Haute Lumière

Commerce · VI · MMXXVI · daylight

La Bourse  /  Volume VI  /  Nº VI.E3

The Translation — One Economy, Two Vocabularies

Volume VI — Governance and the Commons · Extension III of III Nine movements, one dictionary.

---

THE PLATE

---

THE LETTER

Two vocabularies have grown up around governance and they do not speak to each other, which costs both of them.

On one side: boundaries, appropriators, provision, monitoring, graduated sanctions, collective choice, nested enterprises. On the other: shareholders, reserved matters, transfer pricing, internal audit, the disciplinary ladder, special resolutions, segment reporting. A practitioner fluent in the first is routinely told that none of it applies inside a firm. A controller fluent in the second is routinely told that a commons is a different kind of thing altogether.

They are mostly the same objects. Not analogies, not metaphors, not inspirations — the same object, drafted twice, by people who never met, and in several cases with the same arithmetic underneath. This chapter writes the dictionary: sixteen rows, each one object under two names, each with the number or the citation that shows it is one object rather than two that rhyme.

Then it does the more valuable half. Seven pairs look like translations and are not, and a false friend in a governance document is more expensive than an untranslated term, because an untranslated term gets looked up and a false friend gets used. Those seven are set out in full, with what each mistranslation costs.

One of the seven is the sharpest sentence in this whole volume and it belongs in the failure list rather than the table, because it is a place where the two vocabularies agree perfectly and are both wrong.

And there is a direction in all of this that nobody notices until the rows are laid out together, which is the cut this chapter turns on.

— The Editors

---

DISCOVERY

Where somebody has already written the dictionary

Begin with the translations that already exist, drafted by people doing something else, because they are the strongest evidence that the two vocabularies are describing one world.

IFRS 10 translated principle one into accounting. Ostrom's first design principle asks who may withdraw from the resource and where the resource ends. The consolidation standard asks who has power over the relevant activities of an investee — and VI.01 showed that the answer is written in the reserved-matters schedule, where a minority holder with substantive participating rights over those activities can prevent the majority holder from consolidating while merely protective rights cannot. A boundary and a consolidation boundary are the same decision about whose voice counts, and one of them has a standard-setter, a diagnostic, and an auditor who tests it every year.

FINRA translated principle seven into securities law. Minimal recognition of the right to organise — the weakest and most-skipped of Ostrom's principles — is written into the Exchange Act as a filing regime: the industry writes the rules and the regulator may approve, and may abrogate. VI.11 read that as the interface at its best. Read it the other way and it is a commons whose charter with its sponsor was actually drafted, which is exactly the clause VI.02 found missing from the largest digital commons in the world.

Te Awa Tupua translated a guardianship into a fund. VI.07's finding is that a right nobody is paid to exercise is a sentence, and the settlement that worked carries a named contestable fund behind two named appointees. In corporate vocabulary that is a restricted fund with a published draw rate and a remuneration schedule — three instruments a treasurer uses every quarter, doing the work that ten other jurisdictions tried to do with a declaration.

Morning Star translated the peer contract into a document. The Colleague Letter of Understanding (VI.03) is what a commons calls a congruence rule between appropriation and provision, written as a contract between peers with metrics attached, negotiated annually with about ten counterparties. Nobody at Morning Star has read Ostrom. They wrote principle 2B anyway, because it is what the problem requires.

The German cooperative banks translated a risk pool into a federation. One function assigned upward since 1934 and held through a ninety per cent consolidation of the members beneath it (VI.08), which in the other vocabulary is mutual insurance with a levy, and in this one is Ostrom's eighth principle with a balance sheet.

And the Loan Market Association translated governance into a margin. VI.10's covenant works because sustainability-linked lending already has the machinery — agreed targets, an independent verifier, a margin that steps. Nothing had to be invented. Somebody had already built the channel for exactly this shape of claim and pointed it somewhere else.

One pattern, and the chapter is built on it. In every case the translation was performed by whoever needed the thing to be enforceable, and enforceability is what forced the vocabularies together. That is the useful news: you are not proposing a synthesis. You are noticing one that is already in force, in statutes, in standards and in signed agreements, and writing it down so the next person does not have to rediscover it.

---

THE ARITHMETIC

The table, the false friends, and the direction

First, the table. Sixteen rows, each one object.

Twelve of those sixteen rows — 75.0 per cent — already have a recognised accounting or legal treatment on the corporate side: a standard, a statute, or a document an auditor tests. Four do not. That is the whole of what a commons would have to invent, and it is a short list.

Second, the false friends. Seven pairs that are not one object.

One — the excluded party is not the externality. This is VI.11's count and it is the sharpest thing in the volume. Ostrom's principles are eight; after Cox, Arnold and Villamayor-Tomás split three of them they are 11; and the number of them that requires a group to weigh a cost it imposes on somebody outside its boundary is 0. The corporate vocabulary appears to have the answer — externality — and does not: an externality is a term of art in welfare economics with no procedural consequence inside a company, and the routes that do carry consequence, a derivative claim or a members' resolution, are members' routes. Both vocabularies are blind to the same party, and the translation succeeds perfectly, which is precisely why this is a failure and not a row: a table would carry the blindness across intact and leave it looking handled. VI.11's worked instance is land use, where 63 per cent of hearing commenters oppose and the cost of the missing party is put at 3.7 per cent of national output in the published paper — a figure the same authors' working paper had at 9.5 per cent and revised down by a factor of 2.57, which is quoted here in its published form for the reason VI.11 gives.

Two — a graduated sanction is not progressive discipline. They look identical: a ladder, rungs, escalation. VI.04's finding is that the surviving commons are the ones that punish least, because a first sanction must be small enough for a peer to apply today. Corporate discipline is drafted to be defensible to a tribunal in eighteen months, which means its first rung is already outside the room. The severity ratio between a distant enforcement and a peer one is about 50.00 times — same expected penalty, entirely different distribution — and certainty is what was doing the work.

Three — monitoring is not internal audit. Principle four is a compound claim and the compound is the point: monitors audit both the resource and the behaviour, and are accountable to the appropriators or are the appropriators. VI.02 found that the accountability half — not the environmental half — was the single best-evidenced element in the whole meta-analysis. Internal audit reports to the audit committee, which is upward. It satisfies the paraphrase and reverses the principle, and the reversal is invisible because the noun is the same.

Four — an attested governance score is not a credit rating. Same shape: an assessor, a scale, a published grade. A rating converts to a default probability through a calibration table built on observed defaults. A governance score does not, because the likelihood ratio has never been measured for a commons — nobody has assembled the dead. VI.02's sweep across three plausible ratios, at a base rate of 31.78 per cent and prior odds of 0.4658, gives:

  likelihood ratio 1.2     2.54 bits     posterior  73.0784 %
  likelihood ratio 2.0     9.67 bits     posterior  99.7366 %
  likelihood ratio 4.0    19.33 bits     posterior  99.9997 %
  ----------------------------------------------------------
  band                   16.79 bits

The same score on the same institution supports anything from seventy-three per cent to five nines. A score licenses a diagnosis. A rating licenses a price.

Five — consent is not unanimity, and neither is consensus. VI.01's arithmetic: a consent rule, which admits only argued and paramount objections, is a two-thirds supermajority wearing unanimity's clothes, and the difference is billed — precisely — to whoever cannot phrase a preference as a workability argument. Translating consent as unanimous in a shareholders' agreement grants a veto the drafters did not intend and priced at nothing.

Six — the participatory share is not engagement. VI.06 defines s = P / B, the share of total spending actually placed under participatory control, and found New York's eighth cycle at one part in two thousand two hundred and eighty-six. A firm reporting engagement is reporting the room. Turnout is a fact about the room and the share is a fact about the money, and no amount of the first will tell you the second.

Seven — a right appurtenant to membership is not a transferable share. The non-severability that VI.04 traces through English commons law is what stops rights leaking to people with no stake in the resource, and it is also what fixes the holder's horizon to the resource's. Put the two clocks beside each other with VI.11's table: a forest stand turns over in 30.0 years and soil carbon in 100.0, against a delegated charter term of 5.0 to 10.0 — a mismatch of 3.0 to 20.0 times. The share is designed to be sold; the right is designed to be inherited, and a document that treats them as one object has quietly shortened somebody's horizon by a factor of ten.

Third — the cut.

Lay the seven out and something that no single row shows becomes obvious. In all seven — 100.0 per cent — the corporate term is the one whose route ends at an authority outside the parties, and the commons term is the one whose first move is cheaper. The tribunal behind the discipline. The audit committee behind the audit. The rating agency behind the rating. The court behind the share.

Which means the dictionary runs downhill in one direction and uphill in the other. A firm can adopt any commons instrument by writing it into a document it already has — a ladder into the handbook, a fixed forum day into the calendar, a congruence rule into the chargeback, a boundary into the reserved matters. It costs drafting and nothing else. A commons cannot adopt a corporate instrument without buying an outside authority it does not have, and that purchase has a price, which this volume has repeatedly computed without ever putting the figures on one page:

  a trial against a one-day mediation, in elapsed days        913.00 x
  a distant fine against a peer fine, in severity              50.00 x
  a trial against an internal panel, in elapsed days           43.50 x
  a funded guardianship against an unfunded one, in days held   31.80 x
  settling standing against being granted it                    6.25 x
  five years of archive against one year of the room            5.00 x
  the tort system, cost per pound actually delivered            2.17 x
  ----------------------------------------------------------------
  geometric mean                                               23.77 x
  median                                                       31.80 x
  dispersion, largest over smallest                             420.0 x

And the dispersion is the finding. There is no exchange rate between the two vocabularies — a spread of 420.0 times is not a conversion factor, it is a warning that each crossing has to be priced on its own. What there is instead is a direction, and it is uniform. The tort system's row makes it concrete in the plainest possible terms: of every pound the coercive route spends, about 46 per cent reaches the injured party, so the route costs 2.1739 per unit delivered and 1.1739 of that is the price of finding out who was right.

So the practical instruction is the reverse of the usual one. The commons literature is routinely offered to firms as inspiration. It should be offered as paperwork, because that is the direction in which the translation is free.

Now the honest negative, and it is about this table rather than about either vocabulary.

Every count above is a count over a list one person wrote. Sixteen rows and seven false friends, coded once, by one hand, with no second rater and no published disagreements. That is exactly the fault the commons literature spent a decade repairing in itself — Ratajczyk and colleagues wrote a whole paper on the difficulty of coding the commons reliably, and Cox's review was believable precisely because its coding protocol was published and its inter-rater reliability reported.

Size it rather than apologise for it. At 16 rows, the widest half-interval on any proportion — the share of rows a second rater would agree with, for instance — is ±24.5 points. To get that to ±10 points needs 96.04 rows, which is 80.04 more than exist. The claim that 100.0 per cent of the false friends run in one direction is therefore a claim about seven cases, and seven cases is one honest sentence, not a law.

Three exits, and all three make the instrument better than it would have been:

Write the other eighty rows. The corpus is there — every reserved-matters schedule, every co-operative rulebook, every charter in VI.11's table. Eighty rows is a week and it converts a striking observation into a measurement.

Code them twice, blind, and publish the disagreements. The disagreements are the valuable part: a pair that two fluent readers classify differently is precisely the pair a drafter will get wrong.

Publish the rubric before the rows. Same object means a reader can point to one arithmetic identity, one statute, or one document that governs both. Anything weaker is a cousin, and cousins go in a separate column where they cannot be mistaken for translations.

---

DREAM

What becomes ordinary

In the world that has absorbed this, a governance document is filed once and read twice.

A co-operative's rules carry a translation column, and the column is not decoration: beside each clause sits the corporate instrument it corresponds to, so that a lender's credit team, an auditor and a member's solicitor can all read the same page without anybody hiring an interpreter. The clause that says appropriation shall be proportional to provision says, in the next column, transfer pricing basis, documented, and the conversation with the auditors takes twenty minutes instead of a season.

Firms carry the same column the other way. The disciplinary policy notes which rung a peer may apply today and which requires the machinery, because somebody read the severity arithmetic and noticed that the first rung had been drafted for a tribunal rather than for a Tuesday. The shared-services charter names its member classes and its contribution formula, and both words come from a literature the firm has never cited and does not need to.

The four rows that have no corporate treatment have acquired one, because they were a short list and short lists get finished. Somebody wrote the accounting note for a sanctions escrow. Somebody else wrote the disclosure for an enforcement ratio, so that a commons' accounts show what its own government cost beside the value of what it governed.

And the false friends are known by name. A drafter meets the word consent in an agreement and checks which of the two it is, the way a translator checks a word that means actually in one language and currently in another. That check costs nothing and it is the difference between a two-thirds rule and a veto that nobody priced.

---

DESIGN

Building your own, in six steps

One — write the rows for your own institution, not this one. Sixteen rows is the shape; yours will share perhaps half of them. Start from the documents you already have to file, because a row that does not end in a document is a metaphor.

Two — for each row, name the thing that makes it one object. One arithmetic identity, one statute, or one document that governs both sides. s* = h/(1 + h) is an identity. The Companies Act's seventy-five per cent is a statute. IFRS 10's participating-versus-protective distinction is a standard. If the best you can produce is a resemblance, it is a cousin and it goes in the other column.

Three — hunt false friends deliberately, because they will not present themselves. Three tests, and any one of them failing makes a pair a false friend rather than a row: does the first move cost the same on both sides? does the accountability point the same way? and is either side calibrated where the other is not? Those three caught six of the seven above.

Four — price each crossing separately and never average them. The dispersion of 420.0 times is the reason. A crossing that takes a dispute from a panel to a court is 43.50 times; a crossing that takes it from a mediation to a trial is 913.00. One number for both is worse than no number.

Five — budget the crossings you intend to make. A bilingual schedule that names an outside authority without funding anybody to reach it is VI.07's declaration, and the record says a declaration lasts somewhere between a hundred and nine days and however long it takes somebody with a budget to notice it.

Six — code it twice. Two readers, blind, disagreements published. This is the step that will be skipped and it is the one that turns a table into an instrument.

---

DESTINY

How a dictionary stays true

Three things keep a translation honest, and none of them is scholarship.

Every row ends in a document somebody has to file. A row that is only an observation drifts, because nothing contradicts it. A row that is a clause in a filed agreement gets corrected by whoever has to sign it.

The rubric is published before the rows. Same object means an identity, a statute or a document. Published first, the rubric constrains the table; published afterwards, it describes it.

The false friends are maintained with more care than the table. A row that is right is useful once. A false friend that is wrong is a clause in somebody's constitution.

Now the failures, plainly.

It fails by becoming a metaphor generator. Anything can be said to resemble anything. The counter is the rubric, and the discipline of putting cousins in a separate column where nobody can cite them as translations.

It fails by flattening the direction. Once a table exists it is tempting to read it as symmetrical, and it is not: 100.0 per cent of the false friends have the outside authority on the corporate side. A commons that adopts corporate paperwork believing the exchange is free will discover the cost at the first dispute, at somewhere between 2.17 and 913.00 times what it expected.

It fails when the excluded party becomes a row. That is the one entry that must stay in the failure list. Put it in the table and it reads as handled, and the whole apparatus becomes what VI.11 warned it could become: a specification for a very effective cartel, in two languages.

And it fails when nobody re-codes it. At 16 rows the interval is ±24.5 points, and a table that stays at sixteen rows for a decade has not become more reliable by being quoted more often.

---

DELIGHT

What it feels like

There is a specific pleasure in watching two people discover they have been arguing about the same object. A controller says reserved matters and a grower says who has to agree before the water moves, and somewhere in the third sentence both of them stop, and one of them says: say that again.

Then the better pleasure, which is the one that belongs to this chapter. You build a guardian's cost from five lines and land four per cent from a settlement schedule you had not read (VI.07). You price a rulebook against a handful of understandings at a tool library's prices and land one member away from a consortium's answer at a hundred times the scale (VI.02). Two people who never met, reasoning from the same structure in different currencies, arriving at the same place. That is what it feels like when the object is real rather than the vocabulary — and it is the only evidence a dictionary can offer that it is describing something.

And there is a small, slightly mischievous joy in the direction. For a century the traffic has been assumed to run the other way: the commons as a charming antecedent, the firm as the serious instrument. The table says the firm has already written down two-thirds of what the commons literature spent forty years describing — and that the four rows it has not written down are, on the record, the cheapest governance any of these institutions own.

---

OPERATIONALIZE THIS

At the level of finance

The instrument: a bilingual governance schedule, filed once and read by both sides, with the crossings budgeted.

One document. In a co-operative it is the rules; in a company it is the shareholders' agreement or the group policy; in a consortium it is the governance deed of VI.02. What makes it bilingual is a third column, and what makes it an instrument rather than a glossary is a budget line under it.

The mechanics.

• Three columns, not two. The clause · the corporate instrument it corresponds to · the identity, statute or standard that makes them one object. A clause whose third column is empty is a cousin and is marked as one.
• A false-friends appendix, short and maintained. Seven entries is enough to start, and the first four sentences of each are worth more to a drafter than the whole table.
• The four uninvented rows get a treatment. Of the sixteen, 12 already have a recognised accounting or legal treatment on the corporate side and 4 do not. Write the note for those four — the sanctions escrow, the enforcement ratio as a disclosed line, the peer-applicable first rung, the contribution formula as an equation — and you have completed the dictionary for your own institution.
• The crossing budget, and it is the line that makes the rest true. For a consortium appropriating £4,500,000 a year, VI.11's advocate for the party the boundary excludes costs between £45,000 and £135,000 a year at one to three per cent — endowed at VI.07's 4.0 per cent real draw, that is £1,125,000 to £3,375,000, and the solvency test comes out at 1.00 by construction. Beside it sits VI.09's arbitration backstop and VI.04's sanctions escrow. Three lines, and each of them buys one crossing that the vocabulary alone cannot make.
• The levy that funds it. VI.11's worked charter levies £27.67 per member against a right worth £520 — 5.3 per cent, one twentieth of the value of the thing being protected. That is the ratio to take into the room, because it is the one that answers the only real objection.

The balance-sheet treatment. Nothing exotic, which is the point. The advocate's endowment is a restricted fund with a published draw. The escrow is restricted cash applied to the resource, never income. The levy is an operating cost at the payer and restricted income at the recipient, disclosed by name and amount. The enforcement ratio is one disclosed line, E, printed next to V — a commons whose accounts do not show what its own government cost is not governing itself, it is being governed by whoever absorbs the cost — and a firm that discloses the same pair has made its governance legible to a lender for the first time.

The counterparty. Your own auditor first, at drafting rather than at review, because every contested row in this table is a question about the substance of rights, which is a conversation they have annually. Then a lender, who now has three measured numbers rather than an assurance.

The number that decides it. One line per crossing, on the front page:

      funded annual cost of reaching the outside authority
     ------------------------------------------------------   >=  1
        annual cost of the office that has to reach it

Where that holds, you have an instrument in both languages. Where it does not, you have a clause that reads well in one of them, and the record says what that is worth.

The first ninety days.

---

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

1. Where in our own documents has somebody already translated one of these objects without calling it a translation — and who did it?
2. Which of our governance clauses would a lender, an auditor and a member all read the same way today? What makes that one legible?
3. Think of a time two people here discovered they were arguing about the same thing under different names. What did it unlock?

Dream — what becomes possible

1. If our rules carried a third column and everyone outside could read them, what would we be asked for that we are not asked for now?
2. Imagine the four rows with no accounting treatment all have one, five years from now. Which one do we most want written first, and why that one?
3. If a drafter here could check a word against a list of false friends in ten seconds, which word would they check first?

Design — what we build

1. For our most important clause: is there an identity, a statute or a standard that makes both sides one object — or is it a resemblance?
2. Which of our terms is a false friend we are currently using as a translation, and what has it cost us so far?
3. Who would be our second reader — someone fluent in both vocabularies who did not write the table?

Destiny — how it holds

1. What would have to be true for this schedule to still be accurate when everyone who wrote it has moved on?
2. Which crossing have we named but not funded, and what is the annual figure that would fund it?
3. Who is outside our boundary, and what would it take for their name to appear on the front page of this document rather than in its failure list?

---

WORKS CITED

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.

Coase, R. H. (1937). "The Nature of the Firm." Economica, 4(16), 386–405.

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.

Einstein, K. L., Palmer, M. and Glick, D. M. (2019). Neighborhood Defenders: Participatory Politics and America's Housing Crisis. Cambridge University Press.

Hansmann, H. (1996). The Ownership of Enterprise. Harvard University Press.

Hsieh, C.-T. and Moretti, E. (2019). "Housing Constraints and Spatial Misallocation." American Economic Journal: Macroeconomics, 11(2), 1–39.

International Accounting Standards Board (2011). IFRS 10 Consolidated Financial Statements. IFRS Foundation.

Kakalik, J. S. and Pace, N. M. (1986). Costs and Compensation Paid in Tort Litigation. RAND Institute for Civil Justice, R-3391-ICJ.

Loan Market Association (2023). Sustainability-Linked Loan Principles. LMA, APLMA and LSTA.

New Zealand Government (2017). Te Awa Tupua (Whanganui River Claims Settlement) Act 2017. Public Act 2017 No. 7.

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press. Design principles at Table 3.1, p. 90.

Ostrom, E. (2005). Understanding Institutional Diversity. Princeton University Press.

Ratajczyk, E., Brady, U., Baggio, J. A., Barnett, A. J., Perez-Ibarra, I., Rollins, N., Rubiños, C., Shin, H. C., Yu, D. J., Aggarwal, R., Anderies, J. M. and Janssen, M. A. (2016). "Challenges and Opportunities in Coding the Commons: Problems, Procedures, and Potential Solutions in Large-N Comparative Case Studies." International Journal of the Commons, 10(2), 440–466.

Schweik, C. M. and English, R. C. (2012). Internet Success: A Study of Open-Source Software Commons. MIT Press.

United Kingdom. Companies Act 2006, section 283; Commons Act 2006.

United States. Securities Exchange Act of 1934, sections 19(b) and 19(c).

Williamson, O. E. (1985). The Economic Institutions of Capitalism. Free Press.

Note on figures. Every figure in this chapter is computed in lib/verify/VI_E3.py and reproducible with python3 lib/verify.py VI.E3, which prints the table and the false-friends list as coded, so that a reader who disagrees with a row can see exactly what their disagreement moves. The coding is one hand's work, stated as such in the module and in the Arithmetic, and the module computes the interval that fact implies. Crossing costs are carried from the chapters named beside each and are labelled CITED; the geometric mean, median and dispersion computed over them are this chapter's own. The likelihood ratios in the score-against-rating block are assumed, as they are in VI.02, because the whole point of that entry is that nobody has measured them.