Power nobody voted for

Start with the thing that makes this lesson difficult.

In almost every community where power has settled somewhere nobody chose, the people holding it are working harder than anyone else. They answer the emails at midnight. They remember why the rule exists. They are the reason the boiler works. Asking a group to constrain them feels like ingratitude, and it usually is experienced as ingratitude, which is why the conversation gets postponed for years.

So the framing that works is not who has too much power. It is what happens when they leave — and that question can be asked with warmth, out loud, in a meeting, without anybody being accused of anything.

The five, and what they have in common

They are separate patterns with separate fixes, but they share a mechanism: a decision that was never assigned found a home. Nobody took the authority. It was left on the table, and somebody sensible picked it up.

• The clique that decides before the meeting — the real conversation happens in the kitchen, and the meeting ratifies it
• Speakers who fill the room — nothing is blocked, but only some people’s proposals ever get finished
• The founder’s informal veto — no rule grants it, and everyone routes around it anyway
• Decisions with no stated type — the same question is operational or constitutional depending on who is asked
• Authority from charisma — a person is so compelling that disagreeing with them feels like a failure of understanding

If you recognise one, read the others before concluding you have found the whole problem. They co-occur, and they reinforce each other: a founder’s unwritten veto is far more powerful in a group whose decisions have no stated type, because there is nothing to say the veto does not apply here.

Why it is hard to see from inside

Three reasons, and the third is the one that keeps communities stuck.

It looks like competence. Everything above is indistinguishable, day to day, from a group that has sensibly deferred to the person who knows most. The difference only shows up under stress — when the group wants to go a way that person does not.

Nobody experiences themselves as powerful. Ask the founder and they will tell you, sincerely, that they wish other people would take more on. That is usually true. Informal power is not a thing people feel themselves holding; it is a thing other people feel themselves working around.

And the group already knows. This is the part that surprises people. In our experience the pattern is almost never a secret — members will describe it accurately, in private, sometimes for years. What they lack is not perception. It is a legitimate way to raise it that does not read as an attack on someone who has given a great deal.

That is the whole argument for writing things down. Not because a written rule is wiser than a good person, but because a written rule can be discussed without anybody being the subject.

Is it the structure, or is it the person?

Five situations. In each one, decide whether you are looking at something the structure produces — which would happen to whoever was standing there — or something about this particular person. They look identical from inside, and the wrong answer costs years.

1. Over six years three different people have held the money role. Each time, within about a year, the group is complaining that spending decisions get made without them.

The structure — this would happen to whoever was standing there correct answer Right, and this is the cleanest test there is. When the same complaint survives three changes of personnel, the one thing it cannot be about is the personnel. Something about the role — its limits, its spending threshold, what it must bring back to the group — was never written, so each new holder discovers the boundary by crossing it. The person — the last three have all overreached Three people with nothing in common but a job title is not a run of bad luck. If your explanation requires each of them to have independently developed the same fault, the explanation is doing too much work. Both, and the structure has to be dealt with first Defensible, but it concedes too much. There is no evidence here about any of the three as people — only about what happens to anybody who takes the role. Adding a character question invites a conversation that will hurt somebody for no diagnostic gain.

Where this comes from

2. One member routinely talks over others. They have been asked privately three times, by three different people, and agreed each time. Nobody else who has facilitated has done this.

The person — this is about conduct, and no rule will fix it correct answer Right. Asked, agreed, unchanged, three times, and not reproduced by anybody else in the same position — that is conduct, and a decision matrix has nothing to say about it. Communities reach for a structural fix here because it feels less confrontational, and they end up writing a rule that constrains everybody in order to avoid one conversation. The structure — there is no facilitator with authority to intervene Worth fixing anyway, and it will make the room better. But it does not explain this: other facilitators in the same structure do not produce this behaviour, and someone who has agreed three times and changed nothing is telling you something about themselves. Both, and the structure has to be dealt with first Doing the structural half first is exactly how this gets postponed for another two years. A facilitation agreement gives the group something to point at, which is useful — and it is also a way of never saying the thing to the person.

Where this comes from

3. Nothing proceeds if the founder disapproves, though no document says they may decide anything alone. Asked about it, the founder says — sincerely — that they wish other people would take more on.

The structure — this would happen to whoever was standing there correct answer Right, and the sincerity is the evidence, not a defence. Informal power is not something people feel themselves holding; it is something other people feel themselves working around. The founder is describing their experience accurately and it tells you nothing about whether the veto exists. What exists is an authority nobody granted, nobody can limit, and nobody can inherit. The person — they are more attached to control than they admit This is the reading that feels insightful and closes the only door that was open. It requires the founder to be lying about their own experience, which is both unlikely and unprovable — and it turns a question about roles into a question about somebody’s character, which cannot be raised in a meeting. Both, and the structure has to be dealt with first The structural half is the whole of it here. There is nothing on the record about how this founder behaves — only about what the group does in anticipation, which is a fact about the group.

Where this comes from

4. A question about the guest policy went to the whole group last spring and was settled by two people in the autumn. Both times, everyone involved thought they were following the normal process.

The structure — this would happen to whoever was standing there correct answer Right. Nobody bent a rule, because there was no rule to bend: the same question is operational or constitutional depending on who is asked, so both routes were taken in good faith. Until decisions carry a stated type, every borderline case drifts toward whichever meeting is easier to convene. The person — whoever took it in the autumn overstepped They thought they were following the normal process, and so did everybody around them at the time. Somebody who cannot have known they were overstepping has not overstepped; they have found a gap. Both, and the structure has to be dealt with first Nothing here distinguishes the autumn pair from the spring group except which route they took. Treating that as partly a character question would make two people answer for a gap the whole community left open.

Where this comes from

5. A member is genuinely unpleasant in meetings — several people have stopped coming. They are also the only person who understands the accounts, and have been for eleven years.

Both, and the structure has to be dealt with first correct answer Right, and this is the one case where the order matters. The conduct is real and needs saying. But a community that cannot survive a month without this person cannot have that conversation honestly — every version of it is shadowed by what happens if they walk out. Get a second person into the accounts and the conversation becomes possible. Do it the other way round and you are negotiating with someone who holds the lights. The person — this is about conduct, and no rule will fix it The conduct is not in doubt. What is in doubt is whether this group can act on it, and eleven years of sole custody of the accounts is the reason it has not. Naming the behaviour without changing the dependency is how communities arrive at an ultimatum they cannot afford to mean. The structure — this would happen to whoever was standing there The single point of failure is structural and would happen to anybody. Being unpleasant in meetings would not. Reading this as purely structural is how a group talks about succession planning for two years and never mentions why people stopped coming.

Where this comes from

If it is already happening

Stabilisation before prevention. A community in the middle of this does not need a governance redesign; it needs the next three months to be less bad.

Name the smallest true thing. Not “we have a power problem” — “I noticed the last four decisions about money were made between meetings.” A specific, checkable observation is arguable. A diagnosis is not, and it will be experienced as an accusation.

Ask the succession question, not the power question. What happens to this if you are ill for a month? It is a real question, it is kind, and it produces the same information.

Write down one decision. Not the whole matrix — one. Pick the category that keeps going wrong and state who decides it, by what method, within what limit. A community that has never written any of this down will find the first one takes an evening and the second takes twenty minutes.

What prevents it

A decision matrix Decision matrix A table mapping each kind of decision to who may make it, by what method, and within what limits., and the thing communities forget to pair with it: a register of who currently holds each role. The matrix says what a role may do; the register says who is in it today, which is the question a new member actually has.

Then the two rules that make it hold:

• A decision nobody can classify goes to the higher tier. Without that, every borderline case drifts toward the easier meeting.
• The table says how the table is amended. A document that cannot describe its own amendment will be amended informally, by whoever notices first.

Sources & further reading

1. 1 RCOS stress tests — Governance & Power — four of the five patterns this lesson covers, in specification form
2. 2 RCOS stress tests — Culture & Influence — the fifth — authority from charisma, which the standard files under culture rather than governance
3. 3 Freeman, J. — The Tyranny of Structurelessness (1970/1972) — why a structureless group is impossible, and what informal elites form around
4. 4 RCOS Core — Layer 2: Governance & Decision Logic — authority boundaries, the decision matrix, and the safeguards against concentration
5. 5 Conflict, power and repair — the joiner-facing version, for reading a community from the outside

Written by EcoHubs members, with AI assistance for drafting and editing, and reviewed by a person before publication. Facts are checked against the sources listed; anything we could not verify is marked. How this is written