A zero-tolerance commitment is a capital decision taken in the language of values: it abolishes an entire family of risk responses, and the cost of the responses that remain is never put back through the business case that authorised it.
The wording arrives late in the paper and it reads as the least contentious sentence in the pack. There will be no harm to the people in our care. No uncontrolled release, ever. Zero losses among the population we are moving. The board adopts it without division, because nobody in the room is willing to be recorded as having argued for a tolerable rate of harm, and because the sentence appears to cost nothing.
It costs a great deal, and the cost is incurred at adoption rather than at the incident. What the sentence does, immediately, is delete options. Risk management offers four broad families of response, and an absolute constraint removes two of them for every risk it touches. The enterprise is left with the two most expensive families and no mechanism for deciding how much of them to buy.
This matters because the decision was not taken where money is decided. Values commitments are adopted in a governance paper or a policy refresh. Capital commitments go through a business case, with an authorised envelope, an owner and a return. A zero-tolerance commitment behaves like the second while being processed like the first, and the enterprise carries an obligation of indeterminate size that no funding decision ever sized.
The question is narrow and, in ERANORTH's judgement, almost never asked: what did the absolute constraint cost, and in which line of the accounts does that cost appear?
The Strategic Context
The four response families are avoidance, transfer, reduction and acceptance. Avoidance changes the plan so the risk cannot arise. Transfer moves liability at a premium. Reduction lowers probability or impact and leaves a residue. Acceptance absorbs the event, passively or against a reserve. The taxonomy itself, and the discipline of testing all four before choosing, is the subject of Article 39 in this series and is taken here as given.
What an absolute constraint does to that taxonomy is structural and immediate. Acceptance is abolished outright: to accept is to agree in advance that some occurrence is tolerable, which is precisely what the constraint forbids. Transfer is largely abolished as well, and for a reason that is widely misread. Transfer shifts liability, not impact. An insurer can pay for a loss; it cannot make the loss not have happened. Where the constraint is written about the event rather than about the money, transfer no longer answers it.
That leaves avoidance and reduction, which any experienced risk practitioner will recognise as the two costliest families to buy. Avoidance is paid once, in design and scope, and it is rarely cheap because it means choosing a different route, a different technology, a different site or a smaller footprint. Reduction is paid continuously, for as long as the exposure exists, and it never finishes.
This much is visible in the discipline's own teaching material, and where that material stops is instructive: it notes that acceptance has been eliminated as a response, then moves straight to planning the remaining work, without observing that the elimination has just committed the organisation to the two families it has not costed.
What Leaders Commonly Misread
The first misreading is that insurance closes the gap. A commitment expressed as a maximum financial exposure can be transferred. A commitment expressed as a state of the world — no harm, no release, no loss — cannot be, because no counterparty can restore the state. Insurance buys the money to survive the failure of the commitment, which is a different product from compliance with it.
The second is the belief that reduction can satisfy an absolute. Reduction leaves residual risk by definition; that is what distinguishes it from avoidance, and an absolute admits no residue. Follow the logic strictly and every material exposure is pushed toward avoidance: redesign, narrower scope, doing less. Enterprises rarely follow it strictly, and the gap between the absolute they adopted and the reduction they actually bought is where the exposure quietly sits.
The third is the assumption that the usual proportionality test still functions. Risk practice conventionally bounds spending by comparing the cost of a response with the value of the impact it prevents. Declare the impact unacceptable and that comparison stops working: there is no denominator, because the enterprise has said the outcome has no price at which it becomes tolerable. The result is not unlimited spending. It is unarbitrated spending, settled at whatever level an operating manager can defend, with no test to determine whether it is too much or far too little.
Reframing the Issue
The useful reframing is to treat the constraint as an instrument of allocation rather than an expression of intent. Adopting it does three things at once: removes a class of response, commits capital to the design changes avoidance requires, and commits an operating cost stream for the life of the asset or service. Those are the properties of a capital decision.
One extension is consistently underrated. Where an absolute is paired with a long-horizon success criterion — survival, wellbeing or condition measured well after handover — the obligation crosses the project boundary onto operations. Nothing in a project business case funds a state that must persist for years after the team disperses. That cost is carried by a budget that never saw the paper.
Consider a rare earths processing development as a hypothetical. The enterprise commits publicly to zero uncontrolled release of process residues. Acceptance is gone; there is no reserve that makes a release acceptable. Transfer is thin, because insurance restores the balance sheet and not the licence. What remains is avoidance — a different separation route, a different feedstock, a different site — priced in capital, and reduction, priced in containment, monitoring, redundancy and standing response capability, for as long as the plant operates. The business case that authorised the development recorded a product margin. It did not record the constraint as a cost line, because the constraint was adopted somewhere else.
Where the Cost Goes When Acceptance Is Abolished
Into capital, as a design change nobody labels
Avoidance arrives disguised. It appears in the engineering record as a specification decision, a route change, a materials selection or a reduced throughput, justified on technical grounds. Rarely does the paper say the cause was a values commitment adopted at board level in a different quarter. The cost is real, capitalised and correctly spent — and untraceable to the decision that caused it. It is worth noting that many of the specification variables that drive this cost are not controlled by the enterprise at all; how a physical asset's specification is set by parties outside the site boundary is the subject of [Related article: The Specification Was Set Outside Your Fence], and is not pursued here, where the concern is only that avoidance is the family the constraint forces the enterprise into.
Into operating cost, as a subscription with no end date
Reduction is not a purchase, it is a standing charge: ratios, supervision, redundancy, testing, drills, monitoring, the second person on the shift. Because each increment is small and each is defensible, the aggregate is seldom presented as a single figure to anyone senior. In ERANORTH's judgement, most enterprises could not produce, within a week, the annual cost of maintaining their principal absolute constraint.
Into unserved demand, where no accounting system looks
The most consequential landing place has no ledger. Consider a children's respite care service as a hypothetical, operating under a zero-harm commitment to its clients. Acceptance is unavailable, transfer buys only indemnity, and reduction has a ceiling set by the workforce that can actually be recruited. The remaining response is avoidance, and in a service setting avoidance means declining placements: the complex case, the unfamiliar condition, the family whose circumstances the service cannot staff safely. Each refusal is prudent and each is invisible. The cost of the constraint is paid by demand that is never served, recorded nowhere, and reported as a capacity issue rather than as the price of a decision the board took.
Into the reserve that no longer applies
Active acceptance normally means holding a fund against the event. Abolish acceptance and that mechanism is withdrawn for the affected risk class: money that would have sat in reserve moves into the baseline as permanent expenditure. Which fund an event is allowed to draw on, and who holds the authority to release it, is the subject of [Related article: Two Funds, Two Authorities]; the point here is only that an absolute constraint removes an entire class of event from reserve eligibility, and that most enterprises move that cost into baseline expenditure without recording it as a change.
Decision Framework: The Abolished-Response Accounting
Run this before adopting any absolute constraint, and retrospectively for every absolute already in force. It takes a working day per constraint and produces a number.
Step one — state the constraint in testable form. Write it as the state of the world it forbids, and identify the population, asset or condition it protects. If it cannot be written this way, it is an aspiration and should not be adopted as a constraint.
Step two — list the exposures it touches. Every risk on the register whose impact would breach the constraint. Typically fewer than a dozen, and the discipline of naming them is half the value.
Step three — mark each family for each exposure.
| Response family | Effect of an absolute constraint | Where the cost lands |
|---|---|---|
| Acceptance | Abolished — no tolerable occurrence exists | Reserve withdrawn; cost moves to baseline |
| Transfer | Restricted to financial consequences only | Premium paid; the event itself remains yours |
| Reduction | Available but never sufficient alone | Permanent operating cost, rarely aggregated |
| Avoidance | Forced for any exposure with material breach potential | Capital, scope or refused demand |
Step four — price what remains. One capital figure for the avoidance decisions the constraint compels, one annual figure for the reduction it requires, and one estimate of the demand or scope declined because of it. The third figure will be resisted, and it is the one that matters.
Step five — return the total to the authorising body as a funding decision. The governance test is a single rule: an absolute constraint may only be adopted by a body that has, in the same sitting, seen and approved its three-part price. Where an existing constraint has never been through this, it has been adopted without authorisation, whatever the minutes say.
From Strategy to Execution
Immediate. Take the enterprise's most prominent absolute commitment and complete the accounting on it. Expect the annual reduction cost to be recoverable from existing budgets with effort, and the refused-demand figure to be missing altogether.
Medium-term. Change the approval route. Any paper proposing an absolute constraint goes to the same committee, in the same form, as a capital request. Require the risk register to carry a field showing which families are abolished for each exposure, so the restriction is visible where the work is done. Check also what the enterprise's mandate already permits by way of staging, phasing or partial commitment, because permitted flexibility is often the cheapest avoidance available and is routinely overlooked; how executives read past their own permissions is the subject of [Related article: What Are You Already Permitted to Do?], and is not the concern of this article, which begins after the constraint is set.
Long-term. Review absolutes on a cycle — not because commitment to safety or welfare should weaken, but because an unpriced absolute drifts into a tolerance without anyone deciding. The characteristic pattern is that the absolute survives in the policy while the operational target quietly becomes a percentage. That substitution should be a decision with a signature, not an artefact of someone needing a measurable objective.
Signals to Monitor
The clearest signal is the appearance of a rate where an absolute was written. When a policy says zero and the reporting pack says a percentage, the conversion has already occurred and nobody has been asked to authorise it. The second is an insurance renewal that is described internally as covering the constrained event, which indicates that the difference between liability and impact has been lost.
Watch also for capital variations justified purely on technical grounds in a domain governed by an absolute; for a register in which the gravest exposures carry reduction actions and no avoidance option recorded as considered; for growth in supervisory staffing never aggregated into one figure; and for a waiting list or scope exclusion that rises without anyone connecting it to a commitment made in a governance paper.
Questions for the Leadership Team
- For our principal absolute commitment, what is the annual operating cost of maintaining it, and which budget line carries it?
- Which capital decisions in the last three years were caused by that commitment, and does any approval paper say so?
- What demand, scope or work have we declined because of it, and where is that recorded?
- Which body adopted the commitment, and did it see a price at the time?
- Where in our reporting has the absolute already become a rate, and who authorised the change?
- If the commitment were relaxed by a stated margin, what would we stop spending — and does anyone in the enterprise currently know?
Closing Perspective
None of this argues against absolute constraints. There are populations, assets and obligations for which an absolute is the only defensible position, and an enterprise that will not adopt one has a different and more serious problem. The argument is about where the decision is taken and what accompanies it.
An absolute constraint transfers a decision from the risk function to the capital function without anyone noticing the transfer. The responsibility that follows is to make the transfer explicit: to insist that the sentence everyone agrees with arrives at the table with its price attached, so that the enterprise is choosing to buy the commitment rather than discovering, years later, in a budget line nobody can explain, that it already has.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.