The subject in Economics people most often ask me
about is Behavioural Economics. This doesn’t surprise me: of all Economics’
sub-disciplines, Behavioural Economics is the one that seems to have most
successfully captured the public imagination. What does surprise me, though, is how quickly this rise to prominence has
come about. Just 50 years ago, the term ‘Behavioural Economics’ did not exist.
To understand how Behavioural Economics went from non-existence to Economics’
sexiest topic, we must venture back to the subject’s beginning.
Rational Economic Man
Early political
economists sought to understand human behaviour in a way that was scientific
and generalisable. To do this, they needed a simplified and predictable unit of
analysis. What they came up with was Rational Economic Man (REM) – a
theoretical device that sought to distil human nature to its bare essentials. Human
decision-making was reduced to two fundamental tenets: self-interest and
rationality. Self-interest assumed people’s primary concern was maximising
their own satisfaction or ‘utility’; rationality assumed that people were
capable of pursuing this goal sensibly, via decisions that aligned to various axioms
of rational choice.[1]
Figure 1: Rational
Economic Man
Source: Kate
Raworth, Doughnut Economics
REM proved extremely
useful. His simplicity allowed Economists to build models that were
generalisable and internally consistent. A uniform base unit also allowed them
to model the economy from the ‘bottom-up’, in the same way that physicists
could construct laws of motion from the atom. The rationality assumption seemed
to describe an awful lot of human behaviour, including the near-ubiquitous
negative relationship observed between price and demand. Even the self-interest
assumption didn’t seem too restrictive – especially once one acknowledged that people
could gain personal satisfaction from helping others.
When confronted with
charges that he was unrealistic, defenders of REM typically pulled one of two
cards from their sleeve. First, they argued that the more important the decision,
the more the rationality assumption held, since people could not afford to
behave irrationally. Second, when confronted with examples of human error, they
could argue that these errors were largely random. As such, the ‘the pluses
would cancel the minuses’ and mistakes could be safely ignored for general
analysis.
The Undoing Project
In 1969, two Israeli
psychologists met on a University campus. Their names were Daniel Kahneman and
Amos Tversky.[2]
They didn’t know it at the time, but together they would take a sledgehammer to
the credibility of REM and set the stage for the founding of a new discipline:
Behavioural Economics.
Through a series of
experiments, Kahneman and Tversky showed that mistakes were not random,
and that humans systematically and predictably deviated from
textbook rational behaviour. The Dictator Game provides one famous example: in
this experiment, Player A is given a sum of money – say £10 – and told to offer
an amount to Player B. Player B can either agree to the offer – in which case
both players get the agreed upon amount – or decline the offer – in which case
both Players get nothing. What should Player B do? If fully rational, he
should realise that if he rejects the offer he gets nothing, so he should
accept any positive amount of money – even £0.01. Realising this, the equally rational
Player A should offer as close to £10 as possible – say £9.99.
What do you think
actually happens? Unsurprisingly, this prediction is not borne out in reality –
the average offer is around £4 – though interestingly, as Figure 2 shows, this
varies significantly amongst countries.
Figure 2: Player A
Offers in the Dictator Game
Source: Henrich
et al.
Behavioural Education Economics
Education Economics
provides further nice examples of behavioural findings. By focusing on
self-interest and rationality, REM leaves very little room for the roles of
ideas, stereotypes, habits and emotion. In reality, these are all powerful
motivators of human behaviour. For example, in India, low-caste boys were
essentially just as good at solving puzzles as high-caste boys when caste
identity was not revealed (see Figure 3). However, in mixed-caste groups,
revealing the boys’ castes before a test caused low-caste boys to underperform
high-caste boys by 23% (Hoff and Pandey). In a similar experiment, Shih et al.
(1999) found that Asian-American women primed with their Asian identity
produced superior performance on a math test, whereas participants primed with
their female identity produced decreased performance, relative to women in
control groups.
Figure 3: ‘Framing’
Effects on Academic Performance
Source: World
Development Report 2015: Mind, Society and Behaviour
Role models provide another interesting example. In Uganda, students preparing for their national exams were randomly selected to watch the Queen of Katwe: the inspirational tale of a Ugandan slum-girl who went on to lead the Ugandan team at the Chess Olympiad. Students who watched this film performed significantly better in their exams than a control group who watched a non-inspirational film. REM cannot explain this - point to the need for richer and permissive models of human behaviour.
Role models provide another interesting example. In Uganda, students preparing for their national exams were randomly selected to watch the Queen of Katwe: the inspirational tale of a Ugandan slum-girl who went on to lead the Ugandan team at the Chess Olympiad. Students who watched this film performed significantly better in their exams than a control group who watched a non-inspirational film. REM cannot explain this - point to the need for richer and permissive models of human behaviour.
The Future of Behavioural Economics
So far, Behavioural
Economics has proved extremely effective at identifying scenarios where the
canonical model fails. However, Behavioural Economics needs to be more than just
a collection of observations. Without an attempt to unify these under some
general theory, the subject will remain unsatisfactory. I felt this when I took
a Behavioural Economics course last year: by the end, I hadn’t learnt anything profound
about the world that I couldn’t have learnt from just reading the abstracts of
the required reading.
Generalising
behavioural findings is obviously easier said than done, but I think there are
two reasons for optimism. First, they do not have to reinvent the wheel
entirely – they can borrow many ‘tools’ from the orthodox Economists’ toolkit,
such as general equilibrium analysis. Second, it has some precedent – Kahneman
& Tversky’s Prospect Theory provides a great example of a general theory of
human behaviour gleaned from empirical findings. While Behavioural Economics
may never come up with a model of behaviour as analytically convenient as REM,
further attempts at generalisation can only enhance its credibility.
Word count: 1,000
words
References
- Henrich et al., ‘In Search of Homo Economicus: Behavioural Experiments in 15 Small-Scale Societies’
- Hoff and Pandey, ‘Belief Systems and Durable Inequalities: An Experimental Investigation of Indian Caste’
- Kahneman and Tversky, ‘Prospect Theory: An Analysis of Decision Under Risk’
- Riley, ‘Role Models in Movies: The Impact of Queen of Katwe on Students’ Educational Attainment’
If you are interested in Behavioural Economics in general, I
would recommend ‘Misbehaving’ by Richard Thaler, ‘Thinking Fast and Slow’ by Daniel
Kahneman, or the wonderful ‘The Undoing Project’ by Michael Lewis.
[1] For
those interested, these axioms are completeness (all choices can be
compared); transitivity (if A > B and B > C, A > C) and continuity
(if A > B, then situations suitably close to A must also > B)
[2]
The title and opening sentences of this section are taken from Michael Lewis’ wonderful
biography of the pair, which I implore you to read
