Why Dominic Cummings is doomed to fail to reform state inefficiency

I was rather shocked by a story on the radio news last weekend. It was announced that the NHS was about to spend £40m to sort out the login to its IT systems, as staff are wasting lots of time with separate logins to a dozen or more systems. My first reaction was: “Why on earth has it taken them this long to get round to fixing this, if it costs just £40m?”. It was followed by a more depressing thought: they will spend their £40m and still fail to achieve it. And then: “What on earth possessed them to press-release such an embarassing story?”.

It is no wonder that the Prime Minister’s chief of staff, Dominic Cummings, has such a low opinion of civil servants and public sector management. The quality of management is appalling. It always has been. One of my earliest memories was just how awful the nationalised gas, electricity and telephone industries were in the 1960s and 1970s. It was worse than even I knew, as the government had lost control of their finances, so that these industries simultaneously drained the public purse and were starved of investment. And the vast scale of wasted resource on nuclear energy development only emerged long after it happened, with nobody accountable as usual. I do remember Harold Wilson (the PM of the time) having to cancel an absurdly ambitious development of a new bomber for the RAF. Things have only got better since then because the state has shrunk. One of many current fiascos is the Universal Credit system: a fatal combination of a visionary minister with a weak grip on reality (Ian Duncan Smith), and civil service project management skills.

It is not all bad. The London primary schools where I have served as governor are as well-managed as any organisation that I have seen, though some of the regulations they have to navigate, and the “guidance” from national and local civil servants can be mind-numbing. There are other pockets. These might give us some clues as why such poor management happens – it is nothing to do with being part of the state as such – and indeed many large monopoly private businesses can be just as bad. Partly it is about accountability and incentives, and partly about complexity. Alas it is impossible to get these right in many areas. Health services are mostly too big and complex (our NHS isn’t particularly inefficient in global comparisons, though I imagine few others have the nonsense over IT logins); the secrecy required for national security institutions is the enemy of accountability; and so on.

Mr Cummings wants to address this by a massive shakeup of the civil service, as advertised by his blog, which I don’t read, as it sounds too much like an ego-trip, but this article in the Economist gives more detail. The general idea is to change the culture, to allow in more mavericks as well as engineers and scientists, ensure proper accountability for failure, and to prize competence and the ability to get stuff done, as opposed to just being “a safe pair of hands” (i.e. staying out of trouble before handing the job to somebody else). As somebody who has had a career in management, and experienced the public sector at first hand through my two decades as school governor, I really get this. The serial incompetence of the public sector really annoys me. But with all my experience in management I can see at least three sorts of problem with Mr Cummings’s approach.

Firstly management isn’t just about getting stuff done. There are two types of skill required to be an effective manager. 90% of the attention goes on the first type of skill: task-orientation. Success is generally achieved by picking a small number of priorities and concentrating on them. This seems to be Mr Cummings’s obsession, and he is in line with most people who pontificate about effective management. But there is a second type of skill: risk-management. This is about looking for trouble and heading it off. What most people don’t seem to appreciate is that it requires opposite skills to task-orientation. Task orientation requires focus; risk-management requires the opposite – the ability to step back, gather data from all directions and think about the things that are not priorities. You also require strong judgement as to which potential risks to take seriously and which just to keep a wary eye on – but without that broad horizon it doesn’t work. Some of the worst management disasters (including in the British public sector, notably in the NHS) have occurred from areas outside management’s top three priorities. In theory the incorporation of good risk-management skills is perfectly compatible with Mr Cummings’s idea of maverick managers constantly challenging each other. In practice this is unlikely. In his recent blog he has said “I’ll bin you within weeks if you don’t fit”; this does not sound like the sort of culture that welcomes effective risk management.

Secondly there is a clear implication that decision-making is centralised. If you don’t trust public servants in generality, then you are unlikely to delegate much of importance to them. Mr Cummings is as likely to have as low an opinion of government ministers as civil servants. This would certainly be justified in most cases: good management skills are not a requirement for political advancement, and most politicians have had little aptitude for running things. That’s a recipe for everything coming back to Downing Street. And given that the name of the game in Downing Street will be prioritisation, that means that most of the stuff that is referred to it won’t get looked at properly, because it won’t be in No 10’s top three. This will result in either gridlock or things being waved through when they should be challenged.

And the third type of problem is lack of consultation. One of the most tiresome things for people who want to get stuff done is listening to a wide variety of interested parties who might be affected by what you are trying to do. Managers of Mr Cummings’s sort like to short-circuit such time-wasting and have a good chuckle when people complain. Consultation is hard work, and most of the concerns people raise are about making sure their lives don’t get changed too much, sod the bigger picture. But not only is it good for risk management, but it is also good politics. People get less angry if they have been consulted about things, and it is useful to know who the big trouble-makers are likely to be, and whether there are easy ways to head off some of the trouble.

In my recent post on Boris Johnson’s government I suggested that his and Mr Cummings’s approach would lead to cronyism. This is what happened to the academy programme driven forward by Mr Cummings at the Department of Education. Doubtless Mr Cummings views this as a success because he pushed it through further and faster than anybody thought possible. But it achieved little in advancing the quality of education. It was not in fact over-centralised – but the rather extreme model of delegation he adopted meant that the unscrupulous made hay. And his impatience for consultation meant that academies were run by a small number of politically well-connected organisations, and not the sort “bottom-up” local groups that the policy was supposed to empower. And it created a huge stink, which meant that he and his minister were moved on earlier than people though they would be. Now this failure may be because of the lack of competence of education department civil servants (I have certainly seen that claimed by an initial enthusiast disillusioned by the results) – but that was a factor that had to be managed rather than bulldozed through. In the end Mr Cummings and his boss Michael Gove were responsible for a classic public sector failure themselves.

So what should the government do about poor public sector management? Most approaches have drawbacks, which is why the problem is so persistent. Delegated management and accountability – part of the formula with schools – can work very well, but won’t for large, complex monopolies. Holding managers to measurable targets, the big idea of former Labour reformer Gordon Brown, often has perverse results and leads to poor risk management. Outsourcing, favoured by many Tories, has often disappointed: because the process itself needs to be well-managed and not treated as an exercise in buck-passing, and because it often means carving up processes in ways that make the whole harder to manage. Labour claim to have new ideas to improve accountability to service users; these should not be dismissed, but scepticism is warranted. I think regional and district devolution of political responsibility is an important part of the solution, but it could take a long time before this shows results. Politicians are likely to misuse their new-found powers at first, and the centre is also likely to implement it in such a half-hearted way that the benefits will be hard to obtain.

In some ways I wish Mr Cummings luck – especially with the country’s appalling defence procurement processes. But his ideas are both strategically and tactically misconceived and he may well end up making things worse rather than better.