This is where we are – what next?

Reisch and Zhao’s paper ‘Behavioural economics, consumer behaviour, and consumer policy: State of the art’ gives us an assessment of where we are in terms of behavioural public policy; starting with the background to its development, reviewing the basic findings in relation to human behaviour that can make a difference to policy outcomes (intentionally or otherwise), and looking at how behavioural public policy has moved into a variety of domains.

As the authors note, decisions always take place in a context, and the context along with the way the decision is put to the decision-maker matter because people do not behave in the ways economic theory suggests due to bounded rationality, among other things. Implementing behavioural public policy therefore means policy makers need to design the context or choice architecture within which policy is implemented. The initial evidence presented suggests that the effectiveness of nudges is not dependent on a lack of transparency and that people generally approve of them, providing they would support the policy goal.

With behavioural approaches becoming established as a new paradigm for effective implementation of public policy we move on to the stage of establishing its boundaries in terms of effectiveness and what is permissible in terms of implementation approaches. What are the characteristics of decisions where nudges are less likely to work? What limits do we need to keep this an ethical approach? To what extent are individual differences in psychology and circumstances important (ie when might an approach work on one population group but not another)?

A particular difficulty for policy makers and academics can be in understanding individuals’ motivations in some real-world contexts. In applied research you often have to make sure you have understood someone’s motivation rather than just observed and come up with an explanation for their behaviour – if you don’t understand their conceptual model of the world then any ideas for interventions might not have the effect you expect, or might not be generalisable.

To some extent implementation of behavioural interventions benefits not just from the inertia of defaults and status quo in the initial decision, but also in keeping the behaviour going – at least providing effort is not required. Thus auto-enrolment in pension schemes transfers people to a state where they are contributing, and their contributions to the pension fund continue without action. This does not mean new information might not “rock the boat” in terms of continued involvement. What issues arise for passive involvement if the system people are opted into has problems? What if, say, a pension scheme suffers losses on the stock market and members see their funds fall in value, albeit that this may be temporary? People who have fewer resources may find such situations more difficult to ride out in the hope of improvements later and may feel the opportunity cost of alternatives more keenly.

If effort is required to keep a behaviour going (for example, regular exercise or avoiding unhealthy food) then behavioural interventions may need to focus on how to get repeated behaviour established. The habit literature, which specifically explores repeated behaviour, can provide insights here. Habitual behaviour, once established, has an inertia of its own, with work exploring when habits are vulnerable to change and how new ones become established. An element of this can be disruption of environmental cues that trigger the original habitual behaviour and information to support the desired new habit. Commitment contracts are used by people to try to change their own behaviour; they commit to future behaviours, with penalties for non-compliance. Websites such as stickK.com (developed by behavioural scientists at Yale) can provide an easily accessible way to do this. Financial incentives for compliance have also been tried in areas such as weight loss. What is not clear is whether such interventions support or interfere with habit formation as such in each particular case – when are these approaches helpful? This could impact on long term effectiveness.

Work discussed in this paper and in the article in the first issue of this journal by Cass Sunstein begins to address these issues, looking at the issue of manipulation and at why nudges fail. Further work on establishing boundaries can help move to more effective policy interventions more efficiently.