Cracking Government Data Sharing: Practical Priorities for Progress

Effective government data sharing could make public services easier to use and help civil servants make better decisions, but realising these benefits means overcoming some persistent obstacles.

This was a key message from Aker Systems roundtable held at Global Government Forum’s Public Service Data AI conference in London. The discussion brought together senior civil servants to explore what effective government data sharing should look like and the practical steps needed to achieve it.

“Simple and seamless”

Participants at the roundtable described good data sharing as “simple” and “seamless”, allowing people to use services without having to navigate the boundaries between organisations.

For citizens, this means avoiding repeated requests for the same information. For public servants, the ambition is to get “the right data to the right person” quickly enough to improve an outcome. Information should be easy to find and accessible to those authorised to use it, while remaining protected from unapproved access.

Participants also discussed the benefits of combining different sources of information. For example, linking weather and health data could help services understand how weather affects health and give patients information to manage conditions such as asthma. In a separate example, combining procurement and postcode data could also help to establish where government suppliers are creating jobs.

Changing the culture

Several participants identified culture as a greater barrier to data sharing than technology. They highlighted the tendency to treat data as belonging to a department, rather than considering how it could help other organisations deliver public services.

“Do we actually see [ourselves] as one government?” one participant asked.

They described recruiting senior champions to challenge this mindset and encourage questions about data in business cases, including how programmes would monitor their results.

Attendees also flagged conflicting messages that civil servants receive, with security training emphasising the risks of sharing data, while being required in their day-to-day work to collaborate and share information.

However, participants were clear that just encouraging people to share data would not remove their fear of being held accountable for a mistake. They stressed that people need processes and mechanisms to help them assess data sharing requests and make decisions.

A framework for decisions

Attendees discussed how a shared framework and vocabulary could help organisations act faster and more decisively when receiving a data sharing request.

One practical example is the National Security Digital Centre’s Data Sharing Risk Framework which aims to help make decisions more systematic. It separates risks into three categories: impact, compliance and equity, helping teams identify concerns and appropriate controls.

The framework covers the data, the user or recipient, the environment and the purpose. This allows teams to consider who will access information, where they will use it and what they intend to do.

As one attendee noted, sensitivity labels for individual datasets in isolation are inadequate. “It always depends on the context,” they said.

Making sharing viable

A further issue raised by roundtable attendees was that sharing data takes staff time and costs money. Incentives are often not aligned when the organisation doing the work to share data may bear the cost while another receives the benefit.

Some public bodies also need to recover costs through charging, which can make it difficult to provide data to government partners.

Participants suggested a “common business-model framework” across government bodies to make sharing easier.

There was also some frustration about departments ending up buying the same commercial datasets separately and attendees suggested more work to enable collective purchasing and to enable access across government, where licensing allows.

The point was further made that costs need to be understood before wider deployment. A pilot project may demonstrate the value of a dataset, but ongoing access can involve substantial charges.

In addition, participants called for a common way to measure data’s value. One questioned whether government treats data simply as an output of a system, rather than an asset that can be used.

Finding and using information

Even where sharing is possible, the roundtable discussed how teams may not know what data exists or who to contact. One participant called for a central “marketplace” showing available sources and how to request access.

Finding information does not guarantee that another team can use it -participants described different definitions within a single programme and specialist formats that people outside the originating organisation may struggle to understand.

Access procedures can create another obstacle. Participants described often being asked for a detailed use case before being able to access data, limiting the exploration necessary to understand what problems data could really help to solve.

Attendees discussed the benefits of sharing more non-personal information to show results and bringing together lessons from pilots and deployments, providing an evidence base for others to build on.

Priorities for progress

The discussion identified several priority areas and actions that could accelerate data sharing in government:

• Start with the policy problem. Establish what data sharing should achieve and bring the relevant organisations together around that objective.
  ‍
• Create the culture. Use senior champions to challenge the idea that data belongs to individual departments and encourage greater sharing across government.
  ‍
• Give staff a consistent way to assess requests. Use risk frameworks to identify concerns and agree appropriate controls.
  ‍
• ‍Review incentives and business models. Consider the incentives for organisations to share data and explore a common business model framework to address the costs of data sharing.
  ‍‍
• Coordinate purchasing. Avoid departments paying separately for the same data.

• Make data easier to find. Show more clearly what data government holds, who controls it and how to request access.
  ‍
• ‍Use non-personal data to make progress. Identify opportunities to put data sharing into practice with non-personal data to learn what works and build evidence that can inform wider efforts

• Establish a common way to measure data’s value. Assess the benefits of using and sharing data so it is recognised as an asset, rather than simply an output of a system.
  ‍
  
• Pool lessons from pilots. Share what teams have learned about data costs, access and risks so other organisations can build on their experience.

‘The roundtable reinforced that government's biggest data-sharing challenges are no longer purely technical. Real progress comes from combining secure modern data platforms with the governance, culture and incentives needed to share information confidently and responsibly across organisational boundaries’.

 Tracey Woolley, Client Director from Aker Systems

‍