Data Spaces between Co-ordination Challenges, Financial Viability and Public Policy (OECD, 2026)

OECD Digital Economy report, OECD Publishing, Paris, https://doi.org/10.1787/0066f11b-en. PDF ISBN 978-92-64-95733-6. Declassified by the Digital Policy Committee on 10 June 2026. CC BY 4.0. Drafted by Christian Reimsbach-Kounatze with Shunya Muromachi and Maiko Meguro, under the supervision of Clarisse Girot (Head of the Data Governance and Privacy Unit); research assistance by Shihori Maeda. Funded by voluntary contributions of Japan.

Sources: sources/research/2026-10-27-oecd-data-spaces-co-ordination-viability-public-policy.pdf (immutable original) and its Markdown twin 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 (Marker, 2026-10-27, --disable_ocr). Cited against the Markdown. Conversion caveat: the report’s two analytical tables came out damaged. Table 3.1 (“Comparison of data-sharing mechanisms and enablers”) lost all but the first column — the dimension names and the note survive, the cell contents do not; Table 3.2 (2×2 governance configurations) lost its body entirely, leaving only title, source and the surrounding prose. Everything attributed to those tables on wiki pages is reconstructed from the body text, which states the same content in prose, and is marked where it is. The five figures are chart images, not transcribed; their numbers are given in the notes and in the body.

Handed over by Martin Pompéry, 2026-10-27.

What kind of document this is

An OECD committee report — declassified, so it carries the Digital Policy Committee’s imprimatur rather than only the authors’. It is an evidence-and-policy document: it asks whether EUR 1.7 billion of EU public investment in data spaces is buying something that alternative mechanisms could not buy, answers “sometimes, under four conditions”, and then asks why almost none of the funded initiatives work.

Its empirical base is thin and it says so at length (Box 1.1): one voluntary self-reported registry (the IDSA Radar), four case studies, and one Chatham-House-Rule expert workshop. ⟨Inference: this is the most candid OECD document in this wiki about the limits of its own evidence — the measurement gap is promoted to a finding rather than hidden in a methods note, which is itself the report’s sharpest policy move.⟩

The evidence base:

• IDSA Radar (www.dataspaces-radar.org), data as of February 2026.
• Four cases: Catena-X (Box 3.1), the Finnish Maritime Data Space (Box 3.3), Japan’s ABtC (Box 4.1), Agdatahub (Box 3.4) — plus BoostAeroSpace, EONA-X and T-Systems (Box 3.5) in less depth.
• OECD Expert Workshop, “Assessing the Sustainability of Business Models in Data Spaces and the Role of Public Policies”, 11 March 2025, at the Data Spaces Symposium in Warsaw, under the Chatham House Rule (summary in Annex A). Everything cited “(Annex A)” is unattributable practitioner testimony.

The headline numbers

Figure	Value	Where
EU public investment in data spaces	> EUR 1.7 bn (EUR 1 735.4 m) as of early 2025	§1.1, Ch.5; endnote 5
IDSA Radar entries	235 (Feb 2026), up from 87 in 2022	§1.1, Fig. 1.1
…at implementation stage	128 (54 %)	endnote 3
…operational	25 (11 %) — “one in nine”	§1.1
…scaling	0	§1.1, Fig. 1.1
EU funding share, implementation stage	46.1 %	endnote 3, Fig. 4.1
EU funding share, operational stage	~20 %	endnote 3, §3.3.1
Initiatives charging a fee for access	2 of 96 reporting an access model	endnote 3
European data-sharing initiatives with a clear revenue model	15 % (Moonen et al., 2025)	§4.2.1
Initiatives scoring on all three mechanisms	45 of 153 at implementation+operational	§3.3

Radar data quality, stated by the OECD itself (Box 1.1): ~50 % of entries lack geographical focus, 58 % lack funding source, 78 % lack start dates; failed initiatives may still appear as operational; registration is European-biased.

The EUR 1.7 bn figure is given with an unusually frank provenance note (endnote 5): it originates in a funding portfolio compiled by Gaia-X AISBL and first published in Otto and Tardieu (2025), a Le Grand Continent commentary — “both authors hold or held board positions at Gaia-X AISBL and IDSA”. It aggregates DEP 2021-22 and 2023-24, EU4Health, Horizon Europe and Destination Earth with national funding from six member states, includes programmes beyond data spaces narrowly defined (Destination Earth EUR 240 m, an EU4Health allocation EUR 280 m), omits other member states’ investments, and has no published methodology. The OECD reads it as “an indicative order of magnitude rather than a precise audited total”, likely low.

The two-tier definition (§2.3)

The report’s central analytical device. It refuses to pick one definition and instead separates the inclusive one from the one whose value it tests.

Broad: “a data ecosystem in which participants share data through agreed common policies, standards, semantic models and processes, supported by enabling technical and organisational services that facilitate trusted and interoperable data sharing.”

Narrow: “a data ecosystem characterised by (i) a federated or decentralised governance framework; (ii) open, standards-based interoperability services; and (iii) mutual trust and effective control between data providers and data users over how data are used.”

The broad definition is deliberately wide enough to include Snowflake, Databricks, national statistical offices acting as trusted data intermediaries, and SDMX (Box 2.2) — which satisfies every element of it without ever having come out of the data spaces discussion. The narrow definition is what the rest of the report assesses. Unless stated, “data spaces” means the narrow one.

Lineage, §2.1: the term starts as a computer science abstraction — Franklin, Halevy and Maier (2005), “pay-as-you-go” integration, inverting the requirement that schema alignment precede service. Fraunhofer’s Industrial Data Space and the IDS-RAM (Otto et al., 2017) convert it into a governance framework by adding two things: federated custody via certified connectors, and machine-readable usage control. ETSI EN 304 199 now carries the governance formulation in standard form.

Adjacent concepts that do not meet the narrow definition (§2.4.1): the original dataspace concept, data mesh (Dehghani — domain ownership, data-as-a-product, self-serve infrastructure, federated governance), and inter-organisational “data fabric” (Box 2.3: CESMII, USD 70 m DOE + USD 70 m private, 2016, 200+ members, Smart Manufacturing Profiles, i3X API, 100+ crowdsourced companion specifications). The distinguishing line is not decentralisation and not cross-organisational scope — all of these have both — but federated governance, open interoperability, and usage control remaining with the provider.

Functionally equivalent arrangements that meet the narrow definition without the label (§2.4.2): TEFCA (US health, ~500 m records exchanged as of Feb 2026, up from ~10 m in Jan 2025; 11 QHINs, 14 214 organisations, 79 000+ connections, six purpose-bound exchange purposes), DEPA (India, consent-based, Account Aggregator, >270 million linked accounts), Japan’s ODS, Estonia’s X-Road (broad definition; converging on the narrow one via X-Road 8 “Spaceship”).

The four boundary conditions (§3.2, restated in §5.1)

The report’s most portable output. Data spaces have a comparative advantage over bilateral contracts, centralised platforms and PETs only where all four hold:

1. three or more parties;
2. competitive or jurisdictional relationships that preclude trusting a single commercial intermediary;
3. non-excludable collective value that bilateral contracts cannot capture;
4. ongoing data flows requiring continuous governance rather than one-off transactions.

“Where these conditions are absent, simpler data-sharing mechanisms may be sufficient and may not justify the co-ordination costs that data spaces entail.” Necessary but not sufficient: §5.1 adds a second stage — pre-existing digital infrastructure among participants, identifiable anchor firms, clear value propositions.

Full treatment, including what each alternative mechanism does and does not answer: Data space boundary conditions.

The co-ordination diagnosis (§3.1)

Multi-party data sharing is a collective-action problem (Olson, 1965) in which both ordinary co-ordination mechanisms fail:

• Price formation fails. Data value is context-dependent, so standard valuation across participants is structurally difficult; total social value exceeds the sum of bilateral values (Frischmann, 2012). Endnote 1 goes further and names Arrow’s information paradox (1962): a buyer cannot value information without possessing it, and once disclosed it has been transferred — “price negotiation requires revealing the very asset being sold”. In multi-party settings this compounds, because the same data are worth different amounts to each party.
• Centralised co-ordination fails wherever participants are competitors or in different jurisdictions. The case is Covisint (GM/Ford/ DaimlerChrysler, 2000, later Renault and Nissan): suppliers read it as buyer leverage over pricing and inventory, not neutral infrastructure; the auction business was sold, the OEMs withdrew, the exchange model was abandoned by 2004. “The failure was not technological but rather institutional.” Endnote 3 extends this across four technology waves — EDI (buyer coercion, Hart & Saunders 1997), SOA (hub-and-spoke despite the promise, Löhe & Legner 2010), APIs, control towers (visibility only within the focal firm’s immediate partners, Swink et al. 2024): each “enabled data exchange where one dominant party set the terms”, and each failed where governance had to be shared.
• Asymmetric cost-value distribution compounds both. Tier-2 suppliers pay for traceability data whose value lands on OEMs and regulators; farmers pay digitisation costs captured downstream. Confirmed at the Expert Workshop. “Not merely distributional; it is a structural participation barrier.”

For the public sector the absence of price-mediated co-ordination is “the default condition rather than a failure mode” — the limiting case of Frischmann — and the co-ordination problem reappears as institutional fragmentation across mandates and budgets (endnote 4: statutory purpose limitation, plus a payoff structure where officials are sanctioned for wrongful sharing but capture none of sharing’s value).

The resolution is Williamson’s: with neither prices nor hierarchy available, a hybrid form emerges, which “is not a loose amalgam of market and hierarchy but possesses its own disciplined rationale”. Data spaces are one such hybrid. Endnote 6 qualifies the asset-specificity leg honestly: data are nonrival and redeployable; the specificity is in the participation investment — connectors, semantic mapping, governance relationships — which is sunk and ecosystem-specific.

The three mechanisms (§3.3)

Ordered as adoption actually proceeds:

1. Distributed governance — the institutional foundation (§3.3.1).
2. Open interoperability — “the technical infrastructure that participants actually deploy” (§3.3.2).
3. Identity, usage control and provenance — the trust capability (§3.3.3), which comes bundled in the stack rather than being the adoption motive.

“The governance framework determines why participants join; the interoperability stack determines what they actually deploy.”

Only 45 of 153 implementation- and operational-stage initiatives score on all three. The distinctive claim is integration, not any one mechanism: “Governance neutrality is available through public bodies, usage control through bilateral contracts, and open standards through industry consortia. However, no alternative mechanism combines all three.”

Governance (§3.3.1)

In every mature data space examined, a neutral intermediary or orchestrator holds a key role: a state enterprise (Fintraffic), a non-profit association (Catena-X Automotive Network e.V.), an industry consortium (BoostAeroSpace). Guggenberger, Schlueter Langdon and Otto (2025) call data spaces “meta-organisations”. Kernstock et al. (2025): even decentralised-by-principle ecosystems establish central institutions as they scale — “but these institutions must remain structurally neutral”.

The drift is explained, not deplored: as a network grows, the costs of distributed decision making rise faster than its benefits and free-riding incentives appear; a neutral central institution lowers transaction costs and concentrates accountability “but at the price of reintroducing the very governance asymmetry that decentralised principles were designed to avoid”. The DGA neutrality requirement is read as a regulatory response to that tension.

Interoperability (§3.3.2)

Consolidation around EDC: about a third of operational initiatives use Eclipse Dataspace Components, while the FIWARE connector drops from ~20 % overall to zero at operational stage — with the caveat that IDSA runs the Radar and non-IDSA stacks may simply not register. Open-source implementations (EDC, FIWARE codebases) function as the de facto standardisation mechanism. Databricks open-sourced Delta Sharing and Unity Catalog, which “narrows the gap on the interoperability dimension” without adopting distributed governance or provider usage control.

ODS-RAM’s contrasting route is architectural decoupling: three loosely coupled, independently replaceable pillars — ontology and semantic interoperability; data addressability and discoverability; identity and usage control — permitting “detachability” and incremental opt-in adoption (Tsuda, 2026), which the report ties directly to the bootstrapping problem. Mentioned in passing: emerging “dynamic ontologies” using LLMs to automate cross-ontology mappings, and AI-based semantic mediation at runtime — with the caveat that “governance and provenance guarantees required in regulated dataspaces still favour explicitly agreed models”.

Trust (§3.3.3)

Identity and access control are “necessary but not distinctive” — every mechanism has them; data spaces merely federate them. The distinctive capability is post-access usage control: purpose limitation, time bounds, geographic restriction, prohibition of onward sharing, expressed in ODRL, negotiated over the Eclipse Dataspace Protocol, enforced by connectors.

Its limitation is stated plainly: “once data leave the provider’s infrastructure, enforcement depends on the data user’s compliance”. The access interface “is both the enforcement mechanism and the potential point of circumvention”. Provenance and audit logs are the compensating mechanism — they do not prevent violations, they make them detectable, “shifting the problem from technical prevention to deterrence through accountability”.

The adoption data track that shift: access-and-usage-policy building block 35 % of all → 28 % at operational; provenance 25 % → 48 %.

Two design philosophies are contrasted: IDS-RAM/DSP mandates the full stack, including a negotiation layer even for fully permissive “no operation” policies; ODS-RAM makes contracting optional and prescribes only the interface, accommodating TEEs, differential privacy and other means — “binding usage control to a single, institutionally derived protocol can create a bottleneck for market adoption”.

PETs as complement, not substitute (Box 3.2)

Four categories named: federated learning, secure multi-party computation, trusted execution environments (with “Data Control Rooms” packaging enclaves plus policy enforcement and audit), synthetic data. The construction the report endorses:

a data space connector negotiates a contract that specifies not only who may access data and for what purpose, but through which PET the data must be processed. This makes the choice of privacy technology an enforceable element of the data-sharing agreement. The provider no longer needs to trust the recipient’s compliance because the technology structurally limits what can be done with the data.

And the standing objection, in the same box: PET adoption “faces the same collective-action challenges identified throughout this section: additional computational costs, specialised expertise requirements and the need for co-ordination among participants on which technologies to deploy” (Auñón et al., 2024).

In §3.2 PETs are listed as an alternative mechanism that falls short: they “address the confidentiality dimension” but “do not address the institutional co-ordination problem of who governs, who pays and how collective value is allocated”. §5.5 then makes support for both tracks — usage control and PETs — one of the eight policy considerations, with the G7 Industry, Digital and Technology Ministerial Declaration of December 2025 committing to “advanc[ing] DFFT through the promotion of privacy-enhancing technologies (PETs)”.

This is the counterexample the wiki was looking for — see open task 19 and the confidentiality gap.

The bootstrapping problem (§3.4)

The second-order collective-action problem, at two levels:

1. collective investment in shared infrastructure must precede the network effects that justify it;
2. each participant’s private onboarding investment must be justified by returns that may not materialise until the ecosystem reaches scale.

The infrastructure layer has club good properties (Brousseau, Toledano and Eustache, 2024): high fixed costs, low marginal cost per deployment, non-rivalrous within the ecosystem, excludable to non-participants. Olson’s free-riding applies directly; Bisière et al. (2025) formalise the chicken-and-egg; DGA neutrality constrains the cross-subsidisation that platforms use to bootstrap (Rochet & Tirole, 2003).

Frischmann’s refinement is the analytically load-bearing move: free riding does not necessarily reduce contributors’ investment incentives. Bootstrapping binds specifically when expected private returns to anchor participants fall below their investment threshold despite positive collective returns — “when no subset of participants can justify the investment from their own expected returns alone”.

The cases, read against exactly that test, live on the bootstrapping problem and data space business models.

Also recorded here: the “reinventing the wheel” cost (Annex A) — consortia re-debating technical components from scratch — with Baron, Ménière and Pohlmann (2014) on standardisation’s duplicative R&D and vested-interest delay, and the explicit naming of IDSA, Gaia-X, DSSC and sector consortia each developing partially overlapping building blocks as facing “analogous risks”.

The funding critique (§4.2.1, §5.2)

The sharpest institutional finding, and it is about the funder rather than the funded:

• Two purposes of public funding are run together: funding the technical foundation (schemas, protocols, interoperability frameworks — pre-competitive R&D with spillovers) and funding ecosystem creation (conditional on an identified collective-action problem no market actor has the incentive or legitimacy to solve).
• “EU funding programmes were not explicitly designed around this distinction or the bootstrapping rationale… The bootstrapping justification is an analytical overlay by economists (Bisière et al.; Otto and Tardieu), not the stated rationale of the policymakers who designed the funding instruments.” The stated framing is competitiveness, digital sovereignty, single market for data.
• Therefore: “If the instruments were designed for the first purpose rather than the second, the fact that only one in nine entries has reached operational status becomes easier to understand. The instruments were simply not designed for the problem they are now expected, or may need, to solve.”
• Incentive misalignment on top (Holmström & Milgrom 1991; Lerner 2009): where funding rewards observable compliance — consortium formation, architectural alignment, dissemination — over unobservable operational viability, rational actors optimise for the former. Endnote 2 states it as a multi-task principal-agent problem: “compliance is measurable; progress towards operational viability is not.”
• And a dependency effect, from the workshop: continued public support “may remove the pressure to find viable revenue models altogether”.
• 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.”

The policy landscape (Ch. 4)

Regulation as driver (§4.1). Two channels: instruments that mandate or promote data spaces (EHDS; DEPA consent managers) and instruments whose requirements are most readily met by federated governed exchange (EU Battery Regulation; TEFCA). EU instruments named: DGA (Art. 12 neutrality), AI Act (provenance/traceability for high-risk systems), Data Act, Battery Regulation (carbon footprint declarations and digital battery passports), ESPR. Digital Omnibus (Nov 2025) would repeal the DGA as a standalone regulation, fold it into the Data Act, make intermediation registration voluntary, and relax separation from structural (legal) to functional — which “lowers the barrier for integrated platform operators” and may erode the distinctive regulatory advantage data spaces hold over commercial platforms. Extraterritorial reach is explicit: third-country suppliers must comply to reach the EU market, which “transforms regulatory alignment from a European policy concern into a global market access requirement” — ABtC is the proof.

Funding instruments (§4.2.1). EU ~EUR 336 m in Common European Data Spaces across 14 sectors through DEP as of Nov 2025, plus ~EUR 100 m ongoing; the EU Data Union Strategy (Nov 2025) repositions data spaces “from a governance instrument to AI supply chain infrastructure” and introduces “data labs”. Germany: EUR 200 m Gaia-X Förderwettbewerb via Bundesnetzagentur, 11 projects, ~EUR 117.4 m awarded by end 2024, phase concluded March 2025, Gaia-X Hub Germany shifting to onboarding and community co-ordination. France: inside France 2030’s EUR 2.5 bn AI/digital envelope, data-spaces allocation never broken out. Canada: National Data Spaces Pilot Funding Program, Jan 2026–Mar 2028, up to CAD 3.15 m — “the first dedicated national data spaces programme outside Europe and Asia”. China: National Data Administration Trusted Data Space Development Action Plan (2024-2028), 63 pilot projects designated in 2025, target of 100+ trusted data spaces by 2028. Permanent institutions: Sitra (30+ projects, 15 sectors, 100+ organisations), the Dutch CoE-DSC (47 initiatives, 80+ service providers, ~80 % dependent on government funding), Germany’s National Data Institute, and the DSSC itself — a 12-organisation consortium co-ordinated by Fraunhofer under DEP, Oct 2022 – Mar 2026, now DSSC 2, which the report flags as its own illustration of “the tension between the co-ordination function it serves and the project-funding model through which it is financed”.

The demand-side gap (§4.2.2, §5.3). Three instruments, all rare: Spain’s Data Spaces Kit (Red.es, EUR 60 m NextGenerationEU, non-competitive grants up to EUR 30 000 per private data provider and EUR 50 000 per public administration — “breadth of adoption rather than depth of capability”); Japan’s IPA Data Spaces Academy (guides, a CDO handbook, a data maturity framework) attacking informational rather than financial barriers; India’s DEPA, where regulatory design creates a commercially viable intermediary category instead of subsidising one — “creating a commercially viable intermediary category may address the bootstrapping problem more durably than subsidising infrastructure”.

Standards and cross-border (§4.3, Box 2.1). ISO/IEC JTC1 SC38: DIS 20151-1 (concepts), DIS 26450 (Eclipse DSP), DIS 26451 (Eclipse DCP), with 20151-2 (trust frameworks) and 25850 (use cases) upcoming — DSP and DCP appear in both ISO and CEN/CENELEC pipelines, making them “the de facto technical baseline”. CEN/CENELEC JTC25: EN 18235-1:2026 in publication, Part 2 (trustworthiness) in second editorial phase, Part 3 (interoperability) preparing for enquiry, CEN/CLC/TS 18331 (maturity assessment) under vote — the series is “explicitly framed for Data Act Article 33 compliance”. ETSI: DCAT profiles, EN 304 199. IEEE: 3800-2024 is “the only fully published standard in this landscape”, approached from a data trading / marketplace perspective, chaired by the Secretary General of Japan’s Data Society Alliance. Cross-border: the IPA–Catena-X proof of concept (March 2025), PCF data exchange between the Ouranos Ecosystem and Catena-X for EU Battery Regulation compliance, across ODS-RAM and EDC implementations. Brazil: NEO-UFRGS Green Paper (July 2024, under the Brazilian-German Digital Dialogue, GIZ on behalf of BMDV), IDSA Hub at Eldorado (March 2024, first in the Americas), IDSA Competence Center recognition (Dec 2025) — “these steps do not yet constitute demonstrated interoperability”.

Certification (§4.4). Two logics: the DGA’s registration regime (procedural, declares compliance, “does not assess the substantive quality or public benefit”) versus Japan’s “public interest digital platform operator” certification administered by IPA, assessing against substantive criteria — trade-secret protection, security, interoperability, data provider control. ABtC is accredited under it. The workshop treated it as a model for distinguishing public-interest from purely commercial operators. eIDAS 2 / the European Digital Identity Wallet is named as the identity layer data space participation depends on.

Government as operator and anchor (§4.5, Box 4.2). Estonia’s X-Road (operational since 2001, NIIS — Estonia/Finland/Iceland, ~52 000 organisations, ~3 000 e-services, >1 bn transactions/year, implemented in 20+ countries), EHDS (first EU regulation explicitly requiring government-operated data space components: national digital health authorities and health data access bodies), the EU Public Procurement Data Space (DG GROW, Sept 2024, eProcurement Ontology, 250 000+ public authorities, ~EUR 2.5 trillion/year ≈ 15 % of GDP, voluntary national connection, an OECD-implemented TSI project supporting a Member State cohort), the US Genesis Mission (Nov 2025 Executive Order, federated platform across 17 national laboratories; ~USD 320 m DOE AI infrastructure Dec 2025, USD 293 m Request for Applications March 2026; explicitly raises IP attribution for contributed proprietary data), Israel’s inter-ministerial construction-sector framework. Public-sector value “cannot be captured through market-based mechanisms”, so sustainability rests on recurrent budget as “a permanent utility”, not revenue — X-Road as the two-decade proof. Procurement as demand-side driver: the Interoperable Europe Act (April 2024) and the eFTI Regulation (Member State authorities must accept electronically shared freight information via certified platforms by July 2027).

The IP question nobody has answered (§2.1, §4.5)

Flagged twice and left open: usage control governs how data may be processed but “does not resolve who owns the value derived from them”. Where data spaces pool proprietary data for AI training, model ownership, derived insights and downstream patent claims are unallocated. The worked case is the US Materials Genome Initiative — more than USD 1 billion in cumulative federal investment, and “persistent industry reluctance to contribute proprietary performance and testing data”. PETs, federated learning and revenue-sharing coalitions have addressed parts of it; “none has yet produced a widely adopted commercial model”. ⟨Inference: this is the report’s clearest unclaimed space, and it is adjacent to but distinct from the confidentiality gap — the question is not can my input be seen but who owns what was learnt from it. The wiki has no page on it.⟩

The eight policy considerations (Ch. 5)

1. Target investment where the structural condition holds — two-stage assessment: the four boundary conditions, then sectoral preconditions.
2. Distinguish funding the technical foundation from funding ecosystems.
3. Address the demand-side adoption gap — subsidy, capacity building, or an intermediary category that is commercially viable.
4. Support business models around infrastructural value — cost sharing and membership, plus value-added services, not data monetisation.
5. Support trust-enhancing technologies and their integration by design — usage control and PETs; “Neither alone is sufficient.”
6. Adopt certification and government anchoring — designate anchors where none exists; governments lead by example (G2G, G2G2B).
7. Recognise AI-driven data demand as a pathway to viability — “an economic necessity rather than a policy aspiration”, conditional on delivering governed access “at the speed and scale that AI development requires”.
8. Enhance measurement — condition funding on baseline and outcome data; extend collection beyond Europe.

What this is for SINE

⟨Inference, flagged as such throughout the pages it feeds:⟩

• The four boundary conditions are a qualification test SINE can run on its own use cases — and PACT/iLEAP, CCTI’s reinsurance case and the GIZ agricultural-value-chain tender do not score identically on them.
• The report independently reproduces SINE’s Data Commons argument (governance layer as the binding constraint, not the technology) with an n=235 registry behind it, and reproduces the coordination-vs-bargaining distinction in a different vocabulary (price formation failure vs. asymmetric cost-value distribution).
• It is the first document in this wiki that reaches for cryptography and argues its limits rather than not reaching for it, which changes the shape of open task 19.
• “Data monetisation has not proven a viable funding model” is a finding about SINE’s own market, not just about data spaces.

Related pages

• Data space boundary conditions
• The bootstrapping problem
• Data space business models and viability
• OECD
• IDSA · Catena-X · Gaia-X
• Data Commons · Ostrom’s commons design principles
• Coordination failure vs. bargaining failure
• The confidentiality gap · Cryptographic trust infrastructure
• Digital Product Passport

Sources

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 (Markdown twin of the OECD PDF, the citable file for this page).