Digital by default: the digital strategy of the British Public administration

Interview with Mike Bracken on the digital transformation of the State in the United Kindgom

The Government Digital Service (GDS), founded in 2011, has led the charge on transforming the provision of online public services, inspiring similar central and interministerial units and agencies in many other countries, including France. In this interview we talked to Mike Bracken, the man who founded the GDS and served as its inaugural director, for his insights – gleaned from a long career in the public and private sectors – into how organisations are embracing digital transformation.  

What motivated the creation of the Government Digital Service in United Kindgom?

So in 2011 – in late 2010, actually, we had a new government in the United Kingdom. It was a coalition government. It's very, very rare to have a coalition government. And we had a series of characteristics that was very unusual. We'd had a financial crisis in 2009 and '10. So for the first time in years, the government didn't have as much money as it used to do. So efficiency was the big issue.

We had a coalition government, which made consensus a big issue. And also, we had had – like many governments and organisations – decades of consistent failure of big IT projects. At the start of this administration, the NHS IT projects I think had written off around 10 billion pounds over a decade.

So there was no more tolerance for big IT programmes that were going to fail. At the same time, the Internet and the web had become mainstream in people's lives. People inside and outside government were asking themselves, how can government services be so poor and not keep pace with the technology that I have in my pocket and available to me via a browser?

So these conditions were all acting on the environment in 2010-2011. I had spent part of my career helping civic tech organisations such as mySociety and others develop from outside government on the last 20 years, and I was approached by the government and the Civil Service to take the role of the head of digital for the UK government, which I did, but I negotiated that role, and to do that, I believed that we needed to found a new institution.

Martha Lane Fox, an Internet entrepreneur, and Francis Maude, our minister, had written a report and argued that the government should set up a new digital centre. Not an IT centre, a digital centre at the heart of government. It should have domain power and it should really be responsible not to any particular department but to the needs of the users.

And so from the start, we set up a new institution at the heart of government that was focused on servicing the needs of the government digitally but also – sorry to say that again – to focus on serving the needs of users of government services rather than any individual department.

That was a fundamental change for the UK government. So in part we were set up by necessity and in part by opportunity.

What were the main challenges for the Government Digital Service?

The challenges we faced were pretty profound. We were starting from a very low base. Governments around the world had slavishly followed this era of outsourcing and procurement so that there were very few skills inside government when we actually got there.

The first problem was that there just simply wasn't the right mix of skills – digital skills. So, developers, engineers, architects, product people, service, design – those skills were completely absent in many government departments. So we had to reskill the central government quickly and then re-establish a new normal that those people should be seen as important civil servants rather than seen as providers from a consultancy or so on.

The next challenge we faced was literally where to start because the UK government, like many governments, had had 20 years of indiscriminate use of the Web. At the point we started, there were several thousand websites, all of which were pretty poor, none of which had an element that looked the same. And the users were very confused. What the government was effectively saying to its users was, "You, via a browser, have to work out how we have structured ourselves to therefore determine how you can deal with us". The arrogance of that is astonishing, and governments to this day do that.

So the first challenge was to recognise that we weren't servicing the user needs. The user had – users have many needs. There are many types of user: benefit claimant, immigrant, prisoner, refugee. We give them lots of different names, but at the end of the day they're users of government services.

And the very first need that all users have is to know that it's the government that they are dealing with, and how to find it, how to deal with it. To do that, we had to reset the government as a digital brand, and we launched a single domain: gov.uk, and over two years, we moved two and a half thousand Web services and transactions onto a single domain. It's still there today. It'll never be finished. It changes every day but it's all responsive, it's all based on user need.

Setting that user need as our North Star, as the thing that we always follow, then allowed us to widen our brief and look past the websites and services and look at transactions and organisation, design, procurement, [and]technology. So we ended up taking a considerable reform agenda right across the mechanism of government. But we started off with the web estates and the digital estate.

What vision did you had in terms of digital transformation?

Our vision – along with Thomas Moore, whom I've worked with for 25 years – we had a very well understood vision that we were there to reform the mechanics of government. We literally have a government structure, an organisation, which is based on a cabinet system. It's a Victorian model of government. It served us very well when we had an empire. But in a digital era, it is simply out of date. And it precludes us really developing as a nation.

And we understood because we've been working in this area and talking about it for decades, that this is an opportunity to do that. However, we just had so much basic hygiene stuff to do. Sorting out the digital estate, sorting out the websites, making some great transactions, getting to know our users, reskilling the organization... These are basics, and they're no different from any other big environment, whether it be public or private. Getting the HR system reformed, procurement, finance... We did years worth of work on changing business case processes to move them from massive hundred million pound plus business case proposals to small teams leading iterative change.

That doesn't sound much. It takes ages to do that. It's such hard work. But our vision was that all of these reforms independently and concurrently will lead us to a place where the mechanics of the state would be questioned and changed more suitably for a digital age.

What did you learn from your experience as the head of the Government Digital Service ? Do you have any regrets?

Plenty of regrets, things that have gone wrong. I think the big thing to say is that government departments and agencies all think of themselves as special and unique and different because they do something specific to a certain user group or need. And that may be true. The department that administers nuclear reactors and the department that takes away council waste – of course there's nothing in connection with them. But actually in practice, the operations of those departments, how they organise themselves, their buildings, their people, their technology, their skills, and so on, are pretty much the same. They're not advanced. There are a lot of soluble problems.

So the first thing to recognise is that many of the tasks we were taking are not fiendishly complex, they just require hard work that hadn't been done for the last 20 years in some cases. The regrets I think were some of the choices made. Because I think there's an old phrase “you regret what you didn't do”. So even when we had a failure of some of our projects, I don't regret it because we learned from it.

I regret not taking a bigger bet in two areas. The first is, we first did the digital estate and that took really 2012 and '13 – Sorry, 2011 and '12. And at the start of '13, we started off a programme with 25 of the major government transactions: tax, welfare, benefits, motoring – the big stuff. And it took two years. We brought 21 of them in. There are case studies of how you transform not just a service, a transaction, but an institution behind it.

What we didn't do is pick 25 internal services that our colleagues use. So for instance, if you're using a terrible ERP system in the back office of a government department, that needs a transformation. And we didn't do enough of that. That meant that our users of government services are so much happier. That's not just because my opinion, it's a fact, and we have the evidence to show that it's cheaper, that users are happier, (and) it's quicker. What we didn't do is take enough of our colleagues with us to make their daily working lives better. That was the first mistake.

The second mistake is, the vision of the government as a platform, the idea that government can work very differently - we only partially landed that idea. That's simply because we ran out of time. We had to do too much remedial work, fix too much broken stuff, and we didn't get the political sponsorship right for the idea of what does a future digital state look like. Because at the moment, we have a problem that the role models for a digital state are often the city states or the very small countries.

We can talk about Estonia and Singapore and others, but they're very small countries and not even as big as some of the cities in our respective countries. But we have, and had then, an opportunity to redefine what a modern, big nation state looks like in terms of its operation, its skills and its people.

Because this Victorian model of departments or agencies is not really tenable in the world we live in right now. It's going to be increasingly untenable and we haven't had enough political conversation and backing for a different kind of model. In the UK, in the period when I left after the 2015 election, there were senior ministers - senior politicians - from every quarter: Tony Blair, David Cameron, Nick Clegg... these three prime ministers in the last 15 years. They were all saying the same thing, that as a government, we're just simply not set up right.

They were making the observation out of frustration that they couldn't get things done that they wanted to do. But what they're not saying is, "and this is how we need to be set up". We failed to land that literally because we were doing too much fixing an old system. I regret that we didn't do enough. And I'd like to hope we can work with Henri and other people here in France, too, and President Macron and others to try and have that conversation. Because literally the mechanics of our government systems are not fit for the challenges that we face now.

How to organise effectively the digital transformation of the public administration?

Tim O'Reilly who's - Tim's a very good friend of mine, he is a brilliant mind. He's one of the core Internet thinkers. He coined the phrase "coming to the platform". Our variant of that, and what we published in the UK and the platform strategy that has largely been deployed, with notify, verify and pay and gov.uk is I would say a good model for big countries to look at.

But it's very different from the sort of laissez-faire economics of the sort of West Coast Silicon Valley from where the original phrase was derived.

I think the future for any organisation - any government organisation - has got to recognise a simple truth: that one needs a central organisation. And by that, I mean an organisation that is aligned closely to the Finance or Treasury structure, to be able to effect network effects across a federated government.

Governments are federated for very good reasons. To get them to coordinate, to get them to work together, is a very difficult task even with political direction. Without a central digital team that has a view over that, it's virtually impossible. There's no surprise that the countries that are doing this the best are the ones who have got a digital team in the centre with the appropriate powers and skills, but also the appropriate alignment, usually to the prime minister, head of the Treasury or Finance Ministry.

Governments and organisations that take a digital view and put them at the periphery and say, "You're the little innovation thing, we will put you in the window." It's very difficult to affect the mechanics of government from there. You've got to deal with - as Henri was saying - you've got to deal with passport data and tax systems - the very nuts and bolts of the mechanics of government. To do that, you need to be in the centre. Without that, I fear that most government transformations are actually simply innovation labs which show some nice stuff but don't really change how the government really works.

How did you organize your action with the Government Digital Service?

The centre of GDS was a culture of teamwork and creating multidisciplinary teams. The way governments have approached digital change is to approach it like an IT project. And that's problem number one. The next problem is that IT projects have usually been put into multiple years, like five-year projects, with fixed budgets with false certainty. So we know exactly what we'll get in five years' time. That's crazy because in a digital world, user needs change every two or three months. What then happens is, that is then given to an outsourced third party or technology contract. And the end results are not good.

What we did instead was to create small teams, ask them to deliver a new product or service, a first version of it within 6 to 12 weeks. If that product, or that service, did nothing else it tested an assumption. So if an assumption was that a certain group of users need a certain type of service, you could test that digitally. Testing it digitally is a much better way than writing a policy about it because writing a policy is still an abstract concept whereas testing utility with users proves an assumption.

If you prove it or you think you're onto something, move it into a private beta, develop it more, then a public beta and so on. Keep developing it so that at some point you can just turn the old one off. And that's the critical thing about teamwork is that the more people you have in it, the more of these teams you have, getting the right teams working together, if you're using technology the right way and developing very quickly using open source techniques, things that are very open and quick – small pieces loosely joined - things that can be connected together, then what happens is that the speed of the development is massively accelerated. You get the best of your best people - your operations team, your legal, your finance, your development, your compliance – all in one multidisciplinary team and you deliver really quickly.

There are many, many problems with the tech industry. Many problems. Not least how it pertains to government. But one of the things that the big tech companies have got right is these ways of working. It's the Amazon model, no more than one pizza per team, that sort of thing. For government to adopt that, it's literally a no lose option because even if you get the first alpha or beta wrong and you fail, you learn so much more quickly and using teams to turn government into proper learning environments focused on user needs rather than outsourced technology contracts focused on procurement and policy.

That was the biggest change.  I would argue that when you summarize GDS, we did two things: we redesigned and redeveloped public service for the digital age. They work now. And secondly, we stopped the dependency on big IT. How we did that was by having teams.

Did major digital platforms influenced your action?

I think that the big platform players, in fact, the era of platform technology services, affects governments in a variety of way.

First, concentrate on the things that those companies have got right. Their relentless focus on their users, providing a better digital service, being able to react quickly to what users need. How they structure themselves internally, the fact that they've got the right disciplines and skills. Also the fact that they give engineers, product people and service design people the power to make decisions on behalf of the user. They are all things that governments can learn from.

In fact, I would argue any organisation can learn from them. Silicon Valley is not perfect, but these organisations have really developed some excellent ways of working. However, they have obligations under the law. But what they don't have is a set of consistent values that apply to the service design, their thinking and their activities, in the same way that one does if it's a public service.

And that's the fundamental problem because as these organisations are basically providing a new Commons, a new set of common data, platform, services... they're effectively shadowing existing public services.

What's happening is that the inherent values of those public services are being lost in the mix. So they are too powerful, they're not answerable to anybody as an individual, really, and they're dealing on a sort of a framework of regulation which is unfit for cross-national platform plays.

I look at Stephven Quest in the Commission, and he gave an interview recently about the regulatory backlog, about how long it takes for all of Europe to catch up and deal as one with them. And it's inevitable, I think, that we'll get individual nation-states dealing with these platform companies differently - you're already seeing that in France and the UK - in the next couple of years until there's a more broad brush regulation across Europe. But that regulation is coming and it's way overdue.

You worked in public digital transformation in various countries as well as in different types of public administration. What lessons did you learn from your experiences?

At Public Digital, we work with 20 governments, plus also some of the organisations that deal with multiple governments, such as the United Nations, the Inter-American Development Bank, the World Bank, and so on. So we see into a lot of different countries. It's been a real lesson in terms of how to think about countries. All of us, wherever we are, have a mental list of who's winning and who's losing.

We think in terms of development or social indicators or finance or size or volume. The problem in digital government actually is that the poster children for digital governments are often the very small countries - Estonia, Singapore - and as much as they are impressive, they are little more than city states in terms of size.

Most of the most impressive work is done by teams which don't share the limelight within bigger government but who have got some form of political mandate or some form sponsorship or protection within an existing system.

The thing it's taught me to do, with working with organizations in - Daniel Abadi and his team in Argentina, Marushka Victoria Chocobar in Peru. Some of the teams down in Canberra, Australia. I could go on with more of them. Yolanda Martinez in Mexico. Is that there are small teams, aligned to the central government, working either for the prime minister, the president or with the finance minister, who are doing astonishingly good work.

The problem is sometimes they are in countries that we don't believe are winning the global race, to use a phrase our prime minister said. And I think that what we've got to do in countries like France and the UK is recognise a central problem. Because we are big and wealthy and we are experienced and have relatively good functioning governments, that doesn't mean we are winning now.

It means we're ahead because of what we've done in the past. But it doesn't mean we are winning anymore. Because in small and developing nations, unlike some other nations, there are teams that are creating new services, new platforms, new innovation using digital tools, and they're moving to change the state much more quickly and much more beneficially than France and the UK is.

In Peru, a digital team in the central government launched a new transports system, a driving licence system, for one class of drivers in six weeks flat. That's after sorting out its websites and single domain. Now in a country as troubled as Peru, that's quite an outstanding achievement. In Mexico, even a regime such as the Peña Nieto regime coming to an end were able to launch digital birth certificates, a massive drag on economic efficiency in that country. In Uruguay, they rolled out one laptop per child. It's the only country to do that for all people. They have a central agency called Agesic. Argentina, without a doubt, is one of my favouritecountries because the Argentineans have gone straight towards a platform play, argentina.gob.ar. You can see all your services – an increasing number of your services - digitally on one customised platform.

These are radical changes in the mechanics of these states, and they have come about sometimes by necessity, sometimes by design. But modern countries, apart from the UK with GDS, most countries in Europe are not taking these steps. And they think that they don't have to, and we're ahead in the global race. And actually we're not going to stay ahead for much longer if we continue to lack ambition and move as slowly as the slowest institutions that we have at our disposal.

It is about time that we redesign some new institutions digitally first, to allow our machineries of government to move much more quickly.