Community Governance

Governance sounds like the dull part. It is the part that decides whether the rest of it survives.

The failure people expect is ideological — a group discovers it does not really share values and comes apart. The failure that actually happens is procedural. Nobody wrote down how authority is distributed, so for the first two years it does not matter, because everyone is in the room and enthusiasm carries it. Then someone leaves, someone new arrives, money gets tight, and the group discovers it has no agreed way to settle a disagreement. What fills the gap is not a method. It is whoever has the most time and the most history.

This topic is about where authority sits. Two neighbouring ones cover the rest: decision methods is how a group actually decides, and proposals & ratification is how a decision travels from written to in force.

What governance actually covers

Less than people fear, and more than they plan for.

It is not values. A statement of shared values is worth writing, and it settles nothing — you cannot be in breach of a value. What governance covers is the procedural core: who decides what, within which limits, by what method; who may stop a decision and on what grounds; when a decision gets revisited; and what happens when somebody does not do what was agreed.

That last one is the most commonly missing. A rule with no stated consequence is a preference, and every member works out which is which inside a year.

All of it lives in the community’s agreements Agreements The written rules a community has consented to. Distinct from values: agreements say what happens, and what follows when it does not. — the written text you can point at. The test is unglamorous and works on any visit: ask to read them, and ask when they were last reviewed.

The table that makes it real

Most groups have the information. It is scattered across four years of minutes, a handbook nobody updated, and the memory of two long-standing members — so in practice the answer to “who decides this?” is ask Marta. That is not a governance system. It is a dependency on a person, and it becomes visible at exactly the moment that person wants to leave.

A decision matrix Decision matrix A table mapping each kind of decision to who may make it, by what method, and within what limits. is one table: each kind of decision, who may make it, by what method, within what limits.

Decision	Who decides	Method	Limit
Repairs under €500	Maintenance circle	Circle consent	Within annual budget
Annual budget	All members	Two-thirds approval	—
Admitting a member	All members	Consensus	After the full visitor period
Changing this table	All members	Consensus	Two-week objection window

The last row is the one groups forget, and it is the most important: a document that cannot say how it is amended will be amended informally, by whoever notices first.

Building one is also diagnostic. Almost every community that writes this table discovers two or three categories of decision nobody had ever assigned, and at least one that two different bodies each believed was theirs.

The RCOS Standard requires both this and an authority registry — a companion list of who currently holds each role. The pairing is deliberate. The matrix says what a role may do; the registry says who is in it today, which is the question a new member actually has.

Where authority sits

A community of eight can put everything to everyone. A community of sixty cannot, and the attempt is what produces four-hour meetings about a compost rota.

The structural answer is delegation with a written boundary. Sociocracy Sociocracy A governance system organising a group into linked circles, each with its own domain, deciding by consent. is the best-developed version: circles Circle A small team with a defined domain it may decide within, linked to the wider group in both directions. with defined domains Domain The decisions a circle or role may make alone, and the limits it may not cross without asking., linked in both directions so information travels up and down rather than only up, each deciding by consent inside its boundary.

Note that sociocracy is a system, not a decision method. It answers “who is allowed to decide this without asking?” — a different question from “what counts as agreement?”, and groups that treat it as an alternative to consensus are comparing a filing system to a verb. The pair genuinely worth comparing has a page: sociocracy vs holacracy.

A domain is only real when it carries a concrete limit — usually a budget, a number the circle may spend without asking. This is the part most often omitted, and its absence produces the single most common false adoption in the movement:

If forming a circle has not made the general meeting shorter, the domain was never real.

The opposite failure is rarer and more expensive. A circle empowered to sign a five-year lease is making a constitutional decision inside an operational structure, and the group finds out when the bill arrives. Both are fixed the same way: write the limit, and default the unclear case upward to the wider group rather than downward.

Who may stop a decision

This is the single most consequential rule a group writes, and the one most often left implicit.

If any member may block anything for any reason, the community has not adopted a method — it has issued everyone a veto and hoped nobody uses it. That holds until the first person does, and then it holds never again.

A workable blocking concern Blocking concern In consensus, an objection serious enough to stop a proposal. Healthy systems define narrowly who may block and on what grounds. is defined three ways: who may raise one, on what grounds — almost always harm to the group’s stated aim rather than personal preference — and what happens next, because a block that only stops things creates a queue nobody owns.

Its necessary partner is stand aside Stand aside Recording that you disagree with a decision without stopping it. The release valve that makes consensus survivable at any size.: disagreeing on the record without stopping the group. Without it, anyone with a reservation faces only two options — swallow it, or escalate to a block they do not really believe in. Communities that use consensus at any scale almost always have an explicit stand-aside convention, and the ones that do not are usually the ones describing meetings as exhausting.

Is this a blocking concern?

A proposal is on the table: replace the common house roof this autumn, using the maintenance fund. Five people speak. Which of them has raised something that should stop the decision?

1. "I would have chosen a different builder."

A blocking concern This is a preference. It may be an informed one, and it is worth hearing during the proposal stage — but it names no harm to the group or its aim. Treating preferences as blocks is how consensus quietly becomes a veto. A preference correct answer Right. The test is not whether you would have done it differently; it is whether this would damage the group or stop it meeting its aim. "I would have chosen differently" almost never clears that bar.

2. "This spends the reserve we agreed in March to hold for winter heating repairs."

A blocking concern correct answer Right. It names a specific harm, ties it to something the group already agreed, and is checkable — anyone can go and read the March minutes. This is exactly the shape a valid objection takes. A preference It reads like a factual correction, which is what makes it strong. A reasoned objection points at a consequence the group has already said it does not want.

3. "I don't trust the person who brought this proposal."

A blocking concern It may be a serious and legitimate problem — but it is not an objection to this proposal, and blocking here would settle nothing. The roof still needs deciding, and the distrust would surface again next month on a different subject. Neither — this belongs somewhere else correct answer Right, and this is the one groups handle worst. A real concern about a person goes to the conflict process, not the decision. Communities that let it be raised as a block end up relitigating a relationship every time anyone proposes anything.

4. "We agreed last year not to take on new debt, and this needs a loan."

A blocking concern correct answer Right. Like the reserve, it is a conflict with a standing agreement — and the resolution is not to override it quietly but to either change the proposal or reopen the agreement deliberately. A preference A standing agreement is not a preference. Where a proposal contradicts one, the group has two honest options and neither of them is to proceed while pretending the agreement does not exist.

5. "Something about this makes me uneasy and I can't say what."

A blocking concern Not yet — an objection has to be articulable, or nobody can address it and the proposal simply stalls. But dismissing it is the wrong move too: unease often turns out to be right, and it is worth asking for time to find the words. Not yet — this is a stand-aside correct answer Right. Standing aside records the disagreement without stopping the group, which is what makes it possible to be honest about doubt. If the unease later becomes specific, it can be raised properly — and the minutes will show who felt it first.

The meeting itself

Method matters less than most groups expect. Facilitation Facilitation Running a meeting's process without taking a side on its content. The role that decides whether a decision method works at all. matters more.

The facilitator holds the process — whose turn it is, which stage the proposal is at, when a decision has actually been reached — while staying out of the content. Naming the decision sounds trivial and is the thing groups most often fail to do, leaving meetings that end with nobody certain what was agreed.

Rotate the role, even when someone is obviously good at it. A permanent facilitator accumulates informal power nobody granted: they decide what gets discussed and when a thing is settled. That is authority, whatever the agreements say — and under consent, where a facilitator judges whether an objection stands, it is a good deal more than that.

Matching structure to stakes

One more rule saves more meetings than any of the above: not every decision deserves the same machinery.

Sort decisions by what they cost to get wrong. Operational business — the rota, the repair, the order — should be fast, delegated and reversible. Strategic decisions that commit money or change direction deserve a higher bar and a real deliberation window. Constitutional questions about the community’s purpose, its agreements and its membership rules should be slow on purpose. And a decision nobody can classify should default to the higher tier, or every borderline case drifts downward toward the easier meeting.

That sorting into decision tiers Decision tiers Sorting decisions by what they cost to get wrong, and giving each class its own method and timeline. is covered under decision methods, and it is what stops a careful community from spending an evening on the colour of the bathroom tiles.

What breaks

Four patterns, in rough order of frequency.

Agreements stop being read. Not broken — forgotten. New members never see them, the authors leave, and the document describes a community that no longer exists. A review date written into the agreements themselves is the cheapest defence anyone has found.

Founder authority never converts. The people who started it keep deciding because they always have and because they know the history. Nothing is written, so nothing can be handed over, and the group discovers the problem at exactly the moment the founder wants to leave.

The structure is adopted in name only. A group says “sociocracy” and forms circles with no spending authority, so every plan still goes to the general meeting. Or it publishes a decision matrix and then decides the interesting things in the kitchen. The written system and the working system drift apart quietly, and the gap is only visible to someone who arrived recently enough to still be surprised.

The emergency that never ended. This one arrives disguised as competence. Something urgent happens — a burst pipe, a planning deadline, a member in crisis — and somebody sensible decides it, correctly, without the meeting. Everybody is relieved. The trouble is that nothing about the arrangement ends, because nothing about it ever began: no scope was set, no expiry, no review. The next urgent thing goes the same way, and eighteen months later the community has a decision-maker nobody appointed and cannot un-appoint.

Communities that avoid this have generally written the exception down before needing it — what it covers, and the date it stops. It is cheap to do in advance and close to impossible afterwards, because by then the question is about a particular person who has been right several times. A group that has only ever written down its ordinary process has, in effect, left its emergencies to whoever moves first.

Sources & further reading

1. 1 RCOS Core — Layer 2: Governance & Decision Logic — the specification this topic explains in plain language
2. 2 Ostrom, E. — Governing the Commons (1990) — the eight design principles, and the evidence behind them
3. 3 Sociocracy For All — circles, domains and consent, for groups without shareholders
4. 4 Beyond the Vote: Why the Most Resilient Communities are Replacing Consensus with Consent — the longer EcoHubs argument

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