You are on page 1of 17

Behavioral and experimental economics

Prof. Eberhard Feess


Victoria University of Wellington
April 2024

Outline

I. What is Behavioral and Experimental Economics all about?


II. The benefits (and potential drawbacks) of Experimental Economics
III. A first simple game – and what we can learn from it
IV. A second simple game – and what we can learn from it
V. Some mainly robust insights
VI. A typical example for nudging
VII. Another way of nudging: The relevance of irrelevant alternatives
VIII. Own research I: Self-selection to (cooperative) teams
IX. Own research II: Insights on unethical behavior
I. What is Behavioral and Experimental Economics all about?

Until about 1990, standard (“neoclassical”) economics assumed that


• People are fully rational in the sense that they do the best to maximize
their objective functions,
• and that these objective functions are rather simple, including profit-
maximization for companies (makes mainly sense) and simple utility
functions for individuals (money, selfishness).

• Behavioral Economics adds deviations from rationality and selfishness


to the picture, thereby more and more relying on insights from (social)
psychology
• Experimental Economics investigates behavior in neatly defined,
though somewhat artificial environments.
I. What is Behavioral and Experimental Economics all about?

Early Behavioral and Experimental Economics restricted attention mainly


to identifying deviations: sự lệch lạc from both rationality and
selfishness, which is, of course, nor surprising at all.
Many Behavioral and Experimental Economists inadequately harshly
restricted standard economics by ignoring that
(i) no-one ever doubted that people are neither fully rational nor selfish,
and that
(ii) selfishness is often a good proxy: đại diện, ủy quyền for actual
behavior in cases of markets and anonymity (financial crisis!)
This ignorance had a very detrimental impact on the perception of
economics in society as anti-social and unworldly.
II. The benefits (and drawbacks?) of Behavioral Economics

However: Over time, Behavioral Economists showed:


- Deviations from rationality and selfishness are systematically in
specific directions and hence predictable;
- And heterogeneity: tính không đồng nhất in biases, preferences and
behaviour can be linked to gender, age, personality traits…

This is exciting! – It can be used to increase the predictive power of the


impact of institutions, and can be (ab)used to influence behaviour
(nudging).
Since then, more complex preferences and biases are an integral part of
game theoretical modelling.
II. The benefits (and drawbacks?) of Experimental Economics

Some benefits

- An artificial environment enables us to focus on the effects we are


interested in - one experiment, one idea– it is all about reducing
complexity
- Related: One change at a time!
- Random assignment is often the key to disentangling: chia tách
different factors.
- Random assignment allows for very simple statistical tests (which are
backed up with regressions, though)
- Develop hypotheses based on models or insights from the literature
- When is it important to incentivize subjects?
II. The benefits (and drawbacks?) of Experimental Economics

Some (potential) drawbacks

- Change the design and run sessions until you find what you want: The
importance of pre-registration
- Even worse: Outright data manipulation
- Are the amounts too small?
- External validity /vəˈlɪd.ə.ti/ hiệu lực?
- Differences within or across cultures?
III. A first simple game – and what we can learn from it

• Two individuals A and B are matched, but neither know nor see each other.
• Individuals A gets 10$ and can offer individual B any amount they want to.
• If B accepts, both get the respective amounts.
• If B rejects, both get nothing.

For instance:
• If A offers 2$ and B accepts, A goes home with 8$ and B with 2$.
• If A offers 2$ and B rejects, both go home with nothing.

• Which behavior and outcome would the simplest theory, which assumes that individuals care
only about money, predict?

• What is your opinion about what happens in experiments?

• How would you interpret the results and how can we disentangle different motivations?
IV. A second simple game – and what we can learn from it

• Three individuals A, B and C are matched, and each of them gets 2$.
• They can submit their money to a common pool. If so, each amount is doubled, and then they
share the pool equally.
• Hence, if each of them submits their money, the pool consists of 12$, and each of them goes
home with 4$.
• However, if e.g. A and B submit their money, while C does not, then the pool consists of 8$, and
C gets 2+ 8/3 > 4.

• Which behavior and outcome would the simplest theory that assumes that individual care only
about money predict?

• What is your opinion about what happens in experiments?

• What do you think of the real-life relevance of such an experiment?


V. Some mainly robust insights

• People are less selfish than the simplest theory suggests, but many are at least to some degree.
The behavior of other people matters a lot.

• People are influenced by norms, but many violate them when it pays. They also tend to
rationalize their behavior.

• Disentangling social preferences from norms is challenging.

• There is a large heterogeneity across people.

• Cultural and gender differences do not matter as much as one might think.
VI. A typical example for nudging

Imagine that the US is preparing for the outbreak of an unusual Asian disease which is expected to
kill 600 people. Two alternative programs to combat the disease have been proposed.

We will consider different hypothetical scientific estimates of the consequences of the programs
and ask you for your choice.
VI. A typical example for nudging

Choose between the following two alternatives

A:
Exactly 200 people will be saved.

B:
• With probability 2/3, all 600 will die
• With probability 1/3, no-one will die
VI. A typical example for nudging

Choose between the following two alternatives

A:
Exactly 400 people will be die.

B:
• With probability 1/3, all 600 will be saved
• With probability 2/3, no-one will be saved
VII. Another way of nudging: The relevance of irrelevant alternatives

• A harmless example: Wine prices in my favorite hotel in Italy.


• A less harmless example: First-degree, second degree murder, and manslaughter with law
students from top schools.
• Personal opinion on nudging: It is a double sided sword - any smart marketing strategy is based
on nudging. Personally, I believe more in salient information transmission...
VIII. Own research I: Self-selection to (cooperative) teams

Starting point from the literature: The famous trust game

 The famous trust game: A gets e.g. 5$. If A gives the money to B, it is
tripled. If A doesn’t send the money, the game ends. If A sends, B
decides which part of the 15$ they give back, Then, the game ends.

 Game is used to measure trust and trustworthiness. Famous examples:


Mentoring program for children with weak social background.
VIII. Own research I: Self-selection to (cooperative teams)

Our “innovation”
 B is first asked to decide how much money they return in case A sends
the money.
 Then, B can opt-out, and then gets 5$.

Why is this interesting?


 Self-selection: Who opts out?
 Importance from an applied perspective: Should managers prefer self-
selection or random assignment?
VIII. Own research I: Self-selection to (cooperative teams)

 Amount of B significantly negatively correlated with opt-out probability.


BUT: It is U-shaped
IX. Own research II: Insights on unethical behavior

Framed as “lying about the outcome of a lottery”

Selfish lies:
• Observability & “expected” reputation matters
• Groups lie more
• Hardly any gender differences. BUT: Observability matters far more for women

Altruistic and Pareto-improving lies:


• Groups lie less
• Men lie more often – norm based vs. utilitarian behaviour!

Deontologists or consequentialists: Experimental insights into an old question

You might also like