Since its creation in 2010, the UK’s Nudge Unit has blazed a trail that countries all over the world are following – but its founder says that most governments adopting the inventive Unit’s approach have yet to understand its most important innovation, and that societies are missing the real opportunity as a result.
Properly called the Behavioural Insights Team, the small Unit, founded with only seven members, has accelerated tax receipts by $70million a month, persuaded an extra 96,000 Britons a year to register as organ donors, improved attendance at adult education colleges, and improved racial diversity in the police force – all on a shoestring budget.
The successes that have been copied from the USA to Singapore, and Germany to Guatemala, come from employing ‘nudges’, prompts derived from behavioural psychology. Now commonplace, when the Nudge Unit was founded five years ago, the application of behavioural insights in government was derided as trendy blue sky thinking and would not have happened but for its determined and imaginative founder. (See box below on how they broke through.)
The man cleaning up government’s ‘dirty secret’
David Halpern, the experimental psychologist who founded the Nudge Unit, explains the problem he set out to tackle in 2010: ‘The dirty secret of almost all government is that we don’t know whether we’re doing the right thing. Whether what we’re doing is helping the patient or not helping the patient, or even killing him.’
Halpern, whose interests include painting and knife-throwing with his teenage sons, studied Natural Sciences at Cambridge, intending to focus on physics, but found himself fascinated by the famous social psychology experiments of the 1950s, 60s and 70s. He ended up taking a PhD in that subject and moving into the field where those experiments can be conducted on the grandest scale: government.
In 2001, he was brought into the Prime Minister’s Strategy Unit, set up by Labour’s Tony Blair to make the state more effective. He founded the Nudge Unit after the Conservatives won the 2010 election, drawing on his own scientific background and the book ‘Nudge: Improving decisions about Health, Wealth and Happiness’, by US behavioural economist Richard Thaler. Incoming Prime Minister David Cameron was enthused about Nudges and keen to give them a chance – a project much mocked by the British media, until the results started to come in.
To set it up, says Halpern, ‘We gutted the best staff from the Prime Minister’s Strategy Unit’. Those he considered among the most talented and imaginative people working in government were, above all, ‘used to thinking about policy issues from first principles’. For example: Dr Rory Gallagher, whose Cambridge doctorate is in changing behaviour around AIDS and who now runs the Nudge Unit’s international partnerships.
Those partnerships have become truly worldwide, and taken the form of an innovative public-private vehicle. The Unit, which has been ‘spun out’ of government to be owned half by the UK state, half by its employees, has set up offices in Sydney and New York. Equivalents have been founded in Germany, the USA, Singapore, Denmark and Australia. Individual projects have been worked on with many more countries.
And yet Halpern insists that their method, which he calls ‘radical incrementalism’, has still not been truly understood, and that its revolutionary and lasting benefits are still waiting to be tapped.
How it works and where it’s working
The best means of understanding the power of applying behavioural economics in government is to look at the Nudgers’ work on organ donation, for which they conducted one of the largest randomised controlled trials ever held in Britain.
Organ donation saves lives and nine out of ten people support it, but only a third register as donors. To close this gap, the Nudgers added eight different messages (examples below), to some heavily trafficked government webpages; those for renewing a driving licence or vehicle tax. Over the five weeks of testing, these pages were seen by more than a million people, more than 135,000 for each variant.
The first said simply, ‘Please join the NHS Organ Donor Register’. This was the control, the yardstick, against which the others were measured. Some had different messages, such as ‘Three people die every day because there are not enough organ donors’, or added pictures and recognisable logos, such as that of the NHS.
Messages of this type are characteristic of ‘behavioural insights’, but the real revolution, the thing that Halpern calls ‘the Nudge Unit’s greatest legacy’ is not the type of message, but the fact of trying variations on it. It is what science has long known as the experimental method: make a yardstick, test your variants against it, see what works. Vary, test, learn and repeat: this is how government can find out whether it is ‘helping the patient’ or killing him.
Halpern describes that as ‘a change in Weltanschauung [world view], a real paradigm shift,’ and adds that it’s like ‘planting several seeds and then clipping the weaker ones. That’s very different to a world in which you say, “Vote for me because I know what to do” or “This is the best idea, Minister”.’
And it is revolutionary because: ‘It’s very hard for government to admit that something it’s doing might not work, or that it’s failed […] People are often running blind, overconfident, and dispensing massive sums of money.’
In the organ donation project, the most successful variant read: ‘If you needed an organ transplant would you have one? If so please help others.’ It prompted an extra 1,203 people to sign up compared to the control; over a year, that would mean an extra 96,000 donors.
Unexpectedly, one variant, with a picture of an ethnically diverse group coming together, did worse than the control, conjecturally because readers found it gimmicky. As the Nudge Unit said in their report: ‘There were good reasons for believing that the group image would help. If we had not explicitly tested this idea, we may have done more harm than good.’
It works – but we don’t want it
Success, as Halpern has found, does not guarantee support. Ironically, some of the voices who regularly criticise government for waste have turned on the Nudge Unit for attempting to base policy on evidence. The Daily Mail recently ran an opinion piece headlined ‘It’s time to stop these secret tests’. The story described the Unit’s work as ‘alarming’ and ‘Orwellian’, and went on to say the public was being ‘secretly manipulated’.
This is not an isolated opinion. A libertarian magazine in the US, Spiked, has declared a ‘war on nudge’ and when the Unit began working for Angela Merkel’s government, the Germans even had to check that these methods were constitutional, because in Germany, after the ‘perverted science’ of the Nazi era, it is illegal for the state to experiment on its citizens. (The German government found the Nudge methods were constitutional.)
As Halpern puts it, some people have the idea that ‘we must stop government experimenting on us’. He says: ‘This shouldn’t be entirely dismissed. Many of the decisions we make are done in an automatic, unthinking way; that’s where much of the power of these methods comes from.’
And what is the solution to this distrust? For Halpern, it is transparency. He says: ‘If you’re going to do something different and controversial, something different from conventional policy, you need public support. That’s why we’ve written books.
‘When we tried to do something similar ten years ago in the Strategy Unit, it got a bad press, Tony [Blair] walked away from it, and we essentially put it on hold.’
He explains: ‘We’re helping the Australian government, in Victoria, with a forum on obesity. Plate sizes, portion sizes, things like that. People think we’re going to say, “You can’t have chocolate by the checkout”. So you have to show the public the effects. These things have a big impact on behaviour. Then you ask them, What would you like to happen? And then if they say, We do want guilt-free aisles, it’s very powerful.
‘So if governments are going to use these approaches, they have to enhance their capacity to get permission from the public.’
The future of test and learn
In the beginning, Halpern says, the work of the Nudge Unit ‘was like on a wildlife show when you see a lion chasing a gazelle, nine out of ten times it doesn’t catch it. With nine out of ten ideas, we couldn’t land them.’
This year, by contrast, 24 governments came to the Nudge Unit’s conference, more than double last year’s number. Attendance so exceeded expectations that they had to turn people away. Halpern says: ‘This is a very exciting time, but the argument is not won’.
‘Behavioural insights is spreading fast. But what’s much more important is embracing empiricism. And that has much weaker foundations. If you look at what we’ve done and think, “Let’s do something like them,” that’s great, but it’s not experimenting yourself. It’s not embracing this radical incrementalism. That is changing how you do things, it’s a perpetual testing and improving, a kind of restlessness, and there are very few places that are any way down that road.’
Although there may not yet be as many seedlings as Halpern would like, green shoots are springing up around the world as other public service innovators are not only adopting Halpern’s practices, but starting to speak his language of humility and experimentation.
‘Government [traditionally] isn’t comfortable saying, “We don’t know,” says Leon Voon, the 39 year old co-founder of Singapore’s Human Experience Lab. ‘Government is the one who says, “We know.” And that means there’s no freedom to test. Government programmes have got so complex that we should think of public servants not as programme managers but as designers, facilitators, engaged in experiments.’
For Voon, as for Halpern, what the radical new approach boils down to is the willingness to say, we don’t know — but we know how to find out. And the 96,000 extra organ donors a year suggest they are onto something big.
Follow the Behavioural Insights Team on Twitter and send us your suggestions for Apolitical Pioneers to firstname.lastname@example.org.
Extra: How they did it
The Nudge Unit was set up during the worst economic slump since the Second World War. As a result, it was ‘done on a shoestring’ and included a so-called Sunset Clause: if it did not save the government at least ten times its outlay within two years, it would automatically disband.
That annual budget was under £1million ($1.5million), which meant the Unit could not pay for the testing they proposed. And because they had no previous interventions to point to, their credibility was low. That was why, ‘with 9 out of 10 ideas, we couldn’t land them’. Those they could land:
- ‘Were utterly focused and well judged.’
- Had top-level backing. ‘It was absolutely critical to have top-level administrative and political support. It’s no guarantee that a department will listen to a damned word you say, but it does get your foot in the door.’
- Had a backer inside the department. ‘It was often someone not necessarily at the very top, just someone who’s prepared to try.’
- Were cheap. ‘We designed the interventions to be where people were already collecting a lot of data. That’s why there was such a focus on tax. It’s very well suited to this approach because HMRC already has a good system for seeing what revenue has been collected.’
- Were quick. ‘What we offered is very rapid trialling. If you offer someone a pilot, they groan and think it’ll cost millions, take ten years and in the end tell them it didn’t work. What we do is say, “We’ll send out these two types of letter and we’ll be able to tell you in a month which has greater uptake”.’
- Took what they could get. ‘An incoming minister might say, “I didn’t spend this much time in opposition to run a pilot.” So don’t say, “Wait two years.” This is urgent and must be done to scale. But can we at least bake in some variations so we can keep refining it in the future.’
‘My advice is not to go for a massive push. You’ve got to take small steps, build support inside the system and show them practical results. It needs humility.’ And it was by no means certain the BIT would succeed. I was not at all confident. That’s why we had the two-year Sunset Clause. Two years is enough in an innovation view of the world.’