A village can know who opened the gate. The planet cannot know its users in the same way. The small commons benefits from encounters in which conduct can be seen, explanations heard and a reputation carried into tomorrow’s dealings. A global resource joins people who will never meet and authorities that answer to different publics. It is tempting to infer a ceiling: beyond a certain size, self-government must fail. But the work done by familiarity can sometimes be carried by an institution. A record can make conduct visible to strangers; representation can carry an interest into a larger assembly; a shared rule can coordinate decisions made in different places. The question is which of these substitutions can work, and where the connections between them remain too weak.
The evidence for local governance does not establish that all successful users must know one another personally. Ostrom’s account includes institutions at multiple levels and insists on the significance of context rather than one universally sufficient form. Shared expectations, trustworthy information and a capacity to respond to breaches matter because of what they make possible. A face-to-face community is one setting in which they can be developed. It can also be a setting of exclusion, unequal power and habits no one dares to challenge. The problem of scale should therefore be stated in terms of functions: how participants learn what others are doing, how affected interests enter decisions and how commitments acquire consequences.
The relevant boundary is partly a property of the resource. A grazing rule may be effective within a pasture while a water rule made by the same village changes the supply downstream. A protected forest can displace clearing to a neighboring district. A local institution can perform its own tasks well and still omit people affected by the combined result. Enlarging its jurisdiction is one possible response; arranging coordination with other jurisdictions is another. Either requires an account of the connection. Otherwise the success reported inside one boundary can be purchased by a loss left outside it.
Consider an invented river with two upstream user groups and a stipulated limit of fifteen units of combined withdrawal. Each group adopts a local ceiling of ten and enforces it perfectly. If both use their full allocations, the river loses twenty, exceeding the combined limit by five without a single local violation. A basin arrangement could divide fifteen between the groups, while leaving each to decide how its own share is distributed and monitored. An allocation of seven to one and eight to the other would satisfy the stated total if both comply. It would not, merely by adding correctly, establish fairness or secure consent. The example separates the job of coordinating a shared constraint from the different job of governing use within it.
Nested governance assigns connected tasks to institutions capable of performing them. Local users may know where an allocation is being evaded; a basin body may see cumulative withdrawals; a wider authority may provide finance, recognize rights or hear a dispute that cannot be resolved locally. These roles need not form one simple ladder. Water, energy, land and public health can require overlapping centers of decision, each dealing with a different connection. Such a system becomes governable when the centers have intelligible responsibilities and means of coordination. Merely multiplying committees adds neither knowledge nor accountability if no one can determine who must answer when their decisions conflict.
Common information can support this division of work without erasing local knowledge. A shared method of measuring withdrawals lets a basin compare accounts; a local observer may know why an apparently abnormal reading occurred. A common reporting standard makes a claim available for scrutiny, while an appeal allows a user to challenge the way the standard was applied. Independent examination can help when the reporting authority has a stake in the result. The system needs both comparability and correction. A perfectly uniform account of the wrong thing is little improvement on incompatible accounts, and a rule whose errors cannot be contested can turn measurement into another instrument of power.
Capacity also has to travel across the system. The jurisdiction asked to preserve a shared resource may have fewer funds than the jurisdictions enjoying much of the benefit. An agreement that acknowledges the constraint but leaves the acting party unable to meet it can be formally complete and practically empty. Finance, technical assistance and arrangements for sharing costs can make different commitments feasible. They also require governance: a contributor needs an account of the use, and a recipient needs terms that respect its knowledge and permit the work to be carried out. The larger institution has to connect resources to responsibilities rather than merely circulate demands.
The Montreal Protocol’s financial mechanism shows one concrete way to organize that connection. Article 10 provides assistance to qualifying parties operating under article 5.1 for agreed incremental costs of compliance. It establishes a Multilateral Fund, places policy authority with the parties and assigns operational functions to an Executive Committee with balanced representation between the specified groups of parties. Disbursement requires the beneficiary party’s concurrence. These provisions divide authority, funding and implementation among institutions rather than entrusting everything to one planetary official. They demonstrate an institutional arrangement that can be examined; their existence alone is not a measure of every project’s success or a guarantee that another environmental problem can be solved on identical terms.
Multiplicity creates hazards of its own. A burden may migrate toward the weakest rule; incompatible standards may waste effort; several authorities may each assume that another is watching. A powerful participant may choose the forum most favorable to it while affected people struggle to find any forum that will hear them. These possibilities give wider institutions a continuing role in setting common conditions and addressing effects that local bargains leave out. They also make room for evaluation across jurisdictions: an approach can be compared with alternatives, a failing arrangement revised and a successful practice borrowed where the conditions fit. Local experimentation is most useful when its consequences can be observed beyond the locality conducting it.
Ostrom’s polycentric argument gives this possibility its proper scale of ambition. Multiple centers can undertake action, build knowledge and learn from one another while broader agreements develop. This is a way of making progress and improving cooperation under difficult conditions. It does not establish that a collection of local initiatives will automatically achieve a global limit. Leakage, inconsistent policies, free riding and inadequate certification remain problems within the approach itself. To recognize them is to identify the connections that must be strengthened: a common account of consequences, responsibilities that reach beyond an institution’s easiest constituency and wider rules where the affected system demands them.
The reach of governance depends on the connections it can sustain.
The village need not be enlarged until everyone knows everyone. Its relevant achievements have to be understood and, where possible, carried into institutions fit for a larger field of consequences. Some knowledge should remain close to use; some accounts must span the whole resource; some authority must connect the two. The commons can outgrow a single community without becoming ownerless in every practical sense, but the arrangements that govern it must be built, contested and repaired. The volume ends with that work unfinished as a general problem. The next turns to another institution for coordinating many separate acts: the firm, within whose boundary the allocation of decisions takes a different form. The argument remains open.