David Halpern, a behavioural scientist who tries to figure out how government can work better, must feel like a pilgrim in an unholy land.
After a career as an academic, he went to work in Downing Street to be chief analyst at Tony Blair's Strategy Unit and later became the first chief executive of the government's Nudge Unit, which has led a quiet revolution in how we make policy, bringing academic rigour to the analysis of what policies work, how they could work better and why.
In its five years of existence, the unit, officially the Behavioural Insights Team (BIT), has found how making tiny changes can boost the number of organ donors, help the Job Centre get 10 percent more people off benefits and into work, and increase the number of black and minority ethnic people passing a police recruitment exam.
Its founding principle is the opposite of Whitehall and Westminster convention: You do not know what works until you have the evidence to prove it and a lot of what you think is obvious is wrong.
"We wanted to test the hypothesis that, if you introduce a more realistic model of human behaviour into policy-making, you get better policy... A genuinely experimental approach to how you do government. It's quite a profoundly different world view in terms of humility than the normal way policy is done which is 'I've got this great idea. Let's do it to everyone because it's a great idea'," Halpern tells The Huffington Post UK.
SUBSCRIBE AND FOLLOW POLITICS
Get top stories and blog posts emailed to me each day. Newsletters may offer personalized content or advertisements. Learn more
"It's a bold minister who gets up and says 'we don't know what will work but we'll try out a range of things and figure it out'. That's not what the PM will generally say in his speeches and probably won't for a few years."
If Whitehall and Westminster are hostile to this perspective, Fleet Street is positively appalled by it.
In August, the Daily Mail called research into how differently-worded text messages could prompt people to pay up their tax backlogs "state-sponsored manipulation".
Halpern seems comfortable with their outrage, saying the Mail "has run and will continue to run" stories about how outrageous his work is. His wonders what the alternatives would be. "We're not going to experiment? We're going to for national roll out in many areas? We're doing continue with massive legacy spends in areas where we have no idea whether it's effective or not?"
This attitude is standard in academia. Does Halpern ever feel like he's doing missionary work when preaching it to others? Ever the scientist, he stresses that scientists and politicians have "different paradigms".
"I'm there to keep making the case... Quite often, people are going to look at your strangely and say: 'Really? You want to do what?'" But the way government has worked before is like something "from the Middle Ages", he says.
A spending review is looming and the better informed policy makers are, the better they can spend the money that will be left. That could be the difference between having to cut budgets with "an axe or a scalpel", Halpern says. "Are we going to be able, in local authorities and services across the UK, to make finely made judgments about whether we can shift money to things that are more effective and away from ones that are less effective?
"You'd think that'd be a no-brainer... But if you don't have that evidence, you're often working in the dark, blind, it's like something from the Middle Ages, frankly, in terms of our knowledge in many areas."
One of the most controversial pledges of May's general election campaign turned out to be something politicians did not know as much about as they thought. The Conservatives announced they would extend the right-to-buy to housing association tenants. After they won a majority, the scheme attracted fewer applicants than expected, although the benefits of buying the house you lived in at a discount seemed obvious beneficial.
The Nudge Unit was asked to find out what the government was missing. It identified what Halpern calls "frictional factors". It was not necessarily easy for housing association tenants to prove how long they had lived there or in previous homes, which counted towards the discount. People were unfamiliar with the mortgage process. Few of those eligible would have bought a house before and might balk at paying up front for a survey and a mortgage while being uncertain of the benefits, no matter how great they were said to be, especially if money is tight in the first place.
"Of course in the classic model, none of that would matter that much," Halpern says. "Because you'd do the maths and say 'of course it makes sense'. But with real human beings, it matters quite a lot."
Halpern's inspirations are outside policy-making. They include the doctors whose pioneering trials decades ago shot down false preconceptions in medicine and revolutionised the field. But his biggest contemporary inspiration is Dave Brailsford, the coach who does for British cycling what the Nudge Unit is doing for policy.
A relatively unsung, unseen hero of the 2012 Games, Brailsford experimented with how little things, such as whether cyclists took their own pillows with them when competing abroad, could marginally improve performance. "He would test these very small variations and each of them might just give you 1 or 2% increase but if you start doing that in performance across the piece, you have a literally winning team," Halpern says.
"If we can do that in every area of government policy so that every year it gets a few per cent better and more efficient... It's a game changer."
The world is slowly coming around to the Nudge Unit's radically different approach to how you decide and implement policy.
The Nudge Unit was set up in 2010 with a sunset clause that said it would shut it down unless it saved 10 times as much money as it spent. "It was quite a tough bar, but in fact we did it many times over," Halpern says.
Now, it has a contract to work for the five years of this parliament. It has grown from seven people to 65 people. It advises foreign governments - Australia and Singapore were early to seek its counsel and the US has its own 'Nudge Unit', set up in the White House. An event in London discussing the unit's work was attended by 900 people from 24 countries.
The Unit sometimes has fun with how it presents results, such as when it tested the most effective form an online prompt to join the organ donor should take. Displayed to people while filling out their tax returns, many assumed the prompt would work best with a picture. The unit experimented with eight variations and, before displaying the results, asked the audience to vote for which one they suspected would work best. They left the one including a picture until last, which many people raised their hands to vote for.
But they realised how wrong they were when the results were revealed. Including a picture in the message backfires, making people less likely to join the register.
"The point about that, apart from making slight mischief, is you want to help people understand that very often their instincts will be wrong. They shouldn't feel bad about it... The point is, we can't always know. Sometimes these things will throw themselves up as a surprise but it's ok, we don't have to pretend that we are all some all-knowing beings who now how the public will respond in any given situation."
I ask how Halpern himself become comfortable with accepting his instincts can be wrong. "My first field was experimental psychology. So the clue's in the title, right?" Experimentation is hard wired into his discipline, he says, adding that psychology exposes how people are not of "a single self in our being" but the sum of competing instincts and influences. "When you study it for many years, you realise the human mind and brain is much more complicated, with many process are racing against each other... a democracy of mind."
When asked what of the unit's work has surprised him most, he points out around two out of every 10 trials it conducts fail, emphasising how even a team of scientists' preconceptions about what might work can prove false. The biggest surprise to him showed how a particularly minute change could affect one of the most sensitive issues in Britain today.
Police forces are desperate to recruit more black and minority ethnic (BME) officers. An online, centrally marked situational judgment test for potential recruits suffered an unexplained racial disparity – 60 percent of white applicants passed compared with only 40 percent of the rest.
Without evidence, it was suggested knowledge of English might be the issue. The Nudge Unit's lead on crime and one of its psychologists got to work and made a single change. Before the link to click-through to the test, they added a line asking applicants to pause to reflect on why they wanted to join the police. The gap between white and BME applicants' success rate closed.
"It's just stunning... it brings home the sense that quite often, we say something, we're putting someone down a trajectory which we're not aware of... It's just one of those powerful reminders of how important this stuff is."
Halpern has written 'Inside The Nudge Unit', a book about the slow and steady battle to show the people controlling policy how these small changes can make a difference. The point of writing it was to be as open and transparent, he says, adding: "We need the permission of the public".
It might be frustrating that the right-wing press condemns it but the headlines reflect the genuine controversy. Each trial is effectively an experiment on humans. When the Unit advised the German government, they found the country's constitution forbade some of the potential experiments on the grounds it would be unfair to divide a group of people into cohorts at random and treat each one differently.
But Halpern says the benefits of these trials are clear - as clear, it seems, as his frustration with our current system. "More than three-quarters of health life lost are due to behavioural factors. That is not where we spend our money. That is not where we spend our efforts. We're cutting off how many limbs a week for people with diabetes? Is that the most effective way of addressing the issue? Or trying to move upstream where the behaviour really lies?"
The book is dedicated to "the elected". Politicians may have a different approach to scientists' but, after years of telling politicians and civil servants that cherished projects do not actually work or are unsupported by evidence, Halpern calls them "incredibly dedicated, it's almost incredible they do what they do". He is frustrated so few of his academic colleagues engage to address "the big policy questions in a deep way".
"There's a retreat often to theory... 'Let's do a pile of maths and stick it in an American journal', as opposed to really testing your work in the fire of real world context... a much more genuinely testing environment of your own theories." He hopes the work of the unit and the government's creation of 'What Works Centres' will encourage more of to reach across the divide.
"It is quietly revolutionary to bring in this more sophisticated model of the human being and more humility to policy that it brings in its wake. I hope we're making a lot of progress but, boy, we have long way to go," he adds.
A government humble enough to experiment before it implements everything is still a long way off.
When Halpern was writing his book, he saw an article in The Evening Standard about a London borough that had the bright idea of a Scared Straight-style programme. Troubled kids were taken to prisons to be shown the world that awaited them if they did not clean up their act. The authorities had obviously not got the message that such programmes were proven to actually encourage, rather than discourage, children into lives of crime.
Halpern says the attitudes that have become common in the medical profession need to become common in government. No one claims aromatherapy or crystal healing works without expecting to be challenged about it, he adds.
"We need to get to a world where people to ask 'Why do you think that's true?' and look behind it to the method that gave rise to the claim. Will it take two years, 20 years or 200 years? I don't know… Wouldn't it be great to be in a world where the public are protesting because a big programme is being rolled out but it wasn't tested, the Daily Mail saying 'it's outrageous that these welfare reforms have come in but we didn't run a big randomised controlled trial'."
But the paradigm shift - another phrase beloved of scientists - is gradually happening, Halpern says, and the idea that behavioural science is key to designing policy is gaining ground across the public sector.
"Once you've done it, as with a lot of things, it just seems like: why on earth weren't we doing this years ago?
"Our kids will say to us: 'how did you make policy? You just used to sit around and talk about it, look at bit of evidence and then just say 'well, let's do this and we'll do it for everyone?'"
'Inside The Nudge Unit' is published by Ebury and out now.