Data Spaces between Co‑ordination Challenges, Financial Viability and Public Policy

5. Conclusions and policy considerations

Abstract

This chapter draws together the findings of the report and sets out eight policy considerations for governments seeking to support viable data spaces. These concern targeting public investment where a genuine co-ordination problem exists, distinguishing funding for the technical foundation from funding for ecosystems, addressing the demand-side funding gap, supporting business models around infrastructural value, promoting trust-enhancing technologies, using certification and government anchoring to build institutional trust, recognising AI-driven data demand as a pathway to viability, and enhancing measurement. The considerations support OECD work on data free flow with trust (DFFT) and the OECD Recommendation on Enhancing Access to and Sharing of Data (EASD).

Governments in the European Union (EU) alone have invested an estimated EUR 1.7 billion in a data space ecosystem whose value they cannot yet measure. Only one in nine entries has reached operational status. Moreover, not a single initiative has reached the scaling stage that the definitions of the International Data Spaces Association itself associate with viability. Finally, the evidence base on what has been built remains thin (Box 1.1 of Chapter 1). The technology has been demonstrated to work, but evidence is lacking on whether the investment is delivering the promised returns.

Government intervention may be justified where genuine co‑ordination failure exists that alternative approaches leave structurally unresolved. In domains such as supply chain traceability, health data exchange, multimodal mobility and energy system integration, the parties that need to share data are often competitors or jurisdictional rivals unwilling to cede control to a single commercial intermediary. Data spaces address this class of problem through distributed governance, open interoperability and provider-defined usage control (Section 3.3 of Chapter 3). Regulatory requirements across multiple jurisdictions amplify the urgency of solving these co‑ordination problems (Section 4.1 of Chapter 4).

The policy considerations below address how governments can sharpen their approach to targeting, funding, measuring and delimiting data space initiatives.

5.1. Targeting investment where the structural condition holds

Four boundary conditions define the class of problem for which data spaces have a comparative advantage over alternative mechanisms: (i) three or more parties; (ii) competitive or jurisdictional relationships that preclude trusting a single commercial intermediary; (iii) non-excludable collective value that bilateral contracts cannot capture; and (iv) ongoing data flows requiring continuous governance rather than one-off transactions (Section 3.2 of Chapter 3). These conditions are particularly pertinent in cross-border settings, where regulatory divergence compounds the co‑ordination problem and no single intermediary can easily claim cross-jurisdictional authority (Section 4.3 of Chapter 4). Where these conditions are absent, alternative approaches may suffice and may not justify the co-ordination costs that data spaces entail (Table 3.1 of Chapter 3).

The presence of the four boundary conditions is necessary but not sufficient, however. Sectoral conversion rates (Figure 3.1 of Chapter 3) suggest that pre-existing digital infrastructure, identifiable anchor firms and clear value propositions among early participants are preconditions for data space viability, irrespective of initiative volume. A two-stage assessment could help direct public resources more effectively towards use cases where data spaces are both structurally necessary and practically viable. The first stage would examine whether the co‑ordination problem meets the boundary conditions. The second would consider whether the sectoral preconditions are in place.

5.2. Distinguishing between funding for the technical foundation and for ecosystems

Two distinct purposes of public funding merit separation in policy design (Section 4.2.1 of Chapter 4). Investment in information models, protocols and interoperability frameworks constitutes pre-competitive research and development (R&D) with broad spillover benefits. For its part, investment in the creation of actual data space ecosystems is more conditional and depends on whether a specific collective-action problem has been identified that no market actor has the incentive or legitimacy to solve. Most funding programmes were not explicitly designed around this distinction or the bootstrapping rationale identified in Section 3.4 of Chapter 3. The stated rationale centres on competitiveness, digital sovereignty and, in the case of the European Union, the single market for data, rather than on correcting a formally identified collective-action failure (Section 4.2.1 of Chapter 4).

Where funding instruments reward observable compliance (consortium formation, architectural alignment, dissemination) rather than operational viability and participant willingness to pay, rational actors tend to optimise for the former (Lerner, 2009[1]). Furthermore, long-term public funding may also create dependency effects that prevent data space stakeholders from developing viable revenue models, as discussed at the OECD Expert Workshop (Annex A). Where governments fund ecosystem creation rather than pre-competitive R&D, success metrics calibrated to operational outcomes (participant growth, data flows, revenue readiness) rather than architectural compliance may better align incentives with viability.

5.3. Addressing the demand-side funding gap

The dominant policy response has been supply-side funding directed at shared infrastructure. Far less common is funding that addresses the onboarding costs that individual participants, particularly small and medium-sized enterprises (SMEs), face when joining a data space (Section 4.2.2 of Chapter 4). The bootstrapping problem operates at two levels. The collective investment in shared infrastructure must precede the network effects that justify it. Moreover, each participant’s private onboarding investment must be justified by expected returns that may not materialise until the ecosystem reaches sufficient scale (Section 3.4 of Chapter 3). Addressing only the first level while leaving the second unfunded may help explain why data spaces that receive substantial infrastructure investment nonetheless fail to attract sufficient participation.

Three demand-side approaches have emerged across jurisdictions: direct subsidy, public investment and regulatory design. Direct subsidy for participant onboarding costs is illustrated by Spain’s Data Spaces Kit, which subsidises integration of individual firms into sectoral data spaces on a non-competitive basis. Public investment in capacity building and data maturity are features of Japan’s educational programme targeting businesses and SMEs (Information-technology Promotion Agency). Finally, regulatory design can create commercially viable intermediary categories rather than relying on public subsidy. In Data Empowerment and Protection Architecture (DEPA) in India, for example, regulated consent managers generate revenue from facilitating data exchange. Given these multiple approaches, governments may wish to complement supply-side infrastructure funding with instruments that reduce participant onboarding costs. They could do this through direct subsidy, capacity building or regulatory frameworks that make the intermediary role commercially viable.

5.4. Supporting business models around infrastructural value

The value that data spaces produce is predominantly collective and infrastructural, exhibiting properties that Frischmann (2012[2]) associates with shared resources whose social benefits exceed what pricing mechanisms can capture. Business models premised on data monetisation assume bilateral, transactional value. Where value is collective and non-market, these models risk being structurally misaligned with the value proposition they are meant to serve. Empirical evidence from European data-sharing ecosystems is consistent with this misalignment: in a qualitative study across mobility, aerospace, energy, agriculture and other sectors, 8 of 20 initiatives framed themselves as data marketplaces. However, their orchestrators acknowledged that the initial goal had been deeper collaboration rather than transactional exchange (Eustache, Brousseau and Toledano, 2025[3]).

Viable models observed in operational data spaces resemble cost-sharing arrangements in which participants jointly fund the infrastructure and share in system-level benefits (the Finnish Maritime Data Space); membership-based access to co‑ordination infrastructure (BoostAeroSpace); or hybrid models combining public co-investment with participant cost recovery. Prioritising use-case value over infrastructure development is associated with stronger prospects for financial viability (DSSC, 2026[4]), a finding confirmed at the OECD Expert Workshop (Annex A).

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. In particular, funding logic may need to evolve over the ecosystem life cycle, shifting from early-stage subsidies and neutral cost sharing during emergence towards usage-based or performance-linked mechanisms as the ecosystem matures (Eustache, Brousseau and Toledano, 2025[3]). Governments may wish to distinguish between use cases where data monetisation is viable and where value is collective and infrastructural, and to calibrate funding expectations accordingly.

5.5. Supporting adoption of trust-enhancing technologies and their integration by design

Provider-defined usage control is a major capability that distinguishes data space interoperability from both commercial platform interoperability and traditional industry consortium data exchange (Section 3.3.3 of Chapter 3). Without credible guarantees that shared data will be used only as agreed, rational actors withhold participation regardless of the potential collective benefit. Two complementary approaches are maturing in parallel: data usage control (e.g. built on Open Digital Rights Language and dataspace protocols); and privacy-enhancing technologies (PETs) such as federated learning, secure multi-party computation and trusted execution environments. Neither alone is sufficient.

Sustained support for the maturation of both tracks, and for the interoperability layer that allows them to be combined within a single data space deployment, would strengthen the trust foundations on which participation depends. This observation is based on the support of the data free flow with trust (DFFT) Expert Community on OECD work related to PETs, as well as a complementary OECD project on “Trust by Design” in cross-border supply chains [DSTI/DPC/DGP(2026)5]. Governments are already recognising the importance of PETs for data sharing, including in cross-border contexts. The G7 Industry, Digital and Technology Ministerial Declaration of December 2025 commits to “advanc[ing] DFFT through the promotion of privacy-enhancing technologies (PETs)” and calls for “initiatives that support greater understanding of PETs, such as the OECD Global Repository on PETs” (G7, 2025).

5.6. Adopting certification and government anchoring to build institutional trust

Every data space that has reached operational viability features a credible neutral anchor entity. This could be a state-owned enterprise, a non-profit industry consortium or a large firm whose market position makes neutrality credible (Section 3.3.1 of Chapter 3). In some sectors, no natural anchor exists, whether because of participant fragmentation, heterogeneity or jurisdictional dispersal. In such cases, governments may wish to consider designating public bodies or state-owned enterprises to serve as institutional anchors (Section 4.5 of Chapter 4).

Estonia’s X-Road and the European Health Data Space regulation illustrate two models of institutional anchors: long-running government interoperability infrastructure on which data space capabilities can be layered; and legislative mandates that designate government bodies (digital health authorities, health data access bodies) as the institutional infrastructure for data sharing. Deploying data spaces for the government’s own operations – from inter-governmental (G2G) configurations to arrangements extending to businesses and citizens (G2G2B) – reinforces this channel by leading by example. To that end, it demonstrates that the architecture works at scale, legitimises the underlying standards; and lowers the perceived adoption risk for private participants (Section 4.5 of Chapter 4).

Certification is emerging as a complementary instrument. Japan’s “public interest digital platform operator” certification (Section 4.4 of Chapter 4), under which the Automotive and Battery Traceability Centre has been accredited (Box 4.1 of Chapter 4), assesses operators against substantive criteria for reliability and public benefit. This generates a trust signal that may help address the bootstrapping problem. Governments developing data space frameworks may wish to consider how certification or registration mechanisms can provide the institutional trust that the market may not be able to generate on its own.

5.7. Recognising AI-driven data demand as a pathway to viability

AI is emerging as a catalyst for data space viability. The demand for high-quality, domain-specific training data creates concrete business incentives for governed data sharing that previous waves of data space formation lacked. No single actor possesses the full data resources needed to train competitive AI models in industrial sectors, making collaborative data ecosystems an economic necessity rather than a policy aspiration. The EU Data Union Strategy’s repositioning of data spaces as AI supply chain infrastructure, including the introduction of “data labs” to link data spaces to AI ecosystems (Section 4.2.1 of Chapter 4), signals institutional recognition of this symbiosis. Similar institutional recognition is visible in Japan, where the ODS-RAM v2 design philosophy situates the architecture explicitly in relation to agentic AI (Tsuda, 2026[5]).

Government, including through funding programmes and viability assessments, could encourage data space initiatives to articulate their value proposition in terms of the AI use cases they enable, where these exist. Whether AI-driven demand will prove transformative depends on whether data spaces can deliver governed data access at the speed and scale that AI development requires. The current ecosystem still needs to demonstrate this condition via operational use cases.

5.8. Enhancing measurement

Despite an estimated EUR 1.7 billion in EU public investment alone, few initiatives collect baseline data on the co‑ordination problem they aim to solve; fewer still measure outcomes against that baseline in terms that allow comparison across sectors or with alternative data-sharing mechanisms. Governments may wish to condition future public funding on the systematic collection of baseline and outcome data sufficient to evaluate whether the initiative is delivering value that alternative mechanisms could not. The geographic dimension compounds the problem. The evidence base is overwhelmingly European; non-European architectures operating on comparable principles (e.g. Open Data Spaces but also the Trusted Exchange Framework and Common Agreement and DEPA) are absent or underrepresented in the primary quantitative source for this paper. Extending systematic data collection on federated data-sharing initiatives across OECD Members and partners is a precondition for evidence-based policy guidance on an area that is genuinely international in scope.

References

[4] DSSC (2026), “Data Spaces Blueprint v3.0”, https://dssc.eu/space/BVE/357074549/Data+Spaces+Blueprint+v3.0 (accessed on 9 April 2026).

[3] Eustache, L., E. Brousseau and J. Toledano (2025), “The economics of data-sharing: An empirical investigation of data sharing ecosystems”, Forty-Sixth International Conference on Information Systems (ICIS), Nashville, Tennessee, https://hal.science/hal-05542222v1.

[2] Frischmann, B. (2012), Infrastructure: The Social Value of Shared Resources, Oxford University Press.

[1] Lerner, J. (2009), Boulevard of Broken Dreams: Why Public Efforts to Boost Entrepreneurship and Venture Capital Have Failed — and What to Do About It, Princeton University Press.

[5] Tsuda, M. (2026), Why Open Dataspaces: Design Philosophy and Architectural Paradigm, Information-technology Promotion Agency, Tokyo, https://www.ipa.go.jp/en/digital/architecture-guidelines/individual-link/p1o1lf000001xv4n-att/WhyOpenDataspaces_en.pdf.