Tuesday, January 10, 2017

Behavioural economics – a critique of its policy conclusions

By Philip Booth. He is professor of finance, public policy and ethics at St Mary’s University, Twickenham and senior academic fellow at the Institute of Economic Affairs. Philip was formerly the IEA's Academic and Research Director between 2002-2016. 
"Behavioural economics is not a challenge to neo-classical ways of thinking. It sits firmly within the neo-classical framework. It simply gives us a new way of analysing how markets might deviate from some textbook perfect-market, welfare-maximising equilibrium. That is not to understate its importance. It leads to a new set of what people like to call ‘market failures’ and a new set of tools for resolving the problems – tools, by the way, that many of the original authors in this field (such as Richard Thaler) believe should reduce rather than increase the burden of regulatory interventions.

Austrians school economists, on the other hand, believe that markets simply cannot and will not ever follow the perfect market model of the textbook. So, we should not even look for deviations from some mythical perfection which regulators can fix using the contents of the regulatory tool box.

Austrians would describe the situation that confronts us more as follows…

People in markets go around in a purposeful way trying to maximise their welfare, which is not only monetary welfare. How they do that and how their welfare is measured are subjective matters known only by the persons themselves. People have different preferences for loss aversion, sticking with the status quo (or staying with what you know and trust to put it another way) and so on. Markets help us find ways to improve welfare over time. Entrepreneurs find new products and new ways of producing products. Consumers also find new products and services that satisfy their needs better. They don’t necessarily go about this in the way that neo-classical economists assume (rationally comparing costs and benefits), but they do it in their own way. Sitting at home and watching football and not stressing about whether to move one’s bank account may well be welfare maximising for some – for good reasons to do with how we have evolved, by the way.

The question that faces those of us interested in policy is how confident we can be, when we give a regulatory bureau the power to intervene, that it will improve matters. We should not assume that, because some interventions might have a high probability of success, we should favour giving government the power to intervene so that we can have just the good interventions. When you give the power to government to intervene, you get both the good and the bad just as you get both the good and the bad in markets. And from where might that bad come in the case of regulation?

Firstly, things might go wrong simply because regulators lack the information that is dispersed in markets to understand properly the problem they are trying to solve and also to understand the consequences of their actions. Indeed, it should be rather sobering for any regulator that there are close to 0.25 million academic articles that respond to the search term ‘unintended consequences regulation’ listed on google scholar.

Secondly, regulation might not be shaped by the general public interest, but by private interests. This is one of the areas of research for which James Buchanan won his Nobel Prize in 1986. One of the authors of the TV series, Yes Minister, the late Anthony Jay, was very much influenced by Buchanan’s work.

These private interests might be regulators themselves. They might be politicians. Or they might be the regulated industry which may try to capture the regulatory process for its own benefit.

In addition, politicians and regulators may have cognitive biases that lead them to over-estimate the efficacy of government regulation. They are not neutral arbiters.

These are all manifestations of the reality of human imperfection which it is foolish to ignore and the fact that regulators are using behavioural analysis does not change this at all. The giving of powers to a regulator creates a new institutional situation that leads to the bad stuff as well as the good stuff that we would like to see. We should not compare markets as corrected by an assumed perfect regulator with markets and all their imperfections, we should compare the reality of markets before and after the introduction of a regulatory bureau with all the imperfections that come with it.

Not only that, markets – and society in general – can develop institutions to deal with all sorts of difficulties, including those identified by behaviourists, that might prevent people from improving their welfare.

Until the Social Security Act 1986, it was possible for an employer to require employees to join their pension scheme. This was a voluntary paternalistic device which overcame the supposed myopia of potential members. The government not only made that illegal, they retrospectively over-wrote freely negotiated employment contracts that allowed employers to require their employees to join their pension scheme. This ruling was responsible for about 90 per cent of the pension mis-selling crisis in the early 1990s. Eventually, 30 years on, of course, we have auto-enrolment whereby the government requires all employees to be enrolled in their pension scheme regardless of their preferences and to go through a bureaucratic process to opt out. This is despite the fact that many of those being enrolled may simultaneously be in debt and also saving and therefore paying two sets of intermediation costs whilst having no net assets.

In other words, the government is nudging people very hard – using a behavioural mechanism – to do something that a government in the recent past actually made illegal with retrospective force.

A second example comes from life insurance. Large numbers of people (perhaps even the majority of the population) used to have life insurance and savings policies sold by door-to-door salesmen. ‘Know your customer’ and the mis-selling of insurance being compensated has put paid to that because such insurance was never in the customer’s best financial interests. But, what about other possible preferences for consumers? These may have included the self-respect that came from being able to provide for one’s funeral even if this was not the most rational way to save according to economics 101; the ability to talk to somebody each week who themselves talked to a number of other people in the community – in other words it was a mechanism of socialisation; or, it may be a device that families – generally the women in the family – used to discipline themselves (or their husbands) to save: even at zero interest rates that might have been better than not locking away that money.

Of course, regulators are especially smart. And I am not being sarcastic in saying that – it is almost certainly true, they are. However, regulators cannot know people’s preferences except on average – or at the margin. But, regulators’ actions affect the behaviour of all consumers and not simply those of a ‘representative agent’. We cannot treat people as averages.

In these two examples, we find ways in which people themselves solved their own problems using institutions that developed within the market. Regulators tried to second-guess what these people really wanted with catastrophic consequences in both cases. Regulators now realise that behaviour they effectively banned 30 years ago, that had been going on for 150 years, can now be rationalised according to a new-fangled set of theories which are now being used to enact more regulation to try to reverse the trend they set in train in 1986!

So, what is the fundamental question we need to ask? I think it is this. In general, is it better to allow the market and civil society to come up with solutions, experimenting, trying different things, copying what works and so on; or is it better to vest power in a monopoly government or regulator to deal with these problems? On balance, I believe it is better to let market institutions and civil society solve consumers’ problems including those that seem to have behavioural roots.

Firstly, as mentioned above, regulators can simply get things wrong or enact bad policy because they are motivated not by the general interest but by the interests of rent seekers or the regulators themselves.

Secondly, allowing market institutions to deal with the problem allows experimentation. It means that, if something goes wrong with one solution, something else can be tried. When regulators make mistakes, they make big mistakes that affect everybody.

Thirdly, we cannot rely on regulators to know their own behavioural biases and correct them – they will, I believe, systematically over-estimate their own ability to improve market outcomes.

Fourthly, regulators do not know as much as they think they know about the real underlying preferences of consumers. As such, we get the situation of people being nudged in the wrong direction! Indeed, very often further investigation of tendencies that behaviourists posit as being sub-optimal turn out to be reasonable – especially in the long term.

Fifthly, regulators have a tendency to under-estimate the costs they are imposing on others (I believe they do this systematically because they understand the regulation better than any of the millions of people or businesses on which it has been imposed). So nudging into pensions has turned into a mountain of paperwork, for example.

A key feature of a realistic picture of human nature as opposed to the economics textbook is that we need to learn. Information is acquired by making mistakes. That is how we progress – by making errors and learning. When we try to regulate markets, both in conventional ways and through the application of behavioural economics to try to hit some kind of welfare maximising equilibrium, we need to ask whether we are we undermining this learning process and infantilising consumers. Dose action in one area make consumers more lethargic and less able to make judgements in other areas?

This does not mean that I don’t think there is some role for behavioural economics in public policy.

It might be important in public administration. If supermarkets can use what they know about people’s behaviour to increase sales then those who manage the roads or collect the taxes might want to do the same.

The second area where behavioural economics can be useful is as a replacement for other forms of regulation. It is interesting that Thaler and Sunstein often talk about their ideas in this context. I am not a great fan of how they frame things, but they make a lot of good points. When challenged, those authors always argue that they are in favour of nudges instead of regulation that directs behaviour. That is why they call their set of ideas ‘libertarian paternalism’. I have only seen behavioural economics put forward by regulators either as an additional justification for regulation or as a regulatory option when a new intervention is being tried. What about a whole programme of replacing regulation with nudges? Wouldn’t it be nice to see that coming out of the FCA?

We need a bit more humility amongst regulators in general. And, perhaps it is appropriate that this is my last public statement as Academic and Research Director at the IEA. We seem to have gone from banning and having the state directly control large amounts of economic activity when the IEA was founded in 1955 to a situation where we have regulators who use economics 101 supplemented with behavioural economics to try to bring perfection to markets that simply cannot be perfected and perhaps cannot be improved. The end result is more pages of regulation than it is even possible to count. The question we have to ask is whether that approach is raising the fixed costs of business operation, providing a bar to new entrants, discouraging innovation, treating information like a commodity and therefore preventing the communication of tacit information, and preventing the evolution of the very institutions and habits that might help resolve the problems that regulators seek to solve. I will repeat what I said a few moments ago. We cannot choose only good regulations. We can only choose institutions and those institutions will give us both the good and the bad in a mix that will depend on the individuals who control them and the institutional background within which they work. As Hayek said: ‘the curious task of economics is to teach people how little they know about what they believe they can control.’"

No comments:

Post a Comment