La Bourse / Volume VII / Nº VII.04 / Workbook — the student

A woman on a veranda holding a cup of tea, a book open beside her, a lotus garden and pavilion beyond. — Plate VII.04 · Workbook — the studentThe Certificate and the Box.Two great commons, one atmosphere and one ocean, and only one of them has ever been governed successfully. The difference was never the size of the thing. It was whether the rule could be tested on an object that had to pass through somebody's hands.

WORKBOOK — THE STUDENT

Chapter VII.04 · The Ocean and Atmospheric Commons

For the person studying this alone, or in a seminar, with no treaty to negotiate and no fleet to regulate. You are not too far from the subject. The instrument in this chapter is a way of asking a question, and the question works at any scale you can reach this month.

WHY THIS WORKBOOK IS DIFFERENT

The chapter is about two commons so large that no single person can affect either of them. It would be easy to read it as a spectator sport.

It is not, and the reason is the most portable thing in the chapter: the deterrence condition is a general instrument for reading any rule anywhere. p · F ≥ G. Whenever you find a rule that people break, you can ask what the gain is, what the expected consequence is, and whether the arithmetic was ever going to work. The answer is almost always that nobody computed it — and once you can compute it, you can see through a very large number of institutional arguments that are conducted in the language of values and decided by the language of prices.

The second portable thing is the Montreal checklist. Four conditions, run against any change anybody is trying to make. It takes an afternoon and it will tell you, before anyone has spent a year on it, whether a campaign is aimed at a tractable target.

You will do exactly what the negotiator does. You will simply do it on systems you can actually see.

PART ONE — DISCOVERY

Weeks 1–4: find what is already holding

Exercise 1.1 — The rules that hold, and why (90 minutes)

List five rules in your own environment that are followed far more reliably than their enforcement would predict. A university's rules on plagiarism. A shared kitchen. A library's loan period. A code of conduct nobody has ever been sanctioned under. The unspoken rule about who speaks first in a seminar.

For each, write two columns:

What would breaking it gain?	What would breaking it be expected to cost?

You will find that in most of them, p · F is far below G, and the rule holds anyway. That gap is the normative compliance the chapter's last honest negative is about, and it is the most valuable and least measured asset in any institution. Mark which ones you think would survive being formalised, and which would not.

Exercise 1.2 — Run the Montreal checklist on something you care about (2 hours)

Pick a change you would like to see — in your university, your town, your field. Run the four conditions:

Is there a substitute, and who owns it? If the people being asked to stop doing the thing also own the replacement, you have a tractable problem. If they own only the thing being taken away, you have a fight.
How many points must be reached? Count them. Seventeen is a campaign. 1.5 billion is a different discipline entirely.
Is the signal unarguable, and does it answer inside a career? If the evidence of success arrives after the people who must act have moved on, the design has to supply an interim signal or it will not sustain.
Is there a chokepoint where a binary test can be run on a physical thing? A door. A payment. A form. A licence. A border.

Write one page. Most proposals fail two or three of these, and the most useful output of this exercise is a redesign: what would I have to change about the proposal so that it passes three of the four?

Exercise 1.3 — The appreciative interview (45 minutes, with another person)

Find someone who has worked inside a system with real compliance — a nurse, an auditor, an air traffic controller, a lab technician, a pilot, a customs officer — and ask exactly this:

"Tell me about a rule in your work that people actually follow. Not the ones on the poster — one that really holds. What makes it hold? What would have to change for it to stop holding?"

Then stay quiet. Take notes on the mechanism, not the culture. Almost everyone will first answer "professionalism", and almost everyone, if you wait, will then name a specific structural thing: a second signature, a logged action, a colleague who would see it, a licence that could be pulled.

This is the core skill of the chapter. The stated reason a rule holds and the operative reason are usually different, and only the operative one transfers.

PART TWO — THE ARITHMETIC

Weeks 5–8: do the sums yourself

Exercise 2.1 — Reproduce the frontier (90 minutes)

Do not take the chapter's numbers on trust. Run python3 lib/verify.py VII.04 and then rebuild the central result by hand.

  G   =  250 t × 0.20 × $6,000 × (1 − 0.40)   =  $180,000
  p   =  0.05 × 0.50                          =  0.025
  p·F =  0.025 × $100,000                     =  $2,500
  G / (p·F)                                   =  72 ×
  p*  =  G / F  =  180,000 / 100,000          =  1.80

Then plot the frontier F = G / p for p from 0.01 to 1.0 on log axes. Mark the point where F equals one year's gross revenue. Look at the shape. The curve is steep exactly where every real fishery sits, which is why small increases in coverage feel like they achieve nothing: they do.

Exercise 2.2 — Break the model on purpose (60 minutes)

The module ships a sensitivity sweep. Read it, then extend it. Vary the sanction rate from 0.1 to 0.9 and find the coverage at which a $500,000 fine becomes deterrent. Then answer in writing: which of the model's inputs, if wrong, would change the chapter's conclusion, and which would not?

This is the discipline the whole edition runs on. A result that survives its own sensitivity analysis is a finding. A result that does not is a number.

Exercise 2.3 — Compute Γ for a stock nobody has computed it for (3 hours)

Pick any stock managed by any regional fisheries body. The scientific advice and the adopted measure are both public documents. Find them, and compute:

  Γ  =  adopted / advised
  Λ  =  reported catch / adopted        (where the returns exist)

Write down what you could not find. A Γ you cannot compute because the advice is not published as a number is itself the finding, and it is a more important one than any particular ratio. Note the year, the stock, the document numbers, and your sources. You have just done, for one stock, the thing the chapter says nobody does annually for all of them.

Exercise 2.4 — The Montreal arithmetic, done properly (90 minutes)

The Protocol is usually taught as a story. Do it as a sum instead.

  Multilateral Fund disbursed since 1991        > $4 billion
  avoided emissions by 2010                      11 Gt CO2-eq / yr
  one-off transfer per Gt/yr avoided             $364 million

Now find, from primary sources, the cost per tonne of any carbon abatement programme you can get real figures for — a national scheme, a corporate commitment, a compliance market price. Put the two on one page with their denominators stated.

Then write the paragraph that makes it honest, because the comparison is not clean and pretending otherwise would be the exact fault this edition exists to correct: the Montreal figure divides a one-off transfer by an annual avoided flux, and it excludes the private capital cost of the industrial conversion, which was much larger than the Fund. State both, give the range, and say what you could not find. A comparison with its weaknesses named is worth more than a clean one nobody believes.

Exercise 2.5 — The eight orders of magnitude (45 minutes)

  regulated points, Montreal              17 producers
  motor vehicles in use            1,500,000,000 engines
  ratio                               88,000,000 ×

Extend it. Count, as best you can from public sources, the number of points at which a rule would have to bind for three other problems: antibiotic use in livestock, single-use plastic production, and aviation fuel. Rank the three by the count, and say which of the three you would expect a Montreal-shaped instrument to work on.

Then find the one where the count is low and check the other three conditions against it. You have just done, in an afternoon, the screening that most campaigns do not do in a year.

Exercise 2.6 — The honest negative, applied to yourself (30 minutes)

Write half a page on this: where in my own life would introducing a price destroy something that is currently free? Be specific. Then write what you would do instead, given that the behaviour genuinely does need to change.

PART THREE — DESIGN

Weeks 9–12: build the small version

Exercise 3.1 — Find the chokepoint (2 hours)

Take a system you are inside — a student society, a shared house, a lab, a volunteer rota, a group project. Map every point at which something physical or transactional must pass through a single place. A key. A form. A payment. A booking. A submission portal.

Then ask the chapter's question: what test could that chokepoint carry that it does not carry now? One test. Not a culture change, not a poster, not a meeting. A field that must be filled before the thing proceeds.

Exercise 3.2 — Design the smallest possible bond (90 minutes)

Somewhere in your life there is a recurring arrangement that fails because a promise is not collateral. A shared deposit. A borrowed item. A rota shift. A commitment to a study group.

Design the bonded version. What is held? Who holds it? What releases it? What forfeits it? Who decides, and on what evidence?

Then write the honest paragraph: who does this instrument make worse off, and is that the person you meant? A bond is regressive by construction — it is easiest for whoever has the most spare capital. Name what you would do about that.

Exercise 3.3 — The one-table publication (one week)

Take the Γ or Λ you computed in 2.3 and publish it. A blog post, a departmental noticeboard, a letter, a page on a site, a comment on a consultation. One table, sourced, with what you could not find stated plainly.

This is not symbolic. The chapter's Destiny movement says the only thing that has ever moved a quota back toward its science is the advice being visible next to the vote. A table that did not exist last month now exists. That is the whole mechanism, at the smallest scale it comes in.

PART FOUR — DESTINY AND DELIGHT

The term project, and what to carry

THE TERM PROJECT — the compliance reading

Over one term, produce a 3,000-word compliance reading of a single regime. Any regime: a fishery, a chemical, a professional licence, a tax, a data-protection rule, an emissions scheme, a sports anti-doping code. Five parts:

What is working. Named, sourced, specific. Open here, always.
The deterrence arithmetic. G, p, F, with every input sourced and every assumption labelled. State what you could not find.
The frontier. What penalty deters at current coverage; what coverage deters at the current penalty. If the second number exceeds one, say so and say what follows.
The Montreal checklist. All four conditions, scored, with the reasoning.
One instrument. The structure, the counterparty, the trigger, the treatment, and the single number that decides it.

Assess yourself against this: could a person who disagrees with you reproduce every number in it from your sources alone? If not, it is not finished.

Give yourself one constraint that will improve the piece more than any amount of extra reading: every number in it must be traceable to a document a stranger could open. Where you cannot find one, write the sentence without the number. A paragraph that makes a smaller claim on firm ground is worth three that make a large one on soft ground, and the habit you are building is the habit that makes a body of work citable rather than merely persuasive.

SELF-ASSESSMENT — six questions, answered honestly

Can I state the deterrence condition and explain why p is a joint probability?
Can I compute the frontier both ways without the book open?
Can I run the four Montreal conditions on a new problem and get a defensible answer in an afternoon?
Can I explain precisely why the Montreal lesson transfers to fish and not to carbon — naming which condition fails and why?
Did I state what I could not find, every time?
Have I published one number that did not exist in public before?

FOR A SEMINAR — three sessions that work in a room

Session one: the frontier, on a board. Give the room G, p and F and let them find the ratio. Then ask for the fine that would deter, and watch the room discover p* on its own. The moment somebody says "but that's more than one" is the moment the chapter lands, and it lands better found than told.

Session two: the four conditions, on four real problems. Split into groups, one problem each, forty minutes, four conditions scored with reasoning. The groups will disagree about condition 1 almost every time, and the disagreement — about who owns the substitute — is the most productive argument in the whole chapter.

Session three: the Γ table. Each person brings a stock, a scheme or a rule where an advised number and an adopted number both exist in public. Build one table together. Several people will arrive with nothing, because the advice was not published as a number, and that row goes in the table too, marked.

WHAT TO CARRY

Two things, and they will outlast the examination.

The first is the habit of asking what a rule is worth to break before asking why people break it. It will make you unusually calm in arguments where everyone else is discussing character.

The second is the ozone layer. It is the only planetary wound humanity has measured, diagnosed, priced, legislated against and watched heal, and it happened inside one working life, out of parts you could put on a table: a substitute, a fund, a customs test and seventeen telephone numbers. You are allowed to find that encouraging. It is not naivety. It is the single best piece of evidence available about what this species can do when the design is right, and you will meet a great many people who have never looked at it closely.