Dear Mr Burnham,
Normally a letter this long takes me weeks. This one has taken hours, which is the measure of how much the announcements worry me, the plan to "scrap the digital ID policy," and to break up DSIT's mandate across departments.
I'm a former civil servant. I worked at GDS from 2017 to 2022, across Government-as-a-Platform and the early mobilisation of the programme now delivering One Login and the government Wallet. I watched the Data Infrastructure programme die, and I watched GOV.UK struggle to hold departments together through Brexit and COVID. I was a first-hand witness, or an involved party, in most of what follows. That's the main reason I'm writing: I have seen the reality of the gaps that your move will either create, or entrench.
Let me start with the misunderstanding underneath all of it. The government has never been building a "digital ID", because a digital ID can't be built. It already exists. You have one. I have one. Everyone digitally active has one, because every time we do something online we leave a footprint that other parties collect and check. What the government is building – and what the private sector is building through OfDIA's DVS trust framework – is a way to make that checking auditable and governable, and to give a supplementary footprint to the people who are shut out of digital, so they can reach services the same way you and I can.
The "mandatory ID" announcement was, I'll grant you, a very bad call. It announced a new unpopular policy out of the blue, attached it to something that was being built for other purposes, got the description wrong, and this toxic mix is the reason why we're now in this mess. No wonder people were spooked.
But the thing they're frightened of is a central database of unique identifiers, one store that would let whoever holds the key assemble the whole of a person. That fear is legitimate, but it's also not what's being built, and the conversation worth having is about exactly that: whether this infrastructure creates such a store, and why it would be dangerous if it did. "Digital ID", the phrase, is a distraction from the question that actually matters.
What's being built is more important than the phrase suggests. Every organisation is constantly checking whether your data, or mine, is valid, and so are government departments, when DWP needs to confirm something with HMRC or the Home Office. First they have to be sure it's really you at the other end. That is what One Login does. Then they have to establish that something about you is true, and that forces a chain of questions that's much harder to describe in a business case for Treasury or in a parliament hearing: two departments can hold the same person and describe them entirely differently, so they have to agree what a thing means; they have to agree who is allowed to be the authority on it; there has to be a way to issue something and withdraw it when it stops being true; every check has to leave a trail, so that afterwards you can ask who looked at what, and when, and why; and there has to be a lawful route for the information to travel at all.
The Wallet work was forcing all of those questions into the open. So when you scrap "digital ID plans", understand what goes with them. You are not just cancelling the Wallet. You are cancelling the whole of the infrastructure beneath it. (I've written more on why identity is infrastructure, not a product).
That infrastructure is also what would keep us in step with the EU, and this is where I'd beg you to pay particular attention. By 2027, every EU resident will have been offered a EUDI Wallet. Under the eIDAS 2.0 rules, any service provider required, by law or by contract, to use strong authentication for online identification will have to accept those wallets. UK businesses trading with EU customers will have to conform, whatever you decide to do about Digital ID.
Think of the years we spent defending GDPR adequacy. Adequacy is at least a status, something we hold, that can in the worst case be renegotiated. eIDAS interoperability is a capability, and the only way to have it is to build it. We were building it. Stop now and there's nothing left to renegotiate: we simply won't have it, and it's UK businesses who'll pay for its absence.
Don't let anyone tell you the savings will ease the cost of living. Some estimate that the running cost of foundational identity infrastructure is under a tenth of a percent of GDP, with varying figures on the economic value it creates – but even the most conservative point to multiples of that. This is not where the money is.
People say that a digital ID will create a new Windrush scandal. I say that people caught in the hostile environment aftermath were exactly the people who needed to prove their eligibility, and had no means to do it. So they ended up bearing the human cost of this lack of infrastructure. Let me also tell you a few other examples I witnessed first-hand.
Between 2019 and 2020 the government was scrambling through Brexit and COVID at once. Under Theresa May, the government cut Government-as-a-Platform and Data Infrastructure, and scattered data policy across departments with no real digital capability; this had a very direct bearing on what happened next.
Three failures from that time were, underneath, the same failure. Settled status gave more than four million people a status that no one had a working way to check – not councils, not DWP, not employers. The shielding data existed the whole time, sitting in GP records and in secondary care, with no way to join the two; so it was pulled together by hand, sent to councils as spreadsheets they couldn't reconcile against their own records, and each council improvised its own way of reading them. Identity checks on Universal Credit were relaxed because Verify couldn't get people through, then reimposed after the fact. People who had proved who they were correctly were told they hadn't, and left to prove it again.
Every time, the people who paid were the ones least able to prove who they are. The civil service held it together by hand, and did an extraordinary job of it. What was missing was never effort or competence. It was the standing to make departments answer the same question the same way – the standing GDS had been created to hold, and which we lost when data policy was broken up and parcelled out.
I'm not an isolated voice in pointing at this problem. The UK still sits near the top of the OECD's Digital Government Index on everything a citizen can see, while its two weakest scores are precisely the dimensions that measure what lies underneath: Government as a Platform, and Open by Default. On the platform dimension, which covers shared infrastructure including identity, we fell from 1st in 2019 to 7th in 2023. We were first, because we invented it. Now we trail behind. On support for data re-use we score 0.28 against an OECD average of 0.40 – many countries are now progressing towards the Once Only mode; we have to share spreadsheets.
The standing that GDS represented does not live in code, or in policy documents. It lives in agreements, in standards, in cross-departmental relationships that take years to build and one machinery of government to lose. DSIT put data policy and digital delivery back under one roof for the first time since 2017. Splitting them apart again won't move that capability somewhere else. It will end it. Again.
So here is what I'm asking. Before anything is scrapped, talk to the people who can actually tell you what the landscape looks like: the identity platform owners across the Home Office, DVLA and NHS England, the One Login programme, OfDIA. Talk to the watchdogs and the industry too, but afterwards, once you know what's really there. The watchdogs, because their worries are real but hypothetical, and have to be weighed against what's actually happening. The industry, because they have campaigned against the Wallet since day one, hard enough to drive GDS out of working in the open. There is genuine work to do in reconciling the trust framework with the Wallet, but their opposition risks losing sight of the important point.
When the industry tells you to kill the Wallet, they're asking you to get rid of the bus, while you end up dismantling the whole road. Keep the road. And make sure there's still public transport for the people who can't afford a car. And whatever becomes of the department, keep data policy and digital delivery under one roof.
Yours sincerely,
Paola Roccuzzo