How to read an institution in twenty minutes

17 August 2026 · The Law of Stated Purpose

This is the single most portable technique in the series, and it takes about twenty minutes on any organisation you have access to information about. It is also the one most easily misused, so the third step matters more than the first two.

Step one — the stated purpose

Find what the institution says it exists to do. Use its own words: the charter, the mission statement, the first paragraph of the annual report. Do not paraphrase it into something more cynical. Take it exactly as written, in its most generous reading.

Write it down in one sentence.

Step two — the produced outcome

Now write down what it reliably produces. Three rules for this step.

Measure over decades, not quarters. Any institution can have a bad year, and any institution can point to a good one.

Use outcomes it does not choose. Its own metrics are selected, and selected metrics tell you what it wants measured, which is useful information about the institution but not an answer to this question. Look for numbers gathered by somebody with no stake — a census, a regulator, an outside audit, a competitor.

Include what it produces incidentally. An institution that reliably generates a particular career path, a particular class of intermediary, or a particular category of dependent client is producing those things whether or not they appear in the charter.

Write that down in one sentence too.

Step three — the refusal test

Here is where most people stop too early, and where the analysis is actually made.

If the two sentences differ, that is not yet evidence of anything. Institutions are hard, environments change, and people are not very good at running large organisations. Divergence is the normal condition of the world.

The question that separates failure from design is this: did remedies exist, were they known, and were they refused?

You are looking for three specific things.

A known alternative. Somebody, somewhere, has solved this — a comparable body in another country, a pilot programme, an internal proposal. If nobody has ever solved it, you are looking at a hard problem, not a designed outcome.

Awareness. The alternative was raised, studied, reported on. There is a paper trail. It was not merely available in principle; it was on the table.

Refusal, repeated. It was declined more than once, over years, and the stated reasons for declining changed while the decline stayed constant. Shifting justifications for a fixed position is the strongest single signal in this whole procedure.

Where all three hold, the divergence has stopped being failure. The institution is producing what it produces because the arrangement suits somebody with the power to keep it.

Why this is not a conspiracy claim

Nothing in the procedure requires anybody to have planned anything, and this is the part that gets dropped when the technique is quoted badly.

The mechanism is selection, not conspiracy. An institution that produces a particular outcome is comfortable for the people who benefit from that outcome, and uncomfortable for the people who do not. Over decades, the first group stays and rises and the second group leaves. Eventually the institution is staffed entirely by people who sincerely believe in the stated purpose and who reliably decline every remedy, without anyone ever having decided to.

That is a more disturbing claim than conspiracy, not a milder one. A conspiracy can be exposed. Selection has to be reversed, and reversing it means changing what the institution rewards — which is precisely what the people it has selected are placed to prevent.

Where it fails

The technique has a real failure mode, and it is worth naming.

It cannot distinguish design from constraint when the constraint is invisible to you. An institution refusing an obvious remedy may be facing a binding limit you cannot see — a legal exposure, a funding condition, an agreement with a party that is not in the public record. From outside, refusal-under-constraint and refusal-by-design look identical.

The correction is the same as always: go back to step one of the method and ask who else is playing. A refusal that makes no sense given the players you have listed usually means your list is incomplete.

Try it on something small

The technique works better on a body you know from the inside than on a famous one you know from reporting. A professional association, a licensing board, a school district, a standards committee.

Small bodies are where the mechanism is legible, because you can actually find out whether a remedy was proposed and what happened to it. Large ones are where it is loudest and where you will most likely be wrong.

Common questions

How do you tell institutional failure from institutional design?

Three tests, all of which must hold. The divergence between stated purpose and produced outcome persists for decades. A known remedy exists elsewhere. And that remedy has been raised and declined repeatedly, with the stated reasons changing while the refusal stays constant. Shifting justifications for a fixed position is the strongest single signal.

Does this mean someone planned it?

No, and this is the part most often dropped. The mechanism is selection, not conspiracy. An institution producing a particular outcome is comfortable for those who benefit; over decades they stay and rise. Eventually it is staffed by people who sincerely believe the stated purpose and reliably decline every remedy. See the Law of Stated Purpose.

Where does this analysis go wrong?

When a refusal is caused by a constraint you cannot see — a legal exposure, a funding condition, an agreement not in the public record. From outside, refusal-under-constraint and refusal-by-design look identical. If a refusal makes no sense given the players you have listed, your list is probably incomplete.