Rewiring Whitehall: The next steps in digital government

Has technology really changed how government and public services work – and if not, why not?

Rich Corbridge: It’s the transformation bit, the cultural bit, the moving of the massive ships that are the departments that make up government – that’s the difficult bit. The tech has got easier to get right. The legacy IT hasn’t got easier to drag along with us. Big, old organisations drag the legacy behind them, and that makes transformation more difficult.

Whether it’s the data in there, the tin that’s there, the age of it – once it just works, then building the business case to modernise it so that transformation becomes quicker and cheaper, is quite difficult because [the legacy IT] is bomb proof. It works – it does what it’s supposed to do. So, [people think] let’s leave it for the next 20, 30, 40 years, doing what it’s supposed to do. It’s a difficult business case to get through.

Gina Gill: We give ourselves a hard time sometimes about progress. We could and should celebrate success more than perhaps we do. I came from financial services, and they went from branch to phone to online banking, and they would have called that transformation. I would have called it transformation when I was there. But I look back now and think, was it really? We were delivering the same thing, just through different channels.

Where we need to be – where banks have realised they need to be – is to think about the whole service journey, from a user perspective. Think about how you transform the whole thing, and put technology as a secondary consideration. As Rich said, the tech should be the easy pitch. The hard bit is, what is the user need and how do you make this service meet that need – not just add a bit of tech to what you’re already doing?

Craig Suckling: A lot of what needs to shift is the cultural aspect. What does it mean to have a digital mindset? It’s not focusing on building shiny new technology, it’s focusing on what are our user needs, and working from that to identify the best solutions – technical as well as looking at policy and cultural things that can be used to fix and create improvement.

It also requires broadening the aperture to think about not just the role a particular department plays, but about the services that go across departments. As a citizen or business, I don’t care which department you have to go to. I have a particular need, and if that stretches across departments, the solutions for that should do so, right?

We have to be careful, especially in the world of artificial intelligence (AI) today, that we’re not building technology for technology’s sake – that we’re building it with a mindset for longevity so we don’t, in 10 years’ time, have a new legacy that we’re trying to rip up again.

Data plays a very big role as well – data is still the biggest impediment to cross-governmental work, data is still the biggest impediment to collaboration, to productivity across departments. We do need to have a focus to open access to data to help drive that transformation.

These are all good things, but you’re not the first generation of digital government leaders to identify these problems. What’s going to be different this time?

CS: We’re at a point now in which we’re really seeing exponential acceleration of technology in general, and the rate of change and the opportunity that brings is very different to anything we’ve seen in Parliament. If we look more globally around the macroeconomic or geopolitical and environmental space, there’s a lot of change happening that requires us to be more proactive and to understand how to manage that change – and technology has to play a big role.

So, there is an opportunity and a timing and a criticality around the pace of change. We’re also seeing more of an expectation for government to hit the same bar as we see in the private sector, and that will propel us forward as well.

GG: What’s different now, in the six or seven years I’ve been here, is we have a much stronger digital capability than when I arrived in government. There is now upwards of 20,000 digital and data professionals in central government alone. That’s a huge leap forward from where we were even a few years ago. The conversation is happening at a different level than before.

There’s a different mindset developing outside of digital as well, in terms of that focus on users, which is important because otherwise you can become an echo chamber of digital people talking to each other. If you look at DWP, you have Universal Credit, which is not just about the technology – it’s about the service, it’s about the way it’s organised, it’s about the roles, it’s about how teams work together. The concept of service ownership is starting to take hold.

And then the last thing I’ll say is that I met the new secretary of state in DSIT the day after the election, and he listed one of his three priorities as modern digital government, creating better user experiences and giving people their time back. That is different language than I’ve heard since I’ve been here.

RC: Twenty years ago, we were the experts. People brought us into the room to tell them about technology – you were a translator. You go into a room now with a business problem, and you’re no longer the translator, because what we do is being consumerised to such a degree.

Also, the risk has changed. At DWP, there has been a risk around how and when should we do digital. That risk is turned on its head now – what if we don’t do digital? For example, we can’t keep on recruiting people to work in Job Centres because we’ve recruited everybody in this country who wants to do that job. We can’t keep growing and growing in that way, so digital solutions have to be the answer to getting that right. The appetite is there.

We hear a lot of talk about ‘innovation’ in government. It’s a much over-used word that in Whitehall can mean ‘get rid of legacy IT’, ‘spend more with startups’, ‘use tech to cut costs’ and many other interpretations. What should innovation really mean across all of digital government?

RC: The innovations we’ve seen be successful are as much about how the relationship with our business works and putting systems at the centre of what we do. That digital mindset is the lifeblood of facilitating the bright idea, the new innovation.

That’s been important in the past couple of years – to be open to innovation, instead of only the big [IT suppliers] coming in, and only when you’ve proven yourself for 10 years beforehand do you get in. We need to find ways to be more open to the new idea, and then find ways to bring our business with us and how we put citizens at the centre of what we need to do.

GG: If I had a magic wand, data is the thing that affects mindset, the thing that I would fix, because everything else follows. I was talking at an event [recently] and someone asked the question: “Technology is changing all the time, so how do you avoid a situation where you’re solving a big, complex problem and it’s going to take three years during which you’re not delivering technology, which immediately becomes legacy at the time you deliver it?” My answer was that it’s not about the technology and that we need to stop focusing on the technology.

If we think about service delivery, and if we then think about how we create technology in a way that allows us to be continually flexible and continually responsive, that’s where we need to get to. Administrations change and policies change, but our ability to react to that change is too constrained at the moment and that’s a mindset, organisation, cultural, behavioural set of things that underpin being an innovative organisation.

CS: I worry that innovation starts to become a word that has a bad reputation. We’ve all been there where someone pulls out a fancy proof of concept and an executive buys into it wholeheartedly and says, “I’ll have that tomorrow, thank you”.

Innovation needs to be something that addresses forward system problems, and that requires a mindset. It requires diversity of thinking, it requires people who are technologists and data specialists coming together with people in legal, in ethics, in policy and research. So, you’re thinking in a diverse full spectrum around what it takes to actually solve a really meaningful problem, with longevity. That’s one thing that’s important to get right with innovation.

The second thing is, we have to be careful not to create bottlenecks where central teams own innovation. Probably 10 years ago, there was that era of the guys who rocked up in jeans with flat whites and were “the innovators”. We need innovation to be democratised, and to have that, we need autonomy. That has to be something that we empower all teams to grasp. It comes back to mindset. We expect everyone to do this, not a central team to be the innovators, but to have autonomy with uniformity through standards.

GG: We go and see how different organisations work, and we spent some time with Octopus Energy recently. The thing that I left with was the mindset and the empowerment – the autonomy for people to be entirely responsible for, in their case, customer happiness. Within those boundaries they can do a lot and make decisions on their own, in a way that we would find slightly scary.

The story that really struck me was someone in marketing had a conversation with someone in customer services who said, “I had this awful day, 10 people complained about the on-hold music.” So this person from marketing said, “Let’s fix that problem.” So, apparently now if you are on hold, they will play the number one song from when you were 14 years old, so by the time you get to the customer service agent, you’ve got happy nostalgia feelings. It’s a small, simple idea because people are so focused on the customer and they think creatively about how to meet the need.

CS: I was with Gina that day – the next problem they had was people saying, “You answered too quickly. I was enjoying the music.”

We have heard some big claims from politicians about the prospects for AI. Former deputy prime minister Oliver Dowden said in 2023, ‘AI is the closest thing you have to a silver bullet’ for improving productivity in government. We all know that’s not realistic – but what is the reality of AI in government and what are the early lessons learned?

RC: What we’ve done is taken solutions to problems where AI can help. Where we’ve been successful is chipping away at that big productivity [savings] number. We are somewhat set up, culturally, to go and buy a solution that will fix the whole thing. Actually, what we’re seeing is innovations through the use of AI will chip away and take minutes out of the hour – not remove the hour.

Our learning so far is three key things. One, don’t try to find the end-game solution that fixes the whole lot. Build acceptance that 10 solutions that save a few minutes or make us slightly more accurate, bundled together is going to make us a better department.

Two, if you’ve got a load of crap data, you’re going to struggle to make the AI much better than what you’ve got now, and probably not better.

Three, there is still the polarisation of ideas and views of what AI will do, ranging from, “It’s the march of the robots”, to “We can all go home and put our feet up”. We’ll only find out the truth through testing and learning and seeing how it works and seeing what citizens think of it, and seeing how it can free up humans to do the job that humans are better at. What I’m obsessed with is that AI is a tool in our bag of goodies to help solve problems. Sometimes, the conversation becomes so obsessed [with AI] that it becomes the only tool that anybody’s interested in. It’s important that we accept it as one of the things we can do, not the thing we can do.

CS: We’re going through the classic hype cycle with AI, where we’ve had the heady peak of inflated expectations. Across government, we have one of the most valuable and under-utilised assets in data. There’s so much untapped value in our data, but we have to use it in a very trusted, ethical and secure way, and we need to make sure that it is in close proximity for AI as well. The more we can unlock access to data, the more we can deliver on innovation, the more we can help to drive AI – not only across the public sector, but also into the economy.

GG: Let’s focus on the problem we’re trying to solve. I remember getting an email from someone saying, “Gina, I think that we should do AI-powered prison visits.” What does that even mean? This [sort of thing] was constant. Now, I like to see it as an opportunity.

I spent years trying to have conversations with policy colleagues about digital and working more closely, then suddenly we’re the most popular people and everyone wants to understand AI. So, everyone gets a five-minute conversation on AI and then a big, much longer conversation on how we work collectively, together.

We also need to be clear about where our red lines are. We need to be careful about the equality of data, but also amplifying bias – in a government setting, that is particularly important. There’s a lot of things that we need to manage, as well as the opportunity. I’m a little bit cynical [about attitudes towards AI], but the opportunity will grow.

If there is one thing you could change to make government’s digital capabilities better, what would it be?

GG: The thing that I would change is to incentivise people. [For example] we need to incentivise people to collaborate. I came from financial services, where behaviour is driven by incentives – we don’t think about that creatively enough. What if you encourage departments to put together joint funding bids to fix an end-to-end service? What if we said we’re going to prioritise that over individual projects, particularly if it aligns to missions? Missions give us an opportunity to drive some of that thinking.

I don’t know if I’ll ever convince HM Treasury colleagues of this, but what if we incentivise people to save money? [Currently] if you save money, you’ve got to give it all back, but what if you get to keep 50% of it to reinvest? Or if you have gain-share agreements where I say [to a department], “I really need you to fix that problem over there, because it’s causing me a problem over here”? How do I incentivise someone else to solve that problem, when I’m going to save money over here? You can share some of the savings. How do you create that incentive structure?

CS: A mission-based government helps go a long way in setting up the infrastructures for collaboration. Doubling down around incentives, collaboration in government needs to be a place where value is exchanged, but not everybody will benefit in the same way. For example, if I’m sharing data from DWP to HMRC. HMRC might benefit more from that than DWP.

How do we ensure there’s incentivisation for value to be unequal, but it’s for the greater good of what we’re trying to solve. You need incentives to make value exchange an okay thing even if it’s not equal, and to incentivise people to think about that differently.

RC: In our organisation recently, we had a couple of big things that have gone well, and when we sat down afterwards and discussed the lessons learned, the common word was collaboration. We are disincentivised currently as a government. Citizens see government, but they don’t see that we are different departments, and trying to remove that disincentive is difficult.

Data sharing is a prime example. HMRC, as a made-up example, would benefit greatly from DWP passing them that data. DWP pays to make the system run through all the checks and measures so we can pass HMRC that data so they can make savings in their budget.

But we have to pay – our budget is decreased to save money over there, so there isn’t an incentive to do that and it’s not easy to put in place. There isn’t somebody going to sort that out tomorrow, because it’s really hard to get right. We know, as DWP, if we shared bidirectionally with the NHS and Department of Health better data, then the journey for citizens going through [healthcare] would be a lot smoother and a lot less painful.