Why GDS doesn't matter - the questions for UK digital government

The summer parliamentary recess is meant to be a quiet time for anyone writing about government matters, but for those of us following all things technology it’s been an unexpectedly busy period.

Mike Bracken’s surprise departure as digital government chief brought a burst of analysis, blogs, opinion, speculation and doom-mongering across the board. If you want to read the definitive version of why Bracken is leaving, take a look at Computer Weekly’s exclusive in-depth interview to see the words of the man himself.

But many questions remain up for debate about the post-Bracken environment, not least the future of the Government Digital Service (GDS) that he leads, and the government as a platform (GaaP) strategy that he hopes to establish.

Speculation suggests that GDS will see “savage cuts”, as one source put it. Civil service CEO John Manzoni has refused to back GaaP, according to other rumours. Neither is entirely wrong, but neither are strictly true either – the situation is more nuanced than that.

I wrote soon after the election that the jungle drums even at that time suggested the future for GDS was as a smaller, more focused operation, with greater power handed back to the departments. That seems increasingly likely, but the shape and role of GDS in this sort of set-up is as yet unclear.

Cabinet Office minister Matt Hancock made his first response to the post-Bracken rumours in a blog post, affirming his support for the “next phase” of GDS without saying what that next phase might involve.

Hancock’s post also mentioned that he had been discussing digital government with consultancy PWC, a statement that went down like a lead balloon with a lot of the government digital community on Twitter, with its hints of a return to the old Big IT and Big Consultancy models of the past.

For what it’s worth, having talked to a number of contacts and with Bracken himself, this is my take on the current situation and the possible future…

What next?

GDS has been hit with budget cuts – and they have affected its capacity to deliver in the short term – but so far these are in-year cuts, and are in line with what has happened in other Whitehall departments. But also, the Cabinet Office was given £55m in additional funding in the summer budget specifically for “efficiency and reform”, which is mostly being shared between property overhauls and defining business cases for further digital government projects.

The real issue is the spending review, currently underway, which will set the spending parameters and priorities for the next three to five years of the new government. GDS, like every part of the government, has to submit its business case for funding and what it will deliver with that cash.

What about GaaP?

On GaaP, nobody has dismissed it. Bracken says it is “not true” that his boss, John Manzoni, has refused to back GaaP. Manzoni may well have questioned the strategy – that is his job, after all – but the future of GaaP, like so much else, is dependent on the spending review. Bracken is finalising the business case for GaaP, following the financial principles laid down by Manzoni and the Treasury – a piece of work that has taken several months to put together, and has involved input from every major government department. Its future will be decided by the Treasury.

There is no lack of commitment from Manzoni, Hancock, head of the civil service Jeremy Heywood, the Treasury, or Number 10 for the principle of digital government and the savings and improved services it can deliver.

There seems little doubt, however, that Manzoni and Bracken have disagreed on how the digital transformation of government will be delivered. Bracken insists that transformation must be led by a “digital centre” combining digital, technology and data expertise to act as a “lever” to make things happen in departments.

Manzoni – pressured no doubt by departmental permanent secretaries whose instinct as civil service mandarins is to dismiss any central control over their strategy or budgets – prefers a departmentally led approach.

Bracken, with plenty of historical justification, says that departments are not set up to work on cross-government projects such as GaaP, and will naturally revert to their own siloed priorities. That’s not a criticism of departments – just recognition of their natural inclinations and the way the civil service has always worked.

Bracken quit, mostly because after four years of receiving full support from former Cabinet Office minister Francis Maude for building up GDS as the digital centre, he has been forced back to square one, justifying his plans again and again, and seeing the wind changing direction towards pushing digital power back to departments.

No proven model

Much of the debate has got overly wrapped up in the whole government as a platform idea. It’s important to remember that no government in the world has yet fully implemented GaaP – even if some including the UK have created individual platforms. There is no proven model to follow at the scale of the UK government. Bracken has his vision of how to deliver GaaP, but his is not the only vision. There are plenty of people who feel GDS has gone in the wrong direction and that it’s more important to define what GaaP really means in practice before committing to anyone’s vision of what it should be.

In Bracken’s vision, GaaP involves designing and building a series of common platforms and services – such as payments, status tracking, identity assurance, etc – to be used by all departments to avoid duplication and unnecessary cost. He says that cannot be delivered by individual departments alone.

An alternative vision of GaaP, such as that proposed by Mark Thompson, who has long been close to GDS and co-authored the Tory IT manifesto, Better for Less, before the 2010 election, sees GaaP as a set of enabling standards and principles, with the market providing solutions using data and APIs – minimising in-house software development. Think of it as an Amazon or Uber for government – enabling an ecosystem not building in-house software.

Both strategies are feasible, both could work, so could a hybrid of the two, but neither has been implemented in practice in full at scale in government anywhere, so nobody really knows. Someone in Whitehall has to decide which model – or even a different model – is the one to pursue.

It’s a question of leadership and accountability, not budgets or whether one person likes GaaP as a concept or not.

GaaP is going to happen in some shape or form – but it might not be Bracken’s vision of GaaP, and it might not be labelled GaaP anymore, but it’s going to happen. The question is how it will be managed and delivered.

Future of GDS

This is where the future of GDS comes in. Much as Bracken has argued for a digital centre of government, he said in his Computer Weekly interview that the size of that centre is less important. What matters is having some form of central body with the authority to make things happen – and importantly, to make things happen in collaboration with departments. It’s not about the size of the budget, it’s about what is the best way to do things.

For Bracken, that is a central team plus departments. The battle he is fighting is with the mandarins who believe it should be led by departments. There will still be digital government plans – the question is who leads, and who has accountability.

It seems likely that if GDS continues, it will be smaller, more strategic, and perhaps with a bigger role for Liam Maxwell’s Office of the Chief Technology Officer (OCTO) team. I get the impression from sources that there is a regrouping around Maxwell’s Better for Less paper of 2010, which was effectively the Conservative technology manifesto for the general election of that year. There are certainly people close to GDS who feel that more of the transformation GDS has delivered so far has come from Maxwell’s reforms of technology strategy, rather than Bracken’s digital services.

If you look at the recent figures published by the Cabinet Office for 2014/15 cost savings across government, £1.7bn was attributed to GDS-led digital and tech activities. But of that number, about £1.1bn fell within Maxwell’s remit.

The difficulty for digital transformation and GaaP is a Catch 22. The savings potential is enormous – but because nobody has done it before, it’s harder to prove and takes longer to deliver. The reforms Maxwell introduced have brought about real, lasting cost cuts in a short timeframe. If you have a Treasury mandarin on your shoulder, telling you to cut your costs by 40% in the next three years, which one are you more likely to back?

Digital government at a crossroads

Digital government in the UK is certainly at a crossroads, but the real challenge it faces is not unique to government, and applies to any organisation that is adapting to the digital revolution.

There is only so much transformation that any organisation can undertake, until it reaches the point where fundamental reform is needed. For companies, that means business model reform; for government, it means reform of the institutions by which public services and government policy is delivered. Perhaps the most salient observation in Bracken’s interview was when he said: “For most of this period, digital has not been an institutional challenge. Now it is.”

You can read plenty of articles online about the companies that failed because they missed this critical juncture in the digital transformation of their business.

At Kodak, it was when one of their employees invented the first digital camera, and was told not tell anyone about it so as not to affect sales of film.

At Blockbuster, it was the statement to the US stock market saying the effect of the internet on its business had been greatly overstated.

At HMV, it was the board meeting when executives told an external advisor that they believed music lovers would always want to browse through music in a store before buying.

Government, of course, cannot go bust like a Kodak or a Blockbuster. But it can continue to lose engagement with citizens, to foster cynicism and distrust, and to see the quality of public services deteriorate through continual “salami slicing” of operations without changing the way those services are delivered.

The UK government is at that point now.

Bracken is right to say that the next phase of digital transformation – whatever that might involve – is the step that will require fundamental change of government institutions. But that is not the mindset of the civil service.

Civil service inertia

There was a great blog post published recently by an ex-GDS staffer, Andrew Greenway, which described the 26 years of discussion and denials that led to the creation of a central government unit for recording official statistics. During that time, the civil service insisted statistics could be delivered by departments working together. It never happened. It took a major crisis – World War 2 in this case, and the need for better statistics to aid war planning – to make it happen.

This is how the civil service has been for decades. It’s not surprising it naturally resists change – that’s a lot of inertia to overcome. But the people who manufactured and marketed film for Kodak had worked that way for decades too. Their inertia killed the company. The Whitehall departments that wanted to keep statistics recording to themselves prevaricated until there was a war to force them to change.

What will it take for this UK government to understand and implement the radical reforms it needs, if it is to become truly digital? Five years ago, austerity was enough to give power to Mike Bracken’s elbow. It seems that a slowly improving economy has shifted the balance of power back.

GDS doesn’t matter

So, in that grand scheme of things, GDS doesn’t actually matter. What’s needed is political and civil service leadership from the very top for widespread institutional reform of the way government works, thinks and goes about its business.

If that is best delivered by a central team called GDS (or with any other name), then it needs the accountability and resources to do so. If that means a smaller GDS than now, that shouldn’t be a problem – as long as it has the backing to lead the necessary transformation, in collaboration with departments.

If that is best delivered by departments working together, then it still needs someone with the leadership and accountability to make sure departments look and act beyond the narrow confines of their siloed remit. History suggest there is little precedent for this to work.

If, after the spending review is over, GDS no longer exists, that in itself is not a problem if the commitment to reform is in place and is workable and has clear leadership.

If, however, digital transformation is left to individual departments, then savings can still be made and individual services improved in those departments, but the bigger prize of government-wide reform and dramatic improvements in efficiency are very unlikely to be achieved in this parliament.

Eventually, it will happen. Digital transformation is inevitable and unstoppable in every industry and every part of public life. The civil service will change – either through enlightened leadership in the next five years, or kicking and screaming at some point in future when it has no choice. The next few months will tell us which.

Join the conversation

2 comments

Send me notifications when other members comment.

Please create a username to comment.

Mark Thompson's belief that GaaP could make 1½ million public servants redundant and carve £35 billion off the deficit seems to be based on an analogy between Whitehall and a Dutch community care organisation.

This may or may not convince Ministers, but HMT are unlikely to fall for it. Or for his "certainty-ubiquity" graphs of "promising clusters".

GDS did nothing to displace the oligopoly of IT suppliers who too often deliver spectacularly poor value for taxpayers' money. GDS's demise/transformation is not a return to the status quo ante. The oligopoly was never threatened.

It should have been. The taxpayer wants, needs, deserves and pays for better public administration. But after a mere four years the "revolutionaries" ran out of energy/motivation – that is the revelation in your interview with Mike Bracken.

He laid down a barrage of cute phrases like "consistent, not uniform" that can't even have ruffled the feathers of Capgemini and Fujitsu, working on HMRC's ASPIRE contract, for example.

Time now for decision-makers who understand public services, who know what it means to "put the user first" and who require more than a multi-colour picture of promising clusters before they imperil the nation's ability to raise £500 billion a year in tax to fund the state.
Cancel
I echo others in congratulating you on a very thoughtful piece.

The ONS story that you quote is a good illustration of Peter Gershon's "discovery" that the Civil service has two speeds - panic and manana.

The OGC cartoons which EURIM was given to use in its 2009 paper on "Good Practice in Public Procurement" illustrate that the Mandarin's understand well how the process works.

They also know its importance because, until avid Cameron halted the ministerial merry-go-round, ministers came and went even more quickly than fast-track civil servants rotated on their way to the top.

Bracken tried to use support from above to change the resultant processes rather than support from below to use cross-department co-operation and rotation to make it work better.

The route towards successful GaaP is multi-track.

One is to "encourage" the creation of shared services (not just IT) for those departments too small to have in-house expertise. The most likely form of encouragement is to cut off the budgets that would allow alternatives, unless these a manifestly manifestly more efficient.

Another is the "right-sourcing" of "Civil Service Learning" to provide world class cross-boundary learning and career development experiences for future generations of would-be Permanent Secretaries.
Cancel

-ADS BY GOOGLE

SearchCIO

SearchSecurity

SearchNetworking

SearchDataCenter

SearchDataManagement

Close