Monday, 30 January 2017

Faulty by design: the state of think tank thinking

This month, the think tank Reform have published a report criticising public sector commissioning.  The title certainly doesn’t mince its words: Faulty by Design.

Long-term readers of this blog will know I’m generally pretty sceptical of these kinds of reports, and specifically the work of Reform.  Some of this is natural defensiveness.  When I read sentences that state ‘commissioners … do not possess the necessary skills’ (p.11) I feel attacked personally.  But what I want to suggest here is that part of my frustration at these sorts of reports is about something more fundamental.

(Before I move onto the fundamentals, though, I can’t resist highlighting out the oddity of some points in the paper – for example where they disapprovingly note that 72% of local authorities are planning to cut substance misuse treatment budgets [p.23].  I’d love to know what the other 28% are planning to do, when public health budgets are being cut by central government by 20% up to 2020, at which point funding will reduce to zero.  Or when they note that services to address homelessness are ‘commissioned by a plethora of providers’, when a provider is someone who is commissioned, not someone who commissions.)

One of the recurring themes on this blog is that while straightforward and honest thinking and writing – thinking to some purpose – should be the aspiration of all those involved in politics and policy, humans are complex, as is the world they make around them, and so we shouldn’t imagine there are neat, perfect solutions that don’t require compromise.

Without wishing to violate the first of those principles – that I should be straightforward in my thinking and writing – I’d suggest that the ontological model of the Reform writers is na├»ve.

There is a section entitled ‘Not knowing what works’, which is to misrepresent and simplify the mechanics not just of commissioning but public policy more generally.  Public policy is not as simple as identifying a problem and then implementing a solution.  Any issue will be fused with others, and any ‘solution’ will affect not only that one issue but those others too.

And in any case, it’s not that commissioners don’t know what works; it’s that this can’t be represented as an ‘intervention’, or even a set of interventions, that can be managed as part of contracts.  This model of ‘commissioning for outcomes’ imagines a world of unilinear causality (if that’s a phrase): somewhere we you can pull a lever somewhere in the machine that is society, and then monitor and note the effects.  But society isn’t a machine, and policies and interventions aren’t levers.

What ‘works’ in addressing substance misuse, for example, is a complex mixture of housing, employment, relationships, education, and any number of other factors that most people would simply refer to as ‘life’.  No one organisation, no one ‘policy intervention’ can produce the relevant ‘social outcome’, to use the language of the report.

To be fair to the authors, they acknowledge that part of the problem is that the social world is ‘extremely complex’.  But they still conclude the section by suggesting that there is a solution to this, and that it is ‘greater development of the knowledge base and better dissemination of existing expertise’, with the development of ‘what works’ centres seen as ‘a positive step’ (p.14).

That is, the problem is framed as one of ‘knowledge’: if only we did more research, we could develop the magic lever.

Instead, I’d suggest, we’d have better public policy if we realised that striving for perfect knowledge is futile as there are no magic levers, and the question we should be asking cannot be as simple as ‘what works’.

As such, the approach of this report and others can feel like a lament about the fact that the world doesn’t fit into boxes or categories, or that people don’t behave in easily modelled ways.  When the report complains of the reality of STPs (p.35), this isn’t the fault of the idea of STPs, and it’s not something that a policy or structure can address; it’s simply poor management and people not doing their jobs terribly well.  When, on the same page, the authors describe the tension between ‘what works’ and what the voting public want, we’ve got to the heart of the matter.  Policy isn’t, can’t be, and shouldn’t be, simply about ‘what works’.

And just as there no magic ‘intervention’ that ‘works’, there is no ideal structure for public services.  The report laments ‘the cost of fragmentation’ (p.20), but the fact is that the idea of seamless integration, or a definitive structure, is a chimera.  The cake has to be cut somewhere, and there are pros and cons associated with every option.  Anyone who has observed health policy for more than four years or so will notice the incessant back and forth of the size of administrative units.  The grass may always seem greener, but it’s as if people are actually hankering after flowers that even the greenest of grass won’t produce.

A case in point is public health.  The report complains that locating public health departments in local authorities ‘has stood in the way of integration elsewhere in the NHS’, but equally locally its ‘prevention’ function in the NHS would hamper its ability to shape key influences on health that sit within local authorities, like transport, housing, planning, licensing, schools, and so on.

There is simply no ideal policy on this, and yet the report uses the word ‘integrated’ as if it simply means ‘good’ when there will inevitably be choices of what and how to ‘integrate’, and there will be pros and cons to any approach.  Setting aside the fact that ‘the NHS’ doesn’t really exist as an institution for a department to integrate with, we’d have to acknowledge that public health departments can either be ‘integrated’ with local authorities or ‘the NHS’ – unless of course you’re planning ‘integration’ of the whole set of public services.  But rather than cutting the Gordian knot, this would create one, with different strands of complexity inextricably linked to one another.  And even in that extreme example there would be a dividing line: we’d still have to decide what elements of life are ‘public’ and what ‘private’.

But at points in the document it really does seem that the authors imagine a world without boundaries or departments – of otherworldly ‘integration’.  In fact the language is oddly spiritual, suggesting public services should ‘transcend’ current service boundaries (p.25).  I can only assume they are imagining an all-encompassing ‘service’ (or ‘intervention’?), ‘commissioned’ presumably by some overarching ‘public commissioner’ – a leviathan of the police, community safety, social care, healthcare, transport, and so on ad infinitum.

To go back to a cake metaphor, it might seem like I’m having my cake and eating it – that I’m asking the impossible of the report’s authors as I criticise them for failing to provide a solution to a problem I think is intractable by definition.  But that isn’t quite what I mean.  I’m criticising them for noting the complexity of the world, and criticising commissioners for simplifying it, before they go on to simplify it in their own way, which is no more intellectually or practically justifiable.

Life is complex, and can be understood and arranged in an infinite number of ways.  None of these ways is perfect, and the reality is simply hard work on the ground, not a magical policy or structure being delivered by government, policymakers or a think tank.

The report presents the idea of integrating health and social care (which is already an article of faith of STPs and the Better Care Fund) as if it is groundbreaking or will make all the difference, rather than focusing on the fact that this kind of development is simply difficult and requires hard work.  There is no structure or approach that makes it easy.

And this is where my real unease with these think tank reports lies.  They always feel like they are written by ‘policymakers’ or ‘wonks’, rather than people who actually have to commission, design or deliver these services.

There can be ‘integration’, but it won’t come from breaking down departmental boundaries or fiddling with commissioning budgets.  There will never be a single department or budget where savings across every social policy field can be identified and pooled.


Instead, integration must come from the inevitable ‘plethora’ of commissioners and providers sitting together around a table and talking about the multiple needs of individual people, or groups of people, at a range of geographical levels.  The levels and groups of people around a table will never be perfect, but there’s no need for a report explaining this.  There’s just a need for people to talk to each other and get on with the work.

Friday, 6 January 2017

Dry January and sponsored marathon running

On Tuesday, presumably to break me into my first day back at work gently, I was asked to comment on Dry January for Radio 5’s Drive Time programme.  I’d been asked because they were looking for a slightly sceptical viewpoint, and although I was a little concerned that this might not fit terribly well with my Public Health Dorset role, the team were actually supportive of me offering a bit of balance in the discussion, particularly as I’m more probably more moderate than many of the alternative commentators available.  Unfortunately for Five Live, this meant that the discussion perhaps wasn’t the most electric you might hear – though I did interrupt another contributor at one point, for which I can only apologise.

As I say, it’s might not be the most interesting 20 minutes of radio you’ll ever hear, but if you want to listen to it, I’ve copied it to my SoundCloud account here: https://soundcloud.com/will-haydock/dry-january-discussion-radio-5-whole-segment

I don’t want to exactly re-hash the piece here, but I wanted to outline my thoughts a bit more fully and clearly, and encourage a more involved debate either here or on Twitter.  (And something of a higher quality than this bizarre article pointed out to me by James Morris.  If you want a slightly different - and better - discussion of Dry January, you couldn't do much better than listening to this piece, including a contribution from James Nicholls.  It starts at around 15 minutes in.)

The first point I’d make is that it’s good anybody is talking about the possibility of going alcohol free.  Personally, I enjoy drinking and I don’t think health should necessarily be people’s number one priority when choosing how to live their lives.  But I do think we sometimes default to drinking alcohol when there’s no real reason to, and it might be worth reflecting on that occasionally – even if we then choose to carry on regardless.  Dry January opens out that possibility even for those of us who aren’t signed up.  By other people around me mentioning they’re drinking less, or not drinking at all, I might think twice myself (for better or worse, but at least for variety, and for the simple benefit of thinking at all).  And as I’ve pointed out a few times on this blog, lots of research – including my own – can be crudely summed up as proving that we tend to be pretty good at conning ourselves that it’s other people’s drinking that’s problematic, not our own.

And there’s no doubt that Professor Moore was right when he stated on the programme that there are significant benefits for lots of people from giving up alcohol, even if just for a month.

My concern is twofold, really.  What happens in the longer term, and is this something public sector organisations should be spending time (and potentially money) encouaraging?

In terms of the longer term point, my fear is that a month off alcohol doesn’t really fir terribly well with the standard approaches to behaviour change.  Generally, you look for realistic, sustainable change, which tends to mean small, gradual movements.  It could be argued that this isn’t the case with more extreme problems, such as addiction, where physiological detox is the starting point – but that’s generally undertaken where the long-term aim is abstinence.  In any case, Dry January is explicitly not designed for people who are dependent on alcohol.  That doesn’t mean that some of the people doing it won’t be looking to give up alcohol completely in the long-term, but that’s unlikely.

And that’s where my concern comes in: if you’re trying to encourage yourself to develop more ‘moderate’ consumption habits, complete abstinence isn’t a great way to train for that.  I wouldn’t enjoy Dry January.  I would find it hard and frustrating.  But I think I could probably do it (and I don’t think I’m entirely conning myself on this one).  But I wouldn’t necessarily have learnt a great deal about how to drink fewer pints when I do go out on a Friday evening, or how to not finish off a bottle of wine when it’s open – which are going to be the more useful ‘skills’ or habits in the long-term.

And what’s the message that a Dry January sends?  It’s drawing on the longstanding tradition of having a bit of a ‘detox’ in the new year to try to counteract the supposed ‘excesses’ of the Christmas period.  And so there’s a danger it just reinforces that pattern of drink heavily, then make up for it later by having some time off.  In fact, it looks like patterns of drinking are important in determining how harmful a particular level of consumption is.  Spreading the same amount of alcohol over several days or weeks is less damaging to your health than cramming it into one or two ‘binges’ or ‘bouts’.  As I said in the radio piece, it would be safer to drink exactly the same amount over two months, but rather than having December be ‘wet’ and then January ‘dry’, make both of these ‘damp’ – don’t take a feast and fast approach.  (The idea of labelling something 'damp' comes from this piece on 'damp feminism' - though personally I think I'd prefer the more idealistic 'wet feminism'.)

Of course there is some evidence that people do make changes to their drinking behaviour in the light of Dry January, and I would never deny that this might be a good thing for some people.  At the very least, you’ll reap some of those health benefits Professor Moore talks about.

But who is most likely to make these changes?  Well, unsurprisingly, looking at the main evaluation of Dry January in terms of who participated and made long-term changes (rather than Prof Moore’s work on the direct and immediate health effects), the people most likely not only to have low levels of drinking, but to have changed their drinking, are those who were drinking least to start with.

This isn’t really surprising, and isn’t a condemnation of Dry January, but it is something we should be aware of if the campaign is being advocated as something that will achieve public health aims.  Most of us don’t drink more than health guidelines advise, so it follows that for most people, while Dry January might be helping at the margins, that’s not where energy should be focused.  It’s not clear it’s that effective for higher risk drinkers.  And that’s only looking at the people who participate – which is in itself a self-selecting sample.  Again – and for both of these points we simply lack robust evidence – it seems that those most in need of support are those least likely to engage.

And that’s where I have my biggest concern about the coverage given to this campaign.  Is Dry January something public health departments should be encouraging?  Well, this year PHE aren’t pulling out the stops to support it this year, and in fact our local health improvement service is more likely to be running a campaign in February.

Although it wasn’t universally acclaimed, I have a lot of time for the review of evidence on alcohol interventions PHE published shortly before Christmas – and I’d argue that if we’re going to do anything to address alcohol consumption beyond treatment for those with dependency issues, we should be focusing our time and energy on the actions they identify as being evidence based.

That doesn’t mean that individuals shouldn’t do Dry January, and I welcome the development as a potentially useful natural experiment, but it needs a lot more evidence and people undertaking it need to be clear about what they’re trying to achieve with it.  It isn’t a get out of jail free card for previous or future excesses, unfortunately.

But as soon as it’s thought of as Dryathlon, now that’s a different story.  The whole movement makes more sense when I think of it like sponsored marathon running: it’s an excuse to get people to give to charity; it raises awareness of an issue; it’s not really going to engender long-term behaviour change – it’s about undergoing a trial and proving something to yourself (and other people).

(Incidentally, Dryathlon isn’t run by Alcohol Concern.)


So if you’re thinking of giving up alcohol for January, or Lent, or any other time – go for it (assuming you’ve not got signs of dependency that would make it risky).  You can even ask me to sponsor you, and I’ll probably do it out of some form of guilt.  Just make your choice of charity a good one, and remember that I’ll be giving grudgingly as I don’t really understand the point of sponsored events in the first place.  I’m happy just to give to a good cause, without any fancy dress or test of physical endurance being undertaken.