What are the most important lessons for Dominic Cummings and British civil service?

That is Tyler Cowen’s question. I worked for an MP for a couple of years, during which time they were a junior minister for about a year. These are my thoughts with little-to-no research or sustained reflection.

Because I am a Cummings fan, this will focus on the errors he might make in personability and stakeholder management. I chose this approach this because error avoidance is a way of thinking he is trying to introduce into Whitehall. It is worth saying that I am not against his general approach nor would I disagree with much of what he did at DfE. This is just what I see as the shadow of his strengths.

Getting things done means having people do them or get out of the way. People are always getting in the way in Westminster. I was once told by an official that they couldn’t send me some information I needed because they were ‘covered by the Data Protection sphere’. I explained that we all were (it’s the law) but that there are some exceptions for MPs. We went on like this for a few minutes and she hung up. Needless to say, she didn’t make the complaint she said she would, but nor did I get my information. She was wrong, and I think she knew it, but she won. Whitehall is full of these over-mighty people.

Cummings’ approach will certainly make some big differences (look at the acadamies programme) but he risks having people make life difficult for him in ways it is difficult to respond to rationally and therefore to overcome. He has quoted this aphorism more than once – ‘People, ideas, machines — in that order.’ Colonel Boyd. – but it applies equally to how you deal with the blob as to how you bring in superb talent. He knows this: he dealt with the education blob and the DfE. But it might be possible to underrate the ability of stupid and ineffectual people to act as a drag, diversion or delay on him simply because they are there and dislike what or how he is doing things. These over-mighty barons cannot be invaded or executed: they have to be managed.

In business, stakeholder management is as much about assessing personalities as it is about assessing people’s aims and requirements from a project. Something as unscientific, crude and even BS as a Myers-Briggs assessment can at least get you thinking about how to manage to the person as well as the strategy and the situation, and avoid having your plan blown up by someone pompous or over promoted. That’s why business still uses these BS-type measures. They are not right, but they are not entirely wrong and they do help navigate ambiguous problems. He ought to piss people off, fire them, reshape the system and generally fight the blob, but he ought to be careful to do it in a personable way. The physicists who helped him win the referendum can’t help him with this. To draw an analogy, the equivalent of dropping all your spend in the last week having fine-tuned your message for a month beforehand (ala Vote Leave) is akin to sabotaging someone at a meting with a superb plan they have been locked-out of producing. That works well sometimes. Other times it kills the wonderful pan.

My former boss had difficulties in the private office and ended up wasting so much time just dealing with the problems created by the personality clashes brought out by the disagreements. There was fault on both sides, and everyone has gone onto to other jobs very successfully, but it is important for the person looking to make change to be disruptive in the right way. Everyone who I worked with in politics was too interested in having the argument, rather than seeing the argument as a tool to make progress with (me included). This includes the ones who achieved the most and made the most progress.

Cummings has spoken a lot about MPs’ unsuitability for their jobs. They lack statistical understanding, have never worked in alpha organisations etc etc. My boss wasn’t like that, but still had problems that were unnecessary and which I don’t think would drag on for as long in dynamic private sector organisations. Personality disputes were the main reason.

Gove is famously polite and it is sometimes thought to be a front to make him a better political operator. I don’t know either way (and seriously doubt it) but the more people Cummings hires who are combative like him rather than conciliatory like Gove the more he risks picking (and therefore by definition losing) the wrong fight. There are a lot of people incentivised to defeat him one meeting at a time. And they are incentivised by hard things to solve like insane pensions, gold-plated working conditions and implied tenure.

It is even possible that, maybe even entirely unconsciously, the element of Whitehall he is trying to reform/remove will colonise against him and try and drown him in dispute, like bees swarming an intruder. Their manners are excellent. If any minster gets into a lift with any Civil Servant the CS will say, ‘Hello Minister’. It’s like being at school. Obviously most of them are just following the rules. But it was pretty passive aggressive at times. Some of them had to be asked again and again (and again) not to do it. Whitehall mentalities run deep, and there are advantages to that. But that politeness is one small example of the sort of tactics (like the ones Gove is supposed to use) that could allow people to win even when they are wrong, simply because they are the ‘nice’ one.

Going after these people in the wrong way might get Cummings’ tied up in so much paperwork, disputes, tribunals and dithering that he loses more than he wins. His changes will have a long-term benefit, but he needs short-term wins as well.

As a final point, one of the things you learn about Margaret Thatcher from Charles Moore’s excellent books is that while she was famously rude to her colleagues, she was astonishingly considerate to her staff. Civil Servants remembered and liked her twenty years after they worked for her. She mothered them, fed them, encouraged them, allowed them to tell her she was wrong. She believed much of what Cummings’ believes, in broad terms, but her methods were good for error avoidance. It is notable that her failures to be more personable were a big part of her eventual downfall.

You can read my second post on this topic here.