Putting government data on the map

In this blogpost, Special Adviser Gavin Freeguard looks back at what we’ve learned from our recent National Data Strategy policy project, in which we used a crowdsourcing exercise to map data-related organisations and initiatives across UK government.  

Trying to navigate who is responsible for data in government often means being cast adrift in an alphabet soup of initials, institutions and initiatives.

There’s DCMS (the Department for Digital, Culture, Media and Sport) which is responsible for the NDS (National Data Strategy). There’s the new CDDO (Central Digital and Data Office), which includes the DSA (Data Standards Authority), and the older GDS (Government Digital Service) at CO (Cabinet Office). There are the various arms of the UKSA (the UK Statistics Authority), BEIS (Department for Business, Energy and Industrial Strategy), NHSX and the other parts of the health system, all the different cross-government functions and professions (DDaT, analysis, KIM)… the list, and the letters, are never-ending. 

(And can be set surprisingly well to music, it turns out.)

Mapping the government data landscape

Our recent project to crowdsource the UK government data landscape aimed to help people inside and outside government understand the data-related organisations and initiatives across government. We’ve published a short report today, and you can take a look at the crowdsourced document itself here. A huge thank you to the menagerie of anonymous animals, inside and outside government, that helped us fill out the Google Doc to more than 100 pages.

Our report finds there are well over 100 bodies with some responsibility for ‘data’ across government. It argues that understanding these organisations, their initiatives and how they all fit together is essential – not only to help government transform its use of data (Mission 3 of the National Data Strategy), but to unlock the value of data across the economy (Mission 1) and society. We hope it’s of particular use as we approach a Spending Review and Budget, and given all the recent data-related activity and strategies.

I’m particularly pleased we were able to shine some light on the various advisory boards and especially cross-government networks and communities. These are often more informal than the other institutions in the document, but crucial for sharing knowledge and insight across government. (We didn’t include those which may be largely about or of the government but not actually part of it – unconferences like UKGovCamp or Open Data Camp, even events like Data Bites which I run for the Institute for Government.)

Here be dragons

Maps define the territory. In doing so, they force us to label things and set boundaries – which presented us with a number of challenges – and also highlight what we still don’t know.

One challenge we discuss in our report is the difficulty in finding details about some organisations and their initiatives. Some lack any public presence or description – we really shouldn’t have to piece together details from job adverts, obscure minutes or private information – while others appear out of date (GDS’s ‘About’ page points to governance details which haven't been updated in years). Publishing and maintaining information on the key data players is critical, otherwise how will anybody know what the landscape looks like and who to work with within it?

‘Data’ is a small word but it contains multitudes – everything from back office HR data, to national statistics, to the personal data used in public services, to cutting edge technology and techniques like artificial intelligence. A number of commenters asked where we wanted to draw the line as to what a ‘data-related organisation’ was – was it one that had some cross-government responsibility or any organisation that had a data controller? In 2021, every government organisation should be a data organisation, thinking about how it can best use data in its work. In that sense, our ‘well over 100’ is a massive underestimate, since one should expect every government body to have some data-related initiatives. But our exercise tried to focus on those with some influence across government or on particular sectors. While all organisations should have something to do with data, if everyone has responsibility, then nobody does.

A related point – raised by a number of contributors (including here) – is whether we need a taxonomy for thinking about different types of data institution in government. Some organisations have responsibility for using data in policy, others for supporting government’s data infrastructure, others still act as data repositories, and others have wider regulatory responsibilities for data. We still lack a common language for thinking and talking about this (though projects from the ODI and others are making progress in this area). How that relates to government data organisations could be a fruitful next step from this project.

This project – ‘simply’ mapping data-related organisations in government and what initiatives they have underway, though it is anything but simple – takes just one area where we lack apparently basic information about data in government (or at least, lack a basic repository for). There are others.

Our project looked at space, but not time – there are many previous government data reports and initiatives that look a lot like the ones happening now. Which ones succeeded? Which ones failed – and why? As the National Audit Office has shown, we don’t tend to learn lessons from the past. There are some useful resources out there – Jerry Fishenden’s website is one, and I previously started a list of key government data reports (if you’re not suffering crowdsourcing fatigue, please do add to it). But this might be another area ripe for a similar crowdsourcing exercise to the one we’ve just completed, at least – and would probably benefit from something bigger, like interviews with the pioneers and protagonists of previous reform attempts, to distill their wisdom.

As well as current and previous initiatives, there may be work to be done in understanding all the different frameworks, regulations and guidance pertaining to working with data in government. Finding all the places where government bodies publish updates about their work on data would also be useful.

Our project highlights projects bringing together data from across government – the ONS’s integrated data work, the Cabinet Office’s new Information and Data Exchange, projects like ADR UK and HDR UK. But understanding what data flows there are across government, and which departments are sharing what with whom (inside and outside government), remains a major blindspot. NHS Digital’s monthly ‘Data Uses Register’ and CDDO’s register of data sharing agreements under the Digital Economy Act give some detail, but much more is needed. 

And while we now have a list of major data-focused organisations in government, we still don’t really understand what each of them looks like internally – how are they organised to make the most of data? Are there particular posts (Chief Data Officer, Chief Knowledge Officer, Chief Analyst), particular teams, particular ways of bringing the different data professions within a department together – and ensuring they are represented at the most senior levels of that department – that are more effective in encouraging better use of data than others?

Mapping an evolving landscape

Perhaps the biggest question is how to ensure a resource like this is of lasting value (a point discussed by David Durant here). We know our list is incomplete; we know that new initiatives (and organisations) are springing up all the time; we know that people have found this project extremely useful in navigating data in government and all of its complexities. How should something like this be taken forward – as something that continues to be updated, used and useful, rather than gathering digital dust – and who should do it? What other information would be useful to people inside and outside government about data in government?

No project like this can ever be truly complete. A perfect map of data-related organisations and initiatives in government is impossible. But we hope ours has done a helpful job of setting out what we do know and helping set a course for further exploration. If you have ideas for our next voyage, please get in touch with me or the Public Policy team.