Data Sharing Toolkit

Globally, innovation in data-governance is fairly embryonic. Recent years have seen a flurry of new activity, but much of this remains poorly defined or nascent. Few, if any, countries have cracked the code of responsible and effective data-sharing initiatives and governance.

Through an in-depth review of existing literature and more than 50 case studies, participation in five conferences, and consultation with experts in the field, Nesta, in partnership with Smart Dubai, has found the following:

  • Definitional confusion means that decision-makers have no firm framework to help them understand what their options for sharing data are in practice.
  • Each collaborative project involving data sharing is unique, set in a specific dynamic composed of different actors, laws and rules, expectations, levels of expertise, incentives and relations.
  • Pre-defined models and labels are not useful as tools for translating theory into practice.

To respond to the above, we produced a toolkit to provide useful guidance and resources for private and public organisations to prepare for and design data-sharing initiatives, helping them identify the right combination of options for the specific and unique context for the given circumstances.

The toolkit comprises six decision points prompting and guiding discussions about all key elements of a data-sharing arrangement, and a set of 'project foundations' conditions required to move forward with the data-sharing project, providing the overarching legal, technical and relationship considerations.

By applying this framework to the analysis of more than 20 case studies, a number of trends, recurring themes and challenges that commonly arise in data-sharing initiatives have been identified.

The five most important trends are as follows:

  • Defining a clear problem to be solved, or a specific use-case is essential to supporting a successful data-sharing arrangement.
  • Data governance involves at least three layers of complexity.
  • Data-sharing initiatives will vary in terms of scale and lifespan.
  • Ethical considerations must include issues of privacy and bias.
  • Data-sharing initiatives need to include a variety of stakeholders, not only those that hold useful data.