ECO1110F • Nisreen Narker

2024 • Room 2.13 Leslie Commerce

In Unit 4, we will cover…
Section Content
1. Climate Change Conflict
2. Social Interactions
3. Nash Equilibrium
4. Dominant Strategies and the Prisoners’ Dilemma
4.7 Social Preferences: Altruism
4.6 Public Good Games and Cooperation
8. Repeated Interactions
9. Using Experiments to Study Behaviour
10. Cooperation, Negotiation, and Conflicts of Interest
11. The Ultimatum Game
12. Fair Farmers and Self-Interested Students

2
4.1 Climate Change Conflict

3
 Climate change
conflict
• Climate change is a major global issue
• The Stern Review on the Economics of Climate
Change (2006): early action is crucial
• This means taking drastic steps to cut greenhouse gas
emissions, reduce consumption of energy-intensive
goods, using alternative energy technologies
• We cannot continue with business as usual
• Businesses, governments, individuals will pursue
their own self-interests (maximise own profits) at
the expense of others and future generations
• Source: NASA Climate Change 4
Climate Change
• But there’s conflict about which environmental policies to
adopt, which countries should bear the costs
• All countries have different interests
• Countries are at different stages of development – some have
the means to contribute while others do not
• Some countries are contributing more to the issue of
climate change than others
• The issue of climate change is an example of a Social Dilemma

5
Climate change and economics?
We know that climate change is a real
issue facing society today, but what does
it have to do with economics?
The problem of climate change is what
we call a social dilemma.
A social dilemma occurs when we don’t
take account of how our own personal
decisions (e.g. consumption) impact on
other people’s wellbeing (or utility).

6
The tragedy of the commons
In 1968 Garrett Hardin published an article called The Tragedy of
the Commons.
In it, he argues that when a resource isn’t owned by anyone in
particular, it is easily overexploited and degraded when
individuals act selfishly. That is, unless you control access in
some way
These problems are prevalent everywhere:
• The atmosphere being polluted
• Farmers grazing their cows in a communal field
• Traffic jams
• Your roommate leaving their dirty dishes in the sink.
In all of these cases, someone may reap a benefit from some
communal action without personally doing the work – in other
words, we have a free-rider problem.

7
Resolving social dilemmas
We have been facing social dilemmas for years!
More than 2500 years ago Aesop told a tale called Belling the Cat – some mice
have to put a bell on a cat, so that all mice can hear the cat coming.
But for the mouse doing the belling, it is very dangerous – he has to engage in
self-sacrifice or have altruistic preferences (regard for others).
Modern day examples:
• Tribunas de las Aguas (Water court) in Spain to regulate farmer use of water
• The City of Cape Town in their attempt to mitigate the 2017/18 drought and
avert Day Zero.
• How the world encouraged people to “flatten the curve” during the COVID-19
pandemic.
8
The context of Unit 4
In Unit 3 you learnt about making decisions to try and best serve your
own interests.
But are all our decisions made in a vacuum?
Unit 4 introduces the notion that decisions are often made in
conjunction with other people.
We ask what happens to the individuals when they make decisions
interactively; and also, what happens to the society they live in?
Can people work together to avert social dilemmas? How do their
preferences shape their decisions and actions?
Unit 4 introduces the concept of game theory to model social and
strategic interactions, which are key to analyzing and solving social
dilemmas.
9
4.2 Social
Interactions

10
Which side of the road should you drive on?
Nowadays this is a government rule.
But what would happen if when we let
people make their own decisions?
The key concept underlying the study of
game theory is that of strategic
interactions.
Strategic interactions occur when there are
two or more people, and the actions that
individuals take affect one another, and
they are aware of these effects.
We model strategic interactions as games.
4.1 Social Interactions: Game Theory 11
 The basics of game theory: glossary
To start studying game theory, we need to understand some basics first.
There are four terms we need to understand to know what we mean when we
talk about game theory:
• When people are engaged in a social interaction, and are aware of the ways
in which their actions affect others, and vice versa, we call this a strategic
interaction
• A strategy is defined as a complete plan of action which describes the
actions that an individual will take throughout the course of the interaction
• Models of strategic interactions are called games.
• The study of these strategic interactions is called game theory and is widely
used in economics, but also in many other disciplines, including biology,
mathematics and other social sciences.
4.1 Social Interactions: Game Theory 12
 What makes a game?

Games are made up of some key components:

• The players: Who is interacting with whom
• The (feasible) strategies: What actions are
open to the players
• The information: What each player knows
when they are making their decisions
• The payoffs: A representation of what the
outcomes will be for each of the possible
combinations of actions.
Once you have all of these components, you
can start putting games together for analysis.
4.1 Social Interactions: Game Theory 13
 Information and types of games
Broadly speaking, games can be divided into two types:

Simultaneous games are ones where players have exactly the

same amount of information when they make their move –
e.g. rock-paper-scissors; public goods games

NB! You don’t have to move at the same time for it to be a

simultaneous move game; just with the same information!

Sequential move games are ones where one player moves, and
then the next player observes this, and uses this information to
inform their next move – e.g. chess; ultimatum games
4.1 Social Interactions: Game Theory 14
 A basic game
Let’s consider an example:
Meet Amandla and Betty. They face a problem: should
they grow maize or sugar cane?
We assume that Amandla and Betty both know how to
grow both crops, but that they can only grow one at a
time.
Also, Amandla’s land is better-suited to growing sugar
cane, and Betty’s land is better-suited to maize.
The farmers need to decide on the division of labour
– i.e. who is going to grow which crop?
They have to decide this independently, meaning they
cannot meet and discuss their plans.
There are no other people involved of affected by
their decision. 15
 A basic game (continued)
We want to start representing this interaction
somehow.
We need to identify the players.
Betty
• Amandla and Betty.
What are their strategies?
Maize Sugar Cane
• Each of them has a choice – they can either
plant maize or they can plant sugar cane.
Maize
What information do players have when

Amandla
making their decisions?
• They make their moves simultaneously. Sugar Cane

Because this is simultaneous-move, we depict

the game as a table (also called normal form).
16
 A basic game (continued)
Notice that there are 4 different
Betty
things that can happen
represented by the 4 different (Maize, Maize) (Maize, Sugar Cane)
cells in the table. Maize Sugar Cane

Our notation for identifying cells

is Maize

Amandla
(Row strategy, Column strategy)
Sugar Cane

(Sugar Cane, Sugar Cane)

(Sugar Cane, Maize)
17
 Payoffs in a game
We condense each possible scenario into a
numerical outcome for each player – a
payoff.
Payoffs can be ordinal (only the order of
the payoffs matters, not the actual values),
or cardinal (the actual value of the payoff
matters as well)
Payoffs are like utility: they represent a
measure of net benefit received.
Payoffs can represent any type of
preferences you like. We just want higher
payoffs to represent better outcomes for
players. 18
Payoffs continued
• Most games usually require calculations and comparisons of, for
example, gains and losses, costs and benefits, rankings etc.
• Therefore, we use “payoffs”
• This gives players a complete numerical scale with which to compare
different outcomes of the game, based on the combination of choices
of strategies
• The number associated with each possible outcome is the player’s
• payoff for that outcome
• Higher values typically indicate higher payoffs (context is important)
• However, this is not true for the prisoner’s dilemma (you would want a smaller
payoff because payoffs indicate jail time). 19
 Betty
NOTE: The notation I am Maize Sugar Cane
using differs from the
Payoffs in our basic game textbook! The textbook

Maize
notation is non-standard,
and so I am showing you

Amandla
Let’s assume Amandla and Betty only care international standard

Sugar Cane
about profits. notation which you will see
throughout your studies!
We represent payoffs as numbers in the
cells of our game table. They are always Betty
written with the row player’s payoff first,
followed by the column player’s payoff.
Maize Sugar Cane
In our game, this means the first payoff in a
cell tells us Amandla’s payoff, and the
second payoff tells us Betty’s payoff. Maize 1,3 2,2

Amandla
This means that the (ordinal) payoffs for
this game may look something like this … Sugar Cane 4,4 3,1

20
4.3 Nash Equilibrium

21
 Solving a game: The Nash Equilibrium

What outcome do we expect from a game?

The answer to this is called the Nash Equilibrium
of a game.
Technically, a Nash Equilibrium is the combination
of strategies in a game such that neither player John Nash (1928-2015)
has an incentive to change their strategies, given
the strategy choice of their opponent.
Simply put, it is a mutual best response.
In equilibrium, neither player would want to
change their decision after seeing what the other
player chose 22
 Solving for the Nash Equilibrium Betty’s payoff if she
We use the best response method of grows Sugar Cane while
Amandla grows Maize
solving a game. Betty’s payoff if she
grows Maize while
To do this, we consider a situation where Amandla grows Maize Betty
one player takes a specific action, and then
see what the best response of the other
Maize Sugar Cane
player is.
In other words: Choose an action for one
Maize 1,3 2,2
player, and ask “What is the other player’s

Amandla
best response, given this action?”
For example, we assume Amandla chooses Sugar Cane 4,4 3,1

to grow Maize.
Now, what is Betty’s best response?
23
 Solving for the Nash Equilibrium
We now iterate the same logic over and
over again, each time holding one player’s
choice constant and asking what the other
Betty
player would do:
• When Amandla plays Sugar Cane, what
Maize Sugar Cane
will Betty do?
• When Betty plays Maize, what will
Maize 1,3 2,2
Amandla do?

Amandla
• When Betty plays Sugar Cane, what will
Amandla do? Sugar Cane 4,4 3,1

24
Solving for the Nash Equilibrium
Remember that the Nash Equilibrium
was defined as a mutual best
response. Betty

The Nash Equilibrium must thus be

the cell where we circled both Maize Sugar Cane

payoffs!
NE = (Sugar Cane, Maize) Maize 1,3 2,2

Amandla
We also define the Nash Equilibrium
Sugar Cane 4,4 3,1
Outcome (NEO), as the payoffs
associated with the Nash Equilibrium.
In other words, NEO = (4, 4)
25
4.4 Dominant Strategies
and the Prisoners’
Dilemma

26
 If both players have a strictly
Dominant strategies dominant strategy, then the Nash
Equilibrium of the game is also
If a given strategy is always a player’s best known as a dominant strategy
response, then it is called a strictly equilibrium.
dominant strategy because for every Betty

action taken by the opponent, the

dominant strategy’s payoffs dominate all Maize Sugar Cane
other strategies’ payoffs.
In the example game: Maize 1,3 2,2

Amandla
• Betty has a strictly dominant strategy of
Maize. Sugar Cane 4,4 3,1
• Amandla has a strictly dominant
strategy of Sugar Cane.
27
Dominant vs Dominated strategies
• A dominant strategy is one where, regardless of the chosen
strategy of the other player, one’s best response strategy is
constant

• A dominated strategy is one where, regardless of the belief about the

other player’s strategy, it is never a best response

28
Socially optimal cells
We are sometimes interested in whether
the outcome of a game is good for society
overall. Betty

To determine this, we need to get the

total payoff for society. Maize Sugar Cane

In our games, the players are our society, 4 4

so we add the payoffs in each cell Maize 1,3 2,2

Amandla
together to get society’s payoff.
8 4
The cell with the highest payoff to society Sugar Cane 4,4 3,1
is then the socially optimal cell.
In this case, the socially optimal cell is
(Sugar Cane, Maize)
29
The invisible hand game
Adam Smith suggested that markets operate to
bring about the best outcomes in society as if
directed by “an invisible hand”. Betty

Every individual necessarily labours to render the annual

revenue of the society as great as he can. He generally, Maize Sugar Cane
indeed, neither intends to promote the public interest, nor
knows how much he is promoting it... he intends only his
own security; and by directing that industry in such a
manner as its produce may be of the greatest value, he Maize 1,3 2,2

Amandla
intends only his own gain, and he is in this, as in many other
cases, led by an invisible hand to promote an end which was
no part of his intention.
Adam Smith, The Wealth of Nations (1776)
Sugar Cane 4,4 3,1

So, if a game has 1 Nash Equilibrium and this

Nash Equilibrium is socially optimal, we call it
an invisible hand game.
30
An example
Solve for the Nash Equilibrium and Nash Equilibrium outcome of the
following interaction:

Students

Study Don’t Study Follow-up question: Do

either of the players have
a dominant strategy? If,
Hard Test 2,2 1,0 yes, what is it?
Rob

Easy Test 6,4 4,2

31
 Equal payoffs in a game?
Imagine that there is a shortage of sugar in
South Africa, and so growing sugar cane
becomes more profitable for both players. Betty
Specifically, imagine that the payoff from
growing sugar cane increases by 1.
Maize Sugar Cane
How would this change the way we solve for
the NE?
Maize 1,3 2,3
Now Betty is indifferent between her choices

Amandla
when Amandla grows Maize.
This means that both choices are best Sugar Cane 5,4 4,2
responses, and we should circle both of
Betty’s payoffs to indicate this.
The rest of our analysis continues as normal.
32
 Is there always 1 Nash Equilibrium?
So far, we have been examining games where there
has been one Nash Equilibrium.
Is this always going to be the case?
However, there are instances, hat give rise to multiple
equilibria
Between the different equilibria, there is no
guarantee that we end up in the “better” Nash
equilibrium, and society may get “stuck” in an
equilibrium where everyone is actually worse off
Where we end up depends on a number of
things: chance, history, social norms, and more…

33
 The assurance game:
• There are two Nash Equilibria
Multiple equilibria? • Both players play the same strategy
• Both players strictly prefer the same
Some games, like the driving game, have Nash Equilibrium over the other.
more than one equilibrium. This ends up
giving us a game table that looks like this, Driver 2
for example:
When you use best response analysis, you Left Right
find the Nash Equilibria:
• (Left, Left) with NEO (3, 3) Left 3,3 0,0

Driver 1
• (Right, Right) with NEO (2, 2)
Right 0,0 2,2

4.13 Social Interactions: Conflicts in the Choice Among Nash Equilibria 34

 The battle of the sexes game:
Multiple equilibria? • There are two Nash Equilibria
• Both players play the same strategy
Imagine you and your bestie are choosing • Each player strictly prefers a
between 2 activities. You want to go to different Nash Equilibrium.
movies, but you bestie wants to go to dinner.
The game could look something like: Bestie
Notice that the Nash Equilibria are:
• (Movies, Movies) with NEO (4, 2) Movies Dinner

• (Dinner, Dinner) with NEO (2, 4)

So, we see that there are two Nash Equilibria. Movies 4,2 0,0

Also note that you prefer the equilibrium at

You
the movies, while your bestie prefers the Dinner 0,0 2,4
equilibrium where you go to dinner.
35
Self-interest and undesirable outcomes

Things are not always as easy as the

invisible hand game - we do sometimes
end up in social dilemmas with
undesirable outcomes.
Betty
Let’s assume that Amandla and Betty
now have to choose a type of pest
control: a poisonous chemical, IPC Terminator

Terminator, or Integrated Pest Control

(IPC). IPC 3,3 1,4

Amandla
The game table could look something
like the one on the right. Terminator 4,1 2,2

36
 Self-interest and undesirable outcomes
Nash Equilibrium = (Terminator,
Terminator).
But note: Betty
Players are both doing worse than if
they had just opted for IPC instead. IPC Terminator
This game is an example of what we call
a Prisoner’s Dilemma. IPC 3,3 1,4

Amandla
Terminator 4,1 2,2

37
Prisoners’ Dilemma

A Prisoner’s Dilemma is one which

exhibits the following properties:
• Both players have a dominant
strategy
• Both players would do better in
the outcome where they play
their non-dominant strategy
(also called the dominated Notice that if you confess first, you’ll get a
strategy) than they do in the good deal. But if the other prisoner
confesses before you, you’ll receive a
Nash Equilibrium
tough sentence.

38
 The prisoner’s dilemma - context
The prisoner’s dilemma is a well-known set-up in
game theory where there are two robbers who
have been apprehended for a particular crime.
The robbers are in separate interrogation rooms,
and they are told that the police have enough
evidence to convict them for a smaller crime they
committed previously and lock them up for 1 year,
but not enough to convict them for the bigger Prisoner 2
crime now.
However, if the one robber turns state’s witness
against the other, then the one who turned will get Deny Confess
rewarded with no jail time, while the implicated
one will get 10 years.
If, on the other hand, they both implicate one Deny 1 year , 1 year 10 years, 0 years

Prisoner 1
another, the police have enough evidence to lock
both criminals away for 7 years.
Confess 0 years, 10 years 7 years , 7 years

39
 The prisoner’s dilemma in context Sometimes called the
defecting strategy
Payoffs show years in prison (which people Sometimes called the
don’t like), so we can think of these payoffs cooperative strategy
as being negative.
In other words, 1 year in prison is equivalent Prisoner 2
to a payoff of -1.
Nash Equilibrium is (Confess, Confess) Deny Confess
But why not just cooperate with your fellow
prisoner and get out quickly? Deny -1 , -1 -10 , 0

Prisoner 1
Because there’s an individual incentive to
free-ride.
Confess 0 , -10 -7 , -7
This leads to the worst outcome for
society!
[Link]
Ab2BEw 40
Example

41
 Climate change as a prisoner’s dilemma
Imagine people in society have two choices: “Go
Green” or continue with “Business as Usual”.
Let’s say for each person in society that chooses
to go green, everybody get 5 units of utility each. Me

However, it costs the person 6 units to actually

switch to a green lifestyle. Go Green
Business as
Usual
If everyone sticks to business as usual, you each
get 2 units of utility each.
Go Green 4,4 -1 , 5
This would give you a game table like the one on
the right, which is a prisoner’s dilemma (see if

You
you can show it)! Business as
5 , -1 2,2
Usual
The Nash Equilibrium is (Business as Usual,
Business as Usual) – maybe explaining why we
have a climate change crisis to start with.
4.3 The Prisoners’ Dilemma 42
4.7 Social preferences:
Altruism
43