Paper by Stefaan Verhulst: “As data sharing and reuse become central to scientific discovery, artificial intelligence, and public decision-making, the challenge of data governance has shifted from a primarily technical problem to one of institutional design. Over the past decade, a wide range of governance models for data collaboration have emerged-including data trusts, data commons, data cooperatives, data intermediaries, data unions, data sandboxes, and data spaces. These models are often presented as competing institutional solutions to the problem of responsible data sharing. This paper argues instead that such models represent distinct governance responses to different structural challenges within data ecosystems.
Building on the concept of data collaboratives-which I introduced in 2017 as cross-sectoral arrangements for responsible data reuse in the public interest-I propose a purpose-driven typology that identifies seven governance archetypes and the specific coordination problems they address, including transaction costs, power asymmetries, legitimacy deficits, collective governance needs, ownership inequality, systemic uncertainty, and scaling complexity.
I argue that the question is not which governance model is normatively superior, but rather which model is fit for purpose within particular institutional contexts. The paper concludes by introducing a functional theory of data collaboration centered on institutional orchestration, whereby multiple governance arrangements coexist and evolve within polycentric data ecosystems. In this framework, strategic data stewardship becomes essential for diagnosing governance needs and sequencing institutional responses that enable responsible and sustainable data reuse…(More)”.
