Haute Lumière

Commerce · V.04 · MMXXVI · daylight

La Bourse  /  Volume V  /  Nº V.04  /  Ten concept briefs

TEN CONCEPT BRIEFS · Chapter V.04 — Care as Infrastructure

One page each. A reader who reads only these ten pages has the chapter.

---

BRIEF 1 — The Third-Person Criterion

The idea. There is a rule for deciding whether an activity is production or leisure, and it is one sentence long: if you could pay somebody else to do it for you and still get the benefit, it is production.

Margaret Reid set it out in 1934. Someone else can wash your child, cook your dinner, drive your mother to the hospital and sit with her afterwards — so all of those are production. Nobody can sleep for you, eat for you, or enjoy a concert on your behalf, so those are consumption. The criterion is not about effort or about virtue. It is about substitutability.

Worked example. A man spends Saturday building a shelf. Third person: yes, a joiner could have done it. That is production, and it belongs in a household account. He spends Sunday reading. Third person: no. That is consumption, and it does not.

Why it matters. Every satellite account in the world is built on this rule, which is why they are comparable across countries. It is also why the accounts understate: activities that are half care and half company — sitting with a grandparent, being in the house while a teenager does homework — fail the criterion cleanly and get dropped, and so does almost everything simultaneous. A parent supervising a child while cooking records one activity, not two.

You already know this because you have, at some point, priced a Saturday against what a tradesman would have charged, and decided the Saturday was worth it. You ran the third-person criterion in your head and then you ran a replacement-cost valuation, in that order.

---

BRIEF 2 — Measuring Care in Hours

The idea. Care is measured by asking people what they did with their day. Time-use surveys — diaries, in ten-minute intervals, in nationally representative samples — are the primary instrument, and they are why the numbers in this chapter exist at all.

The numbers. The ILO pooled time-use surveys from 64 countries covering about two-thirds of the world's working-age population and found 16.4 billion hours a day of unpaid care work. At an 8-hour day that is 2.05 billion full-time equivalents; over 365 days it is 5.99 trillion hours a year.

Who does them:

Worked example. 1,612 hours against a 2,000-hour working year is 0.81 of a full-time job. A woman at the global average is holding four-fifths of a second job, and it does not appear in any labour force statistic, any tax return, or any pension record.

Why it matters. Hours are the honest unit. Every valuation argument in this chapter is an argument about what to multiply the hours by — and the hours themselves are not in dispute.

You already know this because you have kept a timesheet, and you know that the act of writing down what you actually did with the afternoon produces a different number from the one you would have guessed.

---

BRIEF 3 — Replacement Cost

The idea. Value an unpaid hour at what it would cost to buy the equivalent service on the market. What would it cost to replace this?

Two variants. Specialist replacement prices each task at the wage of the specialist who does it — a chef's wage for cooking, a nurse's for nursing. Generalist replacement prices everything at the wage of a housekeeper or home aide who could do all of it. Statistical offices mostly use the generalist, because it is defensible and because the specialist version produces a number nobody will believe.

The numbers. In the US in May 2023, home health and personal care aides had a median wage of US$16.12 an hour; childcare workers, US$14.60.

Worked example. The UK's household satellite account valued unpaid household service work at £1,240 billion for 2016 — 63.0 percent of a £1,968 billion GDP, with childcare alone at £352 billion, which is 17.9 percent of GDP. That is a replacement-cost figure, and it is the one statistical offices publish.

Why it matters. It is the conservative method, and therefore the one to use when you are talking to someone who does not want to believe you. It is also the method that answers the only question a buyer actually asks: what would this cost me if it stopped?

You already know this because you have worked out what a week of childcare costs when a grandparent is ill, and been surprised by how large the number was for something that had been free the week before.

---

BRIEF 4 — Opportunity Cost, and the Wedge

The idea. Value the unpaid hour at what the person doing it gave up. What did it cost to supply this?

The numbers. Median usual weekly earnings for US women in full-time work in 2023: US$1,005 over a 40-hour week — US$25.12 an hour.

  opportunity / replacement    25.12 / 16.12  =  1.559 x
  opportunity / childcare wage 25.12 / 14.60  =  1.721 x
  the gap                                         US$9.00 per hour

Worked example. US household production was 26 percent of GDP in 2010 at replacement cost — US$3.9 trillion on a US$14.99 trillion economy, which is 242 billion hours, or 2.14 hours a day for every person in the country. Re-price the identical hours at the opportunity wage and the same account reads 40.5 percent of GDP. The choice of method moves a national account by 14.5 points of GDP.

The point most people miss. The wedge is not an error. It is a measurement. Market care wages sit near the bottom of the distribution partly because so much care is supplied unpaid, so the gap between the two methods is a reading on how much of care's value is not being paid for. There is a matching figure on the individual: the net wage penalty for care work is on the order of 7.5 percent, which on a US$38,000 salary is US$2,850 a year, and discounted at 3 percent over 30 years — annuity factor 19.600 — a present value of US$55,900 surrendered for choosing the occupation.

You already know this because you have, at least once, heard someone say they "can't afford to go back to work" — which is an opportunity-cost calculation, run correctly, at a kitchen table.

---

BRIEF 5 — The Satellite Account

The idea. A satellite account is a set of national accounts kept beside the main one, using the same rules, for something the production boundary excludes. It is not an alternative GDP. It is GDP's annex, and it is built so that the two can be added or compared without double counting.

The numbers.

Worked example — and the thing the fall actually means. The US figure drops 13 points between 1965 and 2010, a 33.3 percent relative fall. Less care did not happen. Women entered paid work, and the same activity moved across the production boundary, so it began to be counted. Thirteen points of measured national output appeared without anybody making anything new — which is the sharpest available demonstration that GDP measures a boundary, not an economy.

Why it matters. Once a satellite account exists, care stops being a topic and becomes a series. Series get revised, compared, and cited in budget documents. In England and Wales, unpaid adult care was priced at £162 billion against an NHS England budget of £164 billion — a ratio of 0.99 — and legislation followed a number that size.

You already know this because your firm keeps a management account beside its statutory one, for exactly this reason: the statutory boundary is a legal artefact and the management account is what you actually run on.

---

BRIEF 6 — The Care Employment Multiplier

The idea. A given sum spent on care creates more jobs than the same sum spent on construction or manufacturing — and the reason is arithmetic, not sentiment.

The numbers. Modelled across seven OECD economies at 2 percent of GDP invested. In the US, on a 2014 GDP of US$17.4 trillion, that is US$348 billion, creating about 13 million care jobs against 7.5 million in construction.

  care          37.36 jobs per US$1m
  construction  21.55 jobs per US$1m
  ratio          1.73 x

Later UK work gives 2.7× overall and 6.3× for women's jobs; the Turkish study gives 2.5×.

Worked example — where the ratio comes from, and where it goes. Direct jobs per million is labour share × (1 − import leakage) ÷ average compensation:

Care beats construction 3.25× on the direct term and manufacturing 11.01×. But the published total ratio is 1.73×, because construction pulls a long, largely domestic supply chain behind it and care pulls almost none. Care wins on the direct term and gives much of it back on the indirect term. Quote the total. The direct ratio is the mechanism, not the answer.

You already know this because you know what a hospital's cost base looks like compared with a factory's: one is nearly all people, the other is nearly all bought-in materials and machines.

---

BRIEF 7 — The Price Elasticity of Maternal Employment

The idea. Lower the price of childcare and maternal employment rises. The size of that response is a number, it has been estimated many times, and it is about −0.20: a ten percent fall in the out-of-pocket price raises maternal employment by about two percent.

Worked example — two routes, no shared assumptions.

Route one, quasi-experimental. Quebec introduced universal low-fee childcare in 1997; the rest of Canada is the control. Childcare use rose 14.6 percentage points; the labour force participation of married mothers rose 7.7 percentage points, a 14.5 percent relative increase — implying a baseline of 53.1 percent, which is what it was.

Route two, from the elasticity. A 60 percent cut in the fee, at an elasticity of −0.20, raises maternal employment 12.0 percent; on a 65 percent baseline that is 7.80 percentage points.

  elasticity model     7.80 pp
  Quebec, measured     7.70 pp
  difference           0.10 pp

Why it matters. A meta-analysed elasticity and a difference-in-differences on a real policy change agree to a tenth of a point. That agreement, and not either number on its own, is the reason the labour-supply effect of childcare is one of the better-established results in applied economics.

You already know this because you have watched a colleague's return-to-work date move when their childcare arrangement changed, and you did not think of it as an elasticity, but it was one.

---

BRIEF 8 — The Fiscal Return, and the Inframarginal User

The idea. A childcare programme returns money to the treasury through four channels: tax on the induced worker's earnings, reduced transfers to her, tax on the care workforce's own pay, and consumption tax on both. Whether those exceed the cost is arithmetic, and it turns on one parameter.

Worked example. A quality place at 1:6 with a US$52,000 loaded teacher, staff at 65 percent of cost, is US$13,333 a year. Parent pays US$4,800; the employer pays US$1,437, being its own verified turnover and absence saving; the public residual is US$7,096.

Recovery, per place:

• Fixed, regardless of behaviour: care-worker tax US$2,080 plus consumption tax US$593 = US$2,673.
• Per induced maternal entrant: US$35,562, being earnings of US$38,000 at a 28 percent wedge plus US$2,600 of transfer reduction, multiplied by a 2.5× persistence factor, plus consumption tax.

  break-even jobs per place = (7,096 − 2,673) / 35,562 = 0.1244

One induced entrant per 8.0 places.

The honest part — deadweight. Most users of a universal programme would have worked anyway. The programme pays for a large inframarginal group in order to move a small marginal one. That is why the fiscal case is tighter than the advocacy suggests — and it is not an argument against the programme, because the inframarginal spending buys the child-development return and an income transfer. It is an argument for booking those in the ledger they belong in, rather than smuggling them into the fiscal one.

You already know this because you have run a promotion and known perfectly well that some of the customers who used it would have bought anyway.

---

BRIEF 9 — The Vertical Fiscal Split

The idea. Whether a care programme "pays for itself" depends on which treasury is asking, and that is a question about constitutional structure, not about care.

Worked example. Quebec's programme cost the province C$1.6 billion in

1. It returned 104 percent of that to the Quebec treasury —

C$1.66 billion — and 43 percent to the federal treasury — C$0.69 billion — which had contributed nothing. Combined public-sector recovery: 147 percent.

  federal share of the total return  =  0.43 / 1.47  =  29.3 %

Twenty-nine percent of the entire return was collected by a government that did not write the cheque.

Why it matters. This is not a curiosity. It is the structural reason sub-national governments under-provide care: the payer is the province, the city, the school district, and a large share of the payroll-tax and income-tax return goes upward. The provider of the good is not the residual claimant on its return, and any first-year student of public finance can tell you what happens next.

The remedy is a clause, not a study. Write an intergovernmental transfer into the agreement, indexed to induced employment as measured by the statistical office rather than by either party.

You already know this because you have seen a business unit refuse to fund something whose benefit lands in another unit's P&L, and you know the fix was never more evidence — it was a transfer price.

---

BRIEF 10 — Quality, and Why It Does Not Survive Scaling

The idea. Care investment is not reliably good. High-quality care investment is reliably good, and quality is a cost line.

The numbers.

Perry costs 4.19× average state pre-K and 1.92× Head Start. The randomised Head Start Impact Study found initial impacts of roughly 0.1–0.2 standard deviations, not distinguishable from zero by third grade. Tennessee's lottery evaluation found the treated group scoring lower in maths, science and reading by sixth grade, with more disciplinary infractions and more special-education placement. Quebec's own long-run follow-up found worse self-reported health and life satisfaction and higher criminal activity in the exposed cohorts.

Worked example — where the money actually is.

  1:6  degree-qualified at US$52,000  ->  US$8,667 per child
  1:10 credentialled at  US$38,000    ->  US$3,800 per child
  1:12 at                US$34,000    ->  US$2,833 per child

A factor of 3.06 in the one input the research says carries the effect — and it is the first input a scaling budget cuts. The label survives scaling. The ratio does not.

Why it matters. It gives you a test you can apply to any proposal in ninety seconds. Does it name its ratio, its staff qualification and its turnover target, in numbers, with money at risk against them? If not, the effect sizes it is quoting belong to a different programme.

You already know this because you have watched a service you liked get rolled out across the organisation and arrive somewhere else thinner, with the same name on the door and none of the people who made it work.