In today's Guardian, George Osborne writes about the virtues of "Nudge". He begins:
A few years ago tax collectors in Minnesota were grappling with a problem governments have been trying to crack for decades: how to get people to fill in their tax returns on time. They tried various approaches, from threatening fines to sending out leaflets, but nothing seemed to work.
Then they tried something you'd never come across in a traditional policy textbook. Instead of threatening or cajoling the public, they simply publicised a single fact: most Minnesotans had already filled in their returns. The result? The number of people submitting tax forms shot up almost overnight.
Classic economic theory cannot tell you why this happens, because it's based on the assumption that humans always act in their own self-interest. But as we all know, humans aren't always rational. We do things we know we shouldn't, like eating too much or not saving enough. And, as the Minnesota example shows, we're also powerfully influenced by the people around us.
Mr Osborne is straightforwardly incorrect to assert that "Classical economic theory cannot tell you why this happens." (He is also, incidentally, incorrect to assert that classical economic theory is "based on the assumption that humans always act in their own self-interest", but we shan't go into that very much here.) That people are more likely to do things that other people are doing is a well-known and long-studied phenomenon, and classical economic theory offers many different bases for analyzing it.
For example, in the case of the tax returns, until we do one, we may be unsure how time-consuming and difficult it is to do, and how likely the submission is to be accurate enough to be accepted. That lots of other people have managed to get to the end and had their submissions accepted by the authorities provides us with additional information that we previously lacked as to the (hassle) costs of submission. Again, if most people have not completed their tax returns, then the chances that the authorities will chase and fine me for not completing mine are low; but if almost everyone else has submitted, then the chances that I will be chased and perhaps fined are high. So the fact that many other people have completed their returns increases my incentives to do so. Each of these is an instance of the general (rational) phenomenon of feeling safer in a crowd.
As I have argued before, it is weak simply to assert that, in a situation that we find difficult to explain rationally, the correct conclusion is that it cannot be explained rationally. To do so is to abandon scientific research in favour of story-telling.
But again, as I have argued many times, this does not mean that we should not try to conduct research using bounded rationality assumptions (indeed, I have done so myself and still do). But such research must have its own alternative constraining principles - not simply hope to "let the data speak".
Also, it does not follow that there is nothing in Thaler-style "nudging". Perhaps, though classical economic theory might explain crowding behaviour just as well or even better than behavioural economics does (if behavioural economics really "explains" anything, as opposed to modelling it/accounting for it - I suspect it doesn't, and doesn't seek to) - even if this is true, it could still be the case that the behavioural economics approach, by focusing on certain behaviours, might be more likely to produce certain sorts of potentially innovative and useful policy recommendation.
Thus, I'm not seeking to rubbish the concept of government using small signals and small effects to aim to produce disproportionately large changes in human behaviour. But we should not become carried away with the idea that classical economics is all demonstrably bankrupt and can be replaced by the shiny new ideas behind nudging. Classical economic theory has contributed enormously to human welfare, and continues to do so. If behavioural economics is not simply an intellectual blind-alley - like so many shiny new ideas before it - then its contributions will eventually come, but they will add (if they do) to an enormously successful structure of human thought and endeavour. I'm not expecting economics to disappear any time soon.
