Systems Thinking in the Public Sector: the failure of the reform regime… and a manifesto for a better way: is John Seddon right?

Systems Thinking In The Public Sector by John SeddonI started out expecting I would largely agree with the book, and by the end although my agreement with much of the ideas wasn’t shaken, the way the book argues for them left me disappointed.

Take the book’s bald statement, “I have not seen any evidence that people want choice. I see plenty of evidence that people want services that work.” It’s an extreme factual claim – no evidence at all, indeed (and I guess means he’s not been a reader of David Boyle over the years, ironically given that David Boyle is a fan of Seddon’s approaches). Of course there is a very relevant debate about choice – does it help improve public services?

But John Seddon presents the issue in such a dismissive way that rather than enlightening the reader about the conundrum, he just breezes past it as if it didn’t exist at all – and also leave you wondering what role in his world there is for a member of the public who wants a GP service that works for them, which to their mind includes choice over when to see the GP. Is the idea of people wanting choice over when they see their GP so risible and trivial that it deserves such dismissive treatment?

Or take another example: “The assumption (for which there is no empirical evidence) is that people have different skill sets”. Really? No-one in the public services varies in what skills they have? And there I was thinking that my IT skills are much higher than my plumbing ones, with the evidence of the dripping pipe being a good piece of evidence. Mind you, I am ex-public sector.

There is a good point buried in here, about the way in which public service jobs too often are very specialised. Public sector workers are expected to specialise in very narrow tasks rather than to have the broader problem-solving skills which would better reflect the messy actual demands on public services.

Yet although the point about need greater and broader skills at the front end of public services is returned to elsewhere, the generality expressed with utter confidence that people don’t have different skill sets obscures rather than helps such a discussion.

What say should the public get in public services?

Moreover, there’s a illiberal thread running through the book as it’s not just the idea that people might want choice that it dismisses. Also in the long list of ideas breezily dismissed, for example, is the idea that the public should be asked what matters to them when it comes to setting targets or objectives for public services. “Such surveys can only yield unreliable data and invalid conclusions,” says Seddon in his sweeping dismissal of the idea that a public service perhaps should in part worry about what service the public wants.

In doing so, he again leaves behind an important debate. In this case, two of the crime-related ideas he derides are that the public is concerned about fear of crime – and so tackling that might be advisable – and also that the public likes to see a police presence and so therefore that might be something to try to balance with other calls on police time. Of course, dealing with fear of crime and putting police where the public can see them may well detract from using police resources to stop crime or catch criminals.

But the questions about how do you reconcile those competing demands do not get a look in as John Seddon instead discards as utterly flawed the idea that the public views on priorities matter or that fear of crime might be a significant problem.

Failure demand

Which is all very odd, not only in its own terms, but also when you turn to other parts of the book, where he argues eloquently and convincingly that the best way to understand public services and to improve them is to focus on the overall experience of individuals and how they get treated by different parts of the system.

That leads to the very useful insight of ‘failure demand’, namely how much of the work done by public services is caused by the failure to deal with an issue properly an earlier stage. As a result, apparently efficient services are really nothing of the sort. A call centre that deals with a large volume of calls, at a low cost per call, may look a success – until you then realise how many of the calls are generated by the failure of an earlier call to resolve an issue.

Repeatedly dealing quickly with progress chasing calls isn’t a sign of efficiency, it’s a sign of failure that costs more than a longer call which results in an issue being sorted first time round. Hence the point mentioned above about needing broader problem solving skills at the front line of public services rather than niche specialisms which shuffle people around repeatedly without anyone quite getting to grips with the underlying issues.

When is a target not a target?

That in turn leads to a useful discussion about the problem with traditional targets in the public sector. The distinction between a target (bad) and an outcome measure (good) can appear to a novice rather like a medieval theological debates at times. Whilst Seddon and his supporters are often very critical of targets and dismissive, verging on rude, about the main proponents of targets, they own preferred approaches still involve turning things into numbers where the numbers moving in one direct is bad and in another is good. For example, when I heard John Seddon speak about his approach, he started off with an example of a much improved local council housing service – and he led with two numbers to illustrate how improved it was, namely lower costs and shorter waiting times. Yet on the same occasion he was also very hostile to the idea that numerical targets are useful for improving public services.

The book sets out an approach for using numbers that help understand what is really going on in a system and which leave people in the public service free to work out the best way of providing a service, rather than micro-managing their choice for them. But ultimately the book didn’t persuade me that in practice – especially given the political and media scrutiny and pressure around public services – such measures wouldn’t end up being that different from a good target. They would certainly be better than a bad target, but if the measure tells you something useful about what is happening in a public service, and people are keen to see the public service improve, then it ends up morphing into a target, even if only at the behest of media stories covering the public service. Indeed, Seddon quotes W Edwards Deming approvingly saying, “A system must have an aim. Without an aim, there is no system”. But don’t tell him that sounds to you rather like saying you should have a target…

There is plenty of value in this book then, such as the importance of integrating policy making with administration so that the policies that are set are capable of sensible administration (echoing the point made in Conundrum), even if the author’s sometimes rather implausible sweeping claims and regular dismissiveness about almost everyone else often obscures rather than illuminates.

Got a view on this review? Then please rate it on Amazon. If you are interested in this book, you may also be interested in Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector by David Osborne and Ted Gaebler.

Buy Systems Thinking in the Public Sector: the failure of the reform regime… and a manifesto for a better way by John Seddon here.

UPDATE: John Seddon has been in touch to respond and says,

This review is a little unfair, so I write this to respond to two criticisms and offer Mark a challenge.

Firstly: “The assumption (for which there is no empirical evidence) is that people have different skill sets”. This is taken entirely out of context. I was describing how HMRC was assessing managers as being suited to ‘policy work’ versus ‘people management’, evidencing how far people will go in pursuing the wrong problem. There is no empirical justification for this distinction. It was a specific, not general, argument about skill sets.

Secondly: Choice. I have read David Boyle’s report. It was an attempt to seek ways to extend the idea of choice; the report didn’t question whether choice improved public services. The evidence for choice improving public services simply doesn’t exist. What research there is, is contrary to the view that choice improves services. See for examples here and here.

It is more accurate to see David Boyle’s examples of public services failing to meet the needs of users as just that, failures, rather than as an argument for ‘choice’ as a solution. I gave Mark a pre-publication copy of our report commissioned by Locality, which illustrates why public services fail to meet users’ needs and charts a way of working that not only meets peoples’ needs but dramatically reduces costs. It will be launched in March.

My challenge to Mark: Where is the evidence for choice improving public services?

UPDATE 2: David Boyle has penned a good response to John Seddon’s questions.

4 responses to “Systems Thinking in the Public Sector: the failure of the reform regime… and a manifesto for a better way: is John Seddon right?”

  1. Our team had a timeliness measure with a target – 90 per cent of requests processed within twenty days.

    We removed the target. We started turning around requests faster. Variation also reduced.

    Happy days!

  2. I can completely understand the Seddon approach. When confronted with failing systems and confused reform, the dissenting voice needs to be forthright and there is a risk to this, but one worth taking. Seddon is simply saying ‘No’ and offering a better, safer way. OK we can all throw criticism around but I’d prefer to work on creating a better system and that means a more stringent approach to shift existing mindsets. Gentle balance has achieved nothing so far.

Leave a Reply

Your email address will not be published. Required fields are marked *

All comments and data you submit with them will be handled in line with the privacy and moderation policies.