There has been barely a passing public mention from the government about the need for new or redeveloped IT systems to support whatever trading and immigration regime the UK has when it leaves the European Union in March 2019. One sentence in a Philip Hammond speech has been all there is so far – plus some vague mentions of “technology-based solutions” in this week’s customs union position paper.
And yet, an internal Cabinet Office report produced in 2015 and seen by Computer Weekly, suggests that it’s already too late to develop any new digital services needed by that date.
The report was produced by the Government Digital Service (GDS) to assess its digital exemplars programme – the 2013 plan to digitally transform 25 of the highest volume government transactions.
When the review of those projects was completed in June 2015, 20 services were available to the public, either as a finished product or as a working prototype (known as a “public beta” in GDS terminology). Some had been a bit of a disaster, such as rural payments and the original Universal Credit programme.
But the report found that the average time for a new digital service to reach public use – either fully live or in beta – was two years. The completed exemplars ranged from 1.2 years to three years in duration.
Most of these exemplars were tightly focused transactions – booking a prison visit; registering to vote; lasting power of attorney, for example. The more complex services – passport renewals, rural payments, Universal Credit – were excluded from that two-year average because they hadn’t been completed in time.
So the big lesson from government’s attempt to digitally redevelop its most important services was that you need two years to do so. And we don’t have two years until Brexit.
What’s more, the exemplars report said the average size of the development team for a new digital service was 43 full-time employees – the biggest of those studied needed 350 people.
According to a report earlier this year by the Treasury select committee, HM Revenue & Customs (HMRC) alone may have 24 IT systems that require changes for day one of Brexit. That implies as many as 1,000 digital experts needed if all those services need redeveloping, for just the one large department.
GDS director general Kevin Cunnington has already said that the civil service could need as many as 4,000 extra IT staff just to complete its existing digital workload – and that doesn’t include Brexit.
Where are all these extra digital experts going to come from, at that sort of scale, in time for March 2019?
And so far, we’ve mostly been talking about timescales and resources needed for relatively straightforward services – as rural payments and Universal Credit showed, the most complex systems need much longer and more people.
HMRC is already struggling with its new import and export system for freight handling – and that’s before we know what it will need to do post-Brexit.
The Home Office has been trying to replace its core immigration IT systems since 2003 and hasn’t done so yet – relying still on a 20-year-old application alongside a system developed in 2004 for a trial project.
Cunnington has said that GDS is working with departments to assess their digital Brexit needs – quite how they can make such an assessment when the details of what will happen are still unclear, is a whole other question.
But GDS’s own experience shows emphatically that whatever the conclusion of that assessment will be, it’s already too late to start.