Interview: David Halpern

“My job is to help policymakers in Whitehall and beyond use what we now know about how people make decisions to introduce a more sophisticated model of shaping people and their decisions,” says David Halpern, head of the Behavioural Insights Team. Government, which for so long has tried to push people in the desired direction using financial carrots and punitive sticks, wants to use a far more subtle set of persuasive tools.

Change is afoot on Whitehall. And alongside all the massive change that can’t be missed – all the policy reforms, the central controls and the major reorganisations – there’s an underlying set of subtle changes to policies, documents and techniques. A more personalised way of addressing a tax demand; a new format for NHS prescription charts; a cross-industry agreement to reduce salt content in food – these alterations are small in size, but designed for maximum impact. Welcome to ‘nudge’: a form of behavioural economics that uses subtle interventions to gently push people towards desired outcomes.

Behavioural economics is a concept that, while newly fashionable in UK policymaking, has been in existence for decades. From the 1920s, when it was realised the number of road accidents could be dramatically reduced if white lines were drawn on the road, to the 1990s when charity fundraisers discovered that the inclusion of a pen in their direct mail packs resulted in much higher donation rates, there are countless examples of people being guided in a preferred direction.

Over the last few years, though, a number of events have encouraged the British government to think a little more seriously about the potential of behavioural economics in public policy. In 2002, Daniel Kahneman received a Nobel prize for his work within economics – despite being a research psychologist with no formal economics training. This was followed by the publication of Nudge: improving decisions about health, wealth and happiness, by American academics Richard H Thaler and Cass R Sunstein. Hugely popular, it used real-life examples to demonstrate how choice can never be presented in a neutral way, meaning people are always influenced to behave in particular ways.

Meanwhile, businesses were employing increasingly sophisticated marketing techniques. The energy and mobile phone sectors, for example, were beginning to develop many hundreds – sometimes thousands – of different packages that aimed to ‘nudge’ customers onto higher tariffs. If government wanted to regulate these vast and often confusing markets, it too needed to deepen its understanding and use of behavioural insights techniques.

Finally, there came the acceptance that hard tools such as legislation, regulation and taxation were both unable to achieve desired policy outcomes, and – as budgets fell – expensive to introduce, operate and enforce. If society’s deepest ingrained problems were to be solved, citizens had to be persuaded to change their behaviour and their lifestyles.

The deliberate use of behavioural theory in policymaking began towards the end of the last Labour administration, which in 2010 commissioned the Institute of Government to produce a report, Mindspace: influencing behaviour through public policy. Containing a wealth of evidence, the report was well received and its ideas were subsequently championed by the coalition. Indeed, so keen are David Cameron and Nick Clegg to explore how this theory can benefit policy that in July 2010 they established the Behavioural Insights Team (BIT) in the Cabinet Office.

Transformational theories
The man they appointed to lead the charge was David Halpern: former Oxbridge and Harvard academic, once chief analyst to the Labour government’s Strategy Unit, and currently senior fellow at the Institute for Government. His role is to introduce this new way of working across Whitehall, and to transform at least two (unspecified) areas of policy – and all while achieving a ten-fold return on the cost of his team. This is no mean feat, but Halpern does not seem in the least bit overwhelmed by the task facing him.

“The basic idea is that rather than necessarily going for hard regulation, you can use what look like quite subtle influences or effects; small changes which nonetheless can have a big impact on behaviour,” explains Halpern, who quietly describes the application of this theory as an opportunity “to shape more effective, less intrusive, lower-cost policies”.

He has certainly been busy getting the message out. In addition to workshops and seminars introducing civil servants to ‘nudge’ techniques, his team is running behavioural insights trials with several government departments, including the Department of Health, the Office for Civil Society in the Cabinet Office, the Department for Business, Innovation and Skills, and the Department of Energy and Climate Change.

And so far, so good. The trials, whose objectives include doubling the number of people joining the organ donation register and making homes more energy efficient, have recorded encouraging preliminary results (see box). This in turn has fostered widespread support for the project from senior officials, including cabinet secretary Jeremy Heywood and work and pensions permanent secretary Robert Devereux.
Even the House of Lords has expressed its approval. In the Science and Technology Select Committee’s (STSC’s) report on behaviour change, it recommends that the “cabinet secretary, in consultation with the government chief scientific advisers and chief social scientist, should take steps to ensure that civil servants with responsibility for policymaking have the necessary understanding of the importance of changing behaviour, and can identify the most appropriate people to consult in their own departments about the development of behaviour-change interventions.”

“What I think really helps bring all of this alive, what really persuades people across Whitehall, is when they see the results,” says Halpern. “Take the fraud work as an example. This is about a very small change, almost without cost, which leads to up to a 30 per cent improvement in a given outcome. When you look at that, how can you not be interested?”

Test, learn, adapt
Halpern is referring to a series of eight trials conducted with a number of different departments, agencies and local authorities, which aimed to reduce the amount of money lost to the UK economy each year due to fraud, error and debt – a figure which currently stands at around £38bn. While traditional attempts to combat these problems have assumed that individuals are mainly deterred from committing fraud by the risk of punishment, behavioural science suggests that instead it is people’s strong sense of moral obligation, justice and fairness that prevents them from participating in such activity. With this thinking in mind, BIT designed a variety of trials that, for example, made it as easy as possible for people to pay tax or debt, reminded people that the majority of their peers are honest, and rewarded desired behaviour.

Already, the results are considerable. Some £160m of tax debts were advanced to the Exchequer during the six-week period of one trail while another, which targeted doctors, resulted in over £1m in additional yield. The report that followed, Applying behavioural insights to reduce fraud, error and debt, stated how the results “demonstrate that even relatively minor changes to processes, forms and language can have a significant, positive impact on behaviour, and can often save the public time and money too. Indeed, if trialled on a national scale, we expect that these interventions will save hundreds of millions of pounds.”
No wonder the coalition is so keen to introduce these theories across other areas of government. So will all policymakers suddenly find themselves having to use this new way of working?

Not necessarily. Halpern is keen to emphasise that nudge theory is simply one of the many tools available. “We’re interested in a whole range of possible levers, drawn from behavioural economics, psychology and other areas, so that the policymaker has a wider set of tools to use rather than just mandation, regulation or indeed just spending more money,” he says.

“The practice tends to be: don’t change everything tomorrow. Let’s approach this with a degree of humility. Let’s work up proposals with departments and then go into trials of one form or another. If it works, great; if it doesn’t, it doesn’t matter. What we’ve done is develop an approach called ‘test learn adapt’, where your ability to try alternatives is designed into the system.”

Getting buy-in
Indeed, in its response to the STSC report, the government conceded “there are few instances in which government would argue that nudging alone is likely to be sufficient”. For example, in areas such as smoking or alcohol consumption, the government feels it “useful and right” to also use regulation and taxation to reduce consumption, particularly within younger age groups.

Public acceptability is also key to successful interventions, according to Halpern – who points out that if the public don’t accept a particular measure, the desired change is unlikely to occur. Get people onside and success is much more likely, he says, citing automatic enrollment into employee pension schemes as a prime example: “Members of the public were exposed to evidence and came to the conclusion that it would be better for it to be opt-out, not opt-in.”

However, despite the success of trials so far, not all the commentary about this type of intervention has been positive. In addition to some scepticism from policymakers, Halpern has had to field criticisms that ‘nudging’ people into behaving in a certain way is rather ‘nanny state’ – something he dismisses as a “deep misunderstanding of this approach”. In fact, nudge is “libertarian, in the sense that you leave an individual with choice”, he says.

Other complaints, adds Halpern, focus on how the theory has been developed in labs rather than in the real world. Sceptics argue that should the approach work in practice – which some doubt – it will only gain traction with trivial behavioural changes rather than tackling bigger, meatier social problems.

Failure is an option
Certainly, it is true that some nudges will fail, even with buy-in from the public. In Ipsos Mori’s Acceptable Behaviour report, published at the end of February, the pollsters highlighted a number of occasions when nudging hasn’t been effective. In New York State, for example, legislation was passed that obliged restaurants to publish the calorie content of menu items, in the hope of encouraging people to make healthier choices. Yet according to the report’s author, despite an official estimate that the law would prevent 150,000 individuals from becoming obese, there was no identifiable change in the number of calories purchased after the introduction of calorie labelling (despite almost 30 per cent of people questioned suggesting that the data had influenced their eating choices). This does not auger well for the UK government, which has been pushing for a similar scheme here.

Again, though, Halpern is none too concerned about the criticisms. “The very point of the trials is so that we can conclusively say what we know works,” he says. Nor does he agree that the technique can only create minor behavioural changes: “The big things in life are affected in the same way as the more administrative changes,” he says. “One of the key lessons is that whatever you choose [as a public services delivery system], there is no neutral option. When you understand what the influences are, you set up your choice architecture to make it more likely the good thing will happen rather than the bad.”

Halpern and his team are already building the evidence base to back up his agenda. In addition to BIT’s own trials, the Department of Health has invested £5m over five years in work by the Policy Research Unit on Behaviour and Health. Based at the University of Cambridge and directed by Professor Theresa Marteau, the unit’s aim is to help generate evidence on the effectiveness, value for money and impact on health inequalities of interventions delivered at individual, community and population level. In addition, the coalition has agreed that proper evaluation of behavioural interventions is critical, and within its 2011 Open Public Services White Paper laid out a vision for the establishment of bodies capable of analysing and accrediting interventions.

Despite his calm and considered manner, Halpern is clearly excited about the potential of behavioural economics within policy development, and it is easy to see how others have been inspired. In fact, this is where BIT is proving particularly successful: in building excitement around the potential. The positive results from trials, the numerous seminars attended by hundreds of civil servants, and the publicity surrounding the unit have led to “too many demands to be able to respond to them all”, he says. The main aim now, therefore, is to develop the capacity of government departments so they can apply behavioural economics to policies themselves – effectively doing Halpern and his team out of a job. But then, this was the intention all along. When approached by the government to take on this role, Halpern specifically requested a two-year ‘sunset clause’ limiting the life of the unit. “Quite often, government sets these things up and they’re still around long past their sell-by-date,” he says.

In this case, the unit will continue past its deadline of this summer, because permanent secretaries are still demanding more from it. “That’s fine, but let’s make it an active choice,” comments Halpern. And in the long run, he wants nudge to be so commonplace that there is no longer a need for his dedicated unit. If it does indeed close in the not-too-distant future, he says, that “will possibly be a good measure of my success.”