Just before Christmas I wrote a post questioning whether or not the UK had a strategy for the Government Digital Service's much touted idea of 'government-as-a-platform'. In it, I stated that I liked the idea of the public sector building common platforms with open source tools and flexible APIs, which could be reused and built upon across government - but I also said that I didn't understand when or how we were going to get there.
And following Sprint15 this week, which is Whitehall's annual event to discuss all things digital, I now realise that I was a bit naïve in my writing. What was clear from this year's event is that government-as-a-platform is at the top of everyone's agenda. However, what also became evident, is that government-as-a-platform isn't a project along the lines of what we have seen previously within GDS - it's a movement, an ecosystem, that could fundamentally change the face of the Civil Service as we know it.
And that's not me being hyperbolic.
Whilst government-as-a-platform sounds like an incredibly simple and basic idea, which is actually quite easy to define, I underestimated the complexity in getting this right. What will ultimately look like one joined up government (hopefully), built upon a bunch of common goods and services, is actually a beast that has so many moving parts, different players and an array of problems to solve.
But, I am optimistic that we can get this right. Whitehall just better be prepared for the disruption...
Changing gearsOne of the most common criticisms facing GDS, and something that we hear time and time again, is that whilst they have done a very good job of building some nice looking websites, they haven't really proven themselves in tackling the backend. I myself have wondered whether this is true.
However, what I hadn't considered, was that GDS was having to prove itself. It was having to show Whitehall and the public what it was capable of doing – at the most basic level. It was setting the rules – such as the design principles – and it was managing expectations. It was building its own opportunity, getting itself into a position of power, where it could showcase the exemplar services and then change gear. I believe that this is now what is happening.
Sprint15 was an opportunity to highlight this and the wheels are already in motion. We already have GOV.UK as a platform, and we are close to having GOV.UK Verify – the UK's new identity verification platform. But once the election is out of the way, things will accelerate.
Cabinet Office Minister Francis Maude explained:
Before the last election, we were wasting money because departments thought they were unique and distinctive and convinced themselves that only the most expensive bespoke systems would do.
But it’s nonsense.
If one department needs a service for publishing information to the web, so will other departments – so why should they all pay money to different people to achieve the same end? Why not just build a publishing platform they can all use and share?
That was the principle behind GOV.UK. The same was true of Digital Marketplace and GOV.UK Verify. Common systems that can be used by departments and agencies across Whitehall, removing duplication, minimising risks and cutting costs.
Numerous other services across government that could - and should - share common platforms.
Maude said that some of the platforms that we can expect to materialise after the general election in May – providing there isn't too much of an upset – include a common payments platform, a common system for appointments bookings, with messaging likely to follow.
Director of the Government Digital Service, Mike Bracken, also suggested that status tracking would be another platform to be developed. He took Sprint15 as an opportunity to explain that GDS will no longer be about websites. It's now a platforms game. He said:
Much of what we have had to do is remedial. And yet we offer so many services that transforming them one at a time isn't going to change things quickly enough. We can't go through one at a time, it won't get us there quickly enough, it won't get savings quickly enough.
We do need platforms, because we will get a more joined up government. And it represents a fundamental change to how we deliver public services. It isn't about technology, it's not about websites, they're just a visual indicator – it's really about our values. Government-as-a-platform is nothing if we don't believe in values.
How come our systems and our processes and our technology doesn't have any of the values that we have personally and in our working lives? We have to combine those. We think about our people as public servants, we think about our buildings as places where public services happen, and yet for decades we have not thought about our software as public services. Our software is a public service.
For a visual explanation of what Bracken and Maude are speaking about, check out this video from GDS, which has been used internally to explain the concept:
We get it – so how do we get there?
Well, this is the problem. Whilst there are some incredibly digitally advanced governments out there – such as Estonia – I would argue that none of them have managed to reach the holy grail of 'government-as-a-platform'.
There's so many reasons for this. One being, if we create a platform for government, that platform literally then becomes government. What does that mean for all the siloed departments and organisations within the civil service? A culture shock, that's what. If all of the departments are using common services and goods to carry out their transactions – that suddenly means that they're not very special, they're operating together. That's a very drastic shift from where we are today.
One of the real 'wow' moments from Sprint15 yesterday was when an ICT policy advisor from Estonia said on stage that the Estonian government has built into law that the government cannot ask a citizen for the same data twice. If an agency has the data, and another agency needs it, they have to share it. This is a world where government operates as one entity.Also – the most successful platforms out there in the market today (think Apple and iOS, think Google, think Amazon), they are viewed as incredibly successful because they have opened up their platforms for others to build services and goods upon them with APIs. Apple now has over 1 million apps on its app store, which it could never have achieved on its own. It crowdsourced endless possibilities. This is something government-as-a-platform is aiming for, but one of the problems I see, is that it isn't immediately obvious how this is going to happen.
Tim O'Reilly, an open source expert, was speaking at Sprint15 and he explained that GDS has paved the way for this with its exemplar services, which show people what is possible, but also noted that government is actually in direct competition with the private sector over standards. This is a big problem. He said:
Before Apple opened up the platform they built some beautiful, easy to use apps that showed the capabilities of the platform. Guess what, GDS did that too. They got people used to the expectations of what government services should look like. They did that hard work to start thinking about the underpinnings to make it simple.
If data is the 21st century platform, we need some standards. We need standards for things like identity verification, payment, location, credit history, health history. The problem right now is that those standards are being set by companies like Facebook, Google, Apple, Amazon – and government is in some way in competition with these services that are getting ahead of government's governance. The problem with that is that when someone else sets the standard, you are always having to explain yourself when you want to do it differently.
My main concern on this front – and I hope to be proven wrong – is that there are few 'big' examples of where government data has proven to make a profit for anyone. There are examples of companies springing up through the use of open data, but I'm not entirely convinced – or perhaps it is better to say I don't entirely understand – how government data can be used to create a platform of applications similar to that found on the app store, for example.
One of the main issues I see is data control. Will the government organise data in a way and get control of data so that companies – with approval from citizens – get access to data from across what are currently a number of siloes? O'Reilly tweeted me saying that blockchain could have a role to play, which is very complex, but very interesting.
O'Reilly also argued that the government needs to understand what its role is in all of this to make government-as-a-platform a success. For example, with Verify, it isn't actually carrying out the verification process itself, the private sector is, but government is setting the rules about who owns the data, what they can use it for, etc. These are the conversations that need to happen going forward.
O'Reilly said that it is okay to use the private sector for common services, stuff that government doesn't need to build itself, but it should retain control and “set the rules of the road”.
- SPRINT15 - what Francis Maude learned about government IT from House of Cards Francis Underwood
- Still standing - Francis Maude, UK minister for digital thinking
- Election 2015 - Goodbye Francis Maude, the minister who wouldn't go away - much to many IT providers annoyance
Coming back to my headline, I don't think I previously comprehended the impact that government-as-a-platform could potentially have on the Civil Service. One person speaking at the event that really illustrated the possible impact was Mark Thompson, Senior Lecturer University of Cambridge and joint owner of Methods Group.
Thompson got a huge round of applause during his session when he said that if we got this 'government-as-a-platform' thing right, we could genuinely change the political discussion from 'cuts' to 'getting more for less'. He said that vertically integrated organisations, such as government departments, attract self-service bureaucracies, which are expensive. Government-as-a-platform could change all this. He said:
We don't have any political engagement about how we are going to reconfigure public services and the civil service away from bureaucracy and measuring and monitoring, to stewardship of common consumption. Platform is about common capability, it's about government sending a message that it is going to start to do very common things in very common ways.
There is a cost to this though – the cost is that we have to take a national conversation to the country that says effectively we can't maintain the vertically integrated, bureaucratic jobs in public services any longer because we can't afford them.
If only we could sort out our backend business model. My primary motivator is that we can actually have more public services, rather than less, and therefore the political discourse in this election about public service cuts doesn't need to happen necessarily.
In other words – fewer people on the backend, more people on the front line. That's a huge and scary proposition for the public sector as it stands today, but a discussion that's worth having. It's about nurses and teachers versus paper pushers. This is what government-as-a-platform is all about.
I'm not going to sit here and pretend that I completely understand the notion of government-as-a-platform. A simple idea with a world of complexities behind it.
But from what I heard at Sprint15, I understand that it has huge potential to modernise the civil service. If we can get this right, we will genuinely have the whole world watching us.