The refusal list

Every professional firm publishes the list of what it sells. Almost none writes down the other list: what it will not do — engagements declined, claims never made, shortcuts never taken — regardless of the fee. Where it exists at all, it exists as culture, tested only on the day a refusal becomes expensive.

Our position, written into how this firm works, is that a standing refusal list is a governance instrument, not marketing. The reasoning is mechanical, not moral: a refusal made in advance is cheap, a sentence adopted while no fee is on the table, whereas a refusal under commercial pressure must be produced at the moment the fee is visible, the relationship warm and the rationalisation ready. Case-by-case judgment prices a refusal at its most expensive moment.

The claim is narrow. A written list does not guarantee integrity; without enforcement it is a brochure. What we argue is that the instrument itself — a categorical, written prohibition adopted before the pressure arrives — is what regulators, ethics codes and the behavioural research reach for when judgment under fee pressure cannot be trusted.

Start with the strongest precedent: auditor independence law is a legislated refusal list. Rather than ask statutory auditors of public-interest entities to weigh a lucrative mandate against their audit case by case, the European Union wrote the refusals in advance, categorically.

The United States built the same instrument earlier, by the same logic. Both regimes concede the same thing: the regulator does not trust even a sophisticated professional to decide well while the fee is in view.

Timing is the active ingredient, and the lineage runs back to Schelling: the constraint works because it is adopted before the temptation arrives.

The professional ethics architecture completes the picture. The IESBA International Code of Ethics works through threats and safeguards, but at defined points it writes a categorical prohibition instead — a firm shall not charge a contingent fee for an assurance engagement, full stop.

Our position, graded as such: the same instrument belongs inside an advisory firm, in writing, before the fee arrives. It is worth most where refusing is hardest. In our corridor — European holding boards deciding for operating companies in Abidjan, Dakar or Lomé, payrolls in CFA francs, revenues consolidated in euros — the advisory market is thin, and declining an engagement can cost a relationship built over years rather than a fee, which is the condition under which case-by-case judgment fails.

Our own list is short, written and dated. We do not take contingent success fees on judgments we grade: a fee that depends on the answer is a thumb on the grade. We do not accept engagements whose conclusion is fixed before the evidence is examined — advocacy dressed as analysis. We do not reuse one client's non-public data in another client's work. And we do not make claims graded above their evidence, at any fee, in either language we publish in.

Evidence cards

Enforcement, not drafting, separates the instrument from the brochure. Our list is versioned and dated, so a client can see it predates the mandate; any waiver requires a recorded decision by the Managing Partner in a register our assurance process reads, and an unrecorded waiver is a breach, not an exception.

Sources and limitations

Sources and limitations. The institutional facts in this note rest on the text of Regulation (EU) No 537/2014 Article 5, Sarbanes-Oxley s. 201 with SEC Rule 2-01(c)(4), and the IESBA International Code of Ethics, each cited with its caveat; the behavioural mechanism rests on Ariely and Wertenbroch (2002), which examined no advisory firm. The step from those anchors to the refusal-list rule is the firm's position, graded SEG-2: an interpretation consistent with the cited architecture, not a measured result — we cite no dataset comparing firms with and without standing refusal lists, because we found none meeting our standard. The boundary is explicit: the independence regimes cited bind auditors, not consultancies, and we invoke them as design precedent only; nothing here is legal advice on independence law, and nothing here claims that any adviser without a written refusal list acts improperly. On our reading, a refusal list is necessary discipline; it is not sufficient.

1. The Chair requests the adviser's refusal list, written and dated before the engagement began, as a condition of the next mandate. If it arrives after the question, refusals are being decided case by case, under pressure.

2. The procuring executive asks for the most recent engagement declined under it, anonymised as professional confidence requires. A list that has never refused anything is marketing by other means.

3. The Managing Partner names one waiver authority, opens the register, and budgets the revenue the firm expects to decline. A list without a named authority is a preference, not a control.

4. The board adopts its own short list, in advance and in the minutes. Related-party terms, guarantee extensions, year-end exceptions: a board that writes its own list buys what it should demand from its advisers.