The Digital Centre of Government

Data and Digital in the Rest of A New Government

To government, data and digital are increasingly intertwined. There are separate silos for the “the digital centre of government”, the “national data library”, the UKSA Assembly, and ”smarter data”, but they all need tying together:

Just as tiktok and instagram use data to drive you to watch more and buy things, government wants to use data analysis to change services according to the political priorities of the day, which now includes injecting people with desire altering drugs to increase economic growth, just as Our Future Health founder Sir John Bell CH outlined last year.

Web and apps

There is vast scope to use digital to do good – progress by HMRC since our 2022 paper on the paperwork of new parenthood has removed some of the barriers we outlined (some remain), but DWP still prosecutes people for the consequences of DWP’s own service design failings, and other parts of the state prosecute for information not disclosed to citizens at all. DWP now wants access to medical records to measure and justify their own policy positions (item 6.3.4). 

The “tell us once” service has long been constrained by shifting political priorities. Carers allowance was redesigned by DWP but the new processes didn’t account for consequences that could lead to prosecution. Did anyone go to prison because a 2011-2015 flagship changed only the easy parts of a service? Service design can ignore edge cases or complexity in order to meet a HMT business case or a Ministerial announcement that will move fast and break people. So a summary of our questions:

  1. What input will the review of the carers allowance announced by HMG have into the Panel? 
  2. Which other parts of partly digitised public services are prosecuting honest and blameless citizens for institutional failures of service design?
  3. Who in the hierarchy should bear responsibility for the failures that resulted in honest people receiving criminal records? Does the board feel no one should?
  4. What happens when a “tell us once” style service conflicts with the  primary legislation?

In northern Rail or elsewhere, a lawyer will have said the self-serving change was fine as it satisfied some “user need” – and like others we deal with, the consequences on people were never considered simply because the action is in line with “policy”. “🎼That’s not my department says Werner Von Braun🎶”

Data and UKSA Statistics Assembly

There are few limits on data analyses in Government, and fewer limits on what public services can do with digital dark patterns. Some civil servants believe they deserve more attention than citizens, and civil servants can be sent to endless meetings: the suggestion of “User Needs*” aren’t enough, promises have to be kept if they are to mean something. One of the first actions of this government was to tear up promises made about pandemic data.

Against this backdrop, the UK Statistics Authority are running a statistics assembly and asking for submissions about “user needs” (no asterisk). We said:

  1. Every project should be transparent
  2. Surveys and Admin Data are not similar
  3. A Data Preference Service, because data mining is the new junk calls

While ONS and the statistical system assume that statistical data for policy making and raw data being made for decisions are entirely separate functions, and they are in government, to the citizen the effect is indistinguishable.  The research paper advocating cuts in benefits can have a direct impact on their reduction, even if authors hide behind the comfort blanket of “policy” and “research”. 

DSIT’s role model

Government often assumes that its data is perfect and accurate, treating it as an “official truth” regardless of actual reality. In this context, data serves only to reinforce institutional fictions, perpetuating a narrative among civil servants. The consequences can be severe: if you’re just £1 over or under a limit you are a criminal. This mentality has its roots in the Home Office’s hostile environment towards migrants and continues to spread to other areas of citizen interaction with government. 

Data isn’t enough – an essential component of the system must be digital services and digital service design. Governments prosecute people for what they type into forms and apps, yet changes to those forms can occur as unpredictably as a developer’s whim on any given day. 

“Imitate Tesco” might be the personalisation vision from some within DSIT, but when the Tesco app decides it should recommend something to you, there’s no way to tell it not to – if the reason you regularly buy a product leaves your life, Tesco will keep reminding you they are gone indefinitely, with no way to tell it you no longer need mandatory and intrusive reminders to buy senior cat food, or baby food, or your ex-partners favourite treat. Your only option is to shop elsewhere. Various silos in DSIT see that as their role model.

Tesco will share data with Governments if required or encouraged (or Tesco gets a better deal on something as a result), and citizens have no choice. If you shop at Tesco, having a Clubcard is decreasingly a “choice” because of the price differences. Clubcard culture was brought into Government by the CDEI (as was –  while everyone welcomes Responsible Technology Adoption, it is the Irresponsible Technology Adoption that causes the problem; and all adopters believe, of course, that they are ‘being responsible’). 

The same way more facts turbocharge racism, smarter data risks turbocharging institutional intransigence.

Data silos in government affect each other

The “digital discovery”, the “National Data Library” and the “UKSA Assembly” are each silos with narrow remits. Government operates outside of them and will do whatever it wants with data and digital.

Institutions extend “counter fraud” activities irrespective of outcome. Failure does not deter as more intrusive searches are justified under the guise that fraud hasn’t been found; and finding fraud also justifies expanding the powers.  Ultimately, such activities can become overly focused on speculative pursuits rather than concrete results, devolving to little more than ghost hunting and unicorn farming.

When a hospital makes a decision that it’s not in the public interest for the hospital to try to recoup costs from someone nominally “chargeable” for that care, then the hospital will tell her, but the Department of Health in England wants to copy the data so it can later make a decision to reclaim the money anyway, even if the patient was told that they were not liable for it. The first that anyone would hear about it might be when the Home Office rejects them at a border decision because the data goes from one bit of Government to the other without any knowledge of the patient – because the Department of Health in England doesn’t have an easy way to talk to individual patients to talk about charging. 


That’s the sharp end of digital and data in government, but no one sees it as their responsibility. 

ONS say they want to demonstrate good practice, but instead the Integrated Data Service perpetuates the secrecy by hiding how data is used from those to whom ONS at one time felt they were accountable: the public. Parliament suggested they change it (paras 100/102), the new government may well decide to double down on secrecy.

The signs are not promising; the new data bill was laid in Parliament with a press embargo 9 hours later…

Various silos: