SINE Strategy Wiki

Ostrom's Commons Design Principles & SINE's governance approach

19 min read

Elinor Ostrom (see person page) won the Nobel Prize in Economics for showing that shared resources (“commons”) can be governed sustainably by their users without requiring either full privatization or full state control — provided the governance design meets certain principles (clear boundaries, collective-choice rules, monitoring, graduated sanctions, low-cost conflict resolution, etc.).

The 2022 Data Commons whitepaper states that SINE/Deutsche Bank/Systemiq/TLGG took Ostrom’s principles and amended them for data-sharing specifically, distilling them into three values — see Data Commons for the values themselves.

This page is now grounded directly in Ostrom’s own book, Governing the Commons (1990), rather than in how the 2022 whitepaper paraphrases her — per Martin Pompéry’s request (2026-08-05) to go deeper on the framework details, the underlying scientific methodology, and the “two options” critique specifically. All line numbers below cite sources/whitepapers/2026-08-05-ostrom1990-governing-the-commons/2026-08-05-ostrom1990-governing-the-commons.md.

Why this page exists as its own concept, not just a subsection

⟨Per Martin Pompéry, ingest discussion 2026-08-05⟩: Ostrom’s school of thought should not be filed as mere theoretical “Unterbau” (substructure) underneath the Data Commons concept. It is:

  1. The origin SINE’s standardization work derives from — the Data Commons values (Reciprocity, Security, Openness) are described as a first derivation of Ostrom’s principles applied to data-sharing, not an independent framework that happens to cite her.
  2. The frame SINE uses to govern what it produces — i.e., Ostrom’s commons-governance logic isn’t only the justification for why Data Commons should exist, it is the operating logic SINE applies when designing governance for the standards and networks it builds (e.g. Pathfinder Network).

This makes Ostrom’s principles closer to SINE’s constitutional/operating layer than to background theory — worth tracking as its own concept so future sources (e.g. governance docs, network rulebooks) can be linked back to it directly rather than buried inside the Data Commons page.

⟨Inference: the 2022 whitepaper itself does not spell out point 2 — that SINE uses Ostrom’s logic to govern its own outputs — as explicitly as Martin did in conversation. This page represents Martin’s framing as the current authoritative read of the relationship; the PDF only states point 1 explicitly.⟩

The “two options” critique (Ostrom’s central negative thesis)

This is the argument Ostrom spends the whole book dismantling, and the piece most relevant to why “Ostrom-inspired” governance is a genuine third path rather than a compromise between market and state.

The claim she is attacking: that CPR problems can only be solved by (a) an external central authority (“Leviathan”) imposing and enforcing rules, or (b) dividing the resource into private property. She calls this “the ‘only’ way?” question explicitly (line 294) and states her thesis directly:

“If this study does nothing more than shatter the convictions of many policy analysts that the only way to solve CPR problems is for external authorities to impose full private property rights or centralized regulation, it will have accomplished one major purpose.” (line 1,736–1,737, Ch. 6)

Why scholars fall into the dichotomy — her diagnosis, not just her disagreement: static, fixed-structure models (the tragedy of the commons, the prisoner’s dilemma, Olson’s logic of collective action) are built so that the rules of the game cannot change from inside the model — which means, by construction, “someone external to the situation must change the structure” (line 1,749, Ch. 6). The modeling choice manufactures the policy conclusion; it isn’t independently derived from evidence about what resource users can or can’t do. Ch. 1 makes the same point about the two camps sharing one unexamined premise (line 294):

“Both centralization advocates and privatization advocates accept as a central tenet that institutional change must come from outside and be imposed on the individuals affected. Despite sharing a faith in the necessity and efficacy of ‘the state’ to change institutions so as to increase efficiency, the institutional changes they recommend could hardly be further apart.”

How she demonstrates it, formally: Chapter 1 builds a sequence of increasingly complex games on a shared “Hardin herder” setup (a pasture with carrying capacity L, two herders) — see lines 214–332:

  • Game 1 — no institution, complete information: a prisoner’s dilemma; both herders defect, zero profit (the textbook “tragedy”).
  • Game 2 — a central authority with perfect information and costless, unfailing enforcement: recovers the cooperative outcome. This is the idealized case privatization/centralization advocates implicitly argue from.
  • Game 3/4 — the same central authority but with realistic incomplete information about herders’ actual behavior: sanctioning errors appear, and the equilibrium becomes worse than the unregulated Game 1 (line 284: “the equilibrium of the regulated game has a lower value than that of the unregulated game”). The central-authority solution’s optimality was an artifact of assuming away monitoring costs and error.
  • Game 5 — a “self-financed contract-enforcement game”: the herders negotiate their own binding contract and jointly hire (and pay for) an external enforcer of their own choosing — no government imposition, no privatization, and no assumption of perfect information. This is Ostrom’s formal alternative: “The ‘solution’ of a commons-dilemma game through instrumentalities similar to Game 5 is not presented as the ‘only way’ to solve a commons dilemma. It is merely one way. But this way has been almost totally ignored in both the policy-analysis literature and the formal-theory literature.” (line 318)

Illustrative real-world critiques she gives:

  • Quotes Sugden (1986) on how economic theory treats “the government” as a fictitious, omnicompetent single actor — “like the US Cavalry in a good Western, ready to rush to the rescue whenever the market … fails” — while crediting private individuals with “little or no ability to solve collective problems among themselves” (lines 1,990–1,992, Ch. 6).
  • Contrasts her own field study of California groundwater basins with Rolph’s (1982/1983) analysis of the same basins: Rolph frames government as the only relevant actor and misses that the users themselves negotiated the actual working solution (lines 1,994–2,000, Ch. 6) — a concrete case of the dichotomy causing an analyst to miss what was actually happening on the ground.
  • Nova Scotia’s Port Lameron Harbour inshore fishery (Ch. 5): a locally sound informal governance system was undermined because Canada’s federal Department of Fisheries and Oceans recognized only “private property vs. government regulation” as legitimate options, refusing to recognize the fishers’ own rules — the dichotomy operating as policy, not just theory.

Her alternative, in one line: institutions are “rarely either private or public — ‘the market’ or ‘the state.’ Many successful CPR institutions are rich mixtures of ‘private-like’ and ‘public-like’ institutions defying classification in a sterile dichotomy” (line 302) — self-organized governance, nested/polycentric public-private arrangements, and institutional diversity matched to local conditions, in place of one universal prescription.

The framework: how Ostrom analyzes a CPR situation (Ch. 2)

The chapter most directly about the analytical apparatus, as distinct from the case narratives.

The CPR situation, defined (lines 420–446): a common-pool resource (CPR) is “a natural or man-made resource system… sufficiently large as to make it costly (but not impossible) to exclude potential beneficiaries from obtaining benefits from its use” (line 424). She distinguishes the resource system (the stock — a fishing ground, a groundwater basin) from resource units (the flow appropriated from it — fish, acre-feet of water). Resource units are rivalrous/subtractable (fish one boat catches aren’t there for another); the resource system itself is subject to joint use, which is why CPRs sit structurally between private goods and pure public goods. Appropriators withdraw resource units; providers arrange for the resource’s provision; producers maintain the resource system itself — often, but not necessarily, the same people (lines 424–432).

Bounded rationality, not full-information rational choice (lines 448–484): Ostrom explicitly does not assume appropriators have complete information or unlimited computational capacity. They are “broadly rational individuals who find themselves in complex and uncertain situations” (line 450), engaging in real trial-and-error learning (“By definition, trial-and-error methods involve error, perhaps even disasters,” line 456). Individual choice is modeled via four variables — expected benefits, expected costs, internal norms, and discount rates (Figure 2.1) — and she explicitly frames this as an application of Popper’s advice to keep the rationality principle “an almost empty principle,” placing the analytical weight on specifying the situation rigorously enough that observable variables, not unobservable mental states, can be used to test the theory (line 482).

The three-level rule typology (lines 606–630), her central structural device:

  • Operational rules — govern day-to-day decisions: when/where/how to withdraw resource units, who monitors whom, what information must be shared, what sanctions apply.
  • Collective-choice rules — govern how operational rules themselves get made, by appropriators, officials, or external authorities.
  • Constitutional-choice rules — govern who is eligible to participate in making collective-choice rules, and by what procedure.

Each level is nested inside the one above it: “changes in the rules used to order action at one level occur within a currently ‘fixed’ set of rules at a deeper level,” and deeper-level rules are harder and costlier to change, which is precisely what makes mutual expectations stable (lines 616–620). She also draws a working rules vs. rules-in-form distinction (de facto vs. de jure, line 612–614): a working rule is one people actually expect to be monitored and enforced, regardless of what a formal statute says — “my primary focus in this study will be on the de facto rules actually used in CPR field settings” (line 614).

The three central puzzles any theory of self-organized collective action must solve (lines 522–564) — Ostrom’s framing of what makes self-governance a genuine puzzle rather than an assumption:

  1. Supply — “Institutions are demanded because they enhance the welfare of rational actors. The problem is: Why are they supplied?” (Bates 1988, quoted line 528). Supplying a new institution is itself a collective good, subject to the same free-riding logic it’s meant to solve — “the institution … is subject to the very incentive problems it is supposed to resolve” (line 532).
  2. Credible commitment — appropriators may agree ex ante to restrict their use, but in any later period the short-term payoff to breaking the rule usually exceeds the payoff to keeping it, unless detection and sanction make compliance pay (formalized at line 548). “No one wants to be a ‘sucker,’ keeping a promise that everyone else is breaking” (lines 550–552).
  3. Mutual monitoring — enforcing rules against free-riders is itself costly to the enforcer and a public good to everyone else, so rational self-interest predicts no one bothers (Elster 1989, quoted line 562).

These nest into each other, which is why she treats them as a single chain rather than three separate problems: “Without monitoring, there can be no credible commitment; without credible commitment, there is no reason to propose new rules” (line 564).

The 8 design principles (Table 3.1, Ch. 3)

The empirical answer to the puzzles above — regularities she found across long-enduring, self-governed CPR institutions (Törbel, the Japanese villages studied by McKean, the Spanish huertas, the Philippine zanjeras — see the source page for case detail):

  1. Clearly defined boundaries — both of the resource itself and of who is entitled to appropriate from it.
  2. Congruence between appropriation/provision rules and local conditions.
  3. Collective-choice arrangements — most individuals affected by operational rules can participate in modifying them.
  4. Monitoring — monitors are accountable to, or are themselves, appropriators.
  5. Graduated sanctions — for those who violate rules, scaled to the seriousness and context of the offense.
  6. Conflict-resolution mechanisms — low-cost, locally accessible.
  7. Minimal recognition of rights to organize — external governmental authorities do not challenge appropriators’ right to devise their own institutions.
  8. Nested enterprises — for CPRs that are part of larger systems, governance is organized in multiple layers of nested enterprises.

Related concept invoked in the discussion of these principles: “quasi-voluntary compliance” (citing Margaret Levi 1988a) and “contingent self-commitment” — appropriators comply because they trust others will too and that rule-breaking will be caught, not out of pure altruism.

Her scientific methodology

This is the part of the book that is explicitly about how she builds and validates the argument, not just what it concludes.

Framework vs. model — her central epistemological distinction (Ch. 6, lines ~1,740–1,760): a model is a narrow, fully specified representation valid only where its assumptions hold — fixing a variable (e.g., “no communication among players,” “fixed rules”) doesn’t make a model more general, it narrows its range of applicability: “Apparent simplicity and generality are not… equivalent. Setting a variable equal to a constant usually narrows, rather than broadens, the range of applicability of a model” (line 1,747). A framework, by contrast, specifies which variables are relevant — for CPR situations: benefits, costs, norms, and discount rates as “summary variables,” further broken into detailed situational variables (benefit-side, transformation-cost, monitoring/enforcement-cost, and norms/discount-rate variables — Figures 6.2–6.5) — without asserting a single universal predictive equilibrium. Mistaking a convenient model (e.g., the standard PD game) for the general theory is, in her account, exactly how analysts talk themselves into the “two options” conclusion (see above): the model is fixed-structure by design, so the only way out of a bad equilibrium is an external change to the structure — a conclusion baked into the model choice, not derived from evidence (line 1,749).

Comparative case-study method, explicitly justified by analogy to biology (Ch. 1, lines 396–410): rather than a random sample, Ostrom picks small-scale CPR cases (50–15,000 users, one country, high economic stakes for participants) the way biologists pick a simplified model organism — “not chosen because it is representative of all organisms… chosen because particular processes can be studied more effectively using this organism than using another” (line 400). Because users are strongly dependent on the CPR for income, their self-organization efforts are not half-hearted, so failures are informative: “when self-organization fails, I know that it is not because the collective benefits … were unimportant to the participants” (line 402).

Design-principle validity test (Table 5.2, Ch. 5): having derived the 8 design principles inductively from success cases (Ch. 3–4), she then scores the failure and fragility cases in Ch. 5 (Turkish fisheries, San Bernardino/Mojave, Mawelle, Kirindi Oya, Port Lameron Harbour) against the same 8 principles. No failure case scores more than 3 of 8 — used as evidence that the principles have genuine explanatory/discriminating power rather than being a post-hoc description fitted only to the success cases.

Explicit rejection of the omniscient-analyst / incompetent-user framing (Ch. 2, line 398 area, restated Ch. 6): “As an institutionalist studying empirical phenomena, I presume that individuals try to solve problems as effectively as they can… Instead of presuming that some individuals are incompetent, evil, or irrational, and others are omniscient, I presume that individuals have very similar limited capabilities.” This symmetry assumption — the same bounded rationality for resource users and for regulators/analysts — is what makes the “two options” critique a methodological point, not just a policy preference: models that quietly grant unlimited competence to “the government” while assuming CPR users are trapped are not neutral, they are asymmetrically specified.

Predicting when self-organization happens, not just describing that it does (Ch. 6): a “zero condition” baseline plus six ranked variables favoring self-organization (shared judgment of harm, low heterogeneity of impact, low discount rates, low information/transformation/enforcement costs, generalized reciprocity/trust norms, small stable group), extended across three types of external regime — facilitative, honest-but- centralizing, and corrupt-centralized — each illustrated with a case.

  • “Second-order dilemma” (Bates 1988) and “mutual prescription” — from the California groundwater case (Ch. 4): how a group negotiates cutback sharing without litigating seniority.
  • “Facilitative political regime” — an external authority that enables rather than controls self-organization (home-rule tradition, subsidized litigation, technical assistance) — contrasted with the corrupt/ centralizing regimes in the Sri Lanka and Turkish fishery failure cases.

A contemporary counter-case: enclosure instead of governance

Added 2026-09-10. Ostrom’s central negative thesis is that the two standard prescriptions — Leviathan and privatisation — are both presented as the only options, and both fail on their own terms for common-pool resources. Upright Oy is a live instance of the second prescription applied to a knowledge commons: its inputs are 200M+ scientific articles and half a dozen public statistical databases; its outputs, taxonomy (~800 impact topics), method and benchmark are proprietary, and its public trajectory runs from “opening our model to public exploring and crowdsourcing” (2018) to a patent-pending technique and a gated whitepaper (2026).

⟨Background, restored by the 2026-09-14 lint: every company fact in the paragraph above — corpus size, database count, taxonomy size, the 2018 quote, the patent and whitepaper status — comes from live web research of 2026-09-10 and Upright’s own account of itself. None of it is filed in sources/, and none of it is independently verified; see the evidentiary health warning. The enclosure reading below does not depend on the exact numbers, but they should not be quoted outward from this page.⟩

⟨Inference: worth keeping on file as the control group against Data Commons and PACT — same coordination problem (no shared measure, so capital cannot be directed), opposite institutional answer. A terminology sweep of 41 pages of their material on 2026-09-10 found zero occurrences of commons, Ostrom, collective action, free rider or externality: the institutional question is not merely answered differently, it is not posed.⟩

Their material also supplies an unusually crisp illustration of the credible-commitment puzzle from the technical side — see Coherence as a truth substitute, a commitment mechanism with no monitor.

Ostrom’s provision problem, in a 2026 registry

⟨2026-10-27, from OECD (2026).⟩

Ostrom’s appropriation/provision split — who takes from the resource, and who builds and maintains it — has an unusually clean modern instance in the data-space field, and it runs the opposite way to the fisheries and irrigation cases.

For a CPR, appropriation is the hard problem: the resource is rivalrous, so overuse destroys it, and the design principles are largely about bounding withdrawal. Data are non-rivalrous, so appropriation is nearly free of conflict. What remains is provision, and it is the whole problem. Brousseau, Toledano and Eustache (2024) characterise data-space infrastructure as a club good — high fixed cost, near-zero marginal cost per deployment, non-rivalrous within the ecosystem, excludable to non-participants — which is precisely a provision problem with the appropriation problem removed.

The numbers are what make this worth recording: of 235 registered initiatives, 25 are operational and none is scaling. The field has spent upwards of EUR 1.7 billion of public money and produced, on its own registry’s evidence, very few self-sustaining institutions.

The design principles, scored against the cases

⟨Inference throughout — the OECD does not cite Ostrom, and this mapping is ours.⟩

  • P1 (clearly defined boundaries) and P3 (collective-choice arrangements) are what the report means by “federated governance”, and the mature cases all have them: Catena-X’s association and “10 Golden Rules”, ABtC’s membership, Fintraffic’s mandate.
  • P2 (congruence between appropriation/provision rules and local conditions) is where the failures sit, and the report names the mechanism without naming the principle: “asymmetric cost-value distribution” — tier-2 suppliers paying for data whose value lands on OEMs. P2 says the provision rules must be proportional to the benefits received. The Finnish Maritime Data Space is the one case that designs this in explicitly, via revenue sharing across shipping company, charterer, port and system operator. It is also the one case with measured outcomes (24 % average CO₂ reduction on the Luleå–Oxelösund route).
  • P4 (monitoring) and P5 (graduated sanctions) are the report’s weakest layer, and it says so: “once data leave the provider’s infrastructure, enforcement depends on the data user’s compliance”, with provenance and audit logs “shifting the problem from technical prevention to deterrence through accountability”. That is P4 without P5 — monitors accountable to appropriators, no sanction ladder behind them.
  • P8 (nested enterprises) is visible and under strain: IDSA, Gaia-X, the DSSC and sector consortia each build partially overlapping building blocks, which Annex A records practitioners calling “reinventing the wheel”. Nesting without a settled division of labour.

The external-authority tension, restated

Ostrom’s cases succeed where appropriators craft their own rules and external authorities do not override them. The data-space field inverts this: external authorities fund the institutions into existence, and the report’s finding is that this is associated with reaching implementation and not with reaching operation — EU funding is 46.1 % of implementation-stage initiatives and ~20 % of operational ones. The workshop testimony is more direct: continued public support “may remove the pressure to find viable revenue models altogether”. And the standing caution:

“That many data spaces depend on public funding does not in itself demonstrate that they are public infrastructure. It may instead indicate that viable business models have not yet been found.”

⟨Inference: this is the closest thing to a controlled test of Ostrom’s external-authority claim that this wiki has. Her argument was that externally imposed rules fail because they lack local knowledge and legitimacy. The data-space evidence suggests a different failure channel with the same outcome — external funding, with no imposition at all, substitutes for the self-organisation that makes an institution durable. The Dutch figure is the cleanest statement of the exposure: nearly 80 % of 47 catalogued initiatives depend on government funding. Fintraffic and X-Road are the honest exceptions, and they are honest because they do not pretend to be self-sustaining — they are permanent public utilities on recurrent budget, which is a legitimate answer to the provision problem and simply not Ostrom’s answer.⟩

Cross-references

Sources

  • ostrom1990-governing-the-commons — sources/whitepapers/2026-08-05-ostrom1990-governing-the-commons.pdf (primary; all line numbers above cite the Markdown twin)
  • sine-data-commons-2022 — sources/whitepapers/2022-04-25-data-commons-final.pdf (secondary — how SINE paraphrases/adapts these principles)
  • sources/research/2026-10-27-oecd-data-spaces-co-ordination-viability-public-policy/2026-10-27-oecd-data-spaces-co-ordination-viability-public-policy.md (§1.1, §3.1, §3.3, §3.4, §4.2, Boxes 3.3–3.5, Annex A — the provision-problem section)

Graph View

Backlinks