Sending people persuasive reminder letters to pay their taxes recovered ₤210 million of revenue for the UK government. Getting the long term unemployed to write about their experiences,  increased their chances of getting a job. Placing healthy choices of food –like fruit instead of chocolate- in obvious locations improves children’s eating habits. These are all success stories from UK’s ‘nudge unit’, which were recently featured in the New York Times.  Founded by David Cameron, the ‘nudge unit’, whose proper name is the Behavioural Insights Team, draws on behavioural psychology to introduce small changes to prompt people to make the right choices.

I have nothing against this idea. Indeed I strongly support the approach by which all ideas are rigorously evaluated. The fact that most nudges can be randomised at the individual level means that evaluations of these ideas are relatively quick and easy to do.  But what is new about this approach?  And what are the limitations of this approach?

Ever since governments have existed, they have been seeking to change behaviour. All development policies, programmes and projects seek to change behaviour. If people’s welfare is to improve then the beneficiaries of the welfare measure have to behave in a  different way. Even if we give the poor cash, they have to go out and spend it in welfare-enhancing ways. If we give them food parcels, they have to consume the food.

Some interventions are more explicit about their intentions – behaviour change communication programmes being the most obvious. But just providing people information of how they should behave is just one way in which we seek to change behaviour. We also do it through prohibition and exhortation.  We do it through taxes,  subsidies and other financial incentives. We do it by changing the environment, for example, by building infrastructure. So behaviour change has always been at the heart of development. The failure to recognise this – and so seriously consider whether people will indeed change their behaviour in the way we hope and expect – has been behind many development failures.

Nudges are different in their scale. Small nudges are intended to bring about big changes. But can development be achieved by nudges? Doesn’t development require a big push rather than a nudge?

This isn’t really an ‘either or ‘ option. There is room for both. But we should realise that the major structural changes required for development require bigger pushes than a nudge.  Writing about their experiences may help the long term unemployed get into work, but it doesn’t create jobs. Only policy reforms and technological developments to enhance opportunities and productivity can do that.

And there is room to debate the right balance. A J-PAL study in rural Rajasthan, India, looked at the impact of offering non-financial incentives on the immunisation of children. The intervention involved offering parents a kilo of raw lentils per immunisation, and a set of metal plates for completed immunisation. The results of the study show that offering such modest incentives can significantly increase the uptake of immunisation services.

At a presentation of this study in Delhi, Esther Duflo was criticised for just giving away ‘a few plates to parents’ and ignoring the bigger problem of fixing the dysfunctional public health system . Her response was that such reform would take years. So, what was wrong with giving away a few plates now to save some lives? There is nothing wrong with that. But perhaps what is also not right is that focusing on just this approach will draw attention and resources from fixing the underlying problem.

Nudges have their place. And that place is in identifying and piloting small-scale interventions which, if they are cost effective, can be institutionalised on a sustainable basis. It is just about better public administration.  As for the insights from behavioural psychology, many of these insights simply say that economists’ assumptions about homo economicus are mostly wrong. I think most of us knew that already. If people don’t behave according to theory, you need to fix the theory not the people. That’s a topic for another blog.

So, back to the question:  a nudge or a big push? The answer is both. Both have their place, and both need to be subject to rigorous impact evaluation.