average decline is 41.5 feet and some sections can no longer run a pivot.
Rapa Nui: a deforestation date used as a proxy for a demographic one. The palms went around 1650. The population fell in the 1860s. Two facts, no causal link, one very popular story.
Reddit: engagement. The 2015 blackout registered as engagement.

Why it matters. If there is no circumstance in which your metric rises while the thing falls, you have not understood the metric yet. That is not scepticism; it is the definition of understanding one.

You already know this because you have seen a sales pipeline look healthiest in the quarter before it emptied, for exactly this reason.

BRIEF 6 — Institutional Response Time (`R`)

The idea. L belongs to the resource. R belongs to you: the elapsed time from a measurement crossing a published line to a rule actually changing.

    R  <  L   the commons holds
    R  >  L   the commons falls

You cannot lengthen L — the biology, the hydrology and the contract set it. The entire art of governing a commons is shortening R.

Worked example. Measured response times in this chapter:

  stratospheric ozone      warning 1985 -> Montreal Protocol 1987     2.3 yr
  Sheridan 6, Kansas       petition 2012 -> rule in force 2013        1.0 yr
  northern cod             advice 1986 -> quota actually cut          > 6 yr

Ozone and Sheridan County did not have better people. They had shorter R.

How to compute yours. Take your last three rule changes. For each, find the date of the measurement that prompted it and the date the new rule took effect. Count the months. Average them. Most institutions have never done this and are startled by the answer, which is itself the most useful output.

Why it matters. It converts "we should be more responsive" into a number on a reporting pack that can go down.

You already know this because you can already say roughly how long it takes your organisation to change a policy, and you know that the number has nothing to do with how bad the problem is.

BRIEF 7 — The Tenancy Test

The idea. A commons in which the appropriators cannot change the rules is not a commons. It is a tenancy.

Ostrom's third principle — collective-choice arrangements, meaning most people affected by the operational rules can participate in modifying them — is the one that distinguishes the two. Everything else can be excellent without it, and frequently is.

Worked example. Reddit's volunteer moderators ran monitoring (4), graduated sanctions (5), conflict resolution (6) and nested communities (8) at a scale no paid department has matched, for roughly fifteen years. They never at any point held the right to modify the terms they operated under. When the API price was set at $0.24 per 1,000 calls in 2023, there was no procedure to contest it, because there had never been one. The institution did not fail. It was exercised.

The test, in one question: what is the procedure by which the people who do the work here change a rule, and when was it last used successfully? If there is no answer, you are a tenant, and you should price your investment accordingly.

Why it matters. It tells you which kind of risk you are carrying. Tenancies fail in step geometry — nothing, nothing, nothing, everything — so the correct response is not vigilance. It is an exit that works, held ready.

You already know this because you know the difference between a house you own badly and a house you rent well.

BRIEF 8 — Overdraft Without an Event

The idea. Some commons never collapse. They just get worse, forever, in public, with nobody able to name a date.

Worked example. The High Plains aquifer. In 1965, in United States v. Shurbet, the Fifth Circuit allowed Southern High Plains irrigators to deduct groundwater depletion from taxable income as the exhaustion of a capital asset. The United States government thereby recognised, in law, that the water was being mined. Sixty years later:

  predevelopment storage      3,183.2 million acre-ft
  2015 storage                2,910.0 million acre-ft
  depletion                     273.2 million acre-ft   =  8.58 %
  aggregate life at mean rate         692 years
  Texas mean decline                   41.5 ft

No collapse. No moratorium. No date.

Why it matters. Institutions act on events. A commons with no terminal event has an unbounded lag, which makes it the most dangerous class in the file — a collapse with no date cannot be put in anybody's calendar. The correction is to manufacture the event: publish a line, and let crossing it be the date.

You already know this because you have a slow structural problem in your own organisation right now that everyone agrees about and nobody has scheduled.

BRIEF 9 — The LEMA: a commons that corrected itself

The idea. Correction is not theoretical. In one Kansas county, the people pumping the water voted to pump less, and the arithmetic came out better than the rule required.

Worked example — Sheridan 6. Petitioned by irrigators in 2012 under the Kansas Groundwater Management District Act of 1972; effective 1 January 2013. Ninety-nine square miles. Allocation: 55 inches per irrigated acre over five years — eleven inches a year, roughly 20 per cent below prior use.

Measured by Deines, Kendall, Butler and Hyndman against a matched counterfactual, published 2019:

  intended cut                      20 %
  measured cut, 2013-2017           31 %
  overshoot on the rule             11 points
  crop production                   no detectable loss

Why it matters. It is the load-bearing positive in this chapter. Failures are only instructive if recovery is possible, and here is recovery with a number on it: a self-written rule, self-monitored, that cut the draw by nearly a third at no measurable cost to income.

The honest negative inside the positive. The water table under Sheridan 6 is still falling — the rule changed the slope, not the sign — and the LEMA covers 0.0569 per cent of the aquifer. It is a proof of mechanism, not a solution, and overselling it is how a good finding gets discredited.

You already know this because you have seen a team, given the number and the authority to act on it, quietly beat the target they set themselves.

BRIEF 10 — The Standstill Facility

The idea. The cost of restraint is immediate and lands on named people; the benefit is deferred and diffuse. That is not a moral problem. It is a financing problem, and financing problems have instruments.

The structure. A pre-committed, trigger-indexed reduction with an escrowed transition fund:

A trigger — one published index, one published line, written before the line is near.
A step — graduated and automatic. Crossing reduces the allocation on a stated date. No meeting; the meeting is what R is made of.
A measurer — independent of every appropriator, paid from the escrow, and obliged to publish whether or not the line was crossed.
An escrow — funded in advance out of good-year rent, sized to cover the gross margin on foregone units for the first years after a trigger.

The number that decides it.

       C_standstill   <   V · ( 1 − e^(−ΔF · L) )

Worked example. V = $100m present value of sustainable rent, ΔF = 0.30/yr excess appropriation, L = 6 years of response time. Value at risk = $83.5m. A three-year escrow at $8m a year = $24m. Decision ratio 0.29 — the lag costs about three and a half times what closing it costs.

Why it matters. It gives a treasurer a reason to fund restraint that is not an appeal to conscience. And note Sheridan 6's version: C ≈ 0, because income was maintained. A standstill that costs nothing needs no escrow — it needs a rule the appropriators wrote themselves, which is the cheapest instrument in this volume and the hardest to buy.

You already know this because you have paid an insurance premium you were glad never to claim on, and you did not experience that as a loss.