Haute Lumière
Commerce · V.04 · MMXXVI · daylight
For the person studying this alone or in a seminar. A term of practice, a term project, and a self-assessment. Applied to a life — yours, and the lives that made it possible.
Most economics you will be taught is about things you have never done. This one is not. You have been inside the care economy since before you could speak, on one side of it or the other, usually both, and the reason the subject is difficult is not that it is technical. It is that you are too close to it to see it as a system.
So the method here is deliberately cold at the start. Count first. Feel second. You are going to keep a time diary, run two valuations on the same hours, and get a number that is yours before you read another word of anyone's argument about it. Nothing in this chapter is hard to compute. What is hard is looking at your own week in units.
By the end of the term you will be able to do four things no undergraduate economics course reliably teaches: build a household satellite account from a time diary; convert between replacement-cost and opportunity-cost valuations and say what the gap means; compute the fiscal return on a care programme from first principles; and read an early-childhood effect size and know whether it belongs to the programme being proposed.
Exercise 1.1 — The time diary (one full week, ten minutes a day)
Keep a diary in ten-minute intervals for seven consecutive days. Use a paper notebook or a plain spreadsheet. For each interval record the primary activity and, separately, any simultaneous activity.
Then classify every entry against the third-person criterion: could you have paid somebody else to do this and still received the benefit?
| Category | Your hours/week | Global female average | Global male average |
|---|---|---|---|
| Unpaid care, primary | 30.9 | 9.7 | |
| Unpaid care, simultaneous | — | — | |
| Paid work | |||
| Consumption and rest |
(The comparators are 265 and 83 minutes a day converted to a week.)
The finding you are looking for is in the simultaneous column. Time-use surveys largely drop it, which means every published figure in this chapter is a floor. Write down how many hours of yours are in that column, and keep the number — it will be the most interesting thing you produce this month.
Exercise 1.2 — The two valuations (90 minutes)
Take your weekly unpaid care hours and value them twice.
replacement cost = hours × local home care / childcare median wage
opportunity cost = hours × your own realistic hourly earnings
the wedge = opportunity ÷ replacement
Look up the two wages properly, from the national statistical office, and write down the source and the date beside each. The US benchmarks in the chapter are US$16.12 for a home health and personal care aide and US$25.12 as an opportunity wage — a wedge of 1.559×. Your wedge will differ. The size and the direction of your difference from 1.559 is the exercise. A student with no earnings history has a wedge below one, and working out why that is not absurd is the real lesson in opportunity cost.
Exercise 1.3 — The interview (one hour, with someone older)
Interview one person who did unpaid care work that made someone else's career possible. A parent, a grandparent, a neighbour. Ask three questions and only three:
"What did a normal day look like in that period?" "What did you stop doing?" "What made it possible to keep going — who or what helped?"
Write it up in 400 words. Do not editorialise and do not add a conclusion. You are collecting a time diary from before time diaries, and the third question is appreciative on purpose: it is the one that produces the design information.
Exercise 2.1 — Build a household satellite account (3 hours)
Scale your own diary to a household, then to a hundred households, then check the result against the published account for your country. The UK's is the most accessible: £1,240 billion for 2016 against a £1,968 billion GDP, which is 63.0 percent, with childcare at £352 billion.
Your scaled number will be wrong. Write down by how much and in which direction, and then find the reason — sample composition, the simultaneity problem, your own atypical week, or a wage you chose badly. Finding the reason is the entire skill.
Exercise 2.2 — The multiplier decomposition (90 minutes)
Reproduce this table for your own country, using published labour shares, import content and average compensation:
jobs per US$1m = labour share × (1 − import leakage) × 1,000,000 ÷ compensation
care 0.75 0.02 38,000 -> 19.34
construction 0.45 0.10 68,000 -> 5.96
manufacturing 0.20 0.28 82,000 -> 1.76
Then find the published total multiplier for your country and explain the gap between your direct ratio and the published total one. The US direct ratio for care against construction is 3.25×; the published total is 1.73×. If you cannot explain a gap of that size, you cannot use either number.
Exercise 2.3 — The two routes to the labour-supply effect (2 hours)
Do both, independently, and only then compare.
Route one. Find the Quebec result: childcare use +14.6 pp, married mothers' participation +7.7 pp, a 14.5 percent relative rise. Compute the implied baseline and check it against published Canadian participation data for the mid-1990s.
Route two. From an elasticity of −0.20, a 60 percent fee cut and a 65 percent baseline employment rate, compute the predicted effect.
They should land within about a tenth of a point of each other — 7.80 pp against 7.70 pp. Write two paragraphs on why two agreeing instruments that share no assumptions are worth more than one precise instrument.
Exercise 2.4 — The break-even (one hour)
Compute it yourself, from the inputs:
place cost 13,333 − parent 4,800 − employer 1,437 = public 7,096
break-even j = (7,096 − 2,673) ÷ 35,562 = 0.1244
Then run the sensitivity. Vary one input at a time by ±20 percent — the tax wedge, the persistence multiplier, the staff ratio, the fee cap — and record which one moves the break-even most. Bring the answer to seminar. It is not the one most people guess.
Exercise 2.5 — The vertical split, in your own country (90 minutes)
Find out, for the jurisdiction you live in, who pays for childcare provision and who collects the income tax and the payroll tax on the earnings it unlocks. In Quebec the answer was 104 percent back to the province that paid and 43 percent to the federal treasury that did not — 147 percent combined, with 29.3 percent of the total return landing on a desk that contributed nothing.
Draw the flow as a diagram: one box per treasury, one arrow per tax head, the cost on one side and the return on the other. Then write the transfer clause that would fix it, in no more than five sentences, and say who verifies the induced-employment figure it is indexed to. A clause that lets either party do its own measurement is not a clause.
Exercise 2.6 — The care wage penalty, on a person (45 minutes)
Take the reference figures: a 7.5 percent net penalty on a US$38,000 salary is US$2,850 a year; at 3 percent real over 30 years, with an annuity factor of 19.600, the present value is US$55,900.
Now do it with local wages for a named occupation — a nursery practitioner, a care assistant, a paediatric nurse. Then find one published estimate of the penalty for your country and compare. Write one paragraph on what would have to change for the penalty to close, and be specific about who pays for the closing. This is the exercise that connects the valuation argument to a payslip, and it is the one students remember.
Exercise 3.1 — The proposal test (30 minutes, repeatable)
Find three real care proposals — a manifesto commitment, a corporate announcement, an NGO paper. Apply the ninety-second test:
| Test | Proposal A | B | C |
|---|---|---|---|
| Names a staff:child ratio, as a number? | |||
| Names a staff qualification? | |||
| Names a turnover target? | |||
| Puts money at risk against any of them? | |||
| States its cost per place? |
Then look up the effect sizes each proposal quotes and find the cost per child of the programme those effect sizes came from. Perry is US$27,600 per child-year in 2024 money; average US state pre-K is US$6,600 — a factor of 4.19. Write one paragraph per proposal on whether the evidence it cites belongs to it.
Exercise 3.2 — Design a place (2 hours)
Specify one childcare place in your own city. Ratio, qualification, loaded staff cost, non-staff share, total cost. Then find three payers and give each one a reason that lives on its own ledger:
Nobody in your design is being asked for charity. If one of them is, the design is not finished.
Exercise 3.3 — The appreciative interview, run properly (one seminar)
In pairs, run the discovery protocol on each other's institution — your university is a large employer with a care problem it has probably never costed.
"When has this place made it possible for someone to keep studying or working through a care responsibility? What made that work? What would it take to have more of that?"
Collect conditions, not anecdotes. Conditions are repeatable and anecdotes are not.
Exercise 4.1 — The failure post-mortem, written in advance (90 minutes)
Take your design from 3.2 and write the account of how it failed, dated five years from now. Use the four named failure modes: expansion outran the training pipeline and the ratio gave; induced employment was assumed rather than measured; the account was built as advocacy and read as advocacy; the programme was evaluated on enrolment. Pick the one that would get yours, and write the sentence in the contract that would have stopped it.
Exercise 4.2 — The thing you enjoyed (30 minutes)
Go back through the term's exercises and mark the one you actually enjoyed. Most students mark the time diary or the interview, and almost nobody marks the essay. That is information about what this subject is: it is not advocacy, it is noticing, and noticing is one of the few kinds of intellectual work that is intrinsically absorbing. Write 200 words on what that tells you about the next thing you should study.
Build a household satellite account for one real institution and price its care exposure.
Choose something with a boundary — a hall of residence, a department, a small firm that will let you near its numbers, a community organisation. Then:
8,000 words, or 3,000 words and a working spreadsheet. The spreadsheet is worth more. Every figure must be reproducible by a reader who has only your document, which means every input appears with its unit and its source before any result that uses it.
Score yourself honestly. This is not marked.
| Not yet | Partly | Yes | |
|---|---|---|---|
| I can state the third-person criterion and apply it to a borderline case | |||
| I can convert between replacement and opportunity cost and say what the wedge means | |||
| I can build a satellite account from a time diary and say what it misses | |||
| I can decompose an employment multiplier into direct and indirect terms | |||
| I can compute a fiscal break-even for a care programme from published inputs | |||
| I can read an effect size and check whether it belongs to the programme quoting it | |||
| I can name the four failure modes and the contract clause for each | |||
| I can hold "the abundance is real" and "the margin is thin" in the same paragraph |
Any row still in the first column: go back to the brief that covers it. The briefs are one page each and they are written to be re-read.
Three things worth doing after the term ends.
Keep the diary once a year. One week, every year, for the rest of your life. It takes ten minutes a day and it will tell you more about the shape of your own economy than any other instrument available to you.
Publish one number. Whatever you found in your project, put it somewhere a stranger can read it, with its method beside it. A number nobody keeps is a number nobody improves.
Ask the question in rooms where it is not being asked. When a proposal comes past you — in a committee, a student union, a first job — ask what its ratio is and whether there is money at risk against it. You will be the only person in the room who asks, and it takes nine seconds.