Wednesday, August 8, 2007


Dinner address, Reserve Bank Roundtable, August 8, 2007

‘Economics has not only become boring but also threatens to become irrelevant.
Therefore I do not feel embarrassed about being unorthodox. In fact, I rather enjoy it!’

Frey (2001)

One of my favourite economist jokes is the one that says an economist is someone
who can’t see something working in practice without wondering whether it also
works in theory. There are two professions that possess an intuitive understanding
of the propositions economists have come to call ‘behavioural economics’. They are
the marketers, and the politicians. So what is behavioural economics? It’s
economists satisfying themselves intellectually that there is a logic — as opposed to
a rationality — to the intuitive behaviour of economic agents. It’s economists
laboriously disabusing themselves of the mistaken beliefs they have acquired about
the way agents behave, as a result of their internalising the assumptions on which
neoclassical economics is built.

Behavioural economics is the scientific study of intuition. It involves accepting the
power of intuition — that people are much more intuitive than rational — and
understanding the reason why this is so, which gets back to the way humans have
evolved and, specifically, the way their brains have evolved. Neuroscience tells us
that the primitive, more instinctive and emotional part of our brain often
overrides — or beats to the punch — the more recent, more logical part of our
brain. This leads to a strange dualism in our minds: we’re often motivated to do
things by considerations the more intellectual part of our brain knows to be unwise.
This dualism explains why we have ‘present selves’ and ‘future selves’ which, in
turn, helps explain the self-control problem humans have — a major topic of study
for behavioural economists — and the misprediction of utility, time-inconsistent
preferences, myopia and procrastination that this involves (Stutzer and Frey 2006). I
guess what I’m saying is that, for a full appreciation of the intellectual power and
fascination of behavioural economics, it helps to take in some neuroeconomics
(Camerer et al. 2005).

Let me also say that I use the term behavioural economics to encompass the closely
related field of research into happiness — or subjective wellbeing, if you prefer a
more scientific-sounding label. Lest you feel that happiness is taking your
newly-acquired tolerance of behavioural economics a bridge too far, let me just
point out that happiness is the subject that brought Professor Frey’s name to
international prominence (Frey and Stutzer 2002), and that when Daniel Kahneman,
the psychologist who won the Nobel prize in economics for his role in founding
behavioural economics, had finished with prospect theory and heuristics, he moved
on to the study of utility and wellbeing (Kahneman et al. 1999).

Conventional, neoclassical economics is widely held to be positive, not normative.
But one of the things you soon realise when you study behavioural economics is
that this is the wrong way round. Behavioural economics is the study of the way the
world actually is, whereas conventional economics is the study of the way the world
should be. That is, we’re not rational, but in many circumstances we’d be better off
if we were. The study of self-control problems involves reaching an understanding
that the seemingly irrational things people do in their search for commitment
devices — such as failing to claim tax rebates in fortnightly 'pay as you earn' tax
instalments so as to maximise the size of their annual tax refund cheque, which the
person is more likely to save — have their own logic; that they represent fallible
agents trying to make themselves more rational. The study of self-control problems
also leads you to the view that there may be a new role for economics in helping to
make the world more rational, by imposing prohibitions on certain disadvantageous
behaviour. In fact, governments already do this extensively — and with widespread
public acceptance. It’s just that no one sees it as having anything to do with
economics, and often economists would be quietly disapproving of such

As an economic journalist, I’m supposed to keep my remarks practical, but I do
want to say something theoretical and controversial. I believe that the assumptions
on which the neoclassical model is based pervade the beliefs and policy preferences
of economists far more than most of them realise. Economists generally have a
strong commitment to individualism, freedom of the individual, the benefits of
choice and the value of personal responsibility and, hence, a bias against
government intervention and a desire to keep governments small and taxes low.
This characteristic of neoclassical economics gives it a great affinity with the
libertarian political philosophy, which to me explains why the right wing of
economic rationalism is a lot more heavily populated than the left wing. (Who’s on
the left wing of economic rationalism? Mike Keating, Fred Argy, Bob Gregory and
a few others.)

But my point is this: I believe conventional economics’ commitment to
individualism and suspicion of government intervention rests heavily — more
heavily than most economists realise — on the assumption that economic agents act
rationally. We doubt that governments could ever know better than the individual
how that individual’s income could best be spent. Why? Because we assume the
individual is rational in all things — that she can accurately predict utility, never
does things she comes to regret and never displays time-inconsistent preferences.
When you accept that individuals are far from rational you open up the possibility
that governments may well be better judges of what’s best for the individual. We
assume agents are rugged individualists and are happiest when treated as such,
whereas psychology tells us humans are group animals, whose preferences are
heavily influenced by those around them, who care deeply about what others think
of them, who are anxious to fit in but also conscious of their status within the group
and desirous of raising that status. In other words, I believe that conventional
economics’ exaltation of individual freedom is simply scientifically outdated — a
hangover from the 18th and 19th centuries, when we knew far less about human
behaviour than we do today.

There are two broad approaches economists can adopt towards the lessons of
behavioural economics. One is to use insights from behavioural economics to
reframe essentially unchanged policy prescriptions from conventional economics,
so as to make them more politically palatable. We know, for instance, that people
react differently to essentially the same propositions, depending on how they are
framed. We know that more people would decline consent for a medical operation
with a 10 per cent failure rate than they would an operation with a 90 per cent
success rate. We know from Kahneman’s asymmetric value function, for instance,
that people weight losses more heavily than gains of the same amount. From this,
Richard Thaler (1985) developed four rules for reframing gains and losses:
segregate gains (don’t wrap all the Christmas presents in a single box), combine
losses (because this reduces aggregate pain), offset a small loss with a larger gain
and segregate small gains from large losses.

The second approach we can adopt is to use the lessons of behavioural economics to
change the policies we pursue. I have no objection to the first approach — indeed, I
think it would repay the close attention of econocrats. But I’m more excited by the
second, more radical approach. So let me suggest some very general policy
implications I draw from behavioural economics.

First, I believe that the profession needs to return to its original goal of maximising
aggregate utility rather than maximising consumption possibilities. We now know it
is possible to measure utility — to some extent at least. We also know that revealed
preference is far from foolproof. People are not good at predicting their utility and
they often come to regret their decisions — even to wish someone had stopped them
doing what they did. We know people get locked into behaviours they wish they
could control. Neuroscience makes it easy to see how people’s consumption
decisions can be influenced at a semi-conscious level by advertising that appeals to
their emotions. Among other implications, a switch of emphasis from consumption
back to utility would require economists to abandon their see-no-evil approach to
advertising. Many of the points that follow flow from a recommitment to
maximising utility.

Second, economists need to study consumption. It never ceases to amaze me that
economists can exalt consumption in the way they do and then take so little interest
in it. The happiness literature makes it clear that people find some forms of
consumption more satisfying than others (Seligman 2002; Van Boven and
Gilovich 2003).

Third, economists need to acknowledge the importance people attach to social
status and social comparison. Conventional economics is good at helping the
community maximise its income, but it can do nothing to maximise people’s
relative income. And yet, we know that people are more interested in increasing
their income in relative terms than absolute terms. From a community-wide
perspective, a status race is pointless and wasteful. It’s likely that, as real income
rises over time, a higher proportion of income is devoted to the purchase of
positional goods. Is this why we pursue efficiency? It’s also likely that efforts to
minimise the role of government and limit the growth of taxation have the effect of
allowing people to maximise their spending on positional goods at the expense of
the provision of public goods that would yield them greater utility (Frank 1999).

Fourth, the simple model of labour supply is misleading and needs rethinking. In
practice, economists tend to underplay the one thing the model gets right: that
leisure yields utility. In the unfavourable comparisons of rates of economic growth
and levels of GDP per capita made between America and Europe, there is little
acknowledgement that much of the difference is explained by the Europeans’
preference for leisure over work. On the other hand, the model is quite wrong in
assuming that work yields disutility. The happiness literature makes that clear —
even if it wasn’t obvious. Like you and me, most people derive great utility from
their work most of the time. It follows that much could be done to increase utility by
policies encouraging job enrichment. That is, when your goal is to maximise utility
rather than consumption, you see for the first time that the issue of job
satisfaction — which may be enhanced by such practices as team work or giving
workers greater autonomy — is part of the economist’s brief.

We know, too, that unemployment is a major source of unhappiness in peoples’
lives — or, if you prefer, of disutility (Clark and Oswald 1994; Layard 2003). This
fact creates a conflict between measures to increase efficiency and maximise utility
that reformers rarely acknowledge. This may be partly because their modelling
assumes full employment, but I believe it’s also thanks to a hidden assumption that
the unemployed are to be envied for all their leisure time.

Fifth, policy makers undervalue the utility people derive from security and
predictability. We give too little weight to the utility workers derive from job
security, for instance. We need to learn that efficiency isn’t everything.
Sixth, self-control problems are ubiquitous, but susceptible to policy remedies.
Some self-control problems may be regarded as minor (television watching, for
instance), but many constitute significant social and economic problems: obesity,
smoking, drinking, drug-taking, gambling, speeding, the overuse of credit and the
inability to save. Economists aren’t as conscious as they should be that government
intervention — and often, outright controls — to assist people conquer their
self-control problems and to protect the community from negative externalities are
widespread, of long standing and uncontroversial. Consider all the regulation
governing the consumption, sale and advertising of alcohol and tobacco. Consider
all the controls — speed limits, seatbelts, random breath-testing — that have
succeeded in reducing the road toll. Consider the way employees are compelled to
save 9 per cent of their wages, and how little opposition that relatively recent
measure encountered. It’s clear to me that the public often wants governments to
impose these external commitment devices on it — and that this attitude makes
considerable sense. The insights of behavioural economics should help economists
to be much more receptive to proposals to use intervention to alleviate self-control
problems, including the newly recognised problem of obesity.

Economics doesn’t have to be boring, stuck in a rut and open to the charge of being
based on out-of-date science. But to make economics more interesting and relevant
to the solution of a wider range of the community’s problems, economists have to
be willing to learn new tricks.

Behavioural Economics and Public Policy