Haute Lumière

Commerce · V.07 · MMXXVI · daylight

La Bourse  /  Volume V  /  Nº V.07  /  Workbook — the executive

WORKBOOK — THE CORPORATE EXECUTIVE

Chapter V.07 · Income, Floor, and Dignity

For the person with a P&L, a headcount and a signature. You cannot legislate a national floor. You can price and sign one inside your own firm this quarter, and the arithmetic that justifies it is already in your own turnover numbers.

---

THE PREMISE, STATED COMMERCIALLY

Strip the politics out and what is left is a variance problem you already know how to price.

An hourly workforce on a fluctuating rota faces monthly income that moves ten to twenty percent with no warning. That variance, not the level of pay, is what drives the payday loan, the missed rent, the absence and the resignation. Your firm is already paying for it — in turnover, in overtime premium to cover gaps, in agency cover, in the productivity of a team working around a vacancy — and it is paying for it in lines that are not labelled.

The Floor Facility is the instrument that converts that unlabelled cost into a priced liability with a stated break-even. In the chapter's worked case it costs £603,688 a year across 2,000 people and it is free at a turnover reduction of 10.06 percentage points.

That is the whole proposition. Everything below is how to get it signed.

There is a second commercial reason, and it is the one that survives a change of CFO. Your firm competes for hourly labour against every other employer in the same travel-to-work area, on a wage that is set within pennies by the market and the National Living Wage of £11.44 an hour. A pay floor is the only differentiator in that market that costs you less than a wage rise and is worth more to the recipient. A 50p hourly rise costs £910 a year per full-time head; the guarantee in the worked case costs £301.84.

---

PART ONE — DISCOVERY

Days 1–30: find the variance you are already paying for

Exercise 1.1 — Pull the real distribution (one analyst, one week)

Do not use average pay. Averages are exactly what hides this.

Pull 24 months of net monthly pay by individual for every hourly employee. For each person compute their own median and their own standard deviation as a percentage of it. Then produce one table: the distribution of personal pay volatility across the workforce, in deciles.

You are looking for the top three deciles. In most hourly operations they run at 18 to 30 percent, and they contain most of your leavers. Cross-reference against last year's leaver list before you do anything else. If the top volatility decile is over-represented among leavers, the business case is already written and you are only quantifying it.

Exercise 1.2 — Cost what you already pay (half a day with finance)

Three numbers, and none of them is usually assembled in one place:

1. Replacement cost per leaver, loaded. Advertising, agency fee, induction, training hours, the productivity ramp, and the overtime premium paid while the post is open. The chapter's worked case uses £3,000 for an hourly role; yours may be higher and is very unlikely to be lower.
2. Annual turnover, hourly population only, on a consistent definition. The worked case uses 60.0 percent — 1,200 leavers on 2,000 heads, £3,600,000 a year.
3. Cover cost. Overtime premium and agency spend attributable to unplanned absence and unfilled shifts.

Exercise 1.3 — The appreciative interview (two days, with your own people)

Twenty conversations, on shift, with the question asked exactly:

"Tell me about a month here that went unusually well for you financially. What made it different? What could you do that month that you cannot usually do?"

Take notes on the conditions, not the outcome. You are looking for the sentence that will end up in the board paper, and it is always a specific one about a specific month.

---

PART TWO — THE ARITHMETIC

Days 31–45: price it properly

Exercise 2.1 — Price the guarantee (half a day)

The expected top-up per person per month is the standard shortfall integral, and it is one line in a spreadsheet:

  E[(a − X)⁺]  =  σ · ( z·Φ(z) + φ(z) ),    z = (a − μ) / σ

where μ is the person's trailing median, σ their standard deviation, and a the guaranteed floor.

Worked, from the chapter: median net monthly pay £1,850, volatility 12.0 percent so σ = £222.00, floor at 90.0 percent of median so a = £1,665.00. Then z = −0.8333, Φ(z) = 0.2023, φ(z) = 0.2819, and the factor z·Φ(z) + φ(z) = 0.1133. Top-up is £25.15 per person per month, £301.84 a year, £603,688 across 2,000 people.

Price three floors, not one. Run 85, 90 and 95 percent of trailing median. The cost is convex in the floor level — moving from 85 to 95 percent roughly triples it — and you want the board to see the curve, because the curve is what lets them choose rather than approve.

Sensitivity to volatility is linear: the same floor at 18 percent volatility rather than 12 costs £37.73 per person per month, £905,533 a year, because z is unchanged and only σ has scaled.

Exercise 2.2 — Compute the break-even (one hour)

The number that decides it, and it goes on the front page:

      annual cost of the guarantee            £603,688
  ---------------------------------------  =  ------------------  =  10.06 points
     headcount × replacement cost             2,000 × £3,000

If the guarantee moves annual turnover by more than 10.06 percentage points, it is free. At 5 points it saves £300,000 against a £603,688 cost — net £303,688 out. At 10 points it saves £600,000 and is within £3,688 of break-even. At 15 points it saves £900,000 and returns £296,312.

Present all three. A single point estimate invites a single objection; three points invite a judgement about which is likely, which is the conversation you want.

Exercise 2.3 — Find your own honest negative (one hour)

Write it before anyone else does. Three candidates, and at least one will apply to you.

The floor is too low to change a decision. A guarantee that never binds costs little and does nothing. Check what proportion of person-months actually fall below your proposed floor; if it is under five percent, raise the floor or drop the scheme.

The guarantee is gamed at the rota. If managers can reduce scheduled hours knowing the floor will cover the gap, you have converted a retention instrument into a subsidy for poor scheduling. The control is to hold rota managers' hours budgets gross of top-up and to report top-up by manager.

Selection. The people who value the floor most are the people with the most volatile hours, who may also be the people with the highest turnover for reasons the floor does not touch. Your break-even assumes the guarantee causes the retention improvement. Run it on one site against a matched site so that it does not have to be assumed.

---

PART THREE — DESIGN

Days 46–60: the instrument, in a form the treasurer signs

The structure. The firm guarantees each hourly employee a minimum net monthly pay equal to a fixed share of their own trailing three-month median. Any shortfall is topped up in the same payroll run. It is not a loan, it is not advanced against future hours, and it is never recovered.

The balance sheet treatment. It is a constructive obligation with a reliably estimable expected outflow, so it is provisioned rather than disclosed as a contingency. Carry the reserve at twelve months of expected top-up — £603,688 in the worked case — plus a buffer sized to the historic worst month. Recognise the expense against payroll as it accrues, not as an exceptional item, because an exceptional item is a thing the next CFO removes.

Your auditors will ask whether this is a defined benefit obligation. It is not: the obligation is capped at the shortfall, is computed monthly, and terminates with the employment. Have that answer ready at the first meeting rather than the third.

The counterparty. Yourself, for the first twelve months. This is a self-insured retention, not a product, and it needs no broker and no underwriting. Once you have twelve months of loss data you can buy a stop-loss wrapper above your historic worst month for very little, because you are now selling an insurer a dataset rather than a hypothesis. At that point the whole structure can be syndicated across a sector, and the first firm to assemble twelve months of clean data owns that conversation.

The governance. Three things, all small, all load-bearing:

• The floor is published, as a percentage, to every employee it covers. An unpublished guarantee buys none of the behavioural effect and all of the cost.
• Top-up is reported by manager, monthly, in the operations pack. It is the gaming control and it is also the early warning on rota quality.
• The floor is indexed, to your own median hourly pay rather than to prices, and reviewed on a stated date. An unindexed floor erodes silently, which is how almost every guarantee in history has ended.

What not to do. Do not fold the guarantee into a discretionary bonus pool. Do not make it conditional on attendance, performance or length of service — the moment it is conditional it stops being a floor and becomes a target, and it buys you neither the retention nor the goodwill. Do not means-test it against household income; you will have built a withdrawal band inside your own payroll, and the chapter is largely a demonstration of how badly those behave.

---

PART FOUR — DESTINY AND DELIGHT

Days 61–90: prove it and make it hold

Exercise 4.1 — The matched-site pilot

One site with the guarantee, one comparable site without. Agree and sign the baseline before launch: turnover on a stated definition, absence, agency spend, and overtime premium, over a period long enough to contain normal variation.

The chapter is emphatic on this point for a reason that applies exactly here: an unagreed baseline is not a baseline, it is a future dispute. When the result arrives, anyone who dislikes it will attack the comparison.

Exercise 4.2 — Report against the break-even, not against a target

Each quarter, one line: top-up cost to date, turnover change to date, and the implied position against the 10.06-point break-even. Nothing else. A single line that a reader can check is worth more than a dashboard nobody opens.

Exercise 4.3 — The second owner

One person is a project; two is a practice. Recruit a second owner before you need one — ideally in operations rather than in HR, because the operations director's turnover number is the one the facility moves — and give them the credit for the first result.

Exercise 4.4 — Notice what it feels like on the floor

The behavioural effect arrives before the money does. In Finland, participants' life satisfaction moved from 6.76 to 7.32 on a ten-point scale, and the evaluators attribute most of it not to the €560 but to the end of proving. The same thing happens in a payroll: the month people stop calculating is the month they stop looking.

Ask, on shift, eight weeks in: has anything changed about the first week of the month? You will get the sentence for the board paper there, not in the data.

---

THE FAILURE MODES, NAMED

So you can see them coming

It is quietly not renewed. Reversal is rarely a decision; it is an absence of renewal that nobody has to defend. The defence is that the floor is published to employees and reported in the standing operations pack. Removing a published guarantee requires somebody to explain why, and that is a conversation nobody wants.

It erodes. An unindexed floor falls against your own wages every year without anyone choosing it. Index to your median hourly pay and put the review date in the scheme document.

It is means-tested into uselessness. The moment somebody proposes withdrawing it against household income, you have built a withdrawal band. Compute the resulting effective marginal rate and put it in front of them — in the UK, income tax at 20 percent, National Insurance at 8 and a 55 percent taper compose to 67.6 percent, which is 22.6 points above the 45 percent top rate of income tax. That number ends the proposal faster than any argument.

It is sold as ethics. An ethical proposal is approved when there is money and cut when there is not. A retention instrument with a published break-even of 10.06 points is approved on its arithmetic and survives the next downturn, which is precisely when the people it protects need it most.

---

THE NINETY DAYS ON ONE PAGE

---

BOARD PAPER TEMPLATE

Recommendation. Approve a pay-floor guarantee at 90 percent of trailing three-month median net pay for the hourly population at [site], for twelve months, at a budgeted cost of £[ ], provisioned against payroll.

The number that decides it. The guarantee is cost-neutral at a turnover reduction of [ ] percentage points against a current rate of [ ] percent. At the mid case of [ ] points it returns £[ ].

What we are buying. Not a pay rise. A reduction in the variance of pay, which is what the evidence identifies as the driver of the turnover we are already paying for.

What could go wrong, and the control for each. [The three from Exercise 2.3, each with its control.]

How we will know. Matched-site comparison against a baseline signed before launch, reported quarterly as one line against the break-even.

---

APPRECIATIVE QUESTIONS FOR YOUR LEADERSHIP TEAM

1. Where in this business does someone already have a guaranteed floor — a retainer, a salaried role in an hourly team — and what do we observe about how they work that we would want more of?
2. Think of a month when one of our sites ran unusually well. What was true about how people were scheduled and paid that month?
3. What have our longest-serving hourly people told us about why they stayed, and how much of it is about money at all?
4. If nobody here had to take a shift out of fear, what would we have to change about how we ask for one — and what would improve?
5. What would keep this guarantee in place through a bad year, and who would defend it if the person who built it had left?