The Question

From the documentation of the field

What must a decision-maker get right, structurally, to avoid failure under uncertainty?

This is the one question the field exists to answer. Every other page here is part of the answer, a way of checking it, or a statement of where it stops.

Where the question comes from

The Dossier collects 26 articles, each given an independence rating, within an investigation window of August 2020 to August 2026, about eight major consulting and audit firms and the industry around them: McKinsey, Deloitte, PwC, EY, BCG, Bain, Accenture and KPMG. They cover settlements, sanctions, a government report that contained fabricated citations, a failed restructuring and a banned contractor. They also include surveys, academic work, and defences of the industry, so that the record is not one-sided.

The Dossier reads the pattern as predominantly structural. Its root-cause analysis weights the failures at roughly 65% structural decisions (what a firm sells, how it is paid, whom it serves) and 35% operational execution. The weighting is the proportion observed when the entries are classified by the kind of failure they describe. Two of its better-supported findings show what "structural" means there. A Bloomberg Tax report dated 2022-08-17 put advisory fees above audit fees at the Big Four, which, on the Dossier's reading, gives an auditor a financial reason not to look too hard at a client that also buys consulting (Entry 19). A study of KPMG in Germany found that clients suffered negative market reactions after failures in KPMG's advisory work while KPMG itself did not, which suggests reputation alone may not discipline large auditors (Entry 14).

If the pattern is structural, asking who erred stops being the most useful question. The Dossier describes the pattern, weighs its causes and supplies the reason to ask the question; it stops short of testing remedies or answering it.

Why this question and not a nearby one

"Who is at fault?" asks about people. The Dossier's pattern recurs across eight firms with different leaders and different clients, so an answer that names individuals does not carry from one case to the next. A question about structure can.

"How do I make a good decision?" is downstream. Advice on weighing options, estimating probabilities or resisting bias works inside a structure. A structure can hold capable people making careful choices and still lack any means of noticing that its picture of the world is out of date, or of stopping an action that cannot be undone. The question here is about what the structure must have for good choices to count.

"How do I succeed?" is a different kind of question. The question asks what must be right in order to avoid failure, and that is a question about necessity. Failure under uncertainty has a definite shape that can be stated and proved. Success depends on what a particular organisation is trying to do, so no structural answer to it can be general. Necessity and sufficiency sets out where the field draws that line.

Each word in the question does some work. A decision-maker can be a person, an institution or a machine. "Structurally" means the property belongs to the arrangement, whoever happens to be inside it. "Failure" means a loss that compounds as time passes, not a bad outcome on one occasion. "Under uncertainty" means the future is not known in advance. With a known future, no structure would be needed.

What the field's answer looks like

The research this field rests on answers the question with six capabilities. In a complex environment, a decision structure that lacks one of the six primitives of Ismail's Primitives incurs linear regret (the necessity result, proved in SixPrimitives/Phase3.lean): its loss grows in proportion to time. None of the six can stand in for another, and the six are required together. Every claim is machine-verified in the Lean 4 proof assistant. The proofs use the smallest environments that demonstrate each capability's necessity, and the theory applies wherever the difficulty appears. It is a statement about structures that lack a capability and makes no forecast about any organisation.

Next, read the field definition. To check the mathematics yourself, start with audit it yourself. The paper and the Lean repository are listed on the references page.