Last night I delivered a speech – without any slides for a change – at the Institute for Government in London (it’s embedded below, or you can watch it on YouTube). The opportunity came from a meeting some months ago with Peter Riddell and his team, so I used Martin Donnelly’s speech earlier in the series as the basis of my response. I welcome Martin’s response, as sadly he couldn’t make it, but I hope this allows us to continue our discussion — MB
From policy to delivery – changing the organising principle of the Civil Service
I’d like to thank the Institute for Government for having me here to talk as part of this fantastic series of speeches. The depth of knowledge displayed on the subject of the civil service has been fantastic. So, since I touch on the subject of the past, I want to reassure you that I realise that this isn’t the comprehensive history.
It’s my great privilege to talk to you as a civil servant. I’ve worked in five industries in 15 countries, and I’ve experienced the process of digital transformation many many times.
It’s tribute to the civil service that it talks openly and occasionally critically, and from the inside, about digital transformation in government.
Let me start by defining terms: digital transformation is defined by Wikipedia as “the changes associated with the application of digital technology in all aspects of human society.” It is not simply another name for technology adoption, but a recognition that in the last 25 years the changes we have seen have been networked, user-led and transparent. This is not a speech about the pressures facing us from a new wave of technology, but about the social and political effects of a networked era – and about the effects on the civil service.
Civil servants do brilliant work all over the country. In my opinion, too humbly and without enough recognition. Recently I received an email asking senior civil servants to screen for Ebola at Heathrow. My colleagues are doing this today. This is a fantastic example of civil servants going above and beyond, and it wouldn’t happen in any other organisation.
Yet most of the work the civil service does goes unseen, or at least unheralded. But whether it’s Ebola screens, student loans, renewing your car tax, or a thousand other things, that work is vital to everyone in the UK.
Often that work is harder than it needs to be.
I don’t think anyone disagrees that the civil service needs reform. It’s the nature of that reform I want to talk about today.
The Internet has changed everything. Digital is the technological enabler of this century. And, in any sector you care to name, it’s been the lifeblood of organisations that have embraced it, and a death sentence for those that haven’t. If you take away one thing today, please make it this: government is not immune to the seismic changes that digital technology has brought to bear.
The Internet is changing the organising principle of every industry it touches, mostly for the better: finance, retail, media, transport, energy. Some industries refuse to change their organising principle. The music industry was dominated by producers – the record labels – now it’s dominated by digital distribution – like Spotify and their ilk.
Others, like airlines, have rapidly changed how they work internally, and are organised radically differently in order to serve users in a digital age. British Airways used to have over 80 ticket types, with departments and hierarchies competing to attract users. Now it has a handful, and the organisation is digital first and much simpler. These changes are invisible to the majority, but that’s doesn’t make the changes any less significant.
Twenty five years into the era of digital transformation, the Internet has a 100% track record of success making industries simpler to users while forcing organisations to fundamentally change how they’re structured. These characteristics are not going away. Yet the effect on the civil service has been, until very recently, marginal.
This is because we deferred our digital development by grouping digital services into enormous, multi-year IT contracts, or what we refer to as ‘Big IT’. Or in short, we gave away our digital future to the IT crowd. While most large organisations reversed these arrangements we have only recently separated our future strategy – digital literacy and digital service provision – from the same contracts that handle commodity technology. By clinging to this model for 15 years, we have created a huge problem for everyone involved in delivery and policy.
Today I want to talk about two things.
The first is delivery, because I believe delivery to users, not policy, should be the organising principle of a reformed civil service.
And the second is skills, and why it’s time for the civil service to put digital skills at the heart of the machine.
In a recent speech in this series, Martin Donnelly, the Permanent Secretary at the Department of Business set out his defence of a Civil Service led by generalist policy makers. It’s a vision which many people share. This is the best articulation I’ve heard for the primacy of the policy civil service. Its logic is faultless, its delivery peerless – if you accept the premise that policy is fundamental. I don’t. And I fully expect the Internet will reject that premise, too, given the alternative: putting users and delivery first.
In a digital age, traditional policy-making is largely broken. It is slow, inflexible, unnecessarily complicated, afraid of technology and afraid of change. The cycle of green paper, white paper, draft bill, and secondary legislation is no longer the best way to decide to create or develop new services because user needs are given scant consideration, however necessary the process may be for Parliament.
We’re stuck with a statute book that demands wet signatures, which holds back services whether they be benefits or car registration, preventing them from being as simple as they could be. Our Victorian legacy prevails in our language. Take the simple act of registering to vote, which, until this year, required the “head of the household” to register on behalf of the family… his family, being the intent of the language.
Or take the archaic legalese of a service like accelerated property possession, which at one point asks users whether the property is a dwellinghouse. How many of you live in a dwellinghouse? The law might be centuries old, but the language we use to talk about it shouldn’t be.
The benefits of ditching the paper driving licence have been clear for a decade, but because of a fear of change, fear of failure and, yes, fear of digital, it hasn’t happened. We’ve forgotten the art of the possible.
This is not how things are done any more.
We already have numerous government services so complex that only intermediaries like solicitors and accountants can use them. In a world where everything is becoming quicker and easier, if government doesn’t become quicker and easier too, it will be intermediated away. Not out of existence, but to the point where it’s invisible to the public, where engagement with government services will disappear.
I want to address some of Martin’s points on policy in more detail.
Policy civil servants, he says, must recognise “the political constraints that exist as well as the dangers that go with too rapid decision-making.” And yet many Ministers are frustrated by an inability to get things done. As of this month it’s legal to rip an audio CD for the first time, 23 years since the MP3 codec was approved. That’s a small example, but it shows that, if it’s an anachronism you want, a policy-driven approach is the way to get one.
I understand Martin’s “interest and desire to get things right first time”, but if the first time cost X billion pounds, took Y years, and still wasn’t right, then something’s up. It’s much better to fail fast, fail cheap, and then put things right at a fraction of the cost.
To succeed we must be humble in the face of messy reality. The intent behind many policies and services is for people to adopt or change a particular behaviour, be it registering to vote as an individual rather than as a household, or letting go of the concept of a physical paper tax disc.
But people are complicated. And no policy or service we civil servants think up will ever work in practice the way we thought it would in theory. We must start out humble, and rapidly iterate in response to the messy reality of real users using real services, not critiquing the policies of Whitehall.
It’s not the nuance behind, say, a social impact bond which affects people’s lives, but the steady drip drip drip of inadequate services for which we’re all paying. It’s the duplicate forms, the large minority of transactions which fail or can’t be finished in one go and result in phone calls and wasted visits, the waiting on the phone and in line, and the arcane language which demand much simpler digital services. This simple stuff needs fixing, and in doing it, we can save huge amounts of money.
Sadly, our existing culture in Whitehall fetishises the complex in the name of giving Ministers ever more detailed policy options. Front-line civil servants, struggling to deliver services in the real world, too often lose out to those who relish the theoretical edge case, adding more complexity while burnishing the reputations of those who “display a real grip of policy detail”. People on the front line see the cost of that complexity; the perverse incentives, the failure demand, the confusion, the frustration, the waste of money and human potential.
Look at Carer’s Allowance, where an applicant could suffer a possible 500 questions to qualify for their £61.35 a week – and these are people with some of the most profound needs in society. That’s why one of our design principles is Do The Hard Work To Make Things Simple. By valuing simplicity, Leigh Mortimer and his DWP team at the front line in Preston – working with colleagues in London, and carers’ support groups – got rid of half of these often inhumane questions.
The result: a far more effective, efficient and human service. Now, nearly 60% of carers choose to apply online, up from 28% 12 months ago. Leigh and his team have shown the excellent results of doing the hard work to make services simpler.
That’s how we do things now. It really works.
Carer’s Allowance is one example from the government’s digital transformation programme: the 25 services we’re making simpler, clearer, faster and digital by default. We’re on course to have the vast majority live within 2 years of starting that work, and in some cases, public betas were available within months. These services make things better for millions of people across the country at a fraction of the cost of the previous versions. And they exist because we put users first and got on with the job, learning from our missteps along the way.
Martin also makes an interesting point about transparency. “Transparency within the system”, he says, “should not be equated with public openness of the decision-making process.” Not only is that a enlightening admission about how policy is made, but it’s irrelevant if you focus on delivery, especially if you work in an agile way. Build a core service quickly, then fix it and continually improve it by concentrating on user needs. Then, you not only accept that failure is inevitable, but that it’s desirable, for the reasons I’ve already talked about.
We should say to critics in the media or elsewhere that failure is an essential part of government, just as it is in private enterprise. And the cost of failure should be tiny, dwarfed by its rewards. Every government service now needs to pass a Service Standard assessment. We publish all those assessments, whether or not the service passes. Some don’t. So the teams go back, make adjustments, and try again.
Sometimes we stop the project, but it’s having spent only a few tens of thousand of pounds, not millions – millions because we’ve signed a binding contract. The cost of failure is only enormous if you plan to launch with a big bang on a fixed date in a couple of years time, with the world’s media and public watching – but before you’ve really started the work to understand how to best meet the needs of the people who will use the service. Big bang was fine in 1986. It is a disaster waiting to happen in 2014.
Finally, policy civil servants, Martin says, must “engage with external expertise in policy formulation and execution,” and “encourage technical experts to join, for short periods or a substantial career, flexible structures which make full use of their skillset.”
This is simply wrong.
The answer is not a box in the corner labelled “digital” which you open on the rare occasions you need some Internet. The answer is to flood your organisation with digital people and let them lead. They’re the ones who run successful organisations now, because they’re the ones who know how to.
If the civil service fails to embrace digital transformation, it would be more than a missed opportunity – it would be a catastrophe. According to Her Majesty’s Treasury, digital has already saved over £1 billion since 2011. That could be just the beginning. Delivering public services as cost-effectively as possible is a solemn obligation. That would be just as true in a time of economic plenty. But, as I expect you’ve noticed, this isn’t a time of economic plenty.
We can do better than shaving 5% from a contract re-negotiation here or 10% by switching supplier there. We can cut costs to fractions of what they were – making molehills instead of mountains. But to do that we need to let go of the security blanket of the “safe policy space” and let go of our misplaced belief in big IT procurement.
It’s about much more than making websites. It’s using digital – its immediacy, its flexibility and its interconnectedness – to rethink public services. Yes, it’s about making digital services which are simpler, clearer and faster to use. But more, it’s about changing the way those services work behind the scenes.
And that means starting again. This is a chance to rethink tax, rethink benefits, and rethink how people come in and out of the country. Starting from the beginning to build the services we need will prove quicker, cheaper, and more responsive to what our political masters actually want. It’s better than the alternative: attempting to build on top of old, broken machines.
In his speech, Martin argued for keeping the Northcote-Trevelyan model of the civil service. In 1853, Sir Stafford Northcote and Charles Trevelyan reformed the civil service from one that hired on the basis of “preferment, patronage or purchase” to one which hired on merit. In the beginning it served us well. The late 19th century and, later, the post-war era saw some of the best development and redevelopment in the history of the modern world. There were spectacular successes.
But it would be remarkable – unfathomable – for them to have devised a civil service to outlast mass production, the integrated circuit, Moore’s Law and cloud computing – technology which has rewired the basis of our society, not to mention the challenges of cryptocurrencies, Bitcoin and personal data yet to emerge.
Martin says that the best way to get impartial advice is to ask an impartial adviser, if there is such a thing. These days, sadly, those impartial advisers don’t know enough. Advice must come from the doers – the people at the coal face that understand the shortcomings of their services and who are brimming with ideas for how to make them better.
We must build a civil service with a bias for action and delivery at its centre, in place of an endless cycle of critiques. John Fulton saw that as far back as the 1960s. He thought that perhaps generalists weren’t always best placed to deliver in increasingly technical times – in “the white heat of technology” as Harold Wilson put it. I don’t want to re-open the C.P. Snow debate which preceded that speech, but we have been talking about this cultural separation for over 50 years.
It took until the Citizen’s Charter in 1991 to put some of that right. But meanwhile, between 1969 and 1995, the Internet happened, and the whistles and chirps of modems began to reprogram the brains, skills and expectations of, well, basically everyone.
We have an opportunity now to build a civil service relentlessly focused on delivery. And we’ve already started.
We’ve built GOV.UK, a single online home for government information and services. We’ve built government-wide platforms like the performance platform and Verify – a new way to confirm people’s identity when they interact with government.
We’re already a world leader. New Zealand and Hawaii are both using the GOV.UK source code, published under an Open Government Licence, for their own government websites. And the US government is building its own digital service modelled on GDS.
I’ve already mentioned the transformation programme. Government departments embracing digital are doing great things. Register to vote in the Cabinet Office, apply for student finance in the Department of Business, make a lasting power of attorney, a service delivered by the outstanding digital team at the Ministry of Justice – three services among many which are now digital by default and easier to use than ever before. And thanks to very talented and dedicated civil servants embracing agile, we’re building them much faster and much more cheaply.
Show, don’t tell.
The transformation must go deeper. We must fix the shop floor as well as the shop window. For example:
Another of the new digital services lets you book a visit to see someone in prison. It’s made it much easier for friends and families to visit loved ones in prison, helping to rehabilitate inmates and reduce reoffending rates.
But behind the scenes the prison service is stuck with a decade-old IT system. But because we’re afraid to throw the old machinery away, staff at prisons are forced to copy and paste visit bookings from the new system into their old software. A block on development prevents their integration – in other words, they’re not allowed learn the skills to fix the problem.
The civil service is crippled by a fear of risk, often for good reason, a fear of throwing good money after bad – again, for good reason. That’s understandable given the IT disasters of the past. But those disasters are the failure of policy and big IT. That isn’t how we do things any more.
Take Internet retailers. If they worked like the prison service works right now, behind the scenes you’d have people entering orders into stand-alone databases; typing, printing and posting letters of confirmation; manually taking stock, and entering that data into yet another database. They wouldn’t survive in 2014, much less thrive.
In reality, these are extremely efficient organisations. Products and services share common platforms for signing in, receiving payments, and for tracking stock. They all talk to each other. They’re built that way. Not only does that make life easier for users, it makes the organisation as efficient as possible behind the scenes. Why should government be any different in this regard? That’s what we mean by “government as a platform”. It’s not a catch-phrase – it’s a necessity.
Internet companies don’t use the Northcote-Trevelyan model, and they don’t endlessly tweak policy behind closed doors. Neither does any competitive business in 2014.
We need a civil service with fewer critiques and more makers. We need technicians to put platforms in place, and service managers to build upon them. We need leaders for whom the idea of using real-time performance data is a no-brainer, not an alien concept. And we need user researchers and data analysts to do the hard work of understanding user needs and behaviour to continually improve services. We should be actively involved in the development and delivery of our services rather than spending millions, sometimes billions, man-marking contracts.
Again, we must turn policy-making on its head. Government is often about providing the services no one else can. It’s time for government action to stem from the needs of users and services. The strategy is delivery.
I believe that the messages here, although unpalatable for some, have already been accepted by many in our policy making community. As with all newcomers to government, I am struck by the quality and friendliness of the people, and their eagerness to learn and engage. Our policymakers, with a few exceptions such as Paul Maltby’s teams and the Behavioural Insights team, use pre-Internet tools and techniques, while living a fully digital lifestyle outside of working hours. This simply cannot continue. Our policy teams are as much a victim of the jam-tomorrow, Big IT culture as our front line staff, and it’s vital that we help them adopt new working practices and embrace digital opportunities.
If government doesn’t get simpler it will be intermediated away in forms we don’t desire. This is the crux of the argument in a digital era. Plenty will argue for the alternative – a complex policy centre surrounded by standalone external delivery agents. This view may be attractive, but in practice has lead to perverse outcomes for public service provision: Rent seeking behaviour by suppliers; services in silos which can never talk to each other; technology prices that go up while the market prices are in freefall; the time taken to change or add features stretching into years instead of minutes.
We need a smaller, cheaper civil service, yes, but one that retains the ability to have a direct relationship with the public that it serves with confidence and transparency. The needless separation of policy and delivery is stretching that relationship to breaking point.
Over the years, we’ve had many opportunities to reform. Some we’ve taken. Some we’ve missed. We’re least effective when we’re political operatives. We’re at our best when we deliver. That’s what the civil service is for. And in 2014, delivery and digital go hand in hand.
We are in a moment in time when 5 forces are all acting on Government. The combination of financial pressures, demand for better digital services from users, the need for better digital tools inside the service, and the ongoing challenges of security and procurement all require digital transformation.
Those five forces will inevitably have to be addressed to meet the financial and service delivery challenges ahead. In doing so, I expect them to challenge the organising principles of the Civil Service. I expect us to move from being more hierarchical to being flatter. We’re going to have to be less siloed and more networked. Of course we will have to be less opaque and more transparent. Our operating principles will be less about command and control, and more about user-lead service provision. And in tandem with that we will have a shift at the centre from policy to delivery.
I say this with confidence, because this is what’s happened in every industry the internet has embraced.
The demand for digital transformation is not a policy option. It’s a delivery crisis. And, if you’ll pardon the meme, it’s because Internet.
We can do this if we work together, but we have to crash through this false binary of policy and delivery.
Let’s make things simpler, clearer and faster for users, and let’s embrace rather than resist the inevitable change in our organising principle. Let’s make government services fit for our time.