Introduction to Behavioral Economics

Topics in Economics, ESCP, 2023-2024

Author

Pablo Winant

Published

February 12, 2024

Behavioural Economics

Behavioral economics combines elements of economics and psychology to understand how and why people behave the way they do in the real world. It differs from neoclassical economics, which assumes that most people have well-defined preferences and make well-informed, self-interested decisions based on those preferences. (intro from university of Chicago)

Behavioural Economists

identify actual patterns in agent’s decisions
- measured as deviations to idealized, rational, utility-maximizing economic behaviour
develop theories that can, consistently predict these patterns
patterns and theories are tested empirically in labs

Good introduction: Thinking Fast and Slow

The lab

Behavioural economists recruit participants for their experiments.

they perform some specific tasks or play a specifically designed game
they are usually paid and incentivized (skin in the game)

Experiments

experiments try to recreate conditions so as to expose specific features of human choices
usually one determinant of choice is randomized so as to measure its effect cleanly
more participants lead to better results (if representative)

The Marshmallow Experiment

Instant Gratification

Deviations from rationality

The (neo)-classical view

Agents are perfectly rational…
- they don’t make errors
…and pursue their self interest
- specified by preferences

\[\max_{x} U(x)\]

The ultimatum game

Setup

Two players negotiate over a fixed amount of money (the pie):

player 1 demands a proportion of the pie
player 2 sees the size and player’s 1 demand and either rejects or accepts it
- accept: split is implemented
- reject: both players get 0

hundreds of papers since original one by Guth et al.
consistent results: average offer is 40%, offers below 20% have high chances to be rejected, lots of 50-50
note that the “rational choice” for both agents would be 99-1!
it is an example of other regarding preferences (aka altruism)

What determines the split? What influences altruism?

. . .

Can you devise an experiment to measure the effect of “shame”?

The ultimatum game: measuring shame

Improved protocol

for half of the candidates, exactly as before (control group)
for the other half (treatment group), the two players, don’t see each other don’t have any contact. Only the demand from player one is communicated to player 2.

Result:

-> players are less generous but the result holds

Reflection Effect or Loss Aversion

Result from literature

first lottery: 92% chose B
second lottery: 80% chose A

Intepretation:

agents are risk-seekers when it is about loosing money
agents are risk-averse when it is about winning money

The Certainty effect / The Allais paradox

Result from literature

first lottery: most people choose A
second lottery: most people choose B

Intepretation: people underweigh outcomes that are merely probable compared to those that are certain

Expected utility

Expected utility framework
- simple assumptions on people’s preferences
- agents maximize $U(x)$ where $U$ is increasing and concave
- standard in economics and finance
Concavity implies that agents are risk agents are risk averse

Prospect theory

Prospect¹ theory was proposed by Kahneman and Tversky in 1979
Attempts to describe realistically how humans evaluate losses and gains

Some Behavioural Anomalies

Behavioural anomalies

We classically distinguish two kinds of behavioural anomalies

predisposition towards error
mental shortcuts (aka heuristics)

They have been studied by psychologists for a while

And are being revisited systematically by economists/psychologists
…following the “replication crisis”²

Instructions: You are vice president of product development and are evaluating 8 new product proposals. You have asked two people from R&D (in whom you have equal confidence) to give independent forecasts of the R&D costs.

Project no.	A’s forecast	B’s forecast
1	167,000	272,000
2	274,000	783,000
3	529,000	433,000
4	357,000	866,000
5	146,000	659,000
6	937,000	446,000
7	906,000	811,000
8	483,000	379,000

Project no.	A’s forecast	B’s forecast	Midpoint	Median forecast
1	167,000	272,000	219,500	250,000
2	274,000	783,000	528,500	600,000
3	529,000	433,000	481,000	500,000
4	357,000	866,000	611,500	697,110
5	146,000	659,000	402,500	458,850
6	937,000	446,000	691,500	788,310
7	906,000	811,000	858,500	875,000
8	483,000	379,000	431,000	450,000

. . .

If A and B’s forecasts are unbiased, your best estimate would be the midpoint
If your forecast is above the midpoint, you are implicitly assuming that R&D personnel underestimate the costs
The median forecast shows the typical response of MBA students with 10 years of executive experience: they have excessive optimism.

Compared to the rest of this group, how would you rate your driving skills?

A. Above average
B. Below average

. . .

Over confidence

Overestimating our own abilities

. . .

Hindsight Bias

When you search for or interpret new information in a way that supports existing beliefs as opposed to challenge them. ???

. . .

Setup (Study from Lord, Ross & Leeper, 1979)

48 undergraduates supporting and opposing capital punishment were exposed to 2 purported studies, one seemingly confirming and one seemingly disconfirming their existing beliefs about the deterrent efficacy of the death penalty. As predicted, both proponents and opponents of capital punishment rated those results and procedures that confirmed their own beliefs to be the more convincing and probative ones.

Result

Participants gave higher ratings to studies that confirmed their initial point of view even when studies on both sides had been carried out by the same method. In the end though everyone had read all the same studies, both those who initially supported the death penalty and those who initially opposed it reported that reading the studies had strengthened their beliefs.

. . .

Confirmation Bias

When you search for or interpret new information in a way that supports existing beliefs as opposed to challenge them.

Test:

Write down 3 examples of successes that you have had
Write down 3 examples of failures
What were the causes of the success?
What were the causes of the failures?

Interpretation

Attribution bias occurs when people attribute their success to skills, and their failures to randomness (Taleb 2001)
It is when we attribute praise/blame to the person rather than the situation. However, successful people can also mistake hard work for luck. Timing can be important but so is developing the skills to be able to take opportunities when presented. After all, “the harder you practice, the luckier you get” Gary Player

. . .

Attribution bias

Overestimating how much control we have over events

Example:

People are given a mug worth 10$
Then the opportunity to exchange it for pens worth 11$
Most people don’t take the deal

. . .

A striking example

Do you see other examples? Implications

. . .

Ownership Bias / Endowment Effect Heuristic

People place a higher valuation on an asset, purely by owning it.

Example: We tend to think that footballers make more money than dentists, but only because we don’t see the failures (there’s a selection bias)

Dentistry is a more lucrative profession since it maximises your expected income across all alternative histories
One cannot consider a profession without taking into account the average of the people who enter it, not the sample of those who succeed Taleb (2001)

Availability Heuristic

Judgement based on the ease with which instances come to mind

Question:

What percentage of members of the United Nations are countries in Africa?

Experiment:

Show a number: e.g. 65
Ask participants: “What percentage of members of the United Nations are countries in Africa?”

Result

People who saw a 10 had an average guess of 25%
People who saw a 65 had an average guess of 45%
Correct answer: 28%. There are 54 African countries, and 192 UN member states

. . .

Anchoring Heuristic

Viewing things in relation to an irrelevant comparison point

Don’t forget to add 65 on the board before.

How oil spills harm birds, dolphins, sea lions and other wildlife

Example: Following the Exxon Valdez oil spill, people were asked how much they were willing to pay for equipment that would do the following:

Save 2,000 birds: $80
Save 20,000 birds: $78
Save 200,000 birds: $88

. . .

The almost complete neglect of quantities in such emotional contexts has been confirmed many times

. . .

Affect Heuristic

Definition: Over-reliance on our immediate emotional reaction

Conclusion

So we know that people deviate from rationality in many ways that can be measured and / or related to theory.

. . .

Policy-makers can exploit the knowledge to design better policies.

Nudges

A nudge is a way to change agent’s decision by changing the decision context, without changing the preferences or the constraints in a meaningful way.

. . .

In principle one could use nudges as an alternative to traditional incentives in order to favour green behaviour:

. . .

Does it work?
Is it ethical?

. . .

To be discussed in the next session…

👉 Read the working paper Nudging as an Environmental Policy

Footnotes

Prospect is a close synonym for gamble↩︎
for details, see the (excellent) video from veritasium ↩︎