Three communities share a water store. Each has a reasonable account of what it needs. Together, their requests exceed what the store can continue supplying.

The difficulty is not solved by declaring that water matters. Everyone already knows that. Nor is it solved by drawing a line around each community. The store remains shared, and one group’s withdrawal changes the conditions faced by the others.

The civilisation question is how people turn shared dependence into workable rules without allowing the resource, the community or the least powerful users to disappear from the decision.

This article uses a fictional resource-sharing case to examine that question. All quantities, communities and proposed arrangements are invented for teaching. They are not a water-management plan or claims about a particular country. The focus is institutional: access, responsibility, evidence, allocation, maintenance and revision. Actual resource limits require appropriate scientific and local expertise.

The conceptual background includes Elinor Ostrom’s work on governance beyond a simple choice between markets and states. Her research provides reasons to examine institutional arrangements rather than assuming that shared resources must always fail, or that one ownership label supplies the answer. [1]

First identify what is actually shared

The phrase shared resource can cover quite different situations. A public explanation can be read by many people without the words being used up. A seat in a full reading room cannot be occupied by several unrelated users at once. Water withdrawn from a store is no longer available there for somebody else to withdraw.

Two questions help separate these cases. Does one person’s use reduce what remains available to others? How difficult is it to exclude a potential user? Common-pool resources combine substantial subtractability of use with difficulty of exclusion; pure public goods are characterised differently, with low subtractability and difficult exclusion. These are analytical characteristics, not ownership labels. [1]

A government-owned resource is therefore not automatically a pure public good in the economic sense. Nor does private ownership automatically remove physical interdependence. In our fictional case, dividing rights to withdraw water would not divide the underlying store into unrelated physical worlds.

This distinction matters because a solution suited to copying information may be unsuitable for allocating scarce space or material. Before choosing a governing arrangement, establish what can be used up, what becomes congested, what can be replenished and which people are affected.

Shared ownership is not the same as open access

A resource can be shared under a set of rules. Membership can be defined, withdrawals can be limited and users can have obligations to maintain the system. Open access, by contrast, describes a situation without effective exclusion from use. The two should not be treated as interchangeable.

Ostrom’s lecture challenges the presumption that common-pool resources are necessarily owned by nobody or that users are inevitably unable to organise. [1] That does not establish that all communities govern well. It establishes that their actual rules and capacities need to be investigated.

In the fictional store, saying that the water belongs to everyone leaves important questions unanswered. Who counts as everyone? Does that include a new household, an occasional visitor or a neighbouring community affected downstream? Does membership confer unlimited withdrawal, a limited share, a voice in rule-making, or several different rights?

The useful task is to make those distinctions explicit. A generous slogan can hide a very narrow practical entitlement. Equally, a resource described as collectively governed may have quite detailed and enforceable limits on individual use.

The case: three requests and one physical budget

Assume that the fictional store begins with 1,000 units. It receives 100 units a day. Ten units a day are lost through processes we will treat as fixed, and twenty units must be released for a separately established downstream requirement. Under these assumptions, seventy units remain available for daily withdrawal without reducing stored quantity.

Community A requests forty units, Community B requests thirty and Community C requests twenty. Their combined request is ninety. If all three requests are met every day while the other assumptions remain unchanged, the store loses twenty units a day. After thirty days, it holds 400 units.

The arithmetic is 1,000 + 30 × 100 − 30 × 10 − 30 × 20 − 30 × 90 = 400. In this simplified construction, continuing at the same rates would exhaust the starting stock after fifty days. A real physical store cannot sustain negative inventory; unmet demand or changed behaviour would then alter the situation.

Nothing in that calculation tells us which community deserves less. It identifies a mismatch between claims and the assumed resource budget. Governance begins where arithmetic leaves the allocation question open.

Scarcity is a constraint, not a complete policy

It would be possible to divide the seventy available units equally, giving approximately 23.3 to each community. It would also be possible to reduce every request by the same proportion, multiplying each by seventy divided by ninety. That produces approximately 31.1, 23.3 and 15.6 units.

Both allocations fit the same aggregate budget. They express different views of fairness. Equal shares might look appropriate if the communities were comparable in relevant needs. Proportional reductions might look appropriate if the original requests were already justified. Neither assumption has yet been established.

Suppose Community A contains twice as many residents, or Community C has fewer alternatives, or the recorded requests include avoidable waste. Those facts would change the discussion. Conversely, allowing a group to claim more simply because it previously used more could reward the very behaviour the arrangement is intended to address.

The lesson is not to choose one universal allocation formula. It is to separate the physical constraint from the moral and institutional reasons for distributing access within it. Calling a decision scientific should not conceal a value judgement that the underlying arithmetic cannot make.

Who can speak for the people using the resource?

A meeting of three community representatives is a useful beginning, but not proof of complete representation. Each community contains people with different needs and influence. Some may do the daily work of collecting or managing the resource without holding a formal position.

Our proposed review would therefore ask how representatives are selected, whose experience they are expected to report and how residents can challenge an account that misrepresents them. It would also ask about people affected by withdrawals who are not members of the three groups.

Future users cannot attend today’s meeting. Neither can the resource itself explain that a nominal limit is becoming unsafe. These absences make evidence and explicit responsibilities important. Someone must bring the relevant conditions into the discussion rather than assuming that the loudest current claims define the whole system.

Participation is not a guarantee of agreement. Its purpose is to improve the information and legitimacy of the decision. A rule can be widely discussed and still be poor; a technically sound limit can also be imposed through a process that ignores important burdens. Both dimensions deserve examination.

Consider three governing arrangements without treating any as magic

One possible arrangement places allocation and enforcement in a public authority. Another gives defined withdrawal rights to particular users and permits some form of exchange. A third gives a user association responsibility for agreed withdrawals and maintenance. Combinations are also possible.

For the fictional store, a public authority might provide a clear external decision-maker, but it would need reliable local information and a way to correct mistakes. Defined rights might make some obligations clearer, but their initial distribution and effects on vulnerable users would still need justification. A user association might possess detailed local knowledge, but it could also be dominated by its strongest members.

The 2009 Nobel Prize announcement summarises Ostrom’s contribution as showing that user associations can successfully manage common property. The finding establishes a possibility supported by research, not a guarantee for every association or every resource. [2]

Accordingly, our comparison asks what each arrangement would actually do. Who measures withdrawals? Who checks the measurements? Who can change the rule? Who pays for repairs? What happens when a user refuses to comply? Ownership language alone does not answer these operating questions.

A proposed agreement must be readable as an operating system

Imagine the communities produce an agreement containing a withdrawal ceiling but nothing else. They have specified an outcome without explaining the mechanism that will maintain it. An implementable agreement needs a way to identify users, observe withdrawals, communicate available capacity and respond to breaches.

It also needs an account of routine upkeep. If a damaged gate causes losses, who is expected to notice, who can authorise the work and how will costs be shared? If repairs temporarily restrict access, which obligations take priority during that interval?

The Ostrom Workshop describes the design principles associated with Governing the Commons as characteristics of robust institutions for common-pool-resource governance. It links robustness to sustained resource use and enduring compliance with rules in practice. [3] This article’s proposed questions are a teaching application, not a reproduced checklist or a claim of formal certification.

The useful standard is whether a new participant could understand what may be done, what must be done and what to do when the normal route does not fit the case. An agreement that exists only in the memory of a few experienced people is difficult to inspect and difficult to hand over.

Monitoring must observe the resource and the behaviour

A withdrawal register can show that each user stayed within its allocation. It cannot, by itself, establish that the total allocation remains appropriate. Inflow may have changed, storage may have deteriorated or the assumed losses may have been wrong.

Conversely, an acceptable storage level does not prove that the rules are being followed. A period of unexpectedly high replenishment might temporarily conceal excessive use. To understand the system, our fictional agreement needs information about both the resource condition and the actions affecting it.

The information should be understandable enough for relevant participants to question it. A number with no unit, date or method is difficult to challenge productively. If different parties use different measuring intervals, an apparent disagreement may be a mismatch of definitions rather than misconduct.

Monitoring also needs limits. Collecting the minimum information necessary to manage withdrawals is different from exposing unrelated personal information. A shared resource does not remove people’s legitimate privacy interests. Transparency should support accountability for the resource, not turn participation into unnecessary surveillance.

The monitor also needs accountability

Suppose the person recording withdrawals belongs to Community A. That does not automatically make the records unreliable. It does raise a reasonable question about how the other communities can verify and challenge them. Replacing the recorder with an outsider changes the relationship but does not eliminate the need for accountability.

In a proposed system, calibration records, transparent definitions, occasional checks and a route for disputing errors could help make the measurements inspectable. The appropriate mechanism depends on the resource and the risks; this article does not prescribe technical measuring procedures.

There is a related burden question. Monitoring takes time and resources. Asking volunteers to perform it indefinitely without recognising the work may undermine the arrangement. A governance design should make the cost visible and explain why it is shared in the proposed way.

The broader civilisational lesson is that information does not arrive for free. Reliable knowledge about a shared resource is itself a maintained capability, with operators, standards and succession needs.

Separate a mistake, an emergency and deliberate overuse

Not every breach has the same cause. A user may misunderstand a unit, face an exceptional need, encounter a faulty measurement or deliberately take more than the agreement permits. A single undifferentiated response would make these distinct situations look alike.

Our proposed agreement would identify who investigates, what evidence matters and which response is proportionate to the finding. It should protect against both arbitrary punishment and the repeated use of exceptions to defeat the common rule.

An emergency allocation also needs an explicit relationship to the total resource budget. Granting extra use to one group without acknowledging its source can quietly transfer the cost to everybody else or to the future stock. Compassionate discretion still needs an honest account of consequences.

The aim is not perfect conflict avoidance. It is a process in which disagreements can be handled without destroying cooperation or requiring each dispute to become a struggle over the entire governing system.

Conflict resolution is part of maintenance

Physical assets need repair. Agreements need ways to handle disputed interpretations. If every disagreement is expensive, slow or personally dangerous to raise, errors may remain hidden until relationships have deteriorated.

In the fictional communities, a practical process might begin with clarification of the measurement or rule, then move to a documented review by someone not responsible for the disputed decision. A further escalation could be available when the issue exceeds the association’s authority. These are illustrative design choices, not legal advice.

A good process should make it possible to correct a case without pretending that the original rule was meaningless. It should also make it possible to revise a rule when repeated cases show that the design no longer fits reality.

This connects resource governance to trust and legitimacy. People do not need to win every dispute to value a system. They need credible reasons to believe that evidence can be heard and power can be questioned.

Local knowledge does not remove wider dependence

Suppose the three communities manage the store carefully but an upstream action reduces inflow. Their local agreement cannot solve that problem alone. The relevant physical system is larger than their immediate membership.

This creates a need for coordination at another scale. Local participants may still manage daily withdrawals, while a wider arrangement addresses upstream effects, scientific assessment or disputes between resource systems. Moving some decisions upward need not mean moving every decision upward.

Ostrom’s broader work examines multiple centres of decision-making rather than assuming that a single hierarchy or a single market arrangement exhausts the possibilities. [1] In our example, the important question is which decisions belong at which scale and how those scales communicate.

The danger in either direction is a mismatch. A local body may lack authority over external pressures. A distant body may lack detailed knowledge of everyday use. A layered arrangement should be judged by whether it connects the necessary knowledge and authority, not simply by the number of organisations it contains.

A resource can be conserved while access becomes unjust

Imagine that the fictional store remains stable because Community C receives almost nothing. A graph of stored quantity might suggest success. The social arrangement could still be unacceptable. Resource condition and fair access are separate questions.

Now imagine the opposite: each community receives its preferred allocation for a month while the reserve declines rapidly. Satisfaction during that month would not establish durable success. Present access and long-term viability also need to be examined together.

A useful review therefore records several outcomes without forcing them into one reassuring score. What is happening to the resource? Who receives usable access? Who contributes work and money? Who carries restrictions? Which disputes recur? What options remain for future users?

This is the multidimensional approach introduced in How to Categorise Civilisation?. A strong result on one axis does not automatically cancel a serious failure on another.

Digital commons require a different physical question

It is tempting to transfer every lesson from a reservoir directly to a shared digital library. But copying a document does not normally remove the source copy. The scarce parts of the digital arrangement may instead be hosting capacity, editing time, attention, funding and the labour required to keep information dependable.

As an analytical example, a community-maintained collection might welcome unlimited reading while needing rules for who may edit, how corrections are reviewed and how costs are covered. Its access problem is therefore not identical to allocating a finite quantity of water.

This does not make the comparison useless. It tells us what to compare carefully. Both arrangements need responsibility and continuity. They differ in what use consumes, what errors look like and where the main constraints lie.

The lesson is to transfer questions rather than copy answers. Ask what is scarce, what can be replenished, who contributes and which actions affect others. Do not assume that the same membership or allocation rule fits every object called a commons.

How the agreement should change when the world changes

Return to the fictional water store. The original allocation depended on assumed inflow, losses, downstream obligations and requests. If one of those conditions changes materially, preserving the same numbers may no longer preserve the same outcome.

A revision process should specify which evidence can trigger review, who participates and what remains protected during the discussion. A temporary restriction might be necessary while uncertainty is resolved. It should be recorded as temporary, with reasons and review conditions, rather than quietly becoming an unexplained permanent arrangement.

Changes in membership also matter. New users need a legitimate route into the arrangement. An agreement that can continue only by ignoring newcomers may be conserving its internal peace by exporting the problem elsewhere.

The connection to preservation and change is clear: preserve the capacity for fair, durable cooperation, not necessarily every historical allocation unchanged.

A workshop for students: write two rules and find their consequences

Use the fictional case rather than a real essential resource. First, check the arithmetic and identify the twenty-unit daily gap between combined requests and the assumed sustainable daily withdrawal budget. Keep the physical assumptions separate from the proposed fairness rule.

Next, write two alternative allocation rules. One might divide the available units equally. Another might divide them in proportion to a stated measure of need. For each, explain which facts would be required before calling it fair. Do not quietly invent those facts to make the preferred rule look better.

Then add an exception: one community experiences a temporary increase in need. Decide who can authorise additional use and where the additional units would come from. Explain how the decision will be reviewed. The exercise becomes more revealing when students must account for the effect on the whole system rather than merely write a compassionate sentence.

Finally, design a record containing the date, measurements, allocation, explanation and unresolved questions. That record should allow another group to inspect the decision. The quality of the answer lies in its clarity and honesty, not in claiming that conflict has been eliminated.

The resource and the relationship have to survive together

The three communities began with a shared store and competing requests. A durable solution required more than a number. It required definitions, credible information, a fair account of access, responsibilities for upkeep, a way to handle disagreement and a process for responding when conditions changed.

That is why shared-resource governance belongs inside the study of civilisation. It reveals how physical dependence meets institutions, knowledge, power, culture and time. It also reveals why simple labels can be poor guides. Public, private and collective arrangements each need an operating reality beneath their name.

The strongest conclusion is not that one form always wins. It is that the arrangement must be judged against the specific resource, the actual users, the surrounding system and the evidence of what happens.

A shared resource is well governed only when people can keep using it without hiding the costs, excluding the affected voices or consuming the conditions that make future use possible.

Sources and boundaries of the argument

[1] Elinor Ostrom, Beyond Markets and States: Polycentric Governance of Complex Economic Systems, Nobel lecture, 2009; archived revised draft at Indiana University’s Digital Library of the Commons. Used through paraphrase for resource characteristics, the distinction between common-pool resources and assumptions of ownerless access, and the need to examine varied institutional arrangements. The archived document is a draft, not represented here as the final published text.

[2] Royal Swedish Academy of Sciences, 2009 economic sciences prize announcement. Used for its account of research showing that common property can be successfully managed by user associations.

[3] Indiana University Ostrom Workshop, Ostrom Design Principles. Used for the relationship between institutional robustness, long-term resource sustainability and rules in practice.

The fictional case, calculations, allocation comparisons and workshop are original teaching material. No single governance design is claimed to be universally effective. Resource limits, legal authority and technical operating procedures must be established for the actual setting.

Continue through the Civilisation Atlas

Return to the Civilisation Atlas, examine cooperation between civilisations, or explore how progress and decline should be measured.