Nudge unit boss David Halpern might be the Cameron government’s most influential researcher. He spoke with John Whitfield about bridging academia and policy.
The walk from Charing Cross station to David Halpern’s office retraces his journey through government. First, there’s Admiralty Arch on Trafalgar Square, where as a member of Tony Blair’s strategy unit Halpern sought to inject behavioural thinking into policy, only to be sidelined when a paper on reducing obesity led to newspaper headlines about a coming fat tax.
Moving down Whitehall, one passes Downing Street, where Halpern spent four years in the Cabinet Office running the Behavioural Insights Team, a.k.a. the Nudge Unit. The unit’s formation attracted a fair amount of ridicule but its interventions were successful enough that Halpern was made national adviser to the What Works centres, which aim to bring evidence to public servants. He is now arguably the most influential researcher of David Cameron’s premiership.
Last year, the BIT was spun out as Behavioural Insights Ltd, a social purpose company owned jointly by its employees, government and the innovation charity Nesta. It is housed in an office block on Greycoat Street, 10 minutes further on across Parliament Square.
The task of a social purpose company, it transpires, is to put itself out of business. “That would be wonderful,” says Halpern, when I ask whether the spread of behavioural expertise could mean increasing competition. HM Revenue and Customs, he notes, now has a behavioural team bigger than the original BIT. “We’re not trying to build some giant business empire. If we could live in a world where 70,000 civil servants deeply understand human behaviour—oh my God, I would love to have that problem.”
To hasten that end Halpern has just published a book, Inside the Nudge Unit. It’s an introduction to the ideas behind the use of behavioural science in government and a manifesto for an experimental, evidence-based approach to policymaking. While there have been a number of books from American practitioners of nudge theory—including White House advisers Richard Thaler and Cass Sunstein—this is the first account of its use in the UK.
It’s also a history of Halpern’s time working inside the machine, albeit, one suspects, a PG-rated one. “We continue to work with many bits of government, so you have to be a bit discreet,” Halpern says. He writes, for example, that the BIT found itself up against “Parliament; 70,000 civil servants across the UK; five million public servants; and of course the media and public”, but he declines to name any individuals.
In fact, what the book’s biographical elements most resemble is the story of a successful start-up. This might be partly a bid to surf the cultural energy of the tech industry’s creation myths, in which a small team of geeks disrupts a vast but sluggish incumbent. But it also surely reflects the reality of getting research applied to policy. Halpern’s recipe for influence, while including intellectual rigour, also makes clear that networking, marketing, and the location of your office are all crucial.
That’s a lot to ask of an individual academic who thinks her work might be relevant to policy. What’s needed, says Halpern are “bridge institutions”—like the BIT, the What Works network, and the national academies—that can bring policy and research together. The Greycoat Street sign-in book reveals the BIT’s links to both worlds, showing visits in recent days with people from the Cabinet Office, Home Office and the University of California, Berkeley. “We are weak on these bridge institutions in the UK,” Halpern says. “The academic and the policy worlds are too far apart. It’s unrealistic to think people are going to bump into each other.”
In seeing himself as a conduit rather than an oracle, Halpern positions himself differently to those government advisers whose job it is to wrestle with complex and technical questions so that politicians and civil servants don’t have to. For one thing, he hardly uses the word advice. The only time the concept crops up in our conversation is when he mentions the BIT’s own academic advisory panel.
The aim instead is to design experiments and present their findings in a way that puts the user in control, as with the Educational Endowment Foundation’s toolkit, which summarises educational research findings for classroom staff. This might involve simplification, in the form of mnemonics or star ratings, but not prescription.
Another feature of the BIT is that, while its office was at the centre of government, its work happened downstream, at the level of policy implementation rather than design. An economist might advise on the optimal rate of income tax; the nudge unit tweaked letters to get more people to pay up on time.
This is not new in itself; governments treat research as a tool all the time, which is why they fund work on weapons, crops and medicines. What’s notable about nudging and evidence-based policymaking is that they aim to engineer the practice of government itself. The BIT is a tech company after all—one that sells psychological and social technology.
Like other technologies, this raises the potential for what’s known as dual use—the possibility that a piece of knowledge could be put to either beneficial or detrimental ends. Halpern acknowledges the risk, and it’s been a real issue. In the book, he declares changes to job centre interviews that improved employment rates as his proudest moment at the BIT. But there has also been controversy, which is not mentioned, about the team’s involvement in psychometric tests given to jobseekers.
One aim of Inside the Nudge Unit is to encourage debate about the use of behavioural science in government, says Halpern. “There are real controversies and unease, so you’ve got to be open and straightforward. It’s not for the nudgers to set their own agenda. It’s important that be set by democratically elected governments or directly through the public.”
That said, although the book describes some of the political hurdles to implementing Halpern’s approach, readers could be forgiven for thinking that everything the BIT was allowed to try came off. In the final pages, Halpern writes that the unit’s trials yielded a positive result about 80 per cent of the time, but he doesn’t go into detail about the unsuccessful 20 per cent.
This reluctance to discuss failure was criticised in an otherwise positive report in 2014 from the House of Lords science and technology select committee. “We would emphasise the importance of publishing information on approaches which have not worked, as well as successes,” the committee’s chairman John Palmer wrote in a letter to Cabinet Office minister Oliver Letwin.
That’s been rectified, says Halpern, with the team using the period of purdah before this year’s election to catch up on publishing. “Since that report we’ve put out a huge number of results. There aren’t many bits of the government community that have tried to lay out quite so clearly what it is they do, including things that didn’t work.”
Even so, the team’s criteria for success come from government rather than academia. “In terms of research papers and publications, it’s not a priority for us, because that is not our primary impact,” says Halpern. “Some of the things that we publish and we wrestle with, many of the leading journals would think completely uninteresting. They might say ‘That’s very neat, you did that little tweak and you were able to bring forward £200 million extra revenue or get that many people back into work, but why is it theoretically interesting?’”
Perhaps leaving Downing Street, and the demands of government image management, has made transparency easier. Asked about the effect of the change, Halpern says that as well as opening up a wider range of clients both domestically and internationally, it’s made it easier to retain and develop staff, as people tend to change jobs every 18 months in government.
There’s also the opportunity to move into new areas. I draw Halpern’s attention to a whiteboard bearing a scribbled list of policy topics left over from a previous meeting, including “social care/housing”, “immigration?” and “corruption?”
“These are some areas that we’re trying to think about in new and fresh ways,” he says. “There are really interesting areas that are getting on to the absolute heartlands for economists. As someone whose own discipline is psychology, I quite enjoy that we’re able to revisit classic areas, issues like productivity and economic growth, and bring something new to the table.”
Another word on the board is “obesity”. Halpern is clearly keen for a rematch with the problem that torpedoed his first engagement with government. “In the UK alone we’re spending £10 billion a year on diabetes and cutting off 130 limbs a week as a result. That is a really deeply fascinating issue, and clearly behavioural,” he says.
Halpern’s efforts to marry nudging with New Labour may have—as he put it—flopped, but they taught valuable lessons. “Previous battles get seared on your soul,” he says. He learned to “make sure your early examples are things the government will care about. Don’t over-promise at the beginning, try to quietly do the work and get some results before you start shouting about it.”
Inside the Nudge Unit shows how well that tactic worked. It concludes that the methods that it describes and advocates have taken permanent root in UK government. But as Halpern also says, “we are totally just scratching the surface at the moment. We don’t really know whether most of what is done by goverment and public service is actually effective.”
The closer he gets to the centre of policy and the more disruptive his proposals become, the more scratch-resistant that surface is likely to get. A post-retirement, post-watershed account of what it’s like to come up against the intellectual conservatism of the Treasury, or the lobbying might of the food industry, would be worth reading. Forced to bet, I’d put a small amount on the incumbents seeing off the geeks. But, given what Halpern’s achieved so far, only a small amount.
Inside the Nudge Unit: How small changes can make a big difference is published by WH Allen (£20).
Something to add? Email comment@ResearchResearch.com
This article also appeared in Research Fortnight