Do you prefer A or B in each of the following:
(1) A: 61% chance of winning £5,200 or B: 63% chance of winning £5,000
(2) A: 98% chance of winning £5,200 or B: 100% chance of winning £5,000
Social sciences evolve over time and
Economics is no exception. Being a subjective discipline in many respects,
economic theory is constantly being created, modified, rejected and accepted.
Theories that were widely accepted for centuries can be discredited by a new
data set and entire new sub-disciplines can develop from a single publication.
I therefore find it astonishing that
Economics is so resistant to changes at its very core. Economists refuse to
budge when it comes to criticisms of the foundations of the discipline, even in
the face of damning evidence. At the very centre of most economic theory sits
two things: Expected Utility Theory and a concept of rationality, both of which
are fundamentally flawed. Even in the presence of overwhelming, undeniable evidence,
these two theories form the basis of most economic theory. Undergraduates are
still taught the subject using these concepts as the building blocks for the
rest of their course.
Expected Utility theory is about decision making. It posits that humans are risk averse and that their “utility” (or
happiness) directly depends on their level of wealth. It argues, quite
sensibly, that every additional pound you earn adds less to your happiness than
the last pound. This is intuitive, £100 to £101 has a larger marginal effect on
your happiness than £1000 to £1001.
Rational choice theory views people as selfish, materially interested
agents that maximize their utility. A rational person can consume all relevant
information and accurately assess their costs and benefits to make decisions.
Most importantly, people make consistent choices. Agents understand their
preferences and can aptly make decisions. If an individual prefers ice creams
to chocolates and chocolates to fruit then he certainly prefers ice creams to
fruit.
Obviously this is an unrealistic view of
human nature and economists use these as simplifying assumptions in their
models to understand human behaviour. Unfortunately this one-size-fits-all ideal
is unrealistic. They study “Econs” and not Humans. Of course it makes sense to
assume a Human makes consistent choices, but any exceptions to the rule are
labeled as irrational and
subsequently ignored. Instead, Economics needs to start studying Humans and
adapt their definition of rationality accordingly.
Many economists have challenged the status
quo over the last century. Behavioural Economics is a sub-subsection of subject
that uses psychology in conjunction with economic theory. For example, experiments
have shown that a Human’s utility does not depend on the level of their wealth but on changes in their
wealth. This seems obvious. A millionaire would definitely be more disappointed
to find his wealth stood at half a million than a individual who was previously
bankrupt.
What fascinates me the most is that
Behavioural economists have highlighted many flaws to the convention yet the
discipline is still resistant to change. All exceptions to the rule or predictable irrationalities are simply banished
to the sub-discpline of Behavioural Economics.
This is the equivalent of a mathematician using the rule that 2+2=5 in
all of her calculations and then adding a footnote that says in fact 2+2 may
actually equal 4. Economists prefer
to assume their concept of “rationality” and analyse what would happen in an
ideal world where the oddities of inconsistency do not exist.
Having read Daniel Kahneman’s ‘Thinking Fast and Slow’ I have been enriched
with examples that leave economists dumfounded. Many of the examples below are
directly from his book, which I strongly recommend.
Allais
Paradox
I gave you a question at the start of this
article. If you are like most people then you chose A in (1) and B in (2). You
probably thought I might as well go for the higher amount in (1), after all
there is only a 2% difference in chance. However, in (2), I can have £5,000
with certainty so why risk the gamble at all? This logic seems incredibly fair,
but it is completely irrational
according to economists. The reason being that the chances of winning both
prizes have increased by the same amount (37%) in both cases. If you preferred
A in (1) then a consistent choice would be A in (2). The chance of winning the larger amount has
increased by the same percentage as the smaller amount, so why switch to the
smaller amount?
Behavioural economists associate this
irregularity with the effect of certainty and regret. Humans wildly
underestimate high probabilities in the presence of certainty. The weight
people place on the 2% difference in each case is much larger in question (2).
Furthermore, the regret effect would not be present in (1). If you win nothing in (2) by choosing to gamble you would probably kick yourself for taking the risk
at all. The decisions most people make are of course inconsistent but I would
not describe what the majority do here as irrational. They did not fit the rule
because the rule does not consider the effect of certainty and emotions. The
rule needs to change. This famous example shocked economists when first
introduced, but once again, was written off as an unusual exception to
consistent behaviour.
Interestingly, at the other end of the
spectrum, the idea of overweighting small odds explains why a lot of people
play the lottery and gamble. When the probability of something increases from
0% to 1%, there is a possibility effect.
Humans tend to overweigh the 1% and this leads to risk seeking behaviour. For
example, if I gave you the choice between simply taking £500 now or a gamble
that offers you 50-50 odds on winning £1000 or nothing, then being risk averse,
you will probably choose the certain £500. However, if I lowered the odds, and
offered you £10 with certainty or a 1% chance of £1000, you will almost
definitely take on the gamble. But in both cases, the gamble and certainty have
the same expected value and a risk averse individual should always prefer the
certain money. The possibility of
the large prize and the smaller certain amount leads to inconsistency (you’re a
Human, not an Econ!).
Linda
& Karl
My favourite example that perfectly
illustrates a human’s capacity to be completely illogical is taken straight
from Kahneman’s book. Consider the fictional character Linda below:
Linda
is 31 years old, single, outspoken, and very bright. She majored in philosophy.
As a student, she was deeply concerned with issues of discrimination and social
justice, and also participated in anti-nuclear demonstrations.
Given the description, now rank the following
scenarios in order of likelihood:
1) Linda is a teacher in
elementary school
2 Linda is part of a feminist
movement
3) Linda is an investment
banker
4) Linda is social worker
5) Linda is an investment
banker and part of a feminist movement
Given her description, most people would
not be surprised if Linda were a social worker, part of a feminist movement or
even a teacher and these are usually ranked very highly. The study investigates
the other two scenarios (3&5). Most respondents thought it very unlikely
that Linda was an investment banker. They were more likely to believe that
Linda was an investment banker and part of a feminist movement, which at least
agrees with part of her description. If you were like 90% of the respondents,
you would have ranked 5 above 3.
For those of you who understand Venn
diagrams, you may have already guessed this is utterly illogical. All
investment bankers who are also part of feminist movement wholly belong to the set
of investment bankers. No individual can apply to scenario 5 without applying
to scenario 3 whereas there are obviously investment bankers who are not part
of a feminist movement. The probability of 3 is without question larger than 5
regardless of the description. This is
an obvious failure of human reasoning but it is widespread. Humans tend to
focus on the representativeness of the situation rather than statistics.
Statistics and logical thinking take time and the law of least effort will
force you to jump to the illogical conclusion with confidence. Of course at closer
look, it is absurd to rank 5 above 3.
A similar failure to focus on statistics is
of Karl who lives in the US:
Karl
is very shy and withdrawn, invariably helpful but with little interest in
people or in the world of reality. A meek and tidy soul, he has a need for
order and structure, and a passion for detail.
Is
Karl more likely to be a librarian or farmer?
An intuitive answer jumps straight for librarian.
The description fits the stereotype. Yet people fail to factor in the base
statistics. There are 20 male farmers for every librarian in the US. Even if
you believe that the description above is 20 times more likely to represent a librarian than a farmer, you should still choose a farmer. An Econ would have
used this thought process. However, you did not consider the base statistics,
even if you knew them, because Humans prefer to focus on the fitting story of
representativeness whilst statistics fall by the wayside.
Framing
There
is a disease that will kill 600 people in the next year. The government has two
options:
A) If they adopt procedure A,
200 lives will be saved
B) If they adopt procedure B,
there is a one-third probability that 600 people will be saved and a two-third
probability nobody will be saved
The expected value of saved lives in both
procedures is 200, but most people prefer not to take the gamble and choose
procedure A. There is an aspect of regret if B is chosen and an effect of
certainty in procedure A. However, when the procedures are reworded and
presented to another group the results are startling:
A) If they adopt procedure A,
400 people will die
B) If they adopt procedure B,
there is a one-third probability that nobody will die and a two-third
probability that 600 people will die
The procedures have not changed but they
have been ‘framed’ differently. If you are like most of the respondents, you
probably will prefer the gamble this time. The effect of regret disappears
as option A is now guaranteeing a certainty of deaths rather than lives saved.
This is completely inconsistent. Economics cannot deal with emotions like this.
When the two groups are shown the other frame, most people struggle to decide
which procedure to choose. Which one would you choose given both frames?
Framing is important in reality which is why many food manufacturers write “90%
fat free” rather than “10% fat” or why a doctor will tell you there is a “95%
survival rate” rather than a “5% mortality rate.” Framing also occurs when
using numbers instead of percentages. A dog rescue appeal will not tell you
that 0.1% of dogs are neglected but will instead say that 1 dog in 1,000 are
left neglected. It invokes an image of a single neglected dog and has been
proven to raise larger donations.
Regression
to the mean
Humans fail to understand how important
luck is in our lives, something an Econ would never do. In an Israeli army
training camp that Kahneman visited, the captain was adamant that it was better
to shout at an officer who had performed terribly rather than praise an officer
who had been exceptional. He had found that those he shouted always performed
better next time round and those he praised always did worse. This was a gold
mine for Kahneman.
The captain had made up a casual story in
his head that somehow praise lead to complacency and aggression instilled a
fear of repetition. However, the results he found can be simply explained by
statistics. There are two things involved in performance: skill and luck. There
is no doubt that skills are involved when an officer performs well, but unless
that officer is consistently exceptional, there is always a large amount of
good luck involved. Therefore, when he praised an outstanding officer whose
performance subsequently dropped, this was not due to complacency, but due to
less luck and the same amount of skill. In mathematical terms, this officer had
regressed towards the mean (moving towards his average). When the captain
shouted at recruits and they subsequently improved, this is because they had
better luck and the same skill the following time. They regressed towards the
mean once again. Of course skill is important but people rarely put enough
weight on the value of luck.
Anchors
Was
Ghandi older than 35 years old when he died?
How
old was Ghandi when he died?
Answer the above questions and note your
answer. Unless you know the answer for sure, the first question almost
certainly affected the second answer. If the first question had been was Ghandi
younger than 130 years old when he died, your answer would have almost
certainly been higher for the second question. The first question should have
no importance on the second but your mind was anchored to 35. It has been shown that people tend to focus on the
arbitrary number provided and move away from it until they find a sensible
answer. Anchoring is a clever negotiation technique used by firms, unions and
governments.
Time to change
Expected Utility theory and rational choice
theory isn’t all bad. It provides a framework for the discipline and has some
intuitive suggestions. It provides a base for theoretical work in which we can
slowly adjust assumptions to reach realistic conclusions. Economics needs a
base and it is unrealistic to incorporate all human “irrationality” into every
model.
However, there are some instances, as
described above, that are so predictably irrational
that it is absurd not to adapt our baseline. Given the overwhelming evidence,
only some of which I have summarised here, surely it is time for Economics to
incorporate Humans, even slightly, into their model of rationality. It is no
longer acceptable to label the inconsistent majority as irrational. Economics
is a study of human behaviour and economists must realise that Econs differ
from Humans in many respects. As Ariely once said, Economics “makes the man fit
the model rather than the other way around.” It is time for this to change.
