Haute Lumière
Commerce · VI · MMXXVI · daylight
For the Practitioner of Luminous Economics
Volume VI — Governance and the Commons · Extension I of III Nine movements, one afternoon.
THE PLATE
THE LETTER
The eleven chapters you have just read were written for people with a secretariat. They price a board's decision rule in loaded hours, they endow a guardian, they commission an attestation, they write a covenant into a five-million-pound facility. Every one of those instruments is real and every one of them assumes somebody whose job is the paperwork.
You are the paperwork.
You are writing the rules for a workshop, a tool library, a housing co-op, a grazing association, a shared kiln, a codebase with eleven committers, a piece of land held by four families. Nobody in your group is paid to govern it. Nobody will read a forty-page rulebook and nobody should have to. And the question you actually have is not what does good governance look like — you have read the chapters and you know — but what can twenty-four people afford, in hours, this year.
So this chapter re-cuts Volume VI into that currency. The mechanisms are the same mechanisms: the enforcement ratio (VI.04) · the graduated ladder and its break-even (VI.09) · the design principles with their honest evidential status (VI.02) · the crossover between a rulebook and a handful of understandings (VI.02) · the gameable measures (VI.10). What changes is the unit. Everything below is priced in hours per member per year, and the whole apparatus comes to 1.33 of them.
One result in here reverses the instinct you arrived with, and it is in the Arithmetic with the sum attached. The famous enforcement ratio — the number that says a surviving commons costs under two per cent of what it governs — is not available to you at any budget, and the reason is the most useful thing in this chapter, because it tells you exactly which move to make instead.
You are not a small version of the consortium in VI.02. You are a different instrument with a different constraint binding, and the constraint that binds you is not money.
— The Editors
DISCOVERY
What is already working, at exactly your size
Begin where the record is strongest, and notice what these five have in common: not one of them was built by an institution, and every one of them still runs, or ran until somebody outside it changed the law.
Törbel, in the Valais, from 1483. A village of a few hundred people wrote articles of association for the alp, the forest and the waste. The rule that carries the whole load is one sentence: no citizen may send more cows to the alp than he can feed through the winter. No herd cap. No metering of grass. The right to take in summer is pegged to hay standing in a barn in February, which every neighbour walks past. VI.04 computed what that swap is worth and put it at 13.33 times — and the people who designed it had no word for design.
The Tribunal de les Aigües, Valencia. Eight elected síndics, farmers, unpaid, sitting outdoors at the Apostles' Door on Thursdays at noon. Oral procedure, no lawyers, no transcript, no appeal, sentences spoken aloud and usually delivered in minutes. VI.04 put its enforcement ratio at well under one per cent of the water's annual value. The operating cost of the oldest working court in Europe is a set of chairs and an hour a week.
A New Mexico acequia. The same rule family, crossing an ocean, and paying for itself in labour rather than salary: sixty parciantes, one compulsory day each at the annual ditch-clearing, a mayordomo on a small stipend. VI.04's computed ratio there is 2.13 per cent — the top of the whole band, and it is the case with no salaried enforcement at all, which is the tell.
Buurtzorg's nurses. Teams capped at twelve, holding their own intake, their own rostering, their own hiring, deciding by consent. VI.01 showed why the cap is not a preference: at a realistic objection rate a body of twelve converges and a body of fifty does not. The smallest working unit of self-government in the modern economy is about a dozen people and a kitchen table.
The copyleft licence. One conditional obligation attached to one event: if you distribute, you distribute the source under the same terms. VI.04 found it won in court twice, in two jurisdictions, and won because it is a boundary rule a judge can read in an afternoon.
And the one that is neither old nor famous. In 2012 the irrigators of Sheridan County, Kansas, used a statute from 1972 to write themselves an allocation, and then cut their own water by nearly a third against a rule that asked for a fifth — with no measurable loss of crop (VI.05). Nobody made them. A room of people with a direct financial interest in not doing it wrote the rule that bound them, and it held.
One pattern, and this chapter is built on it. In every case the thing that travelled was a document small enough for one person to hold: a sentence about hay, a day and a place, a day of labour, a cap of twelve, one conditional clause, one allocation. Not a programme. Not a committee structure. The corporation in VI.01 has to convert its intentions into documents because its people change; you have to do exactly the same thing for exactly the same reason, and you can do it before supper.
So the question for your own group is not what should we build. It is: which of those six documents do we not yet have a copy of?
THE ARITHMETIC
What twenty-four people can afford, and the ratio they cannot
Take a real group and price it, because a number nobody can check is a slogan.
Twenty-four members. A shared stock of 400 tools. Put the hire-equivalent value of one tool at £60 a year — every input here is an assumption, labelled as one in lib/verify/VI_E1.py, and a reader who changes it changes everything downstream, which is the point of printing it. That makes the resource worth £24,000 a year. One member's hour is £22.
First, buy the apparatus and see what happens.
VI.09 costed a graduated sanction ladder properly: £3,500 a year of standing cost — training panellists, holding their time, keeping the record — and £330 per hearing. At its observed dispute rate of 0.05 per member a year this group has 1.20 disputes, so:
standing cost £3,500 / yr
hearings, 1.20 x £330 £396 / yr
------------------------------------------------------------
E, bought £3,896 / yr
V £24,000 / yr
E / V 16.23 %VI.04's whole band, across three continents and seven centuries, is 0.49 to 2.13 per cent. This group is over the two per cent design target by a factor of 8.12.
Now solve it the other way. At £3,896 of governance, the resource has to be worth £194,800 a year to reach two per cent, and £795,102.04 to reach the bottom of the band. A practitioner group cannot buy the apparatus of Volume VI until its shared resource is worth about two hundred thousand pounds a year, and almost none of them are.
Here is the cut, and it is the whole chapter.
The surviving commons did not solve this by finding the money. They solved it by never spending any, and the mechanism has a name in VI.04 — monitoring as a by-product of use — which we can now price at your scale rather than admire at theirs.
Suppose you police the rack properly: 400 tools, checked at three minutes each, 12 times a year. That is 240.0 hours, or £5,280. Now arrange the work so the check happens anyway: every tool is signed out and signed back in, and the person receiving it looks at it for three minutes because they are about to use it. 400 loans a year is 20.0 hours, or £440.
deliberate inspection 240.0 h / yr £5,280
the handover, which happens anyway 20.0 h / yr £440
------------------------------------------------------------
reduction 12.0 x
Toerbel's equivalent (VI.04) 13.33 xTwo systems four centuries and a continent apart, one of them about hay and one about hand tools, landing within a tenth of each other. Both inputs were chosen by one hand, so treat the agreement as a thing worth noticing rather than as evidence — but the mechanism is identical and it is the cheapest move available to anybody governing anything.
Put the whole designed system together and price it in the only currency you actually hold:
monitoring, as a by-product of the handover 20.0 h / yr
the forum: 4 sittings x 3 people x 0.5 h 6.0 h / yr
keeping the record 6.0 h / yr
------------------------------------------------------------
E, designed 32.0 h / yr
£704 / yr
E / V 2.93 %
hours per member per year 1.33 hStill above two per cent — and that is the honest shape of it. The designed system reaches the band only once the resource clears £35,200 a year, not £194,800. Below that, E/V is simply the wrong instrument: it is a ratio invented for a commons whose harvest is economically live, and a tool library's harvest is not measured in money at all.
So the practitioner's number is not a ratio. It is hours. The bought apparatus costs 177.1 hours a year. The designed one costs 32.0 — 5.53 times less — and it lands on twenty-four people as 1.33 hours each, once, annually. That is the figure to put in front of your group, because it is the figure they are actually being asked for.
Second: the biggest fine a neighbour can levy.
VI.04's finding was that the commons that lasted are the ones that punish least, and the structural reason is that a penalty large enough to ruin a neighbour cannot be levied by a neighbour. Give that a threshold at your prices. What does it cost somebody to take your sanction outside?
fee to file £120.00
8 hours to prepare and attend, at £22 £176.00
------------------------------------------------------------
cost of contesting £296.00Set your first sanction below £296 and it is not worth appealing, which means enforcement stays inside the room, which is the condition every low ratio in VI.04 depends on. Above it, you have exported your enforcement to somebody with a filing fee, and you are now paying for the thing the whole design was avoiding.
And the deterrence holds, which is the part that surprises people. Take a violation that gains £40:
the peer rung p = 0.90 S = £50 E = £45.00 sd = £15.00
a distant fine p = 0.05 S = £900 E = £45.00 sd = £196.15
------------------------------------------------------------
severity ratio 18.0 x
variance ratio 171.0 x
detection the peer rung needs 80.00 %
detection the distant fine needs 4.44 %Identical expected penalties. The peer rung deters at a detection rate you can see with your own eyes; the distant fine deters only while detection stays above 4.44 per cent, and it is currently holding that by 0.56 points — a margin nobody in the arrangement measures, on a variance 171.0 times larger. A small certain fine is a cost a household budgets for. A large improbable one is a gamble, and somebody near their margin will take it.
Third: your own break-even, which is lower than the book's.
VI.09 found that the break-even for running a ladder is an absolute count — 2.24 disputes a year, for a commune of fifty or a company of five thousand, because the standing cost is fixed. Recompute it with your standing cost, which has no panellists to train:
F, the standing forum 4 x 3 x 0.5 h at £22 £132.00 / yr
c, one hearing 3 people, one hour £66.00
X, one exclusion £880 of contributed labour
+ 0.10 x £1,200 of claim £1,000.00
q, disputes ending in exit without a ladder 0.30
s, with one 0.05
------------------------------------------------------------
(q - s) X - c £184.00
F / N £5.50
lambda* 0.029891 / member / yr
0.7174 disputes / yrSeventy-two hundredths of a dispute a year. VI.09's threshold for a bought panel is 2.24; yours is lower by 3.12 times, for the single reason that your forum is three people and half an hour rather than a trained panel on retainer. The observed rate of 1.20 disputes is already 1.67 times the threshold. The running cost is £8.80 per member per year against VI.09's £34.00 — 0.40 of an hour each.
There is a floor underneath it and it is worth stating: the ladder has to divert at least 6.60 per cent of disputes from exit, or the hearings cost more than the exclusions they prevent. A forum that hears everything and changes nothing is not cheap; it is a slower route to the same door.
Fourth: when a rulebook beats a handful of understandings.
VI.02 put the crossover at eight members, on a consortium price list with an attestation and a secretariat in it. Yours is a different world — a bilateral understanding between two members costs about two hours a year to keep current, £44 — and the answer is almost the same:
n = 6 15 pairs £660 / yr — understandings still cheaper
n = 7 21 pairs £924 / yr — the rulebook wins
n = 8 28 pairs £1,232 / yr
crossover, exactly 6.18 membersSeven. VI.02's figure, from a price list a hundred times larger, is eight. Two calculations with nothing in common except the shape of the problem — understandings grow as n², a rulebook grows as n — arriving one member apart. If your group is four people, write nothing and talk to each other. From about seven, the sheet of paper is the cheaper instrument, and saying so plainly is the most credible line in any proposal to write one.
Fifth, the honest negative, and it is aimed at you rather than at the literature.
You cannot audit your own governance for three and a half years. VI.10's sampling arithmetic is not kind to a small group. Estimating any proportion — what share of disputes reached the third rung, what share of decisions were taken by exception — to ±10 points needs 96.04 decisions, and decisions inside one meeting are not independent draws, so at VI.10's design effect of 2.05 it needs 196.87. A group taking twenty-four governance decisions a year gets there in 8.20 years. Even ±15 points takes 3.65.
That is a real limit and it cannot be spent away. Three exits, and all three are better than the alternative of believing your own impressions:
Audit the by-product record, not the decision record. Your 400 loans a year are a sample of 400, and they give you ±4.90 points on compliance in one year. Getting the same precision out of governance decisions would take 34.17 years. The record that was written for another purpose is the one VI.10 says to read, and at your scale it is not merely cleaner — it is twenty times larger.
Pool with two other groups. Three tool libraries or three housing co-ops sharing a coding sheet make seventy-two decisions a year between them and reach ±15 points in 1.22 years. VI.10's finding that clustering lives in the meeting means sampling across bodies buys precision that sampling across time cannot.
And borrow a base rate rather than inventing one. VI.02 is explicit that the conversion from a design-principle score to a survival probability has never been measured. Score your group by all means; read the diagnosis; fix what it names. Do not convert the total into odds, because the exchange rate does not exist.
DREAM
What becomes ordinary
In the group that has done this, the constitution is on the wall and it is one sheet.
It has a date at the top, in a hand somebody recognises, and eight numbered lines under it. New members are handed the sheet rather than the culture, because a culture cannot be handed over and a sheet can. When a line is changed, the change is dated too and the old sheet goes in the file, so that four years later anybody can see what the rule was when the thing happened.
Nobody watches anybody. The watching is arranged into the work: the tool comes back and the next person looks at it before they use it, the turn passes to the neighbour who is standing there because it is his water next, the commit is visible because commits are. When somebody proposes a rule, the first question in the room is Törbel's question — what would a person have to hide to break this, and could they? — and if the answer is that they could, the rule is rewritten to bind something nobody can hide. This takes ten minutes and saves 240.0 hours a year.
The forum sits four times a year whether or not there is anything to hear, and most sittings last a few minutes and find nothing. That is not waste. It is six hours a year, six pounds of everybody's time, and it is what makes the forum available on the week it is needed without anybody having to be brave enough to convene it. Nobody ever proposes that there be a hearing. The hearing already exists.
First sanctions are small, immediate and public, and nobody mistakes that for weakness, because everybody has seen the arithmetic that says certainty is doing the work. Fines go into a restricted pot that buys blades and belts and a new lock, never into anybody's pocket and never into general funds, so nobody in the room has an interest in an infraction.
And the annual meeting has a number on it. Not a mood: thirty-two hours, one and a third each, and ten pounds forty-five. People know what their governance costs them, the way they know what the electricity costs, and the reason they keep paying it is that they can see the figure and it is small.
DESIGN
The afternoon, and the eight lines
This is the deliverable. It is one sheet and it takes an afternoon, and the order below is the order to write it in, because each line makes the next one cheaper.
The boundary, with a date. Who holds a right here, and what the resource is. Name the date the sheet was written and keep it. VI.04's sharpest finding on this is that a boundary you cannot date is not a boundary but a mood — and the Epping Forest lesson beside it: draft it so that one member's single right is sufficient standing to challenge a breach. Not a majority, not a quorum, not a committee. One.
The contribution rule, as an equation. What each member gives, expressed in the same units as what they take. Törbel pegged cows to hay; the zanjera pegged water to labour; your version might be an hour of maintenance per ten loans, or a subscription per tool-week, or a commit reviewed per commit merged. VI.02's record says the absence of this is one of the four things most strongly associated with failure, and it is the line most groups leave out because it feels ungenerous. Write it anyway. It is what stops the provision side thinning while every other indicator still reads healthy.
The visible proxy. For each rule, the quantity that moves with use and cannot be concealed. Write the proxy in the rule rather than writing the intention. This is the line worth 12.0 times its own length.
The day. A fixed date and a fixed place for the forum, four times a year, sitting whether or not there is business. Six hours a year, all in.
The ladder, with rungs. A word · a recorded word · a facilitated conversation · a finding with conditions · suspension of a specific right · exit. A rung is used twice before the next is available. The first rung's maximum is £296, because above that figure it is worth appealing and your enforcement has left the building.
The escrow. Fines go to a restricted fund applied to the resource, with the schedule published. One sentence, and it is the sentence that keeps the panel disinterested.
The backstop. A named external mediator on a small retainer, invocable by either side — by a member against the group as readily as by the group against a member. That both-ways clause costs nothing and is the whole of its legitimacy.
The amendment rule, set higher than anything it governs. VI.01's warning is the one to obey here: a rule changed by the rule it governs, mid-question, by people who can see which way it will fall, is not a rule. Set the threshold for changing the sheet above every threshold on it, and set it before you need it.
And what not to write. Do not write a score. Do not write a governance dashboard. VI.10's finding is that three of the six measures worth having are gameable by the only people positioned to collect them — participation by convening, latency by starting the clock late, reversal by never formally reversing — and a small group has no independent auditor to route around that. What a small group can do instead is count the by-product: loans out and back, turns taken, days worked at the clearing, tools missing at the annual count. Those records were written for another reason, which is exactly why they can be trusted.
DESTINY
How it holds when you are ill, or bored, or busy
Three things keep a practitioner's charter alive after the person who wrote it has stopped caring about it, and all three are structural rather than moral.
The monitoring is a by-product, so it cannot be cut. A funded watcher is the first line struck in a thin year, and when it goes the rule goes with it within two seasons. A handover check survives because it is not a line at all — it is the way the tool gets from one person to the next.
The forum has a date, so it does not need a convenor. The failure mode of every small group is that the meeting requires somebody to be brave, and bravery is not a renewable input. The Water Court has a Thursday, not a procedure for getting one.
The document outlives everybody, so the rule is inherited rather than renegotiated. Nobody in Törbel signed 1483, and inheritance is far cheaper than consent.
Now the failures, plainly, because the point of naming them is that each has a counter you can write in advance.
It fails when the proxy stops being visible. Move the stock into a locked room, let the loans be recorded privately, mechanise the handover, and the free monitoring evaporates while the rule reads exactly as it did. The £440 becomes £5,280 and nobody has amended a word. The counter is to treat any change in how the thing is used as a change to the constitution, and to look again at the proxy.
It fails when the first sanction is raised. Somebody is angry, the fine goes up, and above £296 the next dispute goes outside — where certainty collapses, cost rises by an order of magnitude, and the ratio that made the whole design affordable is gone. The counter is that the ceiling is written on the sheet as a number rather than as a principle.
It fails when the ladder collapses to its top rung. Under stress, groups stop using the middle rungs because they take three weeks and reach for exit because it takes an afternoon. The tell is in your own count: rung-three outcomes falling while rung-six outcomes hold. The counter is the rule that a rung is used twice before the next is available.
And it fails where the arithmetic never supported it. Below seven members, the sheet costs more than talking to each other, and a group of four with a constitution has bought a costume. VI.09's floor applies too: if the ladder diverts less than 6.60 per cent of disputes from exit, it is a slower door. Both of those are computable before you write anything, which is the reason to compute them.
DELIGHT
What it feels like
There is a specific pleasure in a rule that enforces itself, and it is available to anybody with a rack and a marker pen. The painted outline behind each tool is a governance instrument: you cannot fail to notice the gap, nobody has to inspect anything, and the person who put the saw back in the wrong place puts it right without a conversation ever happening. It cost twenty minutes and a tin of paint and it does the work of a monitor.
Then the quieter one, which arrives on the first Thursday with nothing on the agenda. Three people, a kettle, four minutes, nothing to decide. It feels faintly absurd the first time and then it does not, because the fourth time somebody brings something they have been carrying for a year — and the reason they bring it is precisely that the meeting was already happening and they did not have to ask anyone for it.
And the best of it is the sheet itself. One side of one page, eight lines, a date, and a group of people who can all say what the rules are without looking. Most organisations a hundred times the size cannot do that. You can, not because you are better at governance but because thirty-two hours is all you had, and a constraint that severe produces a document a human being can hold in their head — which is the only kind that ever gets obeyed.
OPERATIONALIZE THIS
At the level of finance
The instrument: a one-sheet commons charter with a sanctions escrow and a both-ways backstop.
Small, cheap, and in a form a credit union, a community lender or a grant-maker will recognise immediately — because what they are being handed is not a promise, it is a governance cost with a denominator beside it.
The structure. In England and Wales, most naturally a co-operative society or a company limited by guarantee, with the charter as the rules rather than as a policy, because policies are amendable by whoever is in the room. Three things belong in the rules and nowhere else: the boundary and its amendment procedure; the forum's fixed day; and the sanction ladder with its first-rung ceiling.
The membership right. A right of use appurtenant to membership, not transferable separately from it. That non-severability is what stops rights leaking to people with no stake in the resource, and it is the clause that makes the register mean something ten years out.
The escrow. Fines are not income. They go to a restricted fund applied to the resource with a published schedule. The moment fines fund the enforcers, every member can see that the enforcers have an interest in infractions, and they will be right.
The backstop, priced. A named mediator on a retainer of £240.00 a year plus £450.00 per hearing actually held, with 0.02 of disputes reaching them: an expected £250.80 a year, or £10.45 per member. It has to prevent 0.2508 exclusions a year to pay for itself, against an exclusion costing £1,000.00. What that money actually buys is not the hearings — it is the compliance in the disputes that never reach one, because the terminal rung sets the price of every rung beneath it.
The whole thing, totalled. 32.0 hours a year, 1.33 per member, and £250.80 of cash — 1.05 per cent of the resource in money and 3.98 per cent counting the hours at their full rate. That is the figure to put on the front page, and it is the figure a lender should be shown, because a borrower whose governance costs a twentieth of what it governs is a borrower whose covenants can be monitored cheaply.
The counterparty. Your own members first, through a subscription set to cover the cash line with a small reserve. Only after two years of a stable figure take the charter outward — and then the ratio is the credit case, not the pitch.
The number that decides it. Two, and neither needs an accountant:
hours of governance per member per year < 2.0
and: can a member apply the first sanction today,
on their own, without anybody's approval?If the first fails, the design is too heavy for the group and the move is to find the proxy you cannot hide — that is what takes 240.0 hours to 20.0. If the second fails, your enforcement is already exported and the hours figure will not hold whatever the sheet says.
The first ninety days.
| Day | Action | Artifact |
|---|---|---|
| 1–7 | Write the boundary and date it; list every right-holder | The dated sheet |
| 8–21 | For each rule, name the visible proxy or drop the rule | The proxy table |
| 22–35 | Compute V in the units the members already use | One line of arithmetic |
| 36–50 | Count the hours the design costs; test against two per member | The hours page |
| 51–65 | Write the six rungs and the first-rung ceiling; open the escrow | Ladder and escrow |
| 66–80 | Name the mediator; sign the retainer with the both-ways clause | The retainer |
| 81–90 | Hold the first forum with nothing on the agenda | Minutes of an empty sitting |
That last line is the one to keep. A forum that first appears on the day it is needed is a tribunal. A forum that has already met once about nothing is furniture, and furniture is what you want.
APPRECIATIVE QUESTIONS
Twelve, for a room
Discovery — what is already working
- What does somebody here already notice without being asked to look, and what is it about how the work is arranged that makes them see it?
- Name a rule in this group that everybody keeps and nobody enforces. What visible thing is it attached to?
- Think of a disagreement here that ended well. How long did it take, what did it cost the person who raised it, and what made that cheap?
Dream — what becomes possible
- If our whole constitution took each of us under two hours a year, what would we be willing to hold in common that we do not hold now?
- Imagine the sheet on the wall three years from now, with one line on it we have not thought of yet. What is that line about?
- If any one of us could challenge a breach alone, without needing anyone else to agree, what would change about how carefully we write?
Design — what we build
- For the rule we most want kept: what would somebody have to hide to break it, and could they? If they could, what is the version that binds something they cannot?
- What is the smallest first sanction a member could apply this afternoon without asking anyone — and what makes the noticing public?
- What is the date on our boundary document, and where is a copy held that is outside any one of our hands?
Destiny — how it holds
- What would we see first if the contribution side were thinning while everything else still looked healthy? Who is watching that?
- If the person who wrote the sheet stopped coming, which line would go first, and what would hold it in the meantime?
- What would we want the next three groups like ours to be able to learn from us — including from the parts that do not work?
WORKS CITED
Acheson, J. M. (2003). Capturing the Commons: Devising Institutions to Manage the Maine Lobster Industry. University Press of New England.
Baggio, J. A., Barnett, A. J., Perez-Ibarra, I., Brady, U., Ratajczyk, E., Rollins, N., Rubiños, C., Shin, H. C., Yu, D. J., Aggarwal, R., Anderies, J. M. and Janssen, M. A. (2016). "Explaining Success and Failure in the Commons: The Configural Nature of Ostrom's Institutional Design Principles." International Journal of the Commons, 10(2), 417–439.
Becker, G. S. (1968). "Crime and Punishment: An Economic Approach." Journal of Political Economy, 76(2), 169–217.
Cox, M., Arnold, G. and Villamayor-Tomás, S. (2010). "A Review of Design Principles for Community-based Natural Resource Management." Ecology and Society, 15(4), 38.
Deines, J. M., Kendall, A. D., Butler, J. J. and Hyndman, D. W. (2019). "Quantifying irrigation adaptation strategies in response to stakeholder-driven groundwater management in the US High Plains Aquifer." Environmental Research Letters, 14(4), 044014.
Glick, T. F. (1970). Irrigation and Society in Medieval Valencia. Harvard University Press.
Kish, L. (1965). Survey Sampling. Wiley.
McEwen, C. A. and Maiman, R. J. (1984). "Mediation in Small Claims Court: Achieving Compliance Through Consent." Law & Society Review, 18(1), 11–49.
Mnookin, R. H. and Kornhauser, L. (1979). "Bargaining in the Shadow of the Law: The Case of Divorce." Yale Law Journal, 88(5), 950–997.
Netting, R. McC. (1981). Balancing on an Alp: Ecological Change and Continuity in a Swiss Mountain Community. Cambridge University Press.
Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.
Ostrom, E. (2005). Understanding Institutional Diversity. Princeton University Press.
Rivera, J. A. (1998). Acequia Culture: Water, Land, and Community in the Southwest. University of New Mexico Press.
Schweik, C. M. and English, R. C. (2012). Internet Success: A Study of Open-Source Software Commons. MIT Press.
United Kingdom. Co-operative and Community Benefit Societies Act 2014; Commons Act 2006.
Wilson, E. B. (1927). "Probable Inference, the Law of Succession, and Statistical Inference." Journal of the American Statistical Association, 22(158), 209–212.
Note on figures. Every figure in this chapter is computed in lib/verify/VI_E1.py and reproducible with python3 lib/verify.py VI.E1, which prints its inputs and its intermediate terms before its results. The group, its resource, its prices and its escalation shares are scenario inputs and are labelled ASSUMED in that module; a reader who changes them changes every figure downstream, which is why they are printed. Figures carried from the eleven chapters of this volume — VI.04's enforcement band and its Törbel reduction, VI.09's break-even of 2.24 disputes and its £34.00 per member, VI.02's eight-member crossover, VI.10's design effect — are labelled CITED with their chapter, so that what this chapter computed and what it borrowed can be told apart on the page.