Plate xii in print

The Planet

the commons with no sovereign: the climate, the oceans, the web of life, shared by every nation and ruled by none, and the question whether a globe can be governed as a village is.

The absence of one sovereign does not relieve the many of governing what they share.

— the argument of this plate, after Dietz–Ostrom–Stern (2003) and Ostrom (2009)

The pasture has an edge. The atmosphere has no corresponding fence at which a village can stop taking account of its neighbors. A community may keep faith with its own rule and still receive harm from conduct far beyond its reach. Enlarging the map brings the conduct into view, but does not produce a single authority able to decide for everyone on it. The global commons is governed among governments, each with its own powers, dependencies, and obligations to people who do not experience the common danger alike. The absence of a world sovereign leaves a difficult task of coordination. It does not leave an empty world. Agreements, public agencies, scientific institutions, local rules, and practical cooperation can already occupy the space in which a single ruler is missing.

The village offers lessons rather than a scale drawing. In their account of adaptive governance, Dietz, Ostrom, and Stern bring information, conflict, compliance, infrastructure, and adjustment into the same institutional problem. They emphasize dialogue, nested arrangements, and institutional variety, rather than expecting one form of organization to work everywhere. A rule must be intelligible to those who follow it and responsive to the resource it governs; a structure must also reach the consequences that escape its immediate jurisdiction. These are demanding conditions, not a list whose every item must appear in identical form before cooperation is possible. On a larger map, the questions recur at more than one level. Who can observe the change? Who bears the cost? Who has reason to object, and where can that objection alter the decision?1

1. Thomas Dietz, Elinor Ostrom, and Paul C. Stern, “The Struggle to Govern the Commons”, Science 302 (2003), argument and conclusion, pp. 1907–1910. Their proposed principles concern adaptive institutions, not a universal recipe.

Even an agreed total benefit can conceal a disagreement over who should act. Imagine two jurisdictions considering a conservation project that only the first can undertake. It costs that jurisdiction four, yields it a benefit of one, and benefits the second by six. Taken together, the gains exceed the cost by three. Taken separately, the first jurisdiction loses three by doing the work, while the second receives six without having to do it. A transfer of four from the beneficiary to the actor would leave the first ahead by one and the second ahead by two. The common interest can, in this example, support a mutually beneficial arrangement. It cannot write or enforce the arrangement. The parties must still agree about the cost, recognize the result, and make the promised payment arrive.2

2. The example assumes known benefits measured in the same transferable unit, no other effects, and an enforceable transfer if agreed. Without payment the payoffs from undertaking the project are (1 − 4, 6) = (−3, 6); with a payment of four they are (1 − 4 + 4, 6 − 4) = (1, 2). It supplies no general welfare weights for comparing actual countries.
⁂

The Montreal Protocol illustrates how an agreement can give some of these tasks an institutional home. Its consolidated text requires reporting on controlled substances, including production and trade. It also establishes financial and technical cooperation for Parties operating under Article 5(1), with a mechanism covering their agreed incremental costs of compliance and a Multilateral Fund under the Parties' authority. The architecture joins an account of what is being done to an account of how specified participants are to be assisted in doing it. The reporting rule is not itself proof that every report is accurate, nor does a funding provision establish that every practical obstacle has vanished. It makes the obligations and the machinery identifiable. What might otherwise remain an appeal to a common interest becomes work assigned to institutions whose performance can be examined.3

3. Montreal Protocol, official consolidated Article 7, especially ¶3, and Article 10, especially ¶¶1–5. These provisions describe institutional duties; no claim about current environmental outcomes follows from their text alone.

Ostrom's proposal for a polycentric approach to climate change begins from the possibility of action at multiple scales. Local and regional efforts can produce benefits, reveal workable practices, and generate information while wider arrangements address effects that smaller units cannot contain. The units need connections, comparison, and monitoring; their multiplication is not by itself a solution. Some tasks demand larger jurisdictions, and independent efforts can fail to take account of one another. The point is to allow learning and action at several levels without treating either a single comprehensive bargain or a collection of isolated local measures as sufficient. A global difficulty can require coordination without requiring that every useful act wait for all coordination to be complete.4

4. Elinor Ostrom, “A Polycentric Approach for Coping with Climate Change”, World Bank Policy Research Working Paper 5095 (2009), “A Polycentric Approach to the Problem” through the concluding section, pp. 32–39.

The account of success must travel as far as the effect being counted. Suppose a rule reduces a harmful activity by ten units inside one jurisdiction, but eight units of the same activity move across the boundary. If each unit causes the same harm and nothing else changes, the reduction for the two jurisdictions together is two. The first authority can report its ten accurately while the report remains incomplete as an account of the common resource. The same difficulty arises when two institutions count the same improvement as if each had produced an additional one. Keeping a shared account requires boundaries, a baseline, and a rule for attributing changes without losing what happens elsewhere. Those choices are part of governing the commons, because a promise whose fulfillment cannot be distinguished from displacement gives the other parties little reason to trust it.5

5. The displacement illustration holds harm per unit constant and counts both locations: −10 + 8 = −2. Actual assessment would also require evidence about the counterfactual, timing, differing harms, and other effects. The arithmetic identifies an accounting question; it does not estimate leakage in any real program.

There is therefore no need to choose, before looking at the work, between the village enlarged beyond recognition and a world left to itself. Some decisions belong near the people who know the resource; others require a reach those people cannot supply. Some promises need finance, others a reliable observation, others a forum capable of settling a disagreement that would otherwise undo both. Their combination will be unfinished and contestable. So was the village's. What matters is whether the arrangements make conduct answerable to its consequences and give cooperation a way to persist when goodwill is strained. The next volume, The Firm, enters another institution built to coordinate work, one that draws a boundary around authority and asks which decisions should be made within it.