Do we need radical civil service reform?

For the last two years I have been a Fellow of Practice at the Government Outcomes Lab (GO Lab) which is part of the Blavatnik School of Government at Oxford University. It has been an immense pleasure and privilege to have had this role, from which I step down at the end of the year.

The GO Lab is doing a great job in bringing together a wide range of thinking on how we better deliver pubic services and address complex societal issues, around a central focus on how we deliver better social outcomes.  This has also been pretty much the raison d’être of ATQ for the last seven years, which is in part why I ended up as a Fellow in the first place.

The Deputy Director of GO Lab, Nigel Ball, last week published this blog with some challenging reflections on whether and how a new government, with a substantial majority and clear mandate to govern for five years, will address the knotty problem of public service reform.  I am extremely loath to summarise Nigel’s complex and nuanced arguments, and would urge you to read his blog in full before you read this one, but his central thesis is that public services have, for several decades, drawn heavily on both private sector management practice and, through outsourcing and privatisation, direct provision.  This has led to a focus on top-down targets and contract-driven accountability for service performance and delivery.  Nigel argues that:

“Many people – and much recent research – would agree that these ideas, though once useful, are running out of steam. A new approach seems to be slowly emerging, with a very different core underlying idea: greater participation with communities. No longer should we view social problems medically, talking of a “treatment” at the right “dosage” which can be measured for its “effect”. Rather, citizens themselves, grouped by a shared geography, interest or need, can participate in designing and delivering public services. No longer should we use top-down contracting mechanisms, making both public and private delivery bodies answerable to government agencies for the services they run. Instead, government and its agents should give away some of their decision-making power, and work together hand-in-hand with communities and civil society organisations. This might mean they hand over more cash no-strings-attached, and trust people to be intrinsically motivated to do a good job. For some of the most evangelistic proponents of the new approach, even the idea that it is possible to “deliver” outcomes is misguided – if we focus instead on working out what a better system would look like, the outcomes we want will appear.”

My initial reaction to these ideas is that they have much merit.  In particular, I agree with the view that services should be designed with and for those they affect and are designed to help, rather than being imposed by those who think they know better; and that rigid contracting and commissioning models, backed by targets that distort delivery, can do as much harm as good. Beyond that, I would like to offer three ‘micro level’ observations and one ‘macro level’ thought.

At the micro level,  I think this sort of empowerment of communities can co-exist alongside some of the good things that have emerged from the prevailing models of the recent past, and that there are a few babies that I would not want to see thrown out with the bathwater.  These reflect both our own project experience and the evidence we have gathered through evaluations of others work.  They include:

• The value of better evidence for what works (preferably supported by a thoroughly worked-through theory of change). Such evidence is far from perfect, and  there are many limitations to how robust it ever can be in a policy/social context, but it is I think better always to start from an evidence base where it exists.
• The outcomes culture. Our consistent experience as both advisors and evaluators is that a focus on outcomes has benefits.  One of the major disappointments of national debate is that it focuses almost entirely on inputs, rather than on outcomes, or even outputs.  Witness endless bid and counter-bid throughout the election about who will build the most hospitals, recruit the most police officers, train the most teachers etc.  I have become a strong believer in the value of identifying, measuring and managing through outcomes, with and without monetisation, and would need some convincing that such outcomes-based approaches are a retrograde step
• Strong performance management. By this I don’t mean the process of managing performance, which I agree can be over-driven by arbitrary targets. But much of the research I have seen shows clearly that any project – in-house or outsourced, contracted or grant funded, outcomes-based or conventionally measured – is crucially dependent on the effectiveness of those charged with its implementation.  At one level this is a statement of the obvious, but it is arguable that one of the biggest failures of the top-down, private sector-style management approaches we have adopted is that they still tolerate far too much poor performance; and one of the benefits of outcomes-based contracts appears to be that they encourage swifter and more radical action to address under-performance, often by strengthening and sometimes replacing existing operational management.

My big picture thought is to wonder whether government should start to treat spending to improve social outcomes truly as investment – and therefore closer to capital spending on infrastructure than current spending on services?  Many of the projects in which I have been involved in recent years have been supported by what is effectively an investment case – that spending on a preventative intervention will avoid or even reduce spending on a later, crisis-driven intervention.  The problem is that such projects are far too small to make much difference, and as a nation we never seem to have the resources to twin-track to scale – to spend enough on prevention to make a real dent in the long-term costs of poor outcomes, alongside the ongoing expenditure needed to deal with those outcomes.

But if outcomes-based contracts can be funded by external investors – and deliver them a return when they succeed – it should not be beyond the wit of government to adopt the same approach, and fund much larger programmes from borrowing rather than from taxation, underpinned by a business case that shows a return on that capital.  There are numerous areas where this approach could be adopted but how about targeted public health programmes, support to families with complex needs, and interventions to reduce substance misuse for starters? Done properly, such investment could be at least as justifiable as spending on infrastructure.

And yes of course I know that the Treasury would, to coin a phrase, rather ‘die in a ditch’ than allow this, but even an old cynic like me should be allowed to dream sometimes.

Since he became Prime Minister Johnson’s chief advisor there has been much attention on the musings of Dominic Cummings.  In one of his long and detailed blogs criticising Whitehall, he observes that:

“Whitehall …….is parochial about its own past. One of the most useful questions one can ask is not only ‘who has already solved this problem?’ but ‘have we already tried to do X and failed?’  In the DfE there is no system to answer this question reliably. Unless you get lucky with an old-timer, you cannot know and because they abolished their own library you can’t even go and study it.”

There is perhaps much to disagree with in Mr Cummings’ musings, but this strikes a chord. I have written before about the impact of lost institutional memory in the context of public sector productivity performance, and a couple of recent ‘lived experiences’ have prompted me to think further about this particular bugbear of mine.

It is perhaps a function of being 50-something years old that I am often by far the oldest person in the room – that old timer that Cummings writes about – and consequently have nearly always come across challenges that others appear to be encountering for the first time.

But it turns out it is not just a function of age – it’s a question of having any kind of institutional memory.

Lived experience #1:

At a recent meeting of around 30 commissioners and providers of children and young people’s (CYP) services, there was a presentation by a government policy lead which outlined what the policy team had found out over the preceding two years about effective CYP interventions. After around half-an-hour of discussion it emerged from the 30 or so professionals in the room that this two years of policy research and development had done little more than identify as best practice what everyone involved in youth work used to practice.

But as this kind of discretionary public expenditure had been cut back since 2008, much of this knowledge had been dissipated: there was no ‘institutional memory’ of youth work best practice and the policy team had to find it out all for themselves again. In fairness, the policy team were from a different Department of State to the one that led on youth work policy (and different again to the one that reduced local government spending on youth work) – but the effect of silos on cross-government learning is a whole other topic.

Lived experience #2:

Senior staff turnover at an organisation I have worked with for over five years means that I am one of only two members of the project team left with some ‘institutional memory’ of the detailed reasons why previous decisions were taken. In fact, and not unusually, both of us operate in external roles to the organisation and are the only ones providing any form of continuity.

What to do about it?

There is a huge amount of energy wasted when organisations find themselves having to invest time and resources re-learning lessons from before. Loss of institutional memory has to be a drag on productivity and innovation.

In my view, one of the keys is to ensure that there is a succession plan so that when staff move on from teams (usually through public sector promotion or role change), there is a successor who knows what’s what in that team or area of the organisation.

I call this my succession obsession and work with all my clients to ensure that they have identified individuals to develop and bring along as a way of ensuring at least some continuity.

Another good practice is to ensure that decisions are fully documented as projects and programmes progress. This is especially for when things go wrong as well as when they go right. Post implementation reviews are also an essential part of good project management and delivery but are often overlooked or done in only a cursory fashion – often because project teams have already moved on to the next challenge.

Finally, maybe part of the answer is to incentivise staff to stay in post longer and reduce some of the incentives to move so regularly in order to advance their careers. Would it be possible only to allow a staff move when they can apply what they have learnt elsewhere? In other words, when they start to become useful institutional memory themselves.