Research and analysis

National Data Library

Published 29 July 2025

Rapid projects support government departments to understand the scientific evidence underpinning a policy issue or area by convening academic, industry and government experts at a single roundtable. These summary meeting notes seek to provide accessible science advice for policymakers. They represent the combined views of roundtable participants at the time of the discussion and are not statements of government policy.

What are the opportunities a National Data Library can provide? What can we learn from past initiatives, including innovations in the private sector and in start up communities?

Meeting notes from a roundtable opened by Baroness Jones of Whitchurch (Parliamentary Under-Secretary of State at DSIT) and chaired by Craig Suckling (Government Chief Data Officer during this period), facilitated by the Government Office for Science and DSIT.

Please note that this took place early on in the National Data Library policy development. Craig Suckling was Government Chief Data Officer at the time but has since left Government.

8 October 2024

1.What are the opportunities a National Data Library (NDL) can provide?

1.1. Public benefit: The NDL can use data to improve public services and policy making. For this to be achieved, government should take a considered, systematic approach to what constitutes, and what the public considers, a public benefit.

1.2. Value exchange: If the public consent for their appropriately anonymised, securely and ethically managed data to be used, they should also expect to get something out of the exchange in the form of improved services or public policy making. End users of data (including businesses, researchers, civil society and the public) should be consulted about what they would most value from this service, to ensure the NDL’s aims are always framed around clear outcomes relevant to users, rather than simply collating as much public sector data as it can.

1.3. Curated data sets: The NDL could improve opportunities to examine Mission-specific or, for example, regional issues and develop targeted interventions to address them. This would most usefully be in the format of curated datasets (structured, indexed and catalogued data on specific topics). For example, if there is an area in the UK where particular minority groups are suffering from health inequalities, better data access and linkage may enable insights to facilitate more effective public policy support.

1.4. Interconnectivity of data sets: The NDL should not duplicate data but rather be focused on linking and deconflicting of key public data assets. This coherent system could deliver public value, for example enabling quicker, better and more automated decision making, with human oversight.

1.5. Inclusion of private sector data: There are opportunities for data users to be incentivised to increase the range and breadth of datasets accessible through the NDL. This would set up a ‘virtuous spiral’ of richer data being available to all, which in turn would enable users to improve the products and services they build using it.

2.What can we learn from past initiatives, including innovations in the private sector and in start-up communities?

2.1. Define a clear mission  

• Clarity and transparency of the NDL’s purpose is critical to public perceptions. It is critical to define who the key users are (including businesses, researchers, civil society and the public), engage with them to understand their needs, use this to inform the NDL design, and demonstrate clearly how the NDL will benefit these groups.  

• The NDL should not be built (or perceived) as a ‘black box’ where vast amounts of data are entered, stored and used. It should be approached and communicated as a network of services to make public sector data more accessible.

2.2. The importance of building trustworthy foundations  

In the design of the NDL, it will be critical to consider: 

• Data privacy, and the public perception of data risks alongside perceptions of benefit. There is a balance to be achieved between consent, discretion and openness.  

• Ethics of data use, especially for potential commercial purposes.  

• Robust governance, both of the programme, and standards for the data exchanged within it. 

• Competence of the technology, including security, resilience, and user experience.  

• Transparency, around data provenance and quality, as well as the uses to which it is being put once shared via the NDL.

2.3. Understand learnings from past initiatives and start small

• Look at the landscape of what already exists and identify what, if anything, can be reused. 

• Avoid single-vendor lock in and spending a lot of money on new technologies. 

• Learn from past data sharing initiatives (e.g OpenSAFELY) which have undergone years of trials, user-testing and other evaluations.  

• The NDL should begin with small, emerging use cases to prove its value, trial the technology, and build trust. Think of this as a service that grows over time.

2.4. Legislative frameworks and devolution  

• Legislative frameworks – and perceived legislative barriers to data sharing – are important and should be considered as early as possible. 

• There is different legislation on data sharing across the UK’s devolved legislatures in Scotland, Wales and Northern Ireland. This should be considered to ensure the NDL is truly national.

2.5. Not underestimating the effort required for remediation of key data assets 

• The NDL will stand or fall on the underlying quality of its data assets. In particular, the interconnectivity of data sets will be one of the keys to delivering the full potential of the NDL.  

• However, improving the quality and operational management of departmental data, and performing the sort of curation required to robustly link data, will require time and resources. It is therefore critical to link data sharing to value, and define the value sharing approach to ensure data-producing departments are appropriately compensated for the effort required to ensure their data can be used by other departments and third parties.

Participants

• Craig Suckling (Chair, Government Chief Data Officer during this time; he has since left Government)
• Alison Park (ESRC)
• Ben Goldacre (University of Oxford)
• Emma Gordon (Administrative Data Research UK)
• George Marcotte (Lloyds of London)
• Helen Margetts (University of Oxford)
• Iain Mackay (Faculty AI)
• Jules Marshall (BBC)
• Nicola Byrne (National Data Guardian)
• Nigel Green (ONS)
• Rohit Dhawan (Lloyds Banking Group)
• Susan Bowen (Digital Catapult)

Observers

• Baroness Jones of Whitchurch (Parliamentary Under-Secretary of State at DSIT)
• Officials from DSIT and GO-Science