Haute Lumière
Commerce · VII.11 · MMXXVI · daylight
For the person studying this alone, or in a seminar, with no country to account for and no balance sheet to sign. You are on the receiving end of the transfer this chapter measures. That is not a disadvantage. It is the only vantage point from which the delivery-date problem is obvious.
The chapter was written for someone who could commission a national account. You cannot. What you can do is something the chapter's intended reader mostly cannot: build the same account for a unit you actually inhabit, and check the national one against your own life.
That is not a consolation exercise. Every serious error in intergenerational accounting is an error of aggregation — a real thing that is true on average and false for almost everybody. £84,280 per head is true. It is also true that about half the cohort receives nothing material, and that the median receipt arrives at 61. A person who has done the arithmetic on their own position can see both at once, and that is a rarer and more useful skill than being able to quote the aggregate.
So you will do exactly what a treasury does, on a system you control, and by the end you will have three things: a personal intergenerational balance sheet, a declared θ you can defend, and a one-page national account you built yourself from published sources. The third of those is a portfolio piece. Very few people at any career stage have made one.
Exercise 1.1 — What you actually inherited (90 minutes)
Not money. Working stock. Make five columns and fill them honestly.
| Category | In a country | In your life |
|---|---|---|
| Produced capital | roads, plant, housing | the tools, instruments, books, machines you did not buy |
| Human capital | skills, education | what you can do that somebody taught you without charging |
| Institutional | law, courts, records | the assumptions you hold about what is possible and safe |
| Financial claims | gilts, pensions | money, support, housing, debt you carry or were spared |
| Natural | soil, air, water, habitat | the condition of the place you live, and its trajectory |
For each line, write who paid and what it cost them. Most people discover, doing this, that the largest transfers they received were the ones nobody described as a transfer.
Exercise 1.2 — The five published lines (2 hours)
Go and get them. For your own country, find and cite:
Divide each by population. You now have the first column of a real account, and it took an afternoon. Note which ones you could not find. That list is data, not failure — it is the coverage statement, and an account without one is an advertisement.
Exercise 1.3 — Find the working case (45 minutes)
Pick one of Norway, Wales, New Zealand or Alaska from the chapter's Discovery and read one primary document about it: the fund's annual report, the Act, the Treasury framework, the dividend statute. Write 300 words on what the mechanism actually is, distinguishing the fund from the rule, the goal from the duty, the payment from the principal. In every one of those four cases the famous thing is not the load-bearing thing.
Exercise 2.1 — ΔM for your country (3 hours)
Find your national balance sheet or comprehensive wealth series. Take two points about a generation apart. Then:
net worth (later) ÷ population (later) = per head, later
net worth (earlier) ÷ population (earlier) = per head, earlier, nominal
× the GDP deflator between the two years = per head, earlier, real
------------------------------------------------------------------------
the difference = ΔM per head
Work the UK version first as a check on your method: £11,800 bn over 67.6 million is £174,556; £2,800 bn over 58.0 million is £48,276, and ×1.87 is £90,276; the difference is £84,280 per head, a compound real rate of 2.47 percent a year. If your method reproduces those, it works.
Exercise 2.2 — The sensitivity habit (45 minutes)
Move your earliest stock figure by ±20 percent and recompute. For the UK, ΔM runs from £66,225 to £102,336 and never changes sign. Write down whether your conclusion survives. Get into the habit now: a result that does not survive a fifth in either direction is not a result, it is a data-entry choice.
Exercise 2.3 — The land switch (90 minutes)
Find your country's land or non-produced-asset share. Recompute ΔM with it excluded. In the UK the answer falls from £84,280 to £18,365 — 78.2 percent of the headline is land.
Then write half a page answering the question properly, in your own voice: is a rise in the price of land an addition to what the next generation has, or a transfer between generations? There is a real answer on each side. Take one, and say what would change your mind.
Exercise 2.4 — ΔN as a lower bound (2 hours)
Find your country's cumulative territorial emissions over the same window. If only endpoints are published, use the trapezoid and say that you did — the UK's 700 and 417 MtCO₂e over 28 years gives 558.5 a year and 15.64 GtCO₂e. Price it at three published carbon values, not one.
Then make your UNMEASURED register: every natural stock you know has changed and cannot price. Do not set it to zero. Write the word.
Exercise 2.5 — θ* and the four cells (60 minutes)
θ* = −ΔN / ΔM
Compute it with land and without, at each carbon value. Build the four-cell table. For the UK: θ* is 0.74 with land and 3.39 without; the cells are +£22,052, −£62,228, −£43,863, −£62,228.
Then answer the question the exercise exists for: what is your θ? Write one paragraph defending a number between 0 and 1, and one paragraph attacking it. This is the single most useful page you will write this term, because for the rest of your life you will be in rooms with people whose θ is different from yours and who have never noticed that this is what they are disagreeing about.
Exercise 2.6 — Your own delivery date (45 minutes)
At the Green Book's 3.5 percent, twenty-one years of waiting is a factor of 0.4856: £84,280 becomes £40,924, and 51.4 percent of the value is destroyed by timing alone. Nothing is lost in aggregate.
Now run it on yourself. When, realistically, will any material transfer reach you? What is its present value at an age when it would change what you could attempt? What would you do differently if it arrived a decade earlier?
Exercise 3.1 — The one page (2 hours)
Everything above, on one page. Five categories, consolidated. ΔM with its sensitivity. ΔN as a lower bound. θ*, the four cells, the UNMEASURED register, and every source with a date. No prose except the switches.
Exercise 3.2 — The hostile read (60 minutes, with one other person)
Give the page to somebody who disagrees with you and ask them to find the place where you chose a switch quietly. Whatever they find, thank them and fix it in front of them. Do not defend anything on this pass. You are practising the one discipline that keeps an account honest over decades, and it is not rigour — it is the willingness to be read by somebody who wants you to be wrong.
Exercise 3.3 — Your personal covenant (45 minutes)
Write one commitment, with a date and a counterparty, referring to a number on your page. Not a resolution. A covenant: somebody else knows, somebody else can check, and there is a date on which they will.
Pick a place you have a real connection to — a town, a campus, a catchment, an island, a neighbourhood. Build its intergenerational balance sheet over one generation.
What it must contain.
Length. Ten pages, of which the account itself is one. The other nine are sources, method and the things you could not find out.
What it is marked on. Not the answer. Three things: whether every switch is visible; whether the UNMEASURED register is honest and specific; and whether a person who disagrees with your θ can still use your page.
Score yourself once now and once at the end of the term. Be exact rather than kind; the point is the difference between the two readings.
| Now | End | |
|---|---|---|
| I can name the five categories and say why they cannot be naively added | ||
| I can consolidate a domestic claim and explain why it is not a transfer | ||
| I can compute ΔM in constant prices per head from published sources | ||
| I can state and defend the land switch on both sides | ||
| I can price a natural position and mark its lower bound honestly | ||
| I write UNMEASURED rather than zero, and can say why that matters | ||
| I can compute θ* and build the four cells | ||
| I know my own θ and can argue against it | ||
| I can explain the delivery-date result to someone in two minutes | ||
| I have given my work to somebody who wanted it to be wrong |
The one that matters most is the last. Everything else on this list is a technique, and techniques transfer easily. The habit of handing finished work to a hostile reader is the only one that will still be improving your accuracy in twenty years, and it is the one nobody is taught.
A term of practice only works if each week ends with an artifact rather than a feeling. Here is the whole of it on one page, so you can see the shape before you are inside it.
| Week | The work | What exists at the end of it |
|---|---|---|
| 1 | Exercise 1.1 — what you inherited, five columns | A page nobody else could have written |
| 2 | Exercise 1.2 — the five published lines, per head | Five sourced lines and a list of what you could not find |
| 3 | Exercise 1.3 — one primary document, one working case | 300 words separating the famous thing from the load-bearing one |
| 4 | Choose the place for the term project | A place, a reason, and a first list of sources |
| 5 | Exercise 2.1 — ΔM, checked against the UK worked example | ΔM per head, in constant prices |
| 6 | Exercise 2.2 — the ±20 percent sweep | A result that survives, or the knowledge that it does not |
| 7 | Exercise 2.3 — the land switch, both ways | Half a page of your own reasoning, with a falsifier |
| 8 | Exercise 2.4 — ΔN and the UNMEASURED register | A lower bound, honestly labelled |
| 9 | Exercise 2.5 — θ*, the four cells, your own θ | The page you will still be using in ten years |
| 10 | Exercise 2.6 and Exercise 3.1 — the delivery date, the one page | The one page |
| 11 | Exercise 3.2 — the hostile read | A corrected page and a named reader |
| 12 | Exercise 3.3 and the project — the covenant | A signed commitment with a date |
The rule for the whole term: never skip week 11. An account that has not been read by somebody who wanted it to be wrong is a draft, however carefully it was made. This edition's own instruments were built after somebody read a finished claim and disagreed with it, and not one of the faults they found would have been caught by running the tool a second time.
A note on hours. Four hours a week is enough and eight is not twice as good. The binding constraint in this work is not effort; it is the interval between doing a calculation and rereading it, because the second reading is where the quiet switch you chose without noticing becomes visible. Leave a day between building a number and publishing it, every time.
Five, in the order they are useful rather than the order they were written.
You are being handed, per head, roughly twice what the generation before you received, about half of it late, some of it not at all, and with a natural position whose sign depends on a parameter nobody publishes.
None of those four facts is a reason to be angry and none is a reason to be reassured. They are a reason to be specific, which is a much more powerful thing to be in a room than either.
And notice what you are now able to do that you could not do a term ago. You can build the account. You can find the switches. You can name the one parameter that decides the sign and hand the decision to the people whose decision it is. Very few people can, at any age. That capacity is itself part of what was handed to you — by whoever taught you arithmetic, and by every person who kept the records you have been reading — and the way it is passed on is by being used in public, on real numbers, where somebody else can check.