Why the public farm always ends up empty

A shared farm runs dry because the benefit of using it is individual and the cost of maintaining it is collective: whoever takes materials keeps all of them, while whoever repairs and restocks invests time of which they only get part back. With that split nobody has to be selfish for the resource to empty; it's enough for each person to take a bit more than they contribute.

It's an old phenomenon, heavily studied and with known remedies. What's interesting is that servers rediscover it on their own, always the hard way.

The scene, which repeats itself everywhere

Someone spends two afternoons building an iron farm and leaves it open. They put up a sign: it's for everyone, look after it.

The first week it works. The second, someone makes off with the entire contents of the chests. Someone breaks a part while trying to improve it. Someone places torches where they shouldn't go and it stops producing. In the third week the person who built it asks in chat who did it, nobody answers, and the farm stays there, broken, like a monument.

The conclusion people almost always draw is about the people: that's just how people are, you can't have anything nice. And it's the wrong conclusion, because the same group of people does manage to keep other things going.

What was thought to be inevitable

For decades the dominant theory said that a shared, depletable resource ends up destroyed unless one of two things happens: someone becomes its owner, or an external authority watches over it and punishes.

Privatise or police. There was no third option.

What Elinor Ostrom found

The political scientist Elinor Ostrom set out to test this in the field, and found that it was empirically false. She documented communities that had been managing mountain pastures, forests, fisheries and irrigation systems collectively for centuries — sometimes more than five hundred years — with no private owner and no police, without exhausting them. In 2009 she received the Nobel Prize in Economics for that work.

Her most useful contribution wasn't saying that it can be done, but describing what the ones that manage it have in common. Among the recurring features:

• Clear boundaries: it's known who belongs and who may use the resource.
• Rules adapted to the place, not copied from somewhere else.
• Those affected take part in setting those rules.
• Monitoring is done by the people involved themselves, not only by an outside authority.
• Graduated sanctions: the first time is a warning, not a ban.
• A cheap way of settling disputes, before they escalate.

None of them is sophisticated. The hard part is having all of them at once.

Servers arrive at the same list, through one mess after another

If you read the rules of a long-running survival server, you'll find Ostrom's list written in other words and without citing her.

Community farms must be published in a list, so it's known what exists and whose each one is: clear boundaries. Only the person who built it or someone in charge may modify it: assigned responsibility. If two people want to use the same one at the same time, turns are set, and if they can't agree a moderator mediates: a conflict-resolution mechanism. The administration reserves the right to reduce the number of animals when server performance requires it: use regulated according to real capacity.

Nobody wrote those rules after reading a Nobel laureate. They were written after a farm ended up empty, which is exactly how the rules of the villages she studied were written too.

The principle: the commons don't fail because of people, they fail because of rules

Step outside the game, because this is what's really worth taking away.

The office kitchen nobody cleans. The shared flat where the toilet roll is always someone else's problem. The work group where there's a task everyone relies on and nobody maintains.

In all of them, the intuitive diagnosis is moral — these people don't pull their weight — and it's almost always wrong. The same people cooperate perfectly well when the split is clear, the contribution is visible and whoever overdoes it gets a warning in time.

Hence a fairly reliable rule of thumb: if a shared resource degrades systematically, look at the design before you look at the people. A failure that repeats with different people isn't a failure of character.

What can be done

Make it known who looks after each thing, even if it's just a name on a sign. Make restocking visible, because what's invisible doesn't get copied. Make the first warning a warning and not a ban. And have a quick way of resolving friction between two people before it turns into two camps.

If there were a place where shared things had rules of use written before the first conflict, and where the person who restocks was as noticeable as the person who takes, I'd like it to be called KobiiCraft.

Now the honest part. None of this is free: all of the above consumes the time of people who are usually doing it unpaid. At a certain size, clear boundaries stop being workable. And the easy way out — banning shared things and letting everyone have their own — genuinely works, at the price of eliminating one of the few reasons two strangers end up in the same place and talk.

A server with nothing in common is more sustainable and colder. Choosing is legitimate; pretending you aren't choosing is not.

The sign that said "look after it"

That sign is the minimal version of a rule: it states a wish and doesn't say who checks anything, what happens if you overdo it, or whom to ask.

It didn't fail because people are bad. It failed because a wish isn't a rule.

Frequently asked questions

Why do public farms and chests run dry on a server? Because the benefit of using them is individual and the cost of maintaining them is collective. Whoever takes materials keeps all of them; whoever restocks invests time of which they only get part back. With that split nobody has to act in bad faith: it's enough for each person to take a bit more than they contribute for the resource to empty. It's the classic pattern of a commons with no rules of use, and it's fixed by changing the rules, not by telling people off.

What did Elinor Ostrom discover about the commons? That the idea that a shared resource always ends up destroyed unless it is privatised or an external authority is put in charge is empirically false. Ostrom studied communities that had been managing pastures, forests, fisheries and irrigation systems collectively and sustainably for centuries, and identified the features common to those that succeeded. She received the Nobel Prize in Economics in 2009 for that work, which showed that self-governance is a real third way.

What rules make a shared resource survive? The features Ostrom found recurring are: clear boundaries on who belongs and may use the resource, rules fitted to local conditions rather than imported, participation by those affected in setting them, monitoring carried out by members of the community and not only by an external authority, graduated sanctions that start light, cheap mechanisms for resolving conflicts and recognition of the group's right to organise itself.

Is it better to ban public farms and let everyone have their own? It works, and it has a cost worth seeing. Privatising everything solves the depletion by eliminating the commons, and with it goes one of the few reasons two people who don't know each other end up in the same place and talk. A server where nobody shares anything is sustainable and also colder. The middle option is to keep what's shared and give it rules of use, which is exactly what the evidence says can work.