We are in an age when big government matters ever more – to pandemics, climate, inequality and more. If we’re to avoid a generation of disappointments, we need to address the (sometimes boring?) question of how government can work better as well as what it needs to do, the plumbing as well as the grand designs.
I recently gave evidence to a public inquiry that shone a bright light on the realities of how governments engage with technology and reform. This was the inquiry into the Post Office Horizon programme that’s set to last throughout 2023. It’s a story with big implications for anyone interested in public policy and it's a cautionary tale. A generation ago many academics gravitated to simplistic views about the virtues of the private sector. Now just as many propagate equally simplistic arguments about the virtues of government, particularly when it comes to shaping technology. The Horizon story shows the limits of both views.
Horizon was a disastrous programme to digitise UK post offices and in particular benefits payments. It started in the late 1990s, commissioned by the government via the Post Office and was undertaken by ICL, then owned by Fujitsu. In the 2010s hundreds of sub-post masters were accused of fraud as the system appeared to show money disappearing from their accounts. Some were imprisoned, and others committed suicide before it became clear that the system had made the errors, not the sub-post masters themselves.
This appalling story of mismanagement goes back many years. I played a small part in it, when I was in 10 Downing Street. I sat on a Horizon Working Group in 1998 that tried to put together a deal to fix the project, and wrote a series of memos to the Prime Minister. In the summer of 1999 I wrote a lessons learned minute which the inquiry had got hold of, which is why I was called to give evidence.
Reading minutes you wrote 24 years ago, often in a hurry, and one on Christmas Eve, is quite unnerving. Some of mine were minutely examined, on big screens in front of the inquiry (the session is available here on Youtube).
I felt pretty anxious that I might have played a part, however small, in a journey towards disaster, and struggled to remember any of the details of meetings quarter of a century ago.
Fortunately for me, and as much because of luck as judgement, my various minutes in retrospect looked quite sensible. Indeed, it appears I was the only person at the time in the heart of government seriously questioning the reliability and technical quality of the system.
The inquiry saw minutes from others – such as Peter Mandelson – assuring that the system worked and had been validated by independent experts, though at the time I never saw these validations. My then colleague Jeremy Heywood screened out my questions about the technology (even though our boss, Tony Blair, rightly responded to my minutes by saying that the viability of the technology was the key issue).
As I pointed out in my testimony this was a classic example of Whitehall making decisions with no one in the room with practical experience. It was insane that politicians were having to make judgements about complex technical and business issues, far beyond their competence or experience.
In the minute I wrote in mid-1999 I made recommendations for major projects many of which were later implemented, primarily focused on professionalising the work and ensuring clear lines of accountability (many others made similar arguments). I also later commissioned a review of PFI which sharply criticised many of its features. In this particular case the design of the PFI meant that the post office in general and sub-postmasters in particular played no role in designing the system. I also advocated methods that parts of the digital world were using at the time – agile, iterative, involving users – but which were still alien to government then.
The inquiry will run for at least a year. In retrospect, although the minutes I wrote were pretty blunt, I wish I had been even more assertive on the shortcomings of the technology. But the big message is to be careful of wanting too much government engagement in technology decisions. I have always favoured some role for government in steering the direction of innovation and for being strategic rather than only responsive. But we risk now over-compensating. Commentators who have never worked in government can easily picture an idealised government holding the reins. But the reality is different.
In an age when big government is returning it’s vital that we have realistic, thoughtful debates about what governments can and cannot do. Many of the most admired public agencies across the world are deeply grounded in professional knowledge. Others function because they face competition and can’t afford to do badly. And others still work well because of profound relationships with their users.
But if we design new structures for governments that miss these lessons – that bring ministers too close to decisions; that forget just how easily departments can mess up their choices; and that give monopoly power to agencies without clear accountability for results – we are likely to face disappointments. The 'how' of government matters every bit as much as the 'what'.
