Haute Lumière
Commerce · V.08 · MMXXVI · daylight
For the person working inside a gainshare arrangement — where a defined share of verified improvement returns to the people who created it. This chapter is the one that most directly concerns your pay, because meaning is the input you supply that nobody has ever put on your statement.
A gainshare is a written promise with four parts: a baseline, a measure, a share, and a period with a verifier. You already know this and you already read your statement against it.
Here is what this chapter adds. There is a second transfer running alongside your gainshare, in the opposite direction, and it is almost certainly larger.
It is the meaning differential — the fraction of wage, δ, that people in mission-driven work demonstrably accept in exchange for work that matters. At the conservative estimate it is 5.5 per cent. On the May 2023 median wage for a home care worker, 33,530 dollars, that is 1,951.48 dollars a year: a benchmark of 35,481.48 against pay of 33,530. Across six hundred people it is 1,170,888.89 dollars a year.
Compare that with what a gainshare typically returns. If your scheme shares a third of a two per cent productivity improvement on the same wage base, it returns a few hundred dollars a head. The differential is the bigger number by an order of magnitude, it runs the other way, and it has never appeared on a statement.
This is not an argument that your gainshare is a trick. It is the observation that you are already party to two transfers and have the arithmetic for only one of them. This workbook gives you the arithmetic for the other.
Exercise 1.1 — Read your scheme against the four parts (2 hours)
Answer in writing, from the document rather than from memory:
Question five decides whether the scheme is worth being in. A baseline that ratchets automatically means every gain raises the bar you are next measured against, so the same effort returns less each cycle and eventually nothing. A well-built scheme holds the baseline for a stated term or ratchets on a published schedule everybody can see coming.
Exercise 1.2 — Your own meaningful-time percentage (one week)
The threshold finding: below 20 per cent of working time on the activity a person finds most meaningful, burnout ran at 53.8 per cent against 29.9 above — 23.9 points, a relative risk of 1.80. Above the threshold, no further benefit.
Count last week. Hours on the work you would choose, over total hours. Write the percentage.
Then note where the hours went instead. This is a gainshare finding, not a wellbeing one, and that is the whole trick of reading it from inside: time a skilled person spends on work that does not need their judgement is waste with a name, and waste with a name is exactly what your scheme pays you to remove.
Exercise 1.3 — What only you can see (two weeks)
You observe things the benchmark cannot. Keep a list for a fortnight:
| What the firm measures | What you actually see |
|---|---|
| Turnover rate | Which people were about to leave and why, weeks before they did |
| Care hours per client | Which fifteen minutes actually changed the outcome |
| Engagement score | Which question everyone in your team answers strategically |
| Incident rate | The near-miss nobody logged and how it was caught |
Every row is a candidate for a gainshare claim, because every row is a measurement the firm would pay to have and does not.
Exercise 2.1 — Compute your own δ (one week)
This is the most valuable calculation in the workbook and it takes an afternoon.
δ = 1 − (your pay / the benchmark)
δ × benchmark is what you are contributing a year. Multiply by your headcount for what your workforce is contributing.Worked, on the chapter's case: 1 − (33,530 / 35,481.48) = 5.5 %, which is 1,951.48 a year per person and 1,170,888.89 across six hundred.
Write down the range as well as the point. The published estimates of a differential like this run from 5.5 per cent for care work after controls to 17.0 per cent for teachers' total compensation. A single point estimate should not carry a pay conversation, and saying so yourself is what makes you credible in the room.
Exercise 2.2 — Compute the break-even, before anyone else does (one day)
You are going to be met with we cannot afford it. Have the real answer ready, including the part that is against you.
break-even turnover reduction = δ / c
At δ = 5.5 per cent and a replacement cost of 16.1 per cent of salary, that is 34.16 percentage points of annual turnover. Against a sector median of 94 per cent, a fall to 59.84. The wage elasticities of separation in the research run about 1.7 to 4.0, delivering 8.79 to 20.68 points — short of break-even by 25.37 to 13.48 points.
Say that yourself, first. The turnover argument alone does not close, and the person who brings it to the table before the finance director does is the person whose remaining numbers get believed. The residual lives in quality, in client retention, in incident and error rates — in exactly the outcomes your gainshare measure could be extended to cover.
Exercise 2.3 — Price what you already supply (half a day)
Meaning supplied to colleagues is real work and it is unpaid. Price a piece of it.
Beneficiary contact moved a fundraising line by 142 per cent in pledges and 171 per cent in revenue, from five minutes. Task significance moved lifeguards 43 per cent on hours and 21 per cent on helping. Judgement returned to a team of twelve took 40 per cent of the care hours out of a client's week.
Now find the equivalent in your own work — the handover you do properly, the context you give the new person, the reason you explain when nobody required it — and estimate its effect on one thing the firm counts. Two lines, with a number. That estimate is a gainshare claim in draft.
Exercise 2.4 — Learn one correction (one hour)
When your engagement score is presented, one move makes you the most useful person in the room.
A quartile contrast — top-quartile teams are 23 per cent more profitable — is a correlation of about 0.22 with its tails cut off. Top minus bottom is 2.5422 standard deviations of the predictor; multiplied by r, 0.559.
And where the survey supplies both the engagement number and the performance number, divide by 1.26: 0.40 becomes 0.317, and the variance explained falls from 0.1600 to 0.1008 — 37.0 per cent of it was the form. Ask, in a level voice, whether the outcome figure came from the survey or from the systems. That one question, asked once, changes how every subsequent number in the room is quoted.
Exercise 3.1 — Propose the measure, not the money (one week)
Never open by asking for the differential. Open by asking for it to be measured, annually, by finance, published with its method.
This is a smaller ask, it is very hard to refuse, and it does the work. An unmeasured subsidy grows quietly; a measured one acquires a committee that argues about its size. Draft it in four lines:
Exercise 3.2 — Ask for the account (two weeks)
The instrument is the Mission Differential Account: δ × W × N accrued annually as deferred compensation, credited pro rata, vesting after a stated period.
On the worked case: an annual accrual of 1,170,888.89 dollars, four-year vesting at a 6.0 per cent discount rate, a present value of 4,057,253.66 and 7,805.93 per worker at vesting.
The three clauses to read before anything else.
Exercise 3.3 — Extend your gainshare measure to the residual (one week)
The churn case does not close on its own. That is your opening, not your obstacle.
Propose that the gainshare measure be extended to the outcomes where the residual actually lives — client retention, incident and error rates, readmissions, complaints, clinical quality, rework. These are the things your work changes and your current measure does not see.
A gainshare that only counts productivity pays you for speed. A gainshare that counts quality pays you for judgement, and judgement is what the meaning differential is buying from you.
Exercise 4.1 — Get it owned by finance (one week)
Ask, explicitly, that δ be computed by finance rather than by people. A number owned by human resources drifts toward a narrative. A number owned by finance drifts toward a method, and a method survives the person who built it.
Exercise 4.2 — Teach three colleagues the calculation (two weeks)
The most durable governance in existence is several hundred people who can check a number.
Teach three people to compute δ. Ask each to teach one more. Within a quarter, the calculation is no longer yours and cannot be lost with you — which is exactly the outcome you want, and it is the same reason your baseline is signed rather than remembered.
Exercise 4.3 — Watch the two lines (ongoing)
One chart, two lines: engagement score, and median offer accepted for new hires.
If engagement rises while the accepted offer falls, the organisation has improved the supply of meaning and the labour market is collecting it. That is the mechanism running backwards, and it is working perfectly.
You are the person best placed to see it, because you are in the room when new people say what they were offered.
Opening. "I want to ask for one measurement, not for money. I would like us to publish, once a year, how our median pay compares with the skill-matched benchmark outside our sector, with the method attached."
If asked why. "Because we are almost certainly below it, and the gap is a contribution our people are making that appears nowhere. I would rather we knew the size of it than guessed. The published estimates for work like ours run from about 5.5 to 17.0 per cent."
If told it is unaffordable. "It may be. That is a separate question from whether we know the number. And I should say the turnover argument does not close on its own — break-even is δ over replacement cost, which is 34.16 points of annual turnover in our case, and the wage elasticities deliver perhaps eight to twenty. The case has to rest on quality, which is where I think our evidence actually is."
Closing. "Publish δ, accrue what we can fund, state the gap. If the number is small, we have lost nothing and gained a benchmark. If it is large, we would rather find out from ourselves."
Why it works. You asked for a measurement, you brought the argument against yourself before anyone else could, and you left the organisation a way to act that does not require it to admit anything. That is how a number gets adopted, and an adopted number does the rest of the work on its own.