Following a recent report questioning the effectiveness of 'nudging' citizens towards policy outcomes, Suzannah Brecknell considers how civil servants can make best use of behavioural insights in their policy designs.

In Arizona’s Petrified Forest national park, rangers tried to protect the rare fossilised wood with signs reading: “Your heritage is being vandalized every day by theft losses of petrified wood of 14 tons a year, mostly a small piece at a time.” Experiments showed, however, that these signs actually increased the number of thefts – the implicit message was that it was okay to steal pieces of wood, since everyone else was doing it. It turned out to be more effective to tell visitors that most of their peers do not steal from the park – confirming the behavioural economists’ theory that people are likely to do what most other people are doing.

The Arizona experiment was carried out by Professor Robert Cialdini, a social influence expert whose consultancy and training firm Influence at Work (IaW) recently used similar insights to make small changes to GPs’ appointment booking processes – reducing the number of no-shows by 30 per cent. The findings were the latest piece of evidence that ‘nudging’ can work.

A nudge is an intervention or system designed to take advantage of insights into human behaviour in order to encourage people to act in ways that benefit society or that individual. The technique is being promoted across government by top officials and ministers as well as a Cabinet Office-based Behavioural Insight Team (BIT) which, according to junior Cabinet Office minister Oliver Letwin, was formed to help promote“non-regulatory means of achieving behaviour change”.

Letwin was giving evidence to the Lords science and technology select committee, which has recently produced a report on these techniques. The report, while positive about behavioural economics, was critical of the government’s use of nudging so far. It said there is not yet any proof that these approaches work on a large scale, concluding that available evidence suggests that “the most effective means of changing behaviour at a population level is to use a range of policy tools, both regulatory and non-regulatory”.

The report noted with concern that several officials suggested ministerial bias towards behavioural insight techniques is resulting in a failure by civil servants to properly consider the use of regulation as a way of achieving policy objectives. Baroness Neuberger, chair of the sub-committee which produced the report, told CSW the committee is “really delighted that government is taking behavioural psychology and behaviour intervention seriously; we just want them not to see it as alternative [to regulation], but as part of a package”.

Most experts on the subject appear to agree with the baroness. At a recent event exploring behavioural insight and policymaking, hosted by the Institute for Government (IfG) and Cialdini’s organisation IaW, all the experts clearly stated that nudge should not be considered alone. “Most nudges require ‘shoves’,” said Paul Dolan, professor of economics at Imperial College.“They require legal or regulatory changes to apply the nudge.” He added that behaviour science can be applied as usefully to laws and regulations as to softer, non-regulatory forms of nudge.

Beyond this concern about using nudge in isolation, Neuberger told CSW that the committee’s biggest criticism is “the lack of evaluation of a big intervention”. If government invests in some kind of intervention, she said, and then doesn’t “evaluate in the end what really works”, or carries out “a not terribly good evaluation so you don’t get enough of an idea, what on earth is the point of that?” Solid evaluation will be crucial in order for government to judge whether nudge is a cost-effective option for policymakers. “Nothing is free,” said Neuberger. “You can say: ‘This is cheap’, but it’s not cheap if it doesn’t work.”

While there is little evidence that nudging alone can create significant changes in behaviour across a population, there is plenty of evidence that behaviour-change interventions can affect individual or community behaviour. John Bromley, director of the National Social Marketing Council (NSMC) – set up in 2006 by the health department to investigate more sophisticated ways of communicating public health messages – says his organisation can point to dozens of international and UK-based projects that “have pretty substantive analysis and evaluation [and] which have had impact”.

If civil servants are to get best value for money out of behaviour-change approaches, says Bromley, they need to begin with a deep understanding of their stakeholders, and apply this insight to produce well-targeted interventions which take account not only of general behavioural insights but also of the social and cultural background of the community they are trying to reach. Often, he suggests, the knowledge is already available within government.

The NSMC itself has a set of resources which departments can access, he says, and communications or marketing teams can also be a good source of research into the preferences and attitudes of departmental client groups.

Another priority for policymakers should be identifying key messengers and ‘change agents’ who can support behaviour change among target groups, suggests Clive Bates, director general for sustainable futures in the Welsh Assembly Government. Speaking at the IfG event, he urged civil servants to look for “enabling agents”, and consider how to incentivise those agents as well as individuals. He gave the example of the European Landfill Directive, which created a fiscal incentive for councils to reduce the amount of waste gong to landfill, and thereby “turned local authorities into behaviour-change agents, [making] them get into a position where they had to facilitate or provide for people to recycle more”.

Cialdini was also at the IfG event, and pointed out that policymakers often wait too long to start involving such potential ‘change agents’ in their work. Rather than engaging the people who will implement behaviour interventions during the research and development stage, he said, “We impose this on them after the fact and say: ‘You’ve been doing this wrong all the time: change!’.”

In Arizona’s Petrified Forest, he said, park rangers resisted the replacement of signs, even though the data showed it would reduce thefts, because they felt undermined by the findings. Without early practitioner involvement, he said, “not only will these principles not work; they won’t get a chance to work”.

A final piece of advice from the IfG event came from Professor Dolan who, like the Lords committee, emphasised the importance of evaluation and piloting. “Try lots of stuff. Try it quickly and cheaply,” he said. “Most of it won’t work, [so] publicise the fact that it hasn’t worked and move people on to the stuff that has worked. Roll that out, rather than setting up these big tankers that head off in the wrong way or start sinking and you can’t do anything about it.”

If much of the advice on making good use of nudge sounds familiar, it may be because it echoes much of the advice on making good policy, no matter what tools you choose. Peter John, professor of governance at Southampton University, recently led a research team comparing the effectiveness of nudge with more participatory, deliberative ways to try and change people’s behaviour. It’s not either/or, he says: “People should continue to use a whole range of instruments… it’s not just a choice between behaviour change and traditional approaches.” Nonetheless, he adds, this whole area of research is “genuinely exciting: there are some quite dramatic effects you can have from thinking carefully about the design of instruments, how the citizen perceives the policy, and how they react to it”.

Nudge on trial: case study

In July the energy department (DECC) announced it will trial a number of schemes encouraging homeowners to reduce energy use. The schemes are based on a set of principles: that individuals tend to stick to default settings; that they are strongly influenced by social norms; and they tend to “discount the future, paying more attention to short term gains or losses than long-term effects”. DECC will also work with other organisations to find the best times and ways to prompt people to think about energy efficiency. The trials are:
-Upfront incentives, such as a month’s holiday from council tax, to buy energy efficiency products;
-Discounts on green products which rise as more people take advantage of the offer, encouraging collective uptake, and providing subsidised loft clearance to remove one of the barriers to installing loft installation;
-Community rewards for collective take-up of the government’s Green Deal policy;
-Giving consumers feedback on energy use, including comparisons with neighbours;
-Reforming Energy Performance Certificates to set out the cost savings of energy efficiency improvements over a number of years, and include recommended ways to improve efficiency.

Click here to see all news and features from Civil Service World

Written by Suzannah Brecknell, CSW