“There’s something special about Behavioural Public Policy” or “There’s nothing special about Behavioural Public Policy” – Behavioural Public Policy Blog

Applied Behavioural Public Policy (BPP) in the UK has morphed from the formula of a nudge plus an RCT, to be a catch-all for applying the theories and methods of  behavioural sciences to public policy.   This magpie-like approach has brought benefits but leaves the field lacking a clear offer in a period of great opportunity and great risk.

Let’s start with the positive (and possibly self-serving) argument: there’s something special about BPP.  It fills a gaping hole in the government toolkit. 

Human and business behaviour is complex.  It varies from situation to situation.  Anyone who has tried to get their mother, brother, friend or child (or even themselves) to change will know the need to understand the individual in question well; know what’s going on in their life at the time; their habits, preoccupations, and motivations; to choose the right moment; and very often to have more than one bite of the cherry.  And most of the time, you won’t succeed by telling them what to do or threatening them with punishment if they don’t do it.  At least not in the long run.  

And yet many public servants are anxious that their organisation’s reading of policy challenges, and tools to drive change, are blunt and lacking sophistication.  Progress is slow and problems that feel solvable persist. 

The ‘nudge’ approach captured the imagination of public servants, including the one writing this, because it offered a richer, more subtle, set of considerations to understand the drivers of problems.  And a wider range of tools to address them.  

By pulling in the methods of robust evaluation, particularly randomised controlled trials, it also addressed a second huge frustration: the merry-go-round of attempted and partially-successful policies perpetuated by lack of evidence on what works.  

When you throw in a set of bright, movitated, advocates with connections to the highest-ranking decision-makers and a gift for communication, plus some eye-catching early impacts for the approach, it’s not hard to see why the movement gained attention – and traction.

Yes, but..

And yet, to many practitioners, that traction is slipping.

Those high profile, silver-tongued, advocates have less prominence and the decision-makers that embraced them moved on.  The eye-catching successes either aren’t forthcoming or aren’t catching eyes now drawn to the latest remedies for public policy’s ills such as data science and generative AI.

Nudges are under attack.  There are accusations of a lack of efficacy, a lack of replicability.  Some suggest they narrow the lens of public policy rather than expanding it.  Others that they represent a sticking plaster that prolongs underlying policy failure and obstructs reform.  

RCTs are falling out of favour, too.  Critics say they zoom in on what is measurable and miss the bigger picture.  

BPP without nudges?

Perhaps as a response to these criticisms, perhaps simply because behavioural practitioners tend, by disposition, to have a wide outlook, or perhaps because the nudge plus RCT approach seemed to work better in some organisations than others, much applied behavioural public policy has turned away from nudges.

I’m not sure there’s a neat label for this approach (which may be a problem – I’ll come back to this).  It sometimes defines itself in opposition to the ‘nudge plus RCT’ approach, which it rejects as unduly narrow.  In a less negative framing, you’ll often hear teams that follow this approach describe themselves as ‘magpies’.  They’ll ‘steal’ a method from here (perhaps a little ethnography from the anthropologists or a regression analysis from the statos) and a theory from there – ‘protection motivation theory’ from the psychologists, for example.   

They tend to enjoy a behavioural framework.  Using the ISM model to draw in material and social factors as well as individual, perhaps, or COM-B to unpack the various components that might block or unlock a desired behaviour in an individual.

A third common characteristic is a drive to shift focus ‘upstream’.  Upstream might mean broadening the focus of interventions beyond individuals to other players like businesses, or organisations, or all the way up to systems as a whole.  Or it might mean addressing different stages of the policy cycle.  Instead of augmenting already-developed policies (this is ‘downstream’), the focus might be on problem definition, uncovering their organisation’s assumptions about the drivers of current and future behaviour, or stepping in when the drawing board is blank in order to bring in non-standard, policy responses.

This approach has proved adaptable.  Behavioural teams have successfully turned their attentions to a huge array of government activity: crisis communications, change management, compliance, service design and performance management to name a few.

Sounds good, doesn’t it?  Enriching public policy, countering the limited, narrow, conventional perspective, without relying on overblown nudges and RCTs.  What’s not to like?

..is nothing special?

What’s not to like is that this form of BPP struggles to be special.  It’s not unique.  There’s nothing to it that isn’t covered by the principal tools by which government makes and delivers policy.

Strategy teams step back and look at challenges in the round, examining the different players, the circumstances they’re operating in, their motivations and their interconnections. 

Policy teams think through the means available to governments to change citizen and business behaviour, looking not just at economic incentives but also social and environmental drivers, as well as other factors, like feasibility, cost, and unintended consequences.  

Communications teams are armed with frameworks to consider the timing, the channel and the messaging to drive change, and are trained in tailoring those messages to sub-groups.  

Digital teams are experts in human-centred design, making interaction with government services seamless – with a focus on building ease of understanding, removing unnecessary frictions and helping citizens and businesses navigate complexity.  They’ve got the E, A and T of EAST covered.

What about discovery I hear you say, or evaluation – those are key parts of the toolkit of a behavioural team.  Well, government often has developed social research capabilities to scratch beneath the surface and get to the underlying drivers of citizen and business behaviour. 

Ahh, but they are over-reliant on what people and businesses tell them – and it’s widely accepted in the BPP world that people are not great reporters of their own motivations and behaviour.  Particularly not future behaviour.  

But here’s the rub, Government also has strengths in disciplines that search hard for revealed preferences to explain behaviour and model how it might change in response to interventions.  That’s what the statisticians, data scientists and economist do.  

Why we should care

Does this matter?  Even if they are duplicating skills, behavioural teams are deploying their corvid-like approach and adding value to the work of government. What’s the problem?

The problem is that the diversity of methods, theories and applications leaves BPP without a clear definition and a clear offer.  This makes behavioural teams vulnerable when resources are tight.  The officials who make decisions on the skills their organisation needs might decide that they can live without behavioural specialists, feeling they’ve got the skills elsewhere.  

This is why a clear definition, or rather, value proposition is important.  If government and academic leaders can’t confidently say what BPP is, or elucidate the value it adds, then they won’t able to advocate for it, to stand up for its value when the money is tight and the critics are loud.   

What might that definition be?

This is harder than it sounds.  Agreement is hard to find. 

Let me set out the crux of the debate.  In one corner, we have a ‘tight’ definition, with roots in the approach set out by Thaler and Sunstein in ‘Nudge’.  A government intervention would be BPP if it i) departed from the assumptions of rational choice theory (ie it’s not a ‘standard’ incentive like a penalty, law, or tax incentive which citizens will respond to if it tilts the cost benefit balance in favour of the desire action) and ii) it is liberty- or autonomy-preserving – it leaves room for individual agency.  

Practitioners find it hard to relate to this kind of definition.  

On the one hand, even those core ‘rational-choice-based’ tools of public policy can be shaped and enhanced by behavioural specialists.  For a penalty, say for failing to submit your tax return, to be effective it will need to be set at a level that catches the eye of its target audience (ie salient), it will require citizens to perceive the threat of enforcement to be credible, for them not to pick up on a social norm of widespread non-compliance, and so on. 

A sophisticated understanding of the decision-making the penalty is intended to shape can and should influence its design, communication and enforcement.  And this is what the behavioural practitioner brings to the table.

The autonomy-preserving requirement also feels awkward to practitioners.  Certainly, it’s one of the things they’ll consider when assessing different interventions.  But they generally won’t see it as a necessary condition for applying their toolkit.  To take a hypothetical example from my own area, online safety, if the Government decides it wants to take a more interventionist role, blocking sites or banning users that previous governments have tolerated in the name of free expression, the behavioural specialists will still feel they have a lot to offer – on how the policy might be communicated or enforced to maximise adherence, or on anticipating how the groups targeted might find work-arounds to new restrictions, for example.

In practice, many behavioural specialists in government have found that they can add value to almost any policy and delivery challenge.  That’s certainly been my experience.  Much policy and delivery thinking remains crude and the behavioural lens almost always enriches the process (what’s behind this lack of progress in the ‘business as usual’ approach is itself an important topic for the BPP community to consider).  That may explain the shift to a much broader definition along the lines of BPP covering any application of social and behavioural science to the development, design, and administration of public policy.

The problem here is that this breadth can leave outsiders scratching their head as to what exactly they’re getting if they invest in behavioural practitioners.  From a marketing perspective, the nudge approach had the benefit of some neat branding – a catchy, simple, memorable name and a definition that was tight enough for non-experts to grasp.  Moving away from this leaves BPP exposed to competition from rivals with (seemingly) clearer offers.

My twopence worth

It’s always easier to find flaws than to stand up alternatives.  I don’t have a tight theoretical definition to offer.  But, as someone who’s job has been to build and grow Behavioural teams in government, I do have ideas about how to package BPP to help it regain the traction it once enjoyed.

First, don’t lose the niche.  BPP needs a hook.  A successor to ‘nudge’, which gained such profile. For me, there’s hope here.  Just as nudges are feeling dated, concepts like ‘choice architecture’ and ‘decision-making’ are becoming more widespread.  No other disciplines are laying strong claim to these.  As behavioural specialists we need to latch on to terms like these and jealously guard them.  They are our ‘in’.

Similarly, the RCT, for all its criticism, remains strongly associated with BPP.  Experimentation (which is broader than RCTs, of course) carries twin benefits of remaining relatively unusual in government and of offering quantified outcomes, which are a key currency when bureaucrats are running cost benefits analyses of which teams to retain and which to disband.  

The risk of anchoring to niches is that they might seem too distant from the core business of an institution and seem ‘nice to have’.  Leaders might reasonable assume they can get away without running experiments or developing capabilities to understand choice architecture.  Satisficing, after all, is the norm in public policy.  So..

Second, use the niche as the way in to doing ‘deeper’ behavioural work.  Deploy a wide range of tools, bring in your frameworks.  This is where the potential benefits are the greatest.  Once you’re in with policy or delivery (or comms, or digital, etc) teams they will quickly see that you can help them with the countless unknowns they are wrestling with.  Taking the tax return example, that’s questions like whether those who aren’t complying are doing it out of confusion, or as a tactical ploy; whether they find the threat of penalties credible; what tone do they find threatening or reassuring in communications, and so on.  Although other teams – researchers, strategists – should be able to shed light on these questions, in practice they often don’t. A combination of fixed institutional perspectives, limited expertise and a simple lack of time and attention means that this space is routinely left open.

Third, to misappropriate a line from a children’s ditty, don’t forget to scream.  This kind of work can be rich and productive and add a lot of insight.  It may lead to new interventions.  But equally it may prevent new interventions being adopted, or policy dead-ends being avoided.  That’s valuable.  But it’s not eye-catching and it can be hard to demonstrate the difference it has made.  You won’t often get a pat on the back for not introducing a new policy, even if it’s the right thing to do.  

Build up case studies and testimonials, publicise the work.  You will often hear behavioural practitioners complaining that those who ‘get it’ and really understand policy and delivery love what they’re doing, but those more senior don’t appreciate it.  Supplement the niche labels (point 1 above) with compelling stories of the value of deeper investigation.  This will both show the value of the work and build familiarity with what behavioural practitioners offer.

From this you should be able to develop a set of value propositions tailored to your organisation, and that’s your ticket to a long-running and successful behavioural team. 

The future of BPP

In the UK in particular, we’re entering a period where budgets will be squeezed and behavioural teams will come into the budget-cutter’s view.  Cut-backs in adjacent disciplines like social research and comms will mean the behavioural specialists added value will be great.  But if BPP’s value proposition is not clear, it’s the one that will be hit.  The risks now are as great as the opportunities.