Can government remember? Is it condemned to repeat mistakes? Or does it remember too much and so see too many reasons why anything new is bound to fail?

While at the beginning of a data revolution, the deluge of data is creating the potential for an ‘information collapse’ in complex administrations: structured information and knowledge is lost in the noise or, worse, misinformation rises as fact.

There are many reasons for this: the technical design of systems, turnover of people, and contracting out. Information is stored in silos and often guarded jealously. Cultural and process issues lead to poor use of technologies. Knowledge is both formal (codified) and informal (held in people’s brains). The greatest value will be unlocked by combining these with existing and emerging tools.

This report sets out how the public sector could benefit from a federated, data-driven approach: one that provides greater power to its leaders, benefits its participants and users, and improves performance through better use of, and structured access to, data.

The report explores examples from the Open Data Institute, Open Banking Standard, BBC Archives, Ministry of Justice, NHS Blood and Transplant, Defence Science and Technology Laboratory and Ministry of Defence.

Key recommendations

To shorten the path between innovation and policy in a way that is repeatable and scalable, the report proposes six areas of focus be considered in any implementation design.

1. Policy ​– providing strategic leadership and governance; framing and analysing economic, legal and regulatory impacts (e.g. GDPR, data ethics, security) and highlighting opportunities and threats.
2. Culture – creating compelling peer, press and public communication and engagement that both address concerns and inspire people to engage in the solutions.
3. Making – commissioning startups, running innovation competitions and programmes to create practice-based evidence that illustrates the challenges and business opportunities.
4. Learning – creating training materials that aid implementation and defining evidence-based sustainable business models that are anchored around user-needs.
5. Standards – defining common human and machine processes that enable both repeatability and scale within commercial and non-commercial environments.
6. Infrastructure – defining and framing how people and machines will use data, algorithms and open APIs to create sustainable impact.