By Professor Sarah Smith, Professor of Economics at the University of Bristol and Centre for Market and Public Organisation

Behavioural economics moved centre stage in UK policymaking with the establishment of the Behavioural Insights Team – or 'Nudge Unit' – within the Cabinet Office. Behavioural economics incorporates elements of psychology into extended economic models to explain why people make seemingly irrational decisions, such as failing to save for retirement, and why small changes to the way people make choices, such as automatically enrolling people in pension schemes rather than requiring them to opt in, can have powerful effects.

According to some, the nudge unit doesn't do behavioural economics in the true sense, but instead does social psychology. But the implications for policy design are similar in both cases and, notably, are in marked contrast with the approaches suggested by standard economics. Take a desirable behavioural outcome – reducing energy consumption, for example, or eating more healthily. The standard economic approach involves changes to prices – through taxation and subsidies – which affect behaviour by changing people’s choice sets. The behavioural approach focuses instead on the way that choices are presented to people, the so-called 'choice architecture'.

Behavioural policies have a number of attractive features. They can achieve the desired outcomes but they typically don’t cost a lot and they don’t impose additional constraints on choices. The success of the Behavioural Insights Team has been to demonstrate, through randomised controlled trials, numerous low-cost ways to improve policy delivery and save money.

For example, tweaking the text of a reminder letter to inform people who failed to pay their tax that most other people had already paid, increased payment rates by over five percentage points. What's not to like? Most of what has been tested by the nudge unit has involved this type of relatively small change to policy delivery. But when it comes to tackling big issues – obesity or energy consumption – there is more work to be done to show that behavioural policies work, and work better than standard economic approaches.

Recent research in collaboration with Rachel Griffith and Stephanie Scholder contrasts behavioural and standard economic approaches in the context of the UK government’s Healthy Start Scheme – vouchers for fruit and vegetables given to low-income households with young children. The behavioural view is that such vouchers will provide a strong signal to increase spending on fruit and vegetables.

But the underlying economic incentive is for only households that currently spend very little on fruit and veg to increase their spending on these specific items. If you are a household that already spends money on fruit and veg, the vouchers are equivalent to an increase in income. Some of this may be spent on fruit and veg, but most of it won’t be (and some may be spent on crisps, cake and chocolate). The evidence supports the standard economic view. There was an increase in spending on fruit and vegetables – equivalent to around one extra portion of fruit and veg a day. So the policy worked. But all of the increase came from low-spending households who had a direct incentive to change behaviour. There was no evidence of any wider nudge effect. The behaviour change was driven by standard economic incentives.

This is only one example but it highlights the need for evidence on when behavioural insights can deliver effective solutions to big policy issues. Nudges may work better when preferences are not so ingrained and when cognitive biases are more prevalent. But economic incentives may provide policymakers with a more proven and powerful tool for encouraging healthy eating; the focus should be on how to design the most cost-effective standard economic incentives, as well as on how to design effective nudge policies.

From the ESRC magazine Britain in 2015