David Eaves
My name’s David Eaves. I’m your Chair today, and I’m very excited to be here. I – before we just get going, I just – I do want to acknowledge, like, here we are in Chatham House on what may be, as my friend just put it, the most consequential week in geopolitics in the last 40 to 50 years. And I know a fair number of people are maybe feeling nervous, the energy in the building is a little bit intense at times, and yet, here we are, we’re going to talk about “digital public infrastructure” as though nothing else is going on in the world, and I just don’t want to pretend like that’s the case. I think there are ways in which this topic does touch very clearly on some of the broader events that are going on in the world, and I would welcome people to try to engage with us on that and just acknowledge that that’s going on.
I am unbelievably excited to be here with this panel that we have with us. So, I have with us John Lloyd, from – joining us from Digital Public – yes, we can go ahead – and David Lawrence, on the far end, from the Centre for British Progress, and Rowan Wilkinson from Chatham House, who’s joining us, as well. And what we’re going to do is just, kind of, explore a little bit about what’s going on in the DPI space more generally, and then what’s happening here in the UK specifically on this front. And so maybe just to open up, I’m just going to provide maybe a little, like, context for people in the room. And fir – maybe first, just how many people, when we say the words ‘digital public infrastructure’ or ‘DPI’ raise your hand if you have any idea about what we’re talking about. This is, like, a very interesting – so it’s like 50%, for those online, about 50% in the room.
And so, for those for whom this is new, digital public infrastructure really is a reference to a shift from the way a lot of technology has been built, which has really, kind of, focused on silos. So, when you have a government service, you really focus on building it – you’ve recreated the old paper bureaucracy in a, like, in a hardened silo way, and instead of starting to think about how do you decompose what are maybe core elements of those services into horizontal, kind of, platforms? And central among those are identity, since anytime you want to use a government service, or many services across society, you have to be able to assert who you are. Payments, since the exchange of value is pretty central to both the functioning of government and the functioning of the economy. So, how can I provide money from one person to another person? And then the movement of data. So, if we have data that’s been collected about people, how can we move that data more easily around?
And the thing that makes, I think, digital public infrastructure so interesting is not just that it’s digital, we have that, and not that it’s infrastructure, we actually have a lot of that, but it’s the public element that makes it interesting. What is the public governance that ensures these systems are run in a public interest as opposed to in purely private or proprietary interest? So, with those, kind of, opening comments, maybe I’d first just like to ask each of our panellists, how did you – how do you come into this conversation of digital public infrastructure? So, maybe we start, like, in the middle. John, why don’t you talk to us a little bit about you – how you come into this space.
John Lloyd
Sure. So, I might just borrow something from you, and who’s familiar with what a digital public good is? Okay, we’re seeing, like, maybe a tenth of the room here in person. So, digital public goods are open solutions, open-source. They can be software, they can be datasets, they can be content or AI models, that adhere to the digital public good standards. So, it’s a set of nine indicators, things like privacy first, do no harm by design, attaining the Sustainable Development Goals, and because of their open-source nature, it means that they can be reused, adopted and adapted in many different contexts.
They aren’t necessarily just for digital public infrastructure, but the way that I came into the DPI space was a lot of countries are implementing DPI at the moment, and they’re using digital public goods to do that. And so, along with Co-Develop, who’s one of the funders in the DPI space, and a whole bunch of supporting organisations, we launched a campaign called “50-in-5.” And the goal of that is for 50 countries in five years to be working together to implement safe, inclusive and interoperable DPI. We believe that one of the best ways to do that, for countries to work together and to co-operate and collaborate, is by using digital public goods for their DPI. So, that’s, kind of, how I got into the space.
David Eaves
And Rowan, you’ve just written a report here at Chatham House, one of the reasons we’re gathered here. Want to talk to us a little bit about your research journey here, what you have found interesting in this space, and what you want to share with the world about what you’re learning.
Rowan Wilkinson
I think, first, the key starting point is that technology is an infrastructure and is increasingly being seen by governments as a core critical infrastructure that helps to enable citizens’ access to services. Now, digitalisation of government services, I believe, is an inevitability in modern – in mod – for modern states. And the, kind of, core question for us here is how can we improve adoption that is, as John said, kind of, open, interoperable, with public interest in mind, but how can those digital tools also help to enhance different geopolitical advantages, as well? So, whether that’s thinking about digital sovereignty and how the tools you pick can help to navigate foreign tech providers, or whether that’s thinking about security and resilience, you know. What tools are there that help to protect citizens, for example, in times of war? These tools are not just, kind of, simple public modernisation tools. They are a lot more than that, and they have large, kind of, geopolitical ramifications, too.
David Eaves
Yeah, I love this opening statement you make where you say “technology is infrastructure,” and I think part of that’s understood, but there’s actually even more, like, if you work for a development bank, I think an even more subversive statement, frankly, is “software is infrastructure,” right? Most people really think of infrastructure as things that are probably too heavy, but if you were to drop on your toe, it would hurt a great deal, but, you know, bridges, even in the digital space, cables, you know, telecommunications. But this notion that actually the code itself is infrastructure poses a challenge because the way we fund that, the way we develop it, the way we maintain it, looks very different than the way we do physical infrastructure, and the threats and the opportunities around sovereignty, cybersecurity, are also very, very different than they might be for physical infrastructure. And so, it challenges us for how we’re thinking about this space.
It also creates new models for thinking about how we’re going to digitise government, which brings us to our last panellist, David, who’s really thinking about in the UK context, you know, what is the state of digitisation and what are we going to do? And how, you know, how do you look at DPI and what’s going on in the UK? Like, what’s the conversation you’re thinking about in terms of how we could be digitising more services here?
David Lawrence
Great, thanks David, and I’m not a DPI expert, so I do slightly worry that I would have been in the 50% raising my hand or not raising my hand for not really knowing what it is. But I do know a fair amount about the UK state and its ability and sometimes, often, inability, to get new projects going, including digital projects, and I think it’s a really live question for this government. A lot of you will have seen Labour’s initial big announcement on digital ID, which was supposed to be this huge, you know, authentication system that was mandatory and would apply to all parts of interaction with government, and then immediately got huge backlash for this and then have since backtracked. And I think it’s, like, a perfect lesson in how not to do DPI in terms of the announcement and the comms and the subsequent lack of public buy-in.
But the UK has often done it well, as well, and we were just talking earlier about gov.uk, which if you’re in the UK and you’ve had to interact with any part of government, chances are you’ve had to log in at some point to gov.uk to sort out your taxes or childcare or maybe not. Cer – yeah, I’m trying to remember which things are covered and which things aren’t, but huge parts of government are on that, and that’s generally seen as quite a successful initiative.
And I think this government’s also aware that this current moment is politically interesting in that there is huge public dissatisfaction with the way public services are run. People aren’t happy with the NHS, with education systems, with local councils, the general feeling that they’re not being listened to, that stuff doesn’t work, that quality has fallen compared to the past. And I think that’s felt across lots of European countries and lots of the Western world. But then, in addition to that, there is basically zero willingness to pay more for public services. Tax rates are about as high as the government can politically get away with, potentially even higher than they really want them to be. So, I don’t think there’s buy-in for, you know, Blair-era style, big public extra money spent on this stuff, but there is huge dissatisfaction with it.
And so I think technology is potentially the solution to that political problem, because if you can actually make these services work better for people, if people feel more listened to, if people feel like they are getting something back when they’re using technology, it’s easy, seamless to log in, quick to authenticate, you’re not on hold on the phone to ages to sort out a parking fine or whatever it might be, then you get more acceptance among the public for – and more satisfaction in public services. But also, you’re potentially saving huge amounts of money, as well, for the taxpayer, and you can put that into other things, and maybe you can put that into more human interfaces where and when you want them. So, you know, ideally, those who want to use more technology in their interaction with the NHS can do that, and that frees up resources for those who are unable to or who don’t want to.
So, I think there’s this really interesting political moment. I don’t think the government’s managed to really harness that, and I think they made a mistake in making digital ID the thing that they were going to do, rather than, say, trying to digitise the NHS, and I know they’re still interested in that. There’s some really interesting stuff happening in DSIT, the Department for Science, Innovation and Technology, and Labour’s hired someone called Emily Middleton, who’s fantastic, to lead on this. But I feel like they need the, sort of, the one big thing like gov.uk, and that ambition just isn’t there, and I worry that they’ve been spooked by what happened with digital ID.
David Eaves
Yeah. I think what’s so interesting is there are some other countries that have had, like, the one big thing that has been a trigger on the digital public infrastructure front that’s allowed them to proceed. So, maybe, you know, kind of, famously, India has its Aadhaar and UPI. So, Aadhaar is a biometric identity system that allows Indians to sign into public services, into banks, and UPI is a payment rail that allows payments to happen between, kind of, any two devices. And Estonia famously has X-Road, which allows data movement across.
I guess for me, like, the risk here is twofold. One is, whatever you build, if you’re relying on someone else, are you really handing the, kind of, keys to the functioning of your society to whoever is running that? And what – how do we manage that risk? And then, even if you aren’t, if you’re doing it yourself, there’s maybe great expense to, kind of, recreate – you know, for the UK Government to recreate all this functionality itself and not copy what others have done, would also take great expense when we don’t have a lot of money and maybe we don’t have all the technical expertise. How do we square these dilemmas? Rowan, do you have any thoughts?
Rowan Wilkinson
Yeah, I mean, there’s lots of different, kind of, models that governments can take to procuring and implementing DPI approach to digitalisation. One is build it yourself, like, as we’ve seen with Gov UK One Login, which has been very successful, as David’s pointed out. The other is procurement, right? Are you going to procure a private vendor to integrate a system within the government service already? And the third is, are you going to utilise private actors? They might be, kind of, foreign private services, as well. The UK Government has a very mixed approach in terms of adopting private actors, procuring private actors, as well, for different services. And I think this really feeds into the, kind of, growing debate around what digital sovereignty is, what are the trade-offs? And I think it is a, sort of, impossible, wicked problem for a lot of governments.
Sovereignty can never be absolute when talking about technology. The tech ecosystem is far too mixed and involves some – far too many players, and far too many – a lot of the players are the two superpowers, as well, right? So, I think that an approach to sovereignty, especially for lower and middle income countries, needs to really balance what is the value, what is the cost effectiveness of perhaps procuring from a foreign provider compared to doing it ourselves? And, kind of, what is – yeah, is that trade-off one that we’re really willing to take?
And the final point I will make is just on sectors. Some sectors are more crucial and more, perhaps, vulnerable to certain security problems, and governments really need to focus on what their red lines are, for example, if it’s going to be in defence or in energy and, kind of, think very critically about which companies they might be procuring from for that.
David Eaves
John, how are you seeing – like, how is the DPGA helping governments square this problem of, kind of, sovereignty, reuse, expending too much on trying to rebuild everything themselves? Like, what’s the theory of change you have?
John Lloyd
Yeah, like any implementation of technology, I think governments will – countries will often say, like, “Well, our context is really unique,” and for the most part, that’s true. But what they all need is 95% there in the form of open solutions that already exist and are already being implemented in other countries. Sovereignty, like, it was an issue before our current geopolitical situation, but of course, it’s much more to the fore. And through the use of open technologies, what we’re seeing is that countries are able to maintain their sovereignty, having ownership and agency over the data that they’re collecting and controlling, while being able to implement solutions faster and cheaper than with a proprietary solution or by building it themselves.
For example, we’ve seen in the Philippines, they’ve been implementing MOSIP, which is a digital ID program, and Mojaloop, which is a DPG as well, for payments, and what they’ve seen is a dramatic amount of people have now entered the financial system, otherwise without access. But more crucially, they’ve – over a ten-year period, they’re seeing 95% cost reduction in the implementation of their open-source tools, compared to when they were locked into a proprietary vendor. This doesn’t even necessarily mean that they need to be doing the customisation themselves. For example, in Rwanda, they partnered with a local technology company to implement their payment switch called ‘WiredIn’ and that was a commitment by Rwanda to build out their local tech ecosystem. And now, because of the success of what they’ve done with their payment swish – switch, that company, WiredIn, is now providing services to other countries in the East Africa community, their neighbours. So, they’ve created a bit of an industry on the back of their DPI implementations, as well.
Of course, a really good example, as well, is Pix, which is Brazil’s payment switch. They had 170 million separate users, which is essentially every adult in Brazil in the last financial year, which is just incredible, and that means lower fees, or no fees at all. And more crucially, with all of these systems is it’s not excluding the private sector, rather, DPI is providing the rails for the private sector to plug in. And I was just thinking, like, on my way here, imagine if the UK had amazing DPI.
Like, I bought a house a couple of years ago, and anybody here who’s bought a house knows how much of a nightmare it is in this country. Like, the documents you have to put together, all the checks, all the farting around. I had to go to the bank for half a day because they couldn’t do the transaction over the phone, and then I had to wait in a line to send my own money to my Solicitor and got charged a fee for it. Like, imagine having all those documents available, you just send them, your verification, it’s all verified online, you send the money straightaway, and a process that right now takes months could take minutes. That’s not a fiction, that happens in other countries right now that have strong and effective DPI.
David Eaves
Yeah, one of the things I like to do is – I don’t know if there’s an Indian national in the room who has a DigiLocker on their phone. If they have, I’d love for you to raise your hand, if you were willing to. I think one of the things that’s coming out of these stories is it’s amazing how much of this is really a Global South story, and how unaware people in the Global North are about what is happening in the Global South and the revolution that’s taking place. So, I regular – I’m a Te – I’m faculty, I teach, I often have an Indian student in the class, I’ll get them to stand up and I’ll have them open their DigiLocker, which is a credential app that they have. And then they’ll stand up and I’ll be like, “Okay, show everybody what the credentials are,” and they’ll pull up and they’re like, “Here’s my university diploma, here’s my driver’s license,” and they’ll just swipe through.
And, like, most people – I mean, Europe’s still arguing over how they’re going to implement their wallet. Now I’m Canadian, we don’t even ha – we don’t even argue about what the wallet will look like, we just – it’s not even on the radar as a possibility at the moment. And in some American states, they’re bragging right now, like, there’s four states that I think now have a driver’s license that’s starting to move into a wallet. Argentina’s launched their driver’s license in a wallet in 2012. India’s had, you know, a couple of hundred million people with DigiWallet credentials now for five, six, seven years, and they’re – like, that number just keeps rising, rising, and it just hit 500 million. Like, this is a reality in the Global South, and I just don’t think people in the North are even aware.
And so, one of the questions I have is, like, David, is there – I don’t want to put you on the spot, and we didn’t talk about it upstairs, but, like, are you looking around for models outside the UK? Like, where are you taking inspiration from as you think about what the UK be – could be doing differently?
David Lawrence
Yeah, 100% we are, and we ought to be doing more of it, but we definitely are, and just on your example just now from India, my family’s half Chinese and my Chinese uncle, when he visited Britain, couldn’t believe how we pay for parking. Because he said that in a big city in China where he lives, he says, like, “I just use my watch. I use my watch to pay for parking,” ‘cause he – you know, it’s an Apple Watch that’s linked to WeChat or whatever, so he can do that very easily. Whereas we still use coins, or if you don’t have coins, sometimes the coin thing doesn’t work, there’s, like, a pay-by-phone option, where you pay your number and then you read out out loud in a public place your credit card number that obviously you have on you. So, it’s getting slightly better in the UK, but yeah, lots we could learn from the Global South and China.
I think there’s, like, a few pros and cons of being the UK right now. I have this, sort of, general theory, it’s not an original theory, but that now’s, like, a really bad time to be a medium-sized country when it comes to sovereignty. ‘Cause if you’re the US or China, you can try and do this – or even Brazil or India, you try and do this on a national scale, you say, you know, “We’ve got hundreds of millions of people, we’re going to roll this out nationally, we’re going to try and own as much of the supply chain as we can, and we’re going to get loads of data,” and that creates really nice feedback loops, so, like, this stuff just gets better really quickly. If you’re a really, really small country, you’re not going to expect to own the entire tech anyway. You’re, kind of, used to using other countries’ tech for pretty much everything you do, so you’ll probably have fewer qualms about everything being on your soil, and, you know, anything you manage to build on top of that is, kind of, a bonus.
I think for countries like the UK, France, Germany, Italy, maybe Australia, Canada, I think we’re in this, like, really awkward middle position where we, kind of, expect that we ought to be able to do this stuff ourselves, but we actually can’t, and we often don’t have the expertise and the ability to do that, and so I think that’s a real challenge. And I don’t really have an answer there, other than to say that I think the UK is not alone as, like, a medium-sized country in struggling with this. A few things I have observed where it has gone well in the UK, can I jump onto those things now?
David Eaves
Yes.
David Lawrence
Is this a good place? So, I think one thing is – and I think this applies to lots of different kinds of projects and – including DPI, but where you can, starting small and iterating and scaling. This is, like, so obvious to anyone who works in tech, it is really not obvious to the Treasury and the way that they price things and fund things. So, the Treasury’s cost-benefit analysis doesn’t look at, you know, potential exponential growth of things if they go well or potential killing of projects if they don’t go well. They want the fully costed from the start, you know, this is the cost-benefit analysis of this project. So, there’s not really room for someone who is high agency in the UK Government to say, “Hey, I’ve got this idea, I’m not really sure about it, I want to, sort of, start small, and if it goes well, then we can scale it to other departments or other bits of government.” There’s not so much room for that. I think that’s true in lots of civil services, but it’s definitely true in the UK.
And I think, secondly, relatedly, like, where you can taking things outside of the existing Civil Service infrastructure, at least in the building phase, can be super, super helpful. So, where this has gone well in the UK, and it’s not quite DPI, but I think that it’s examples of really good digital innovation, is in ARIA, which is our Advanced Research and Invention Agency, which basically, looked at the existing research funding model in the UK, wasn’t happy with it, set up something new that had completely different rules around hiring talent, around team structure, around sign-off. So, you’re exempt from all of those usual Civil Service processes, and I think that could be really important for certain kinds of DPI, that if you want to build something really well, actually, you’re going to need that freedom over talent.
Another example is the AI Safety Institute, which is – or now the AI Security Institute in the UK, which was the first of its kind, globally respected for expertise on AI safety and security, but exempt from Civil Service hiring practices. They don’t need to have the same salary bands. They have a lot of freedom over which areas they research. The government gave them money, appointed really good people at the top, and then just let them get on with it. And I think other bits of government that’s build – that are building DPI could learn from that.
And then, thirdly, I think being willing to kill projects that don’t work and take risks and acknowledge that some things might not work out. Again, I think that’s something government really struggles with. So, Darren Jones, who’s the Prime Minister’s Chief Secretary, who’s responsible for Civil Service reform, gave a speech earlier this week, which some of you may have seen, which was titled “Move Fast and Fix Things,” so a play on “move fast and break things.” Which I thought was, kind of, clever, like, I could see what he was getting at, but I also thought it, kind of, misses the point. Like, we don’t want to break things for the sake of breaking things, but we break things because that’s a natural consequence of moving quickly and of trying lots of different things, that lots of them are going to break, lots of things aren’t going to work, but hopefully from that process, you get the things that do work. And actually, if you go in with a move fast and fix things approach where you don’t expect things to break and you’re not willing to take those risks, then you’re unlikely to capture the upside either.
So, I think there’s a weird thing in the Civil Service where it’s almost, like, the opposite of entropy, where it’s really easy to set stuff up and really hard to kill it, and really hard to, like, allow bits of government to just, like, die off when they aren’t working. So, that’s, again, not really posing a solution here, but I think where the government is able to move fast and, like, is willing to break things and is willing to take the hit from Treasury officials or the media or whoever it might be for something that didn’t work out, I think that’s going to be really, really important.
David Eaves
Yeah, I love this comment about the financing. Diane Coyle at Cambridge University and I have written a report called “The Economics of Digital Public Infrastructure,” and part of this is recognition that actually governments are terrible at funding infrastructure, period. Like, they really want to find a killer use case that will justify the expense, and infrastructure, the problem with infrastructure is often you don’t have a dominant use case. You just have a long tail of use cases that collectively are very, very significant, but a single one is not large, and so, it’s hard to build a business case unless you’re willing to look at spillover effects or market shaping effects, and not only within individuals, the public sector, but also across the private sector. And so, we have a framework that we’re working with the Treasury on, helping them try to rethink that a little bit. We’re also working with the World Bank on that.
I guess, like, those types of nitty-gritty tales matter a lot, and I guess, like, you know, Rowan, what are you discovering are the, kind of, capabilities that others have that maybe the UK doesn’t have? Like, what did you learn in, like – as you looked at DPI systems? What struck you about what other people were doing right that you would encourage people, like, here in the UK or elsewhere, to do more of?
Rowan Wilkinson
Hmmm. I think I’ve heard so many excellent examples where countries are co-operating on this exact problem, on DPI, in a way that I find it hard to visualise the UK doing. So, for example, the Government of Belgium is currently working with Guinea to set up a digital birth registration process, right? And it’s, like David said, it’s very much starting small-scale, it started with local municipalities, working with hospital data initially and the Birth Registry Office, as was the model in Belgium. And then very quickly they worked out that actually that’s not the way that birth registry is often done in Guinea. There might be a religious leader that attends the birth. There might be, you know, a, kind of, Community Midwife that attends the birth, and so you need a, kind of, flexible, iterative, adaptive process in order to work out what the final product is.
And, you know, I think it’s a great case that has shown that that is possible and it does happen, and Guinea now is scaling up across multiple districts to help provide, ultimately, better access to public services from birth until death. And I, yeah, agree with David’s point of scalability. Starting small and then scaling up is, kind of, the crucial part that I think the government is slightly off the mark on at the moment.
David Eaves
Yeah, I think there’s an inclination on governments to believe that when something matters, you go straight to scale, and, in fact, one of the challenges of infrastructure – Mike Bracken and I wrote a piece and we was, like, cer – like, “Stacks don’t beget services, services beget stacks,” and if you don’t start focusing first on the value for people…
Rowan Wilkinson
Yeah.
David Eaves
…you have a supply side approach where you just build something, and then you learn that actually it doesn’t meet the demand and nothing goes anywhere. And so, you actually have to start with demand and meet it and then use that to increment and build your infrastructure out of.
Rowan Wilkinson
Yeah.
David Eaves
John, how about you? You’re working with governments all around the world, trying to get them to, kind of, work off of shared code base. What are the practices you’re seeing that are essential that you’d advocate, say, for the UK Government and for other middle powers to adopt?
John Lloyd
Yeah, interesting on middle powers, given Mark Carney a couple of days ago. I think the best practices are really – when you’re thinking about DPI, it’s not a technology problem, and I think that’s the first hurdle that a lot of people, kind of, stumble on. Like, we’re not talking about technology here. DPI is more of an approach. It’s a change in the way that you think. It’s looking at whole of government thinking. So, that requires changes to policy, your governance, in addition to that technical side. And the countries that we – that I interact with that are doing best, sort of, have issues, not even so much with funding as their top problem. It’s – their main issues are around capacity, and even then, it’s not even necessarily technical capacity that they have challenges with. It’s figuring out those more tricky issues around governance, around things, like, what we’re talking about with use cases, starting small.
I think a lot of the countries that I work with are not necessarily high trust, and so, in order to, kind of like, prove that the DPI is adding value, picking a use case that people are going to be most – like, making their lives convenient is really what you’re aiming to do here. And through that convenience, you build trust because people are, like, “Oh, this is handy to me.” Like, a digital driver’s license or something like that, which in the United States, they have REAL ID, you can show that at the airport when you’re going through instead of your passport or carrying your driver’s license around. That, kind of, thing, it makes life easier, it encourages adoption, and yeah, I think, like, that’s certainly what we’re seeing a lot of with those governments. And then – I can’t remember what I was going to say, so I’ll just stop talking.
David Eaves
The – it’s – I’m glad you brought up this situation of, kind of like, low trust context situations, and I think people often have countries in mind when you mention those words, but it turns out those countries include the UK, right? You know, it’s, like, here they wanted to launch a digital ID and there wasn’t the trust. Like, does that qualify as low trust? There wasn’t enough trust for them to do it, and so what would you have done differently, David, if you had been trying to launch a digital ID? How would you have gone ab – like, do you think it would have been a good idea to do, and if so, how might you have done it differently? Like, is it just there’s not enough trust here, we can’t do it, or, is it you’re, like, no, actually we just tap into it in a different way?
David Lawrence
Yeah, really good question. So, I think, in general, the UK has higher trust than a lot of, sort of, Western European countries around digital stuff in general and are, sort of, quite pro-tech, but you’re right, the digital ID thing went, like, spectacularly badly. A few things I would have done differently. Firstly, I wouldn’t have made it about catching illegal migrants. I think the government, sort of, did this weird, like, 4D chess thing, where they were, like, the public really care about migrants and they don’t really like digital ID or their data being tracked, so what if we made the, like, digital ID thing the, like, way to catch illegal migrants and crack down on, you know, small boat crossings?
And I can – yeah, I can see how they got there, but ultimately, what people heard was their worst fears about digital ID, which was that it was going to be the state overreaching and using it to catch you out, to, like, get people in trouble. Rather than the case that we actually should be making for digital ID, which is quite simply that this is going to make your life better. This is going to make your interactions with the NHS, with schools, with, whatever it might be, parking, this is going to make all of that a lot easier. Because you’ve now got this thing on your phone, like you have loads of things on your phone that you use to pay for things to get around, that’s going to make your life easier.
And I would have started with making it voluntary, small-scale, something that people can download on their phone and begin using for a host of government services. And then I think if it’s designed well, then people will naturally start recommending it to each other and using it more for more things, and you can gradually roll it out for more things, and if it’s not designed well, then you iterate and you design it again, you make it better. But instead, they led with the mandatory thing, and they led with the catching migrants thing, and I think those – that was a mistake.
I can understand why they rolled – rode back on it, given the failure, but I think it’s, kind of, a shame ‘cause I now worry that any form of digital ID is less likely to work and less likely to get public buy-in, because it – the comms around it were so disastrous. So, I don’t really know where you go next from that. I probably would just have a really clear use case, start small and then, sort of, naturally get buy-in from that, and maybe call it something else, because I think people are quite sick of that digital ID debate. But yeah, those would be my thoughts.
David Eaves
Please.
Rowan Wilkinson
Can I make an additional point on that? Like, one thing that we’ve deser – observed with lots of other DPI projects is that they have clear leadership, clear political leadership, and clear political backing, as well. And whilst I think, you know, we did hear from the PM a statement on digital ID, I think that somebody, kind of, sitting in DSIT with a very clear mandate, with very clear, kind of, deliverables on perhaps a specific aspect of public service was lacking, and is, kind of, something that I think we would have liked to have seen here.
If you take AI, for example, there have been clear champions in government that have really pushed the agenda forward in a way that I don’t think we’ve seen, kind of, technology be pushed in the UK for a long, long time, and that does seem to be lacking when talking about digital ID, digital payment systems and data sharing initiatives.
David Eaves
Yeah, this is very true about the political leadership, you know, you look at places like Togo…
Rowan Wilkinson
Yeah.
David Eaves
…really clear leadership during the pandemic crisis that led them to be able to do payments really quickly.
Rowan Wilkinson
Hmmm hmm.
David Eaves
India, with Aadhaar, very clear leadership there, as well. You ca – you – and Pix, as well, you know, a very, very subtle leadership from the bank, with a very, very interesting team driving that. Let me ask you, all three of you, I think a much more, kind of, challenging question. So, my own sense is, particularly in a world of AI, where, you know, deepfakes are going to become easier, like, is – it is very hard for me to imagine that you’re going to get out of the 21st Century without a digital ID of some form. Like, just in order to validate who you are to a bank. Like, already the banks are getting very, very nervous. Like, now, they don’t want you to do transactions over the phone because they’re very worried that you can replicate voices. You’re going to have to have some deeper way of authenticating yourself if you want to do transactions virtually.
Same with payments, you’re – we already have digital payments. You’re not going to get out of the 21st Century without some sort of digital payments, and the ri – the threat here of that getting monopolised or leading to massive rent extraction because of a dominant player from network effects is very, very real, and data exchange is the most for grabs. But the government is the holder of huge amounts of important data. So, I often, kind of, posit, like, if you’re not getting out of this century without those three things, the only question we’re really having right now is a question of governance. Is the governance going to be private, is it going to be public, or is it going to be something different?
And I guess to each of you, I’m curious, like, what do you see the governance is going to look like, maybe, kind of, generally around the world? You know, is there some models? And then specifically in the UK, what is the governance of these things going to look like if you’re – if it’s an inevitability? David, do you want to take a shot?
David Lawrence
I don’t think I have a good answer, except that I’m – it’s hard to see how you, yeah, as you said, how you exit the 21st Century without some, sort of, digital authentication identity, and it’s also hard for me to imagine a world where the state isn’t involved in that. Like, I don’t want the conversation about identity to be left to…
David Eaves
There are private…
David Lawrence
…Sam Altman, right?
David Eaves
…happily waiting in the wings.
Rowan Wilkinson
Yeah.
David Eaves
You know, like, that’s…
David Lawrence
Like, yeah, should Sam Altman decide what makes someone a human, basically? Like, that sounds like a grand and philosophical question, but if he’s – or, you know, OpenAI, or whoever are the only company that has built the, sort of, gold standard human verification, this is how I prove I’m not an AI and I am who I say I am, then you’ve basically, outsourced that to the private sector, which is a huge thing to outsource, right?
Rowan Wilkinson
Yeah.
David Lawrence
And quite a fundamental thing, and everyone – if that’s what we’re going to do, like, that’s a conversation that we need to have publicly and democratically. If we’re not going to have the state do that, then – and, you know, there are downsides having the state do it, as well, like, I worry about that too, in some ways, but that, whatever process, deliberative process, needs to be thought about there.
I also wonder, and I – this is a lack of technical expertise, so it’s, kind of, a question to everyone else, but, like, I wonder how you – if there are ways to separate out the verification and encryption from the, you know, say, like, the payment system so that you don’t have the same network effects. So, maybe you do have some network effects where everyone wants to use the same service ‘cause it’s cheaper, but not because that’s the only 0ne – that’s the one you have to use, otherwise, you don’t have the same access to everyone else on the same system. But that’s, yeah, that’s a question that I’m sure has a technical answer.
David Eaves
John, how about you? What do you see as, kind of, being essential governance in the next century in this space?
John Lloyd
I think a best-case scenario would be a multi-stakeholder governance situation where you have all the players in your society contributing to the governance of your solutions, of your DPI stack. That means, like, consulting not just with the private sector, but also civil society, community groups, people like that, in order to address those concerns about inclusion, I’m thinking about aspects like safety, as well. I think that would probably be a best-case scenario model. I think open-source gives you a, sort of, a leg up when it comes to that type of governance model. In some countries, like, Kyrgyzstan, we know that they purposefully chose X-Road, which is originally developed in Estonia, as their data exchange model for security reasons. So, they’re trying to stamp out corruption, they also have a big scary neighbour next to them that is constantly trying to hack them. But it also meant that it wasn’t just in the control of one government department or one person. It was, sort of – it was owned, because it’s open-source, it’s transparent, auditable, all of those kind of things.
So, I think what we’ve seen is that those models seem to be working pretty well in other countries and contexts, and I think it could be working pretty well in the UK, as well.
David Eaves
Yeah, I always like to observe that X-Road has been generously pen tested by the Russians. It’s one of the reasons why people have confidence in it. How about you, what do you see as essential governance in this world?
Rowan Wilkinson
Yeah, it’s a really good question. I think the key thing to recognise, as you said at the very beginning, is, like, we’re in a massive shift in terms of what is – was previously geopolitically “normal.” And Mark Carney’s speech, I think he said that “nostalgia is not a strategy,” and I mean, I would agree, but I think that from these, kind of, major shocks can come some kind of governance. And I think DPI is definitely a third way, showing that other things are possible, and I do think governance in the space is, like, starting to be, sort of, built. If states don’t do that, I do believe that private players will fill the void. You only need to come up in the Tube and see that everyone’s using Apple Wallet now to pay for things. You know, ultimately that technology will set the standards which then governments will have to adopt.
But just briefly, one, kind of, core – I mean, a policy recommendation I suppose in terms of governance, is on the data, the datapoint. So, the OECD has this Data Free Flow with Trust initiative, which is focused very much on national security and, kind of, data sharing between intelligence – the intelligence community. But I do think there is a model where it could be applied to DPI in order to help, kind of, countries to share, for example, open-source solutions or things that are working well for them. I just think that we need to see governments taking initiative and doing that so that they’re better able to co-operate and work together on pushing forward the DPI initiative.
David Eaves
John.
John Lloyd
Yeah, and just on that point, because there’s a, sort of like, increased desire for sovereignty, but also simultaneously, this increased desire to co-operate and collaborate with one another, and they’re not at loggerheads with one another. Like, you can still maintain ownership and control over your government’s data, your – their technologies and all that kind of thing, but you can also work with other countries. And, like, I mentioned 50-in-5 at the top of this call, it’s – call –meeting. It’s…
David Eaves
It’s a call for some people.
John Lloyd
Yeah, true. Oh, man, online life, it – and the campaign’s only just turned two. We have 32 countries participating, and the one thing that they all have in common, it’s not about income level, geography, or even digital implementation status, or where countries are, it’s about having that shared mindset, about willing to co-operate and collaborate, share lessons and best practices, implement their DPI faster and cheaper. And we do have countries in Europe participating, like Norway, France, Estonia, some of the DPI leaders in Estonia and Singapore, as part of that, and we have good representation in Africa, Latin America, and Asia. So, what we’re seeing is those countries are able to maintain control over their DPI systems, but what they share is having this shared mindset. Like, taking the DPI approach to their implementations, and willing to work together in order to have that rising tide floats all boats situation.
David Eaves
So, I’d love to open up the floor to questions. So, if anybody has a question, please feel to raise your hand. I might initially just go online. Miranda Sharp has a sharp question for us, where she’s a little bit annoyed. She’s like, “Why aren’t we talking about funding?” And I think this is actually a very fair comment. Is – are these systems just going to be centrally funded by a pot of money out of general tax funds and be free for everybody to use, or do we have to involve the private sector and private-public partnerships? Should they be fee-based? Like, is this all toll roads where those who use should pay more? Like, what’s the vision for how this infrastructure should be working, and what are you observing how different governments are tackling, kind of, the ongoing expense of running this if they’re not going to let the private sector just do it themselves?
Rowan Wilkinson
I can take that.
David Eaves
Please.
Rowan Wilkinson
Yeah, I mean, sustainability and funding is, kind of, one of the core problems that a lot of people come to me and talk about. I think a lot of countries have adopted using a mixed approach to start these projects, to pilot them. So, that might be with philanthropies or other government backers or private companies, and that seems to be quite a good way to initially get it off the ground. But then, I mean, yeah, the problem still remains what you do with the system once you have it up and running and how you back it, I think is still an ongoing question.
David Eaves
Yeah, I mean, like, places like Pix, like in Brazil, the Pix payment system is broadly free, although I think B2B pays a very small percentage of the transactions to help cover the costs, but they basically, said, “Oh, running this infrastructure costs us the same as running the previous system, and so we’re just…
Rowan Wilkinson
Hmmm hmm.
David Eaves
…going to absorb it.” But it’s very opaque, like, in India, it’s very unclear what the ongoing costs of running UPI are or Aadhaar are and what – and where that money’s coming from and if that’s sustainable over some period of time. Any other thoughts, like, David?
David Lawrence
So, I’m less worried about this than perhaps, was it Miranda who asked the question is? I think a few reasons. I think one is that fro – as you said earlier, this is a huge Global South success story, and if it’s working in those countries, who are far more fiscally constrained, quite often, than Western countries are, then I think for countries like the UK, this primarily isn’t going to be a, can you afford it question? I think it’s primarily going to be a, do you have the public buy-in and state capacity to do it, question?
And I think, secondly, because where I’ve seen DPI initiatives talked about and, kind of, started in government, it’s often been actually as a cost-saving measure. It’s because we are spending inordinate amounts on, you know, whatever it might be, NHS bureaucracy or our social welfare system or state pensions, that there is a desire to find more technological solutions. And yet the biggest cost by far for Western governments, I think for most governments, but especially in the West, is just staff costs. It’s just paying people to do stuff, and I think it’s actually going to be quite appealing to countries wanting to save money if they – if there are, sort of, digital ways of doing that.
And then also there’s just the scale of it. This stuff is often – because a lot of it’s software, that means that the actual scaling costs can be quite cheap to just roll it out on a larger scale, like, once you’ve iterated and found the version that works, and the initial, sort of, startup costs of starting smaller projects can be quite low. So, yeah, sorry, that maybe isn’t a great answer for the Global South, but I think for countries like the UK, I actually don’t think money is the main barrier, and in fact, a need to save money and reduce government public spending might actually be a motivation in favour of doing some of this stuff.
Rowan Wilkinson
But I mean, just to build on that, as well as, kind of, helping to co – to save costs, a lot of these initiatives, I mean, UPI being one of them, has shown to help generate, like, generate the – more, sort of, investment in the economy. It moves, kind of, informal markets into formal markets.
David Eaves
Yeah.
Rowan Wilkinson
So, I think there’s endless, kind of, use cases that have shown that it does a lot to, kind of, inject some kind of, yeah, energy into the economy, as well.
John Lloyd
I totally agree, and I think it – that increased efficiency, when you’re thinking about DPI as a project, like, there’s, of course, the cost of it, but then the benefits of that are often cumulative and often aren’t immediately realised. But in Kyrgyzstan, with the implementation of their X-Road system, they already have saved 21 million hours of people’s time. I just think about that example I gave about buying a house and having to go to the bank and all the documents I had to collate and all the back and forth between my Lawyer, like, that was costing me personally a ton of time. Imagine all those savings acro – on a national scale, the benefits that that would have, the DPI, kind of, pays for itself at that point.
Rowan Wilkinson
Hmmm hmm.
David Eaves
Yeah, I think just two observations. First, David, I think the much more sobering point, as you say, “Oh, lower income countries, if they can do it, surely we can,” I actually think that’s the barrier. That we’re wealthy enough to afford to make bad decisions.
David Lawrence
Yeah.
David Eaves
Like, it’s actually – it’s – we’re wealthy enough to actually recreate things over and over and over again, ‘cause that feeds the budgets of various public servants and people who then can be more powerful and have more – have a bigger budget and more staff, and also can choose to be, like, “Oh, I’m slightly different from you, so I want something slightly different so that meets our needs better.” If we can afford to make the, kind of, relatively wasteful choice…
David Lawrence
Yeah.
David Eaves
…that actually makes it harder for us to figure out how to do things at scale. And this is the part that also blows my mind, which is that the real benefit of technology is the almost infinite capacity for scale at a ever-diminishing marginal cost, and yet we constrain everything at the service, or if we’re really lucky, at the ministerial level. So, we bound any market to what a Ministry serves, when actually if you had the whole country as the market, the marginal cost would drop significantly more. So, we’re recreating inefficient markets with everything we produce in government over and over and over again.
David Lawrence
Yeah, the – I think the Economist Duncan Robinson calls it ‘affluenza’ like, Britain’s…
David Eaves
Affluenza, yes.
David Lawrence
…affluenza. I think it applies to, like, lots – like, we could – because we can afford to just have, you know, really high standards, bat tunnels for, you know, our rail lines, that’s what we end up doing, and then we end up building this stuff far more expensively than we have to.
David Eaves
Any questions in the audience? Yes, you madam, the…
Member
I don’t know if there’s a mic. Thank you so much. So, yeah, just a question. To me, it feels, like, a little bit, maybe, of an elephant in the room, and I completely agree, you know, in terms of feeling the pain of the inefficiency and the lack of basic things, like, digi ID or payments and so forth here in the UK. But I know you mentioned this correlation of the South being very much a success story, and the European North, Canada, I guess the US, kind of, falls into this, as well, and the UK, now fail – falling behind.
Now, I completely get the, kind of, state budget gravy train with these countries argument, but is there also not another argument, is that this correlates, this lack of digitisation is correlating with established liberal democracies? And is there perhaps not something around, you know, if the public – should the public just be convinced? Or is it actually that in all these countries the digitisation, especially the identity management piece, have not progressed because there are maybe valid concerns, especially in light of what is happening in the US, for example, at the moment, in terms of, you know, a more, let’s say it, authoritarian approach and how that will play out, or China, we see what happens there and how the technology is used? I’m just curious about your thoughts on this.
David Eaves
Good question. Anybody have thoughts on this?
David Lawrence
I have some thoughts. I mean, someone should do some proper research into this. Chatham House is probably the place for that to happen, but I – so I hope it’s correlation, not causation. I think there is correlation. I think one reason why it might be just correlation, not causation, is ‘cause I think there could be some leapfrogging going on, where some of the more established, industrialised countries, like the UK, do also happen to be older democracies, but I think that makes it harder places to, sort of, introduce completely new infrastructure.
And that applies to any infrastructure. Like, that’s true for electrifying the rail, it’s true for 5G networks. All of that stuff is easier to build in a country like China, where you’re building all of this stuff for the first time. And if you look at other – I think there will be examples of democracies that are, sort of, more recently industrialised, maybe Estonia is the, like, obvious go-to example, where actually there has been success, and I think they’ve been able to do the leapfrogging while also being a democracy. And so, I would want to look at countries like that, but as I said, someone needs to do a fuller analysis.
David Eaves
I would just say, I think I actually have the almost inverse concern, in some ways. I actually think a lot of the early movers in the Global South have been gen – have often been relatively democratic. Whatever one thinks about, you know, India is definitely a democracy, one can contest about the way Aadhaar was created, but it’s still a democratic state. That Brazil is also fairly – has fairly robust democracy, Argentina, as well. Estonia, which was poor when it started, we forget how poor Estonia was when it started its journey. What I actually much more worry about is as the model is proven out and the costs for adopting it drop, other forms of government will take interest in it, and what are we going to do about that? And what are the norms that we want to have around this infrastructure as those countries start to take more and more interest?
And, of course, the Chinese have been also very effective in this space and are very happy to provide people with infrastructure like this, and so, I think – I do think – I think your question’s exactly right. I’m not sure it comes in the right direction, but I think you’re actually exactly the right question. Yes, you, sir.
Creon Butler
Thanks. Creon Butler from Chatham House. So, I – just on the – sorry, just on the problem that you have in the UK, so the – having been a Civil Servant, one big problem is the Daily Mail, you know. So, the problem is that, you know, the – so the – doing things that don’t work, the Daily Mail is completely intolerant of this. And so, my – one of my questions, and I – and maybe this is partly – I mean, I think your analysis of what’s going on in the Global South is correct. There are many – probably many other things, like, you know, they’re ignored by the private sector, you know, say, the Central Bank in the Global South typically has a much bigger role than it would have in the North, and so on, but they may also not face what British Civil Servants face in terms of an incredibly aggressive media. So, one of my questions is, how do you tackle that, how do you – you know, to enable us to do these things?
And then the second thing is, I know why the Treasury wants big, fully costed projects, because it’s had so many experiences of things starting small and just creeping, you know, particularly in the defence sector, they’ve just become enormous.
Rowan Wilkinson
Hmmm.
Creon Butler
And they’ve seen that happen, for example, with our big rail project, that, you know, it wasn’t properly specified at the start, and therefore, it’s gone up vastly in terms of cost. So, you’re absolutely right, they shouldn’t be doing this, but how are you going to convince them, given all those past experiences?
David Eaves
I…
Creon Butler
Especially now with the Treasury.
David Eaves
Maybe first I’ll go to you, David, just because I feel like you championed initially the incremental approach, and I think there’s good defences, but I’d love for you to respond.
David Lawrence
Yeah. I mean, I don’t have a solution to the Daily Mail, sadly. I mean, I suppose a few things I would say. I mean, one is that I think Politicians do worry too much about it, relative to, like, maybe 20 years ago, it made more sense, today, most people aren’t getting their news from the Daily Mail, they’re getting it through social media. Probably the average Politician thinks a bit too much about what, you know, tabloids might say. And then, yeah, I think the – to the other point – well, and to the first one, I think the incrementalism is key, ‘cause actually, I think it’s hard to attack – I think it’s hard – if you announce your intention as a national digital I – I know I keep coming back to that example, but if you announce it as national digital ID, or indeed highspeed rail, or massive new nuclear st – power stations or whatever, and then it – costs overrun, behind schedule, that’s, like, you know, very easy for any newspaper to attack, and to some extent, rightly so, right? Like, in a democracy, you should attack things that go over budget or don’t happen on time. But a small team of Civil Servants, like, sitting in a room trying to build some software, doesn’t cost much, and, like, iterating and testing, like, beta versions of it, that feels like, not super interesting for the broadsheet, the broads – for the tabloids or the broadsheets, but, like, that may well be where the benefits come from.
And in terms of the Treasury stuff, like, I think I – yeah, I – and I know you’re twea – speaking to the Treasury and I’m glad you are, but this ought to be treated differently, I think, from, say, a lot of big physical infrastructure projects. Because well, firstly, ‘cause I think a lot of the upfront costs are a lot lower, so we’re not talking about something like HS2 or a new runway or whatever it might be. And then, secondly, I think you get so much retu – like, the marginal cost of scaling is so much lower, and you get that, sort of, immediate feedback, which allows you to, you know, build things much bigger, much quicker, much more cheaply. So, I’m left – less worried about it in terms of software, in the way that I would be very worried about, like, an MoD project or a big transport project.
John Lloyd
And quite a concrete example of, I think, overcoming the Mail problem and this incremental idea that you’re speaking about, is a colleague of mine is based in Colombia, and they’ve just introduced their payment switch in Colombia. She was in a taxi with her mum and went to go pay using the new payment switch, and her mum’s like, “No, no, no, no, no, no, don’t do that. I read” – and, of course, it was some, kind of, disinformation online, and so it took the younger person to be, like, “No, it’s fine, watch,” bang, just demonstrating how it happens.
So, I think having a proven case that at least some portion of your society trusts as your launch, and then proving that it works, it takes all the steam away from your Mails or your angry trolls online, who are going to be anti whatever the government is doing. I think doing your one thing and, like, absolutely nailing it, that builds that trust back up and really is the best counterargument to anything that the Mail will put out.
David Eaves
Do you want to add something, Rowan?
Rowan Wilkinson
Yeah, I mean, on the Daily Mail point, I think that, kind of, the economic benefits of this can go a long way, especially to a, sort of, Mail audience, and any audience for that matter. So, I think, like, focusing on the, kind of, hard evidence that this works and this is cost effective, I think is a good messaging for the government to take. Then on the point of Treasury, I mean, we’ve not really spoken – it’s not particularly, like, sexy, but procurement and procurement terms, right? You know, these, kind of, problems, long run projects, and so on, or increasing costs, maybe the maintenance costs keep going up and up, all of that stuff can, kind of, be nipped in the bud at the contracting phase. And I think that, yeah, there’s, kind of, more work to be done on making sure that when, for example, the UK Government is, yes, seeking to develop some forms of DPI, that they’re really looking carefully at the small print.
David Eaves
Let me just close. I think one of the things I don’t want people, particularly public servants, to believe, is, like, that we shouldn’t be big and ambitious. We can, and iterating is not about not being big and ambitious, it’s about scaling and learning before you make big announcements, so that you de – you fundamentally de-risk and you don’t jump too early. So, even, like, Aadhaar, used by 1.4 billion people today, when they first launched it, they didn’t launch it as a general ID. They launched it as an ID for a single benefit, and then they launched it in a single city, then in a single state, and then nationwide, and then built it up. Like, even that was iterated. So, I would want the British Government to be ambitious and to have big plans, but to iterate and learn in order to de-risk and then go public once they’ve learnt. I think all too often right now, our ambitions on infrastructure is to just go big before we know anything, and then stumble and learn as we go, and that’s what doesn’t work.
Alright, with that, I think we’re at time, but we’re going to continue at the reception. I just want to thank so much our panellists for joining us tonight and carving out time and sharing their ideas, and Rowan for her amazing report, which you can definitely download off the Chatham House webpage, which we will share a link online for. But thank you so much for joining us tonight, and we’ll see you upstairs. Thank you [applause].
