Nothing but nudges: behavioural economics comes of age

By Stephen Easton

Tuesday October 11, 2016

Nudges have taken the public sector by storm, expanding all over government. One of their chief promoters around the world is confident they aren’t a passing fad, but warns against complacency and overuse.

Nudges are the next big thing in public services, and Australian public servants are among the most enthusiastic adopters of the behavioural sciences.

It’s been a bumper year for the Behavioural Insights Team, an unusual company that started life in 2010 as an appendage of the United Kingdom Cabinet Office and is now exporting nudge theory to the world. It’s set up shop in Australia, the Americas and most recently Singapore, two years since it became a business owned jointly by the UK government, staff members and Nesta, an innovation-focused charitable foundation.

Rory Gallagher, who came to work for the New South Wales government as a civil servant on secondment in 2012 to help set up Australia’s first nudge unit, is now BIT’s Asia-Pacific managing director. He still works with the NSW agency a couple of days a week and according to the report, it’s been the company’s “longest and most impactful” partnership in its brief history.

The NSW administration “took a punt, to some extent” by signing up as BIT’s first international partner, Gallagher told The Mandarin. “But they’ve done a huge amount since then; they now have a team of about a dozen people applying this across the sector.”

“There’s lots governments and agencies interested, but NSW has put the investment and the infrastructure around trying to do this really well,” he said. Having got in early, NSW is now experimenting in more complex areas like recidivism, domestic violence, hospital discharge and traffic congestion.

Rory Gallagher
Rory Gallagher

According to the BIT’s latest annual update, NSW will publish results from “pioneering” nudge trials later this year that include “some impressive wins” as well as data on some interventions that had little or no effect.

Barack Obama was first to start employing the ideas set out in the influential 2008 book, Nudge, by appointing one of its two authors to run the federal Office of Information and Regulatory Affairs in 2009. But the UK one-upped the Americans the next year with its dedicated nudge team, a move later replicated in Washington, and has since picked up the idea and run with it.

The long list of experiments in the BIT report shows nudge theory has continued to grow in popularity and is being applied more widely across government. There’s chapters on economic and social policy areas as well as administrative applications like “reducing fraud, error and debt” across the UK government. Civil service chief Sir Jeremy Heywood comments that “almost every government department is now using behavioural science to help improve the way that public services are run or policies are delivered” in a foreword.

“A few years ago, many of our results were concentrated in just a few areas, primarily tax, health, employment and a little bit in the environment,” said Gallagher.

“Now you can see it’s used across things like justice, and in the education space, where a few years ago lots of people were very nervous about running experiments or doing anything in schools.

“But we’re now seeing [nudges in schools] completely mainstreamed in the UK and increasingly in Australia, as well as in quite complex areas like productivity in the economy and trying to understand how markets operate and how firms make decisions.”

Nudging students

The education chapter reports school kids are being nudged to get better grades, as you might expect. One intervention put them through exercises and training designed to improve “non-cognitive skills” like perseverance:

“First, ‘deep practice’, in which students learn how to set goals, concentrate completely on their work and then actively and feedback. Second, learners were taught that frustrations are to be expected along the way when learning and that it is not only talent that matters, but effort too.”

Another set of exercises were designed to overcome the fear a negative stereotype will come true — known as stereotype threat — which can lead to “a self-fulfilling prophecy where underperformance confirms the stereotype” while a third simply sent supportive text messages and information to parents and friends about upcoming tests and other occasions.

Education departments spend a lot of time thinking up new ways to engage parents in schools as they are such a big factors in varying educational outcomes. According to the report, the simple texting trial showed an 11% increase in attendance with a sample size of 745.

“There’s also some work that we’re doing linked to that which we’re just starting to see the results from around trying to get them to engage in the actual subject matter,” said Gallagher, who has a particular interest in education.

Early indications suggest outcomes are “slightly more mixed” in that area, he adds, speculating it might have to do with the fact that school subjects get increasingly difficult for parents.

VicHealth gets on board

The Victorian and Commonwealth governments are the latest to set up dedicated nudge units in Australia and as the report notes:

“Even where there aren’t dedicated units present, public sector organisations across Australia are drawing on behavioural insights to inform policy design and delivery.”

BIT reports its Sydney-based operation has doubled its staffing in the last year alone and has run about 30 introductory workshops for over 500 public servants in state, local and federal agencies across Australia and New Zealand.

Australia gets its own chapter in the annual update, listing the work of the NSW team plus several new experiments in the Victorian health sector. One successfully nudged more women to attend their breast cancer screening appointments by encouraging them to write down the date. Another related trial found a much bigger increase when interventions incorporated a chance to win an iPad for oneself, or give one to a friend.

At federal level, BIT has been helping the Department of Employment build a behavioural economics capability and is already working with it to design and run a few trials, presumably to encourage positive job seeking behaviour.

The report is filled with interesting snippets about evidence-based policing trials involving behavioural approaches from West Midlands Police, and another interesting BIT report from the UK which showed people are dishonest (or just bad at estimating) when asked how much they eat in terms of calories.

National surveys showed caloric intake going down in the UK, but obesity rising. It wasn’t surprising that people’s survey responses were inaccurate, but the research used new methods to confirm how much people were really eating, and proved beyond a doubt it couldn’t be a lack of exercise to blame for the growth in obesity, as some had suggested. The disparity between the survey results and actual total consumption was also growing.

“There’s a wider lesson that people aren’t always able to report accurately what their behaviours are,” said Gallagher. “And that goes way beyond calorie counting; there’s lots of places across government where we rely on these big surveys and quite often they work very, very well. But in some areas they will be either over- or under-estimating whatever it is we’re looking at.”

Nudging into the future

The idea that policy just needs a nudge is clearly attractive to governments and, with the usual caveats about it not being silver bullet, Gallagher is convinced the expansion will continue.

He believes the movement is mining “huge untapped potential” for “dramatic” improvement in the outcomes government wants to see in most public services, from small changes to systems and the way information is framed.

One challenge is scalability.

“So, I think we’ve got lots and lots of great results now — proof that this does work, the proof of concept,” said Gallagher confidently. “I think we could focus more in the next few years on taking that to scale to make sure that these don’t just remain trials; they actually do get rolled out across the system.”

He enthuses about the need to influence individual decision-makers to take up the approach, and wants to see behavioural economics trials have more influence on policy development. His example is a trial conducted with VicHealth at the Alfred Hospital in Melbourne, testing out a simulated sugar tax “in an empirical real-life setting” using vending machines.

Such a policy is controversial — plenty of people won’t support it on principle, regardless of its efficacy in reducing sugar consumption — but Gallagher hopes a growing capability in behavioural economics research in the public sector can at least  inform such debates with real-world information on how basic incentives and discincentives actually work.

It’s doubtful a sugar tax even counts as a nudge in the original definition of the term, which specifies the interventions “must be easy and cheap to avoid” and cannot be mandatory: “Putting fruit at eye level counts as a nudge. Banning junk food does not.”

The interesting results summarised throughout the report clearly demonstrate that often, slight but significant upticks in key outcomes can be achieved through extremely cheap, simple and scalable interventions.

But with the growing popularity of nudges, governments must not get complacent about the need to base each nudge project on legitimate research, and to continue monitoring over time to make sure they keep working.

Gallagher sees “a slight danger in this whole approach” from the fact that successful nudges often seem obvious, with a heavy dose of hindsight. Public servants might start thinking they can design nudges and roll them out without testing.

“That is definitely a risk, that it does start to get so widely used and so superficially used that it’s not having the effects that we want,” said Gallagher. “That just means we need to keep engaged with the literature and keep up the quality of what we’re doing as well as trialling to make sure these things do work over time.”

He suspects the same old nudges will indeed lose their effect over time, and those effects might also be diminished if multiple agencies are trying similar approaches.

“If every government department was sending you something saying ‘eight out of ten people do X and Y’, it seems to me inevitable to some extent, some of that will wear off,” he said.

Little is known about how the effects will change over time but Gallagher says the field needs to maintain scientific approaches, embrace technological advances and work hard to “keep it fresh” so citizens keep responding.

About the author
Newest Most Voted
Inline Feedbacks
View all comments
james horton
james horton
5 years ago

…now add data, lots of it, and opaque analytics, and we begin to head into uncharted waters wrt to individual agency and capacity for democratic participation. The trajectory is a given. We would do well to not blindly stumble into this future. A good start might be to check out Karen Yeung’s work on “hypernudge” is one of an all too few attempts to bring together a cross disciplinary perspective… &

Geoff Edwards
Geoff Edwards
5 years ago

If well-established insights from psychology and sociology can become assimilated into economics by badging them ‘behavioural economics’, then it’s all to the good.

The essential resource for effective
public sector professionals