What’s wrong with best practice?

Daniel Thornton

Best practice used to be best practice, but increasingly the argument is made that best practice isn’t necessarily best practice at all. This post does a through job of explaining why that might be.

It’s pretty clear that in complex real world systems, attempting to specify the steps towards an outcome with complete precision is unlikely to be helpful. It’s also pretty clear though, that many tasks and processes do have a substantial technical element for which there are best (or at least better) ways of doing things. The value of checklists – of structured compliance with a predetermined sequence of actions – has been clearly demonstrated for pilots and surgeons despite (or perhaps even because of) the fact that there is substantial variation in the context in which tasks are performed.

There are also more subtle – but no less real – forms of best practice. The shift in many areas of activity from basic competence to real expertise comes from the acquisition of tacit knowledge. Best practice is thus what best practitioners do – which doesn’t mean that what they do can be readily codified and copied, both because distillation of that kind is hard, and because the subtlety of judgement which experts bring is almost certain to be lost in the attempt. So perhaps the problem with best practice is not that people try to find it and apply it, but that they conflate adaptivity to complex systems with process compliance.

The argument of this post is that it’s worse than that, that best practice is an intrinsically unhelpful concept. In the specific context of organisational change – which is the starting point for the post – that may be so (though even there it is not meaningless to talk of best practitioners). But perhaps a better conclusion would be that for all its risks and limitations the idea of best practice shouldn’t be wholly abandoned. There is best practice on best practice which is worth understanding and developing.

Nesta Radical Futures brief

Sam Villis – OneTeamGov

A picture of a One Team Gov T-Shirt on a stage with One Team Gov bannersAfter 432 posts suggesting strategic reading, the 433rd is an odd one out, with a first suggestion for some strategic writing (or vlogging).

As a contribution to Nesta’s work on radical visions for the future of government, OneTeamGov is crowdsourcing ideas. Contributions are invited from people working in governmen, responding to one of two questions:

What does your work look like in 2030, what has changed and what has stayed the same?
In the next 10 years what would need to change for you to be able to do your best work on behalf of citizens?

It’s tempting to respond in part with Charlie Stross’s observation that

The near-future is comprised of three parts: 90% of it is just like the present, 9% is new but foreseeable developments and innovations, and 1% is utterly bizarre and unexpected.

That is perhaps a way of linking the two questions together. 2030 is to 2019 as 2019 is to 2008, and just as the bureaucrats of 2008 would not find themselves in a wholly alien world if they were to wake up in 2019, so the world of 2030 may well not be as radically different as some might wish. That brings the focus to versions of the second question – not just what would need to change, but what is the path to changing it, which would give us radically better government in 2030?

Follow the link at the top of this post to contribute your thoughts to the mix.

Making the centre hold: what works?

Tharman Shanmugaratnam – Institute for Government

Tharman Shanmugaratnam, who is Senior Minister and Coordinating Minister for Social Policies in the Singapore Government, gave the opening address at the Institute for Government’s tenth anniversary conference last week. The text of his speech [pdf] is at the link above, the video is below.

It’s something of a tour de force, drawing not just on Singapore’s own experience but on evidence and examples from around the world. But what is perhaps most striking is the level of integration of the policy thinking – education, housing, health and more, each seen as facets of the others, and each set in the context of broad social challenges. It is interesting both for the content and for the political and institutional context which makes the content possible. Singapore has some distinctive characteristics, of course, and not everything is or should be replicable or scalable (the management of ministerial careers through generational planning, is just one example), but the challenge of joined up government comes across as less insoluble than it is often perceived to be, with some clear examples of the gains to be had from doing so.

New platforms for public imagination

Kathy Peach – Nesta

Who gets to think about – and so to define – the future? At one level, self-evidently, we all do. But doing it systematically and thoughtfully is a luxury generally restricted to a few specialists, and translating into democratic decision making happens spasmodically, if at all.

This post is a survey of approaches and initiatives aimed at ‘democratising futures’, ranging from games to citizens’ assemblies and from the wisdom of crowds to the opening of expertise. As the post recognises, there is little hard evidence of what works well either in process terms or, more fundamentally, in terms of real world consequences. But that is an argument for doing more, not less, and there are some useful pointers to how that might best be done.

The Other Half of the Truth: Staying human in an algorithmic world

Sandra Wachter – OECD Forum

The problem of AI bias, once ignored, then a minority concern, is now well into the stage of being popularised and more widely recognised. It’s a debate to which Sandra Wachter has herself been an thoughtful contributor. The fact that AI can replicate and reinforce human biases is of course critically important, but it risks obscuring the fact that the seed of the bias is unautomated human behaviour. So AI which is no more biased than humans should be seen as a pretty minimal target, rather than the ceiling of aspiration.

This post is a manifesto for doing better, for rejecting the idea that the new ways need only be no worse than the old. It’s not about specific solutions, but it is an important framing of the question.

Becoming Prime Minister

Catherine Haddon – Institute for Government

Ministerial time and attention is the scarcest resource in government, prime ministerial time and attention doubly so. An impossibly hard job is then made harder by the circumstances in which people come to it and by the absence of meaningful preparation. This paper is a wholly sensible – and rather timely – attempt to help make the transition easier and the assumption of power more effective. Potential prime ministers would do well to read and act on it.

At the same time, though, it is an implicit acknowledgment of despair. The paper shows a system which works perilously close to the margins of not working at all and a concentration of responsibility and expectations for which preparation is not just inadequate but which it is hard to see how it could be made adequate. None of that is going to change any time soon, of course, so the need for this kind of pragmatic incrementalism is very real. But there is a much bigger and much more difficult debate waiting in the shadows behind it.


Andy Brogan – Easier Inc

This post is a double winner. The post itself, by Andy Brogran, has some important insights, but it is prompted by a presentation by Mark Smith which is a compelling account of what happens when you think differently about the delivery of public services and is well worth watching in its own right.

Brogan’s post includes a particularly clear and succinct account of why a process standardisation model borrowed from manufacturing is particularly unhelpful when thinking about the design of services – manufacturing works to deliver standard outputs from standard inputs through a standard process. But public services do not start with standard inputs and so cannot create value by applying standard processes to deliver standard outputs – and indeed the attempt to do so risks making thing worse, not better.

That serves to frame an account by Mark Smith of the work he has led at Gateshead, breaking into established processes to work out needs and root causes. An overdue debt can be a trigger for enforcement action, at risk of triggering a further downward spiral, or it can be a signal of an underlying need which, if recognised can be addressed. This is a combination of powerful, human examples and pragmatic approaches to understanding and meeting needs:

‘How much of what we do can we do to you?’ That’s not a great question. ‘What does a good life for you look like, how might we help you with that?’ That’s a better question.


Rethink government with AI

Helen Margetts and Cosmina Dorobantu – Nature

Much of what is written about the use of new and emerging technologies in government fails the faster horse test. It is based on the tacit assumption that technology can be different, but the structure of the problems, services and organisations that it is applied to remain fundamentally the same.

This article avoids that trap and is clear that the opportunities (and risks) from AI in particular look rather different – and of course that they are about policy and organisations, not just about technology. But arguably even this is just scratching the surface of the longer term potential. Individualisation of services, identification of patterns, and modelling of alternative realities all introduce new power and potential to governments and public services. Beyond that, though, it becomes possible to discern the questions that those developments will prompt in turn. The institutions of government and representative democracy are shaped by the information and communications technologies of past centuries and the more those technologies change, the greater the challenge to the institutional structures they support. That’s beyond the scope of this article, but it does start to show why those bigger questions will need to be answered.

Where system change meets agile practice; the multidisciplinary primordial soup

Catherine Howe – Curious?

alt="diagram showing circles of influence and control in a system" It’s been clear for a good while that the boundary zone between an agile project and a less than agile host organisation is often rife with friction, incomprehension and frustration. The value of reducing the friction is obvious; the nature of the best lubricant rather less so.

There are various more or less mechanical ways of approaching this – treating it essentially as the alignment of two models of governance. This post comes at it at a rather different angle, with a strong emphasis on finding approaches which deliver psychological safety for those involved and which recognise the different (and ideally complementary) value of different perspectives. Agile projects should carry on being agile, but the right way of thinking about systems is systems thinking. That sounds ludicrously trite, but is both less obvious and much harder than it sounds. As ever, Catherine Howe provides thoughtful guidance through the complexity.

Introducing Missing Numbers: a blog on the data the government should collect, but doesn’t

Anna Powell-Smith – Missing Numbers

Sometimes what is missing can be as telling as what is present. The availability of data drives what can be said, what can be understood and what can be proposed. So the absence of data can all too easily lead to an absence of attention – and of course, even where there is attention, to an absence of well informed debate and decision making. So there is something important and powerful about looking for the gaps and trying to fill them. This new blog is trying to do exactly that and will be well worth keeping an eye on.

Government as a Platform, the hard problems: part 3 – shared components and APIs

Richard Pope – Platform Land

The third part of Richard Pope’s strategic musings is as thought provoking a reflection as the earlier two, though this post is perhaps making a rather different point from the one the headline suggests.

It opens with the idea of small pieces loosely joined, which remains one best descriptions of the web and of quite a few other things besides and leads smoothly into the idea of shared components and services across different governmental organisations. On the face of it, that makes a lot of sense in terms both of system efficiency and of delivering coherent services. Institutional and power dynamics within governments don’t make that easy. The level of trust within governments can be surprisingly low to those who look from the outside and see something monolithic. There’s a whole host of reasons for that, but one of them is the lack of recourse if things do go wrong, with an understandable reluctance to place reputation and service quality on a foundation which is not itself robust. And so the post arrives at the fundamental question, which is not to do with components or APIs at all, other than as the visible symptom of a much deeper issue:

The question governments therefore need to answer is this: what are the appropriate characteristics of institutions capable of operating shared infrastructure for the greater good rather than the priorities of a thematic agency, while remaining accountable?

One answer is to create new institutions whose job is to do that – GDS in the UK is an example. But if, as this post suggests, that model is under threat, that may be a sign it might not be the optimal approach either. So the problem remains very real, the search for solutions continues. Networked government – of small pieces, loosely joined – remains elusive.

Computer says no

Ellen Broad – Inside Story

Ellen Broad is a technologist who is expert on policy, an ethicist who is an incisive thinker, a writer who has published an important and well argued book. She also, as it happens, is a woman. We all know that that last statement should be irrelevant to the preceding three. We all know that it isn’t.

But we can all too easily persuade ourselves that everything is more or less all right, that blatant discrimination is a thing of the past, that even though there are rotten apples, the barrel is sound. This article calmly and comprehensively demolishes that easy optimism. It is a very deliberate breaking of self-imposed silence, a discarding of the assumption that quiet compliance will somehow make things better:

If there is no “right” way as a woman to speak about gender issues — if there is no “right” way for a woman to take up space, to take credit — then silence won’t serve me or save me either. The only way forward from here is to start speaking.

In this essay, Ellen has spoken. She needs to be heard.

Mind the Gender Gap: The Hidden Data Gap in Transport

Nicole Badstuber – London Reconnections

Algorithmic bias doesn’t start with the algorithms, it starts with the bias. That bias comes in two basic forms, one more active and one more passive; one about what is present and one what is absent. Both forms matter and often both come together. If we examine a data set, we might see clear differences between groups but be slower to spot – if we spot it at all – skews caused by the representation of those groups in the data set in the first place. If we survey bus passengers, we may find out important things about the needs of women travelling with small children (and their pushchair and paraphernalia), but we may overlook those who have been discouraged from travelling that way at all. That’s a very simple example, many are more subtle than that – but the essential point is that bias of absence is pervasive.

This post systematically identifies and addresses those biases in the context of transport. It draws heavily on the approach of Caroline Criado Perez’s book, Invisible Women: Exposing the Data Bias in a World Designed for Men, illustrating the general point with pointers to a vast range of data and analysis. It should be compelling reading for anybody involved with transport planning, but it’s included here for two other reasons as well.

The first is that it provides a clear explanation of why it is essential to be intensely careful about even apparently objective and neutral data – the seductive objectivity of computerised algorithmic decision making is too often anything but, and why those problems won’t be solved by better code if the deeper causes discussed here are not addressed.

The second is prompted by a tweet about the post by Peter Hendy. He is the former Transport Commissioner for London and is currently the chairman of Network Rail, and he comments

This is brilliant! It’s required reading at Network Rail already.

That’s good, of course – a senior leader in the industry acknowledging the problem if not quite promising to do anything about it. But it’s also quite alarming: part of the power of this post is that in an important sense there is nothing new about it – it’s a brilliant survey of the landscape, but there isn’t much new about the landscape itself. So Hendy’s tweet leaves us wondering when it becomes acceptable to know something – and when it becomes essential. Or in the oddly appropriately gendered line of Upton Sinclair:

It is difficult to get a man to understand something, when his salary depends upon his not understanding it!


What we’re learning about how the sector understands “digital”

Cassie Robinson

Hard on the heels of Mike Bracken’s pithy definition of digital transformation comes this longer post, exploring multiple meanings of digital through an extended metaphor of running a café. The two don’t contradict each other, but what this longer account draws out is that what counts as digital improvement is different at different stages of organisational and technical development. It offers nine questions to which digital is the answer, or is at least part of it. The eighth is organisational design and transition:

If I were to design the organisation again today, from scratch, how would I design it, based on the new needs and expectations people have, and how the wider context we are in is also changing?”

That comes pretty close to Mike’s one liner – which is a reminder both that digital change isn’t necessarily transformational and that digital as an unqualified descriptor doesn’t describe anything very precisely, a point further emphasised by Anthony Zacharzewski’s tongue in cheek tweet in response:

Digital is Facebook, right?

Own It!

Matt Jukes – Notbinary

Hertz wants a new website. They do a deal with Accenture to produce one for $25 million. It all goes horribly wrong. And it ends up in court, which is the only reason anybody else gets to know about it. Nobody is particularly surprised.

That’s the point from which this post starts, rapidly homing in on the question of the expertise Hertz should have had, but didn’t have, on the client side, and the delusion of outsourcing the product owner role to the supplier. That’s not really about the contractual relationship (and so is just as important when outsourcing is not at issue), it’s about the nature of the product owner role, which this post captures beautifully.

One thing which comes out particularly clearly is that treating product ownership as an ‘agile’ role, rather than as an organisational role can contribute to the misjudgement at the heart of the Hertz/Accenture dispute. That confusion can be seen in other contexts too – in the UK government, for example, treating product ownership (or in their language, service ownership) as one of the digital professions risks introducing a version of precisely that skew (which doesn’t, of course, mean that it necessarily does in practice) and so makes it even more important to focus on the core attributes of the role, without their being submerged by the process – important though that is for other reasons.

Transactions v Interactions

Matt Ballantine – mmitII

Transactions are bounded, manageable and measurable, which makes it makes them appear easy to improve and easy to tell whether some forms of improvement have been achieved. The temptation which results is to focus on improving transactions and, more subtly, on making more things more like transactions. There are some advantages in doing that, which is why it is so often done. But the more things are transactions, discarding the rich messiness of human variability, the less they are interactions.

That’s the starting point for a post which turns out to be about something rather different: how do organisations best support collaboration and effective team working. The connection is that while transactions are readily scalable – that’s almost the point of them – interactions aren’t. So in a world where organisational change tries to make things more transactional in contexts where interaction is vital, working out how to scale and systematise interaction without killing it off in the process becomes particularly important.

Digital transformation is a leadership problem

Mike Bracken – UCL IIPP Blog

‘Digital’ is an increasingly problematic word, ‘digital transformation’ a doubly problematic phrase – in both cases because they are used as much to obscure as to illuminate. One of the difficulties is that ‘digital’ is slapped on as a random qualifier of all sorts of things, to the point where it doesn’t (and probably shouldn’t) add much meaning at all. Arguably the underlying problem comes from the fact that ‘digital’ started by being a word which said something about technology, and has become a word which says something about it’s not quite clear what.

Mike Bracken has neatly sidestepped all that with a definition which might actually be useful:

digital transformation is the act of radically changing how your organisation works, so that it can survive and thrive in the internet era

As the post goes on to note, transformation may deliver some technology change; technology change does not deliver transformation. But confusion on that point is understandable, because it is often the case that transformation becomes visible externally precisely as a technology based changed – and there is some faint irony in the fact that GDS and its first flagship product, gov.uk provide an example of how that confusion can be created.

Four ideas to build more inclusive innovation systems

Alex Glennie – Nesta

Innovation is often discussed as a very abstract process. People are, of course, involved, but they are slightly impersonal people, components of the innovation machine, not fully rounded agents. That’s a caricature, but not a wholly unfair one, and the caricature makes it less obvious than it might otherwise be that we need to look beyond the roles to the people who play them.

This post addresses that head on, focusing on the four specific ideas flagged in the title. None of them sound dramatic or revolutionary – and that is perhaps the point. They won’t solve the problem, but they are all useful ways of doing things better, and they are all well worth folding into approaches to innovation.

A brief introduction to digital government

Eddie Copeland

Digital government is both something and nothing. It is something because there’s no denying that people use the phrase and think they mean something by it. Digital technologies are a distinctive driver of change in modern societies and economies and they make things possible that would be otherwise harder or not feasible at all. Filing cabinet government. carbon paper government and even mainframe government are all very clearly different from digital government. And yet digital government is nothing. Precisely because digital is everything, calling something digital doesn’t add much to understanding, but draws attention to the technology, which is simultaneously vitally important and not the thing which really matters.

But given that the phrase is not going to go away, some clear-headed thinking about what we should understand by it and how we should apply is highly desirable. And that is precisely what this paper offers. It’s a draft open to public comment, so very much still evolving, but it’s already a useful overview. It neatly and successfully avoids the traps of equating digital with technology and digital uniquely with service design, but is slightly less successful in avoiding the suggestion that digital is done by digital teams.



pilots looking and pointing at a 747Fifty year old documentaries are not the staple of Strategic Reading, but this one is an intriguing insight into thinking about the future – a future which has of course become our past and present. It was made just a few months before the Boeing 747 went into service: the physical plane was very real, the implications for wider service design were very uncertain.

There’s an intriguing clip early on where the head of the Royal Aircraft Establishment observes that over the then half century of passenger flight, planes had doubled in size every ten years. The 747 fits that trend perfectly. Extrapolating to the next century, muses the Director, that would imply planes carrying 10,000 people. And you can hear in his voice both confidence in the trend and doubt about its implications, battling for dominance.

Half a century on, we know that it was right to doubt – the 747 was in some important ways on the inflection point of the growth curve. Planes haven’t got faster or higher or more comfortable since then, and they haven’t, with the brief and apparently aberrant exception of the Airbus 380, got any bigger.

That brings Herb Steins’ great line to mind, ‘If something cannot go on for ever, then it won’t’, and that in turn is an important reminder that in thinking about the future, it is unwise to assume that exponential change is unconstrained and indefinite.

Meanwhile, stay with the documentary to stumble across the world’s only vertical take off passenger plane, developed at the same time as the 747, but never quite attaining the same dominance of the skies.