Resilience & Why Communities Fail

Where the ninety per cent comes from

You will meet the claim that ninety per cent of intentional communities fail. It appears in newspaper articles, in books, and from people who have lived in communities for decades.

We went looking for its source. The nearest thing to a foundation is Rosabeth Moss Kanter’s Commitment and Community (Harvard University Press, 1972), a serious piece of sociology that surveyed 91 communal projects founded in America between 1780 and 1860. Kanter needed an operational definition of success and chose one: a community counted as successful if it lasted at least 25 years. On that measure only a handful qualified.

Look at what has to happen for that to become a modern failure rate:

• A study of nineteenth-century American communes becomes a claim about communities today.
• A study of a specific window, 1780 to 1860, becomes a permanent rate.
• Lasting fewer than 25 years becomes failing, which would make most businesses, most marriages and nearly every voluntary organisation a failure.
• A sample of 91 self-selected historical cases becomes a rate for a category that had not been invented yet.

There is a second source, and it is the one most people are half-remembering. Diana Leafe Christian’s Creating a Life Together (2003) states that most aspiring ecovillages and community groups — “probably 90 percent” — never get off the ground; their envisioned communities are never built, because the group cannot find land, runs out of money, or is mired in conflict before there is anything to be in conflict about.

That is a real observation from somebody who edited Communities magazine and interviewed scores of founders, and it is explicitly hedged. It is also about groups trying to start a community, not communities that exist. A group that never built anything did not fail as a community; it never became one.

So there are two ninety per cents in circulation — a nineteenth-century longevity threshold and an estimate about forming groups — and they have been welded into a sentence neither author wrote. If you are joining a forming group, Christian’s version is the one that applies to you, and it deserves to be taken seriously. If you are looking at somewhere that already exists, neither figure tells you anything.

The useful half of Christian’s argument is not the number. It is her conclusion that the successful tenth had all done the same five or six things right, and the rest had made the same handful of mistakes — a claim about pattern rather than rate, which is what the section below offers.

The six patterns

1. A vision that was never written. Everyone agreed at the start, so nobody wrote it down. Four years later the group discovers it never agreed whether this is a housing project with shared meals or a life’s work, and both parties can point at the same warm memories.

2. No agreed decision method. Without a stated method, the first genuinely contested decision has no legitimate way to close.

3. Money without agreements. Not a lack of money — a lack of terms. Who owes what, what happens when someone cannot pay, what a stake is worth on the way out. Communities skip this out of kindness; it becomes distrust among friends later, with interest.

4. Founder authority Founder transition Turning the authority that built a community into something anyone can hold or inherit. Communities that skip it cannot disagree with their founders or continue without them. that never converted. The founders carried the project and were right to. What kills communities is when that informal authority is never turned into something anyone can hold, limit or inherit — so the group can neither disagree with them nor continue without them.

5. Conflict with no process. Not conflict. Conflict without a route. The dispute has nowhere to go, so it goes everywhere.

6. Drift. The slow one, and the most common. Nothing breaks. The agreements Agreements The written rules a community has consented to. Distinct from values: agreements say what happens, and what follows when it does not. stop being read, then stop being taught to new members, then describe a place that no longer exists. Shared meals thin from three a week to one to occasionally. Nobody decided any of it.

Drift deserves its own note because it is the only one that does not feel like anything while it is happening. The others announce themselves.

Which is also why vigilance is the wrong defence against it. Nobody notices a slow forgetting by paying closer attention; there is no moment to notice. What catches drift is duller and works without anyone being alert: dating the agreements, and keeping the old versions. A community that can say which version is in force, and when the last change was made and by whom, has turned a feeling into a check anybody can run.

The difference that makes is mostly social. Without a current text, raising drift means saying “I think we have lost something”, which is a criticism of everyone present and is heard as one. With it, the same observation is that the practice and the document disagree — a discrepancy, pointed at rather than levelled, and the sort of thing a group can look into without anyone having to be at fault. Communities that have been through it tend to describe wishing they had had that conversation four years earlier, and the reason they did not is almost never that nobody sensed it.

Endings that are not failures

Treating every ending as a failure is unkind and analytically useless, and it is the habit that makes the ninety per cent figure feel plausible.

Completion. Some communities form to do something — house a group through a stage of life, hold land until a trust could take it. Finishing is not failing.

Deliberate dissolution. A group decides, while it still can, that this is not working, sells up, pays everyone out under an agreed exit clause Exit clause The written terms on which a member leaves, or the community winds itself up. Agreed while everyone is still calm, which is the only time it can be., and goes. Executed well that is a governance success, and far better than a slow attritional decline where the last three members hold a building and each other’s resentment.

Transformation. A commune becomes a housing co-operative. The original thing ended; something continued.

People leaving. A household leaving is an event, not a verdict.

The endings genuinely worth avoiding are the ones where nobody chose — the money ran out, the founders left with nothing transferred, one conflict ran until the place emptied. What those share is not bad luck. It is that a decision was available earlier and nobody made it.

Sources & further reading

1. 1 Why communities end — the lesson in full, with a quiz on telling a warning sign from a false alarm
2. 2 Kanter, R. M. — Commitment and Community (Harvard University Press, 1972) — 91 projects founded 1780–1860, and the 25-year threshold everything traces to
3. 3 Christian, D. L. — Creating a Life Together (New Society Publishers, 2003) — the "probably 90 percent" of forming groups that never build anything
4. 4 Altus, D. — The Value of Community: What Defines Success? — a Communal Studies scholar questioning longevity as the measure; the 90% figure appears in the comments. Archived — the Foundation for Intentional Community has since removed the original.
5. 5 RCOS Core — Layer 6: Evolution & Adaptation — versioning, time-bounded experiments, and the learning log that makes drift visible
6. 6 Communal Studies Association — the academic literature on communal societies, historical and contemporary

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