A leader's dilemma: Get a grip? Or let go?

This week I’ve had some great conversations with former civil service colleagues about how to do change. I also attended the launch of a report on a vision for public services in 2020, and a seminar by Brian Dive, author of ‘The Healthy Organisation’.

I’ve found myself chewing on the leadership dilemma of ‘controlling’ or letting go. This isn’t a new dilemma for me. It haunted me for most of my time as the director of a caseworking operation.

My default preference is towards letting go. The more I look at things systemically, the more attracted I am to devolving authority and power, to trust the frontline.

The theory goes like this. From the centre you get the birds’ eye view, from the balcony, and you’re able to make strategic decisions, for example changing organisational priorities, or responding to external changes in context. But you don’t get the detail and local variation, and in not getting the detail, you inevitably mess up implementation. So it’s best to let the detail be handled closer to the frontline and allow sufficient flexibility for first line managers to do that.

That’s for performing the job you know about and are trying to manage well. Throw in a dose of chaos theory, and add to your organisation’s context the idea that you can’t predict everything that’s going to happen to your clients and your staff. Then you need even more local adaptability and autonomy to deal with sudden shocks. (This is the Ralph Stacey way. I haven’t read Managing the Unknowable, but have had it recommended)

There are lots of examples where ‘the centre’ failed to do this well and local implementation hasn’t been sufficiently nuanced for local circumstances. The target setting that happened under the Prime Minister’s Delivery Unit in the early 2000’s had some beneficial effects but some detrimental ones, particularly around dealing with variation. (Review of ‘deliverology’. I also like this article from the Systems Thinking Review about benchmarking, showing how that doesn’t deal with variation either.)

In addition, there’s good psychological research that upholds the idea that the more autonomy you have in your job the happier you are in it - and therefore the more ‘discretionary effort’ you put in to the job.

Certainly this approach should be more beneficial for partnership working and collaboration. When I allocated cases to casework team leaders on a locality basis, and then gave the autonomy to team leaders to contact the relevant local authority directly, rather than trying to manage relationships with about 20 key local authorities myself, suddenly our stakeholder management improved vastly. We were able to plan our caseworking more effectively to deal with local authorities’ anxieties in a more bespoke way that we could never have managed if we were planning centrally.

So my prejudice towards letting go was comfortably confirmed by the 2020 Public Services Trust at the Royal Society of Arts. I went to the launch of their new report on Tuesday. In particular they argued for a shift in culture towards coproduction, and a shift in power towards devolution.

The shift in culture they asked for was for public servants to value themselves only to the extent to which they engage and enrol citizens, families, communities and enterprises in creating better outcomes. That all sounds a bit jargony. But I understood it to mean, blurring the line between the ‘state’ and the ‘community’ so that, (to quote a song with an entirely different intention!) ‘sisters are doing it for themselves’. If you think about it, a child isn’t just taught by a teacher. A child has to be up for learning for that to work. And it really helps if the parents are involved too. This is co-production. But, as Stephen Dorrell MP said at the launch, none of this can be delivered unless there’s a change in culture, public servants need to see their role differently. And this is the hardest and most important thing.

The shift in power was basically anti Whitehall.

“Our Whitehall model cannot deliver the integrated and personalised public services that citizens need. We need to invert the power structure, so that services start with citizens.”

Departmental silos should be replaced with citizen-centric and place-centric organisation structures, and with much more autonomy for local authorities and other local bodies.

Hooray, then, for ‘letting go’, trust, complexity and localism.

But, but, but. One of the members of the audience (who were in general very select and impressive - there were more than a couple of Chief Executives and key British opinion shapers in the room) at the RSA on Tuesday night asked, ‘Is localism more expensive because you can’t get economies of scale?’ Systems thinkers often snort at the idea of economies of scale and say this is a myth, given the waste that gets created by dealing ineffectively with variation and detail. (A long piece on it by systems thinker John Seddon is well argued.) However, my experience of running a caseworking directorate did throw up this shadow side of devolving power. I gave more autonomy to team leaders. This vastly improved relationships with stakeholders and our client group. But after a couple of years, I also found a lot of double working and inefficiency had perpetuated within teams. Best practice from one team wasn’t being picked up and implemented by another team. In fact, in the worst cases, the autonomy had allowed bad practice to flourish.

There are plenty of people who work in the centre of Government who have this experience over and over again. They feel like they’ve set the direction, they’ve made the mission clear. But what they have designed isn’t implemented in the way they had envisaged.

How about this for an example. Different Whitehall departments have in the past had very different approaches and some autonomy in setting pay for a new recruit into the Senior Civil Service (SCS), or indeed in offering a package on voluntary redundancy. In so doing, we’ve probably seen a lot of inflation in average SCS pay. Mostly to attract ‘experts‘ from the private sector, particularly in IT and change management. (Who wouldn’t want an aboslute expert in charge of all the benefits and pensions data of the whole nation, at the Department for Work and Pensions (DWP)? But even with that thought, it’s hard to swallow that the Chief Information Officer at DWP is paid around £250,000, which is about £50,000 more than his boss, the Permanent Secretary at DWP. Full high earners list. Perhaps we needed some tighter central control, or even accountability to parliament, for these kind of appointments?

That’s a ‘spending public money’ example. There are examples all over the place of how it looks like too much freedom have resulted in things have going horribly wrong. The case of Baby P is extreme and it feels almost inappropriate to invoke it, but it shows the worst kind of consequences. And there are standard policy solutions to this kind of situation. Minimum standards. Inspections. Guidance. Regulations. Audit. Tight performance management. Central databases to get organisations who deal with the same client to share data.

The question is, how much, when you devolve autonomy, can you trust those you devolve it to, to do the right thing? What is the ‘both/and’ answer here? Clearly it’s good to devolve. But you can’t let go so much that things go wrong.

I’m reminded of something the good people at Harthill consultancy said on the training I attended last year. You can’t ask an organisation to behave from a strategist action logic (i.e. looking at connections, at the big picture, integrating lots of different ways of seeing the world, multi-perspectives) if it hasn’t got it’s ‘bottom floor’ of basic systems and processes right first. A fellow pupil on that course described it as choosing to behave ‘like a War God’ when you’re dealing with people who need direction and clarity. The ultimate in situational leadership, maybe? The bottom line is that maybe there’s a very important place for the rules, the standards, the frameworks, the audits and inspections and standard operating procedures.

That’s the both/and. “It depends on the context”. My question is, when do you know you’re overdoing the ‘getting a grip’ and you should choose to let go?