Elinor Ostrom Proved the Commons Can Govern Itself

For most of the twentieth century, economics had a settled answer to the question of what happens to a resource everyone can use. It gets destroyed. Fence it off into private property or put a government in charge of it, the textbooks said, or watch it be stripped bare. Elinor Ostrom — an American political economist who became, in 2009, the first woman to win the Nobel Memorial Prize in Economic Sciences — did something the theorists had mostly not bothered to do. She went and looked. She found Swiss villages that had been grazing a shared alpine meadow since the fifteenth century without wrecking it, and she spent the rest of her career working out why.

I keep coming back to her because of the shape of what she did, not only the content. A discipline had reached a confident conclusion about how people behave. She tested it against the world, patiently, case by case, for decades. The world disagreed. That is a rarer story than it should be.

Who she was

Ostrom was born in Los Angeles in 1933 and died in Bloomington, Indiana, in 2012. She put herself through UCLA, and when she wanted to do a doctorate in economics she was told her high-school maths was not strong enough — she had been steered away from the advanced classes. She did her PhD in political science instead, finishing in 1965, and her dissertation was about groundwater: how pumpers drawing on a shared basin under Southern California had negotiated their way out of depleting it. The subject of her whole life was already there in the first piece of work.

In 1973, with her husband Vincent Ostrom, she founded the Workshop in Political Theory and Policy Analysis at Indiana University. The name is worth noticing. Not a centre, not an institute — a workshop, a place where things get made by people working alongside each other. It became one of the most productive research programmes in the social sciences, and it was built on collecting cases: forests, fisheries, pastures, irrigation canals, police departments, anything where a group of people had to share something that could run out.

In 2009 she shared the Nobel Memorial Prize with Oliver Williamson. She was a political scientist who had never taken the economics doctorate she was discouraged from, and she won economics’ highest honour for showing that one of its foundational parables was, at minimum, badly incomplete. I’ve written separately about that parable and why it is better understood as the tragedy of the commons as a myth than as a law of nature, so I won’t re-fight that argument here. What I want to set out is the positive thing Ostrom built in its place.

The method: go and count things

Ostrom’s great advantage was that she treated “can people manage a shared resource?” as an empirical question rather than a theoretical one. The theorists had modelled a pasture with no history, no neighbours, no gossip, no ability to talk to one another, and no memory. Real commons have all of those. So she and her colleagues built a database of hundreds of real common-pool resource systems, documented consistently enough that you could compare them, and asked what separated the ones that endured from the ones that collapsed.

The cases in her 1990 book Governing the Commons: The Evolution of Institutions for Collective Action are specific and unglamorous, which is exactly their power. The village of Törbel in the Swiss Alps, where in 1483 the residents formalised an association to govern the grazing meadows, forests and waste lands, with a rule that no one could send more cows to the summer pasture than they could feed through the winter — an elegant piece of design, because it ties your claim on the shared thing to a burden you have to carry privately. The huertas of Spain, irrigation systems around Valencia, Murcia and Alicante governed for something like eight centuries by tribunals of irrigators, one of which still meets weekly outside the cathedral in Valencia to settle water disputes. Common lands in the Japanese villages of Hirano, Nagaike and Yamanoka. The zanjera irrigation associations of the Philippines.

Some of these systems have run for a hundred years, some for the better part of a thousand. None of them were privatised. None of them were run by a distant state. They were governed by the people who depended on them, under rules those people had made and could change.

The theorists had modelled a pasture with no history, no neighbours, no gossip, no ability to talk to one another, and no memory. Real commons have all of those.

The eight design principles

Out of this comparison Ostrom drew eight design principles — features that showed up again and again in long-enduring commons and were missing or broken in the ones that failed. They are the most useful thing in her work, and the most often mangled, so here they are plainly:

1. Clearly defined boundaries. Both the resource itself and the set of people entitled to draw from it are clearly bounded. Everyone knows what is in the commons and who the commoners are. Without this you cannot tell a member from a stranger, and nothing else in the list can function.
2. Rules fitted to local conditions. The rules about how much you can take, when, and with what tools match the actual ecology and economy of the place — and the effort or money each member contributes is proportionate to what they get out. A rule designed for one valley’s rainfall will quietly destroy the next valley over.
3. Collective-choice arrangements. Most of the people affected by the operational rules can participate in changing them. This is not democracy as decoration. It is the mechanism by which rules stay fitted to conditions as conditions change, and the reason members treat the rules as theirs rather than as an imposition.
4. Monitoring. Somebody is actually watching the resource and the behaviour of the users — and crucially, the monitors are either the members themselves or are accountable to them. Not an outside inspector who answers to a ministry. Someone the commoners can fire.
5. Graduated sanctions. Members who break the rules face penalties that start small and escalate with the severity and repetition of the offence. The first transgression earns a warning and a bit of embarrassment, not expulsion. Harsh first-strike punishment tends to go unenforced, because nobody wants to ruin a neighbour over one bad week.
6. Cheap and accessible conflict resolution. There is a fast, low-cost, local forum for settling disputes between members, or between members and officials. The Valencia water tribunal is the model: it meets in public, it is quick, and it costs nothing to bring a case. A commons where disagreement can only be resolved in a distant court is a commons whose rules are effectively unenforceable.
7. Recognised right to organise. External authorities — governments, above all — do not challenge the community’s right to make its own rules. This one is easy to overlook and is often the fatal missing piece. A perfectly functioning local system dies the moment a state declares that only the state may allocate the forest.
8. Nested enterprises. For larger resource systems, governance is organised in multiple layers, from small local units up through larger ones — appropriation, monitoring, enforcement and conflict resolution each handled at the level where they make sense, rather than all pushed to the top.

Later research has broadly held these up. Meta-analyses across large numbers of documented cases have found the principles well supported, with some refinement — boundaries, for instance, turn out to do two distinct jobs, defining the resource and defining the membership, which are worth separating.

What the principles are not

Here is where I think she is most often misread by people who admire her, and it matters. The eight principles are not a recipe. They describe conditions associated with success, drawn from systems that survived. Assemble all eight and you have not guaranteed a functioning commons; you have removed some of the common causes of failure. Ostrom herself spent her later years warning against what she called panaceas — the urge to take a finding that worked in one context and prescribe it everywhere. That warning applies to her own list as much as to privatisation or state control.

The other honest caveat is scale. Almost every case she documented was small: a village, a valley, a few hundred families who could recognise each other. Monitoring works partly because you are watched by people whose opinion you care about. Graduated sanctions work because shame is a real currency between neighbours. Conflict resolution is cheap because the forum is walkable. Strip out the smallness and the mechanisms weaken. The eighth principle — nesting — is her answer to this, and it is more of a research direction than a solved problem. Scaling a commons is genuinely, unresolved-ly hard.

Assemble all eight principles and you have not guaranteed a functioning commons. You have removed some of the common causes of failure.

Why this is urgent now

The commons that matter most today are not pastures. They are digital: open-source software, Wikipedia, scientific literature, standards bodies, the shared pool of human knowledge that now trains machines. And the arguments being made about them are the old arguments in new clothing. This resource is too important to be left unowned. This resource is being abused, so it must be enclosed. It is the same logic that drove the enclosure movement, applied to code and data instead of to fields.

Read the eight principles against any large open-source project and they light up. Contributor agreements and maintainer lists are boundaries. Governance documents are collective-choice arrangements. Code review is monitoring by accountable monitors. The escalation from a review comment to a revoked commit bit is graduated sanction. Issue trackers and RFC processes are cheap conflict resolution. Foundations and licences are the recognised right to organise — a licence is precisely a claim that the community, not an external authority, sets the terms. Forks and working groups are nesting. The projects that endure tend to have these. The ones that burn out tend to be missing several, usually monitoring and conflict resolution, and they collapse into one exhausted maintainer.

Where the analogy strains is data. A pasture is rival — grass your cow eats, mine cannot. Data is not; copying it costs nothing. So the classic failure mode of overuse does not apply, and the real risks are different: enclosure by whoever can scrape fastest, and the extraction of value from a pool nobody can meaningfully withdraw from. Ostrom does not hand us an answer there, but she hands us the right question, which is not “who should own this?” but “who gets to make the rules, and are the people affected among them?” That question is the thread running through almost everything about how technology gets captured: capture is usually not theft of a resource, it is the quiet transfer of rule-making away from the people who depend on the thing.

What I take from her

Ostrom’s deepest finding is not the list. It is that institutions are made, not found — that ordinary people, given the right to organise and enough time, are capable of inventing governance of considerable sophistication, and have been doing it for centuries in places economists were not looking. The alternative view, that shared things require an owner or a policeman, is not a neutral observation about human nature. It is a claim that always happens to justify handing control to whoever is already powerful.

You can see the same argument playing out in living form in the Kerala cooperative model, where large-scale collective ownership works not because people there are more virtuous but because the rules, the monitoring and the dispute forums were built and kept in local hands. That is the Ostrom pattern exactly: not a nicer species, better plumbing.

The last thing worth saying is about her character as a researcher. She was told her maths was not good enough for economics. She ended up correcting economics with field notes, databases and an enormous amount of patience. When someone tells you a settled result about how people behave, the useful question is always the one she asked: has anyone actually gone and checked?