1

ECON191 (Spring 2011)

11 & 13.5.2011 (Tutorial 11)

Chapter 12 Monopolistic Competition and Oligopoly (Part 2)

Price competition with homogenous products: The Bertrand duopoly model
[Simultaneous move price setting duopoly]

Bertrand model: Oligopoly model in which firms produce a homogeneous good, each
firm treats the price of its competitors as fixed, and all firms decide
simultaneously what price to charge.

Each firm assumes its rivals will keep their price level constant when it changes its own
price.
The model
Residual demands for Firm 1:

) ( ) , (
1 2 1 1
p D p p D = if
2 1
p p <
) ( ) , (
1 2
1
2 1 1
p D p p D = if
2 1
p p =
0 ) , (
2 1 1
= p p D if
2 1
p p >

Residual demands for Firm 2:

) ( ) , (
2 2 1 2
p D p p D =

if
2 1
p p >

) ( ) , (
2 2
1
2 1 2
p D p p D = if
2 1
p p =

0 ) , (
2 1 2
= p p D if
2 1
p p <

The Bertrand equilibrium
Bertrand equilibrium: if the two firms have identical marginal cost equal to c, then the
Bertrand equilibrium price is equal to c.
p
1
= p
2
= c (stable equilibrium)

Suppose p
1
> p
2
> c
Firm 2 captures the whole market and earns profits
Firm 1 will match or undercut firm 2s price

p
2
> p
1
> c
Firm 1 captures the whole market and earns profits
Firm 2 will match or undercut firm 1s price

p
1
> p
2
> c
Firm 1 will match or undercut firm 2s price
The adjustment will continue until p
1
= p
2
= c

Any pair of unequal prices cannot be the Bertrand equilibrium
The situation p
1
= p
2
> c is not stable.
As long as there is profit for the one who can capture the whole market, both firms will
have incentive to undercut the other firms price.

The outcome in Bertrand equilibrium is the same as the perfectly competitive market.
No firm can earn profit.
2

Representing the Bertrand duopoly game by algebra
Market demand: ) (
2 1
q q a p + =

Cost functions: c(q
1
) = cq
1
, c(q
2
) = cq
2
Marginal cost = c for both firms

As p
B
= c in the equilibrium, ) (
2 1
q q a p
B
+ =
B
p a q q = +
2 1

c a q q = +
2 1


Since the two firms will share the market when their prices are equal, therefore,
2
2 1
c a
q q
B B

= =
0
2 1
= =
B B
t t

Price competition with differentiated products
Market shares are now determined not just by prices, but by differences in the design,
performance, and durability of each firms product.
In these markets, more likely to compete using price instead of quantity

Example:
Duopoly with fixed costs of $20 but zero variable costs. Firms face the same demand
curves:
Firm 1s demand:
2 1 1
2 12 p p q + =

Firm 2s demand:
1 2 2
2 12 p p q + =

Quantity that each firm can sell decreases when it raises its own price but increases when
its competitor charges a higher price

Firm 1s maximizing problem: 20 2 12
2 1
2
1 1 1
0
1
+ =
< s
p p p p Max
p
t
FOC: 0 4 12
2 1
= + p p

4
12
*
2 *
1
p
p
+
= (Firm 1s reaction function)

Firm 2s maximizing problem: 20 2 12
2 1
2
2 2 2
0
2
+ =
< s
p p p p Max
p
t
FOC: 0 4 12
1 2
= + p p

4
12
*
1 *
2
p
p
+
= (Firm 1s reaction function)

Solving the two reaction functions for the equilibrium, it yields,

4
*
2
*
1
= = p p


8
*
2
*
1
= = q q


16
*
2
*
1
= + q q


Profit for Firm 1 = Profit for Firm 2 = 12
3



Collusion in price competition
They both decide to charge the same price that maximized both of their profits
Firms will charge $6 and will be better off colluding since they will earn a profit of $16
(Try to verify!)

Representing price competition (differentiated products) by algebra
The two firms are facing the same demand curve: q
1
= a bp
1
+ cp
2
and q
2
= a bp
2
+ cp
1
Also assume MC = 0
Firm 1s maximizing problem:
2 1
2
1 1 1
0
1
p cp bp ap Max
p
+ =
< s
t
FOC: 0 2
2 1
= + cp bp a
b
cp a
p
2
*
2 *
1
+
= (Firm 1s reaction function)

Firm 2s maximizing problem:
2 1
2
2 2 2
0
2
p cp bp ap Max
p
+ =
< s
t
FOC: 0 2
1 2
= + cp bp a
b
cp a
p
2
*
1 *
2
+
= (Firm 1s reaction function)

Solving the two reaction functions for the equilibrium, it yields,
c b
a
p p

= =
2
*
2
*
1
,
c b
ab
q q

= =
2
*
2
*
1
,
c b
ab
q q

= +
2
2
*
2
*
1


Representing the Collusion model in price competition

Total profit:
2 1
2
1 1 2 2 1 1 2 1
p cp bp ap q p q p + = + =
+
t
2 1
2
2 2
p cp bp ap + +

The maximizing problems of the two firms are as follows:
Firm 1:
2 1
2
1 1 2 2 1 1 2 1
0
1
p cp bp ap q p q p Max
p
+ = + =
+
s s
t
2 1
2
2 2
p cp bp ap + +
FOC: 0 2
2 2 1
= + + cp cp bp a

Firm 2:
2 1
2
1 1 2 2 1 1 2 1
0
2
p cp bp ap q p q p Max
p
+ = + =
+
s s
t
2 1
2
2 2
p cp bp ap + +
FOC: 0 2
1 1 2
= + + cp cp bp a
Solving for the equilibrium, it yields
) ( 2
*
2
*
1
c b
a
p p

= = ,
2
*
2
*
1
a
q q = = , a q q = +
*
2
*
1

4

Competition VS Collusion: The Prisoners Dilemma
Dominate strategy: when one strategy is best for a player no matter what strategy the
other player uses.

We will explain these concepts with the classic example of Prisoners Dilemma.

Example: Prisoners Dilemma

The story:
Ann and Bob have been caught stealing a car. The police suspect that they have also robbed
the bank, a more serious crime. The police has no evidence for the robbery, and needs at least
one person to confess to get a conviction.
Ann and Bob are separated and each told:
(i) If each confesses, then each will get a 10 year sentence.
(ii) If one confesses, but the other denies, then he will get 2 year and his accomplice will
get 12 yrs.
(iii) If neither confesses, then each will get a 3 year sentence for auto theft.

We will represent the prisoners dilemma with normal form.

Bob
Confess Deny
Ann
Confess -10, -10 -2, -12
Deny -12, -2 -3, -3

Is there any dominated strategy for Ann and Bob?

Lets consider Ann,
If Ann expects Bob to confess, then Ann should confess. (10 > 12)
If Ann expects Bob to deny, then Ann should confess. (2 > 3)
Ann gets a higher payoff with confess than deny no matter what she expects Bob to do.
If Ann is rational, she will confess.
Formally, we say that deny is strictly dominated by confess.
Or we say that confess is a dominant strategy for Ann.

By the same way, we can find that confess is a dominant strategy and deny is dominated
strategy for Bob.

In the prisoners dilemma, if both players are rational, they will choose to use their
dominant strategies, Confess. The Nash equilibrium for this game is (Confess, Confess)
with a payoff of 10 for each player.

We find that the payoff for both players will be much better {3, 3} if they both choose
deny, however in the prisoners dilemma the NE is (Confess, Confess).

Individual rationality does not imply socially optimal outcome in the prisoners dilemma

5

Nash equilibrium
Nash equilibrium: a collection of strategies, one for each player, such that no player can
improve his situation by choosing a different strategy that is available
to him, given that all other players stay put.

In other words, the strategy (s
1
*
, s
2
*
) constitutes a NE if given player 1s strategy s
1
*
,
player 2 finds it optimal to choose s
2
*
, and given player 2s strategy s
2
*
, player 1 finds it
optimal to choose s
1
*
. (Best response)

When the NE is reached, there is no incentive for any player to deviate from it. No player
can benefit or increase his/her payoff by deviating from the NE.

For example, Ann would not deviate if given Bob uses his dominate strategy confess.
Deviation would lower her payoff to 12 given Bob stay puts.


Example: Cartel/ Collusion (Please refer to T10)

Firm B
Honor Agreement Break Agreement
Firm A
Honor Agreement 72, 72 54, 81
Break Agreement 81, 54 64, 64

In the duopoly model, price is lower than the monopoly price.
Incentive for the two firms to collude
They get into agreement to set a higher price, and produce less (monopoly output) in
order to have higher profit

In this duopoly game, Break Agreement is a dominant strategy for both firms.
NE: (Break Agreement, Break Agreement) with profit of 64 to each firm.

This game is a prisoner dilemma
They can both get a higher profit of 72 by following the cartel.

The cartel is not stable, they will have incentive to cheat and deviate from the agreement
Both firms will cheat and ends up in the Cournot equilibrium














6

Chapter 12: Problem 2
Consider two firms facing the demand curve P = 50 - 5Q, where Q = Q
1
+ Q
2
. The firms cost
functions are C
1
(Q
1
) = 20 + 10Q
1
and C
2
(Q
2
) = 10 + 12Q
2
.
a. Suppose both firms have entered the industry. What is the joint profit-maximizing level
of output? How much will each firm produce? How would your answer change if the
firms have not yet entered the industry?

If both firms enter the market, and they collude, they will set MR = MC1 to
determine the profit-maximizing output. (MC
1
< MC
2
)

MR = 50 - 10Q = 10 = MC
1
Q = 4, P = $30

The question now is how the firms will divide the total output of 4 among
themselves.

Since the two firms have different cost functions, it will not be optimal for them
to split the output evenly between them. The profit maximizing solution is for
firm 1 to produce all of the output so that
The profit for Firm 1 will be: t
1
= (30)(4) - (20 + (10)(4)) = $60
The profit for Firm 2 will be: t
2
= (30)(0) - (10 + (12)(0)) = -$10
Total industry profit will be: t
T
= t
1
+ t
2
= 60 - 10 = $50.

If they split the output evenly between them then total profit would be $46 ($20
for firm 1 and $26 for firm 2). If firm 2 preferred to earn a profit of $26 as
opposed to $25 ($50/2) then firm 1 could give $1 to firm 2 and it would still
have profit of $24, which is higher than the $20 it would earn if they split
output.

Note that if firm 2 supplied all the output then it would set marginal revenue
equal to its marginal cost or 12 and earn a profit of 62.2. In this case, firm 1
would earn a profit of 20, so that total industry profit would be 42.2.

If Firm 1 were the only entrant, its profits would be $60 and Firm 2s would be
0
If Firm 2 were the only entrant, then it would equate marginal revenue with its
marginal cost to determine its profit-maximizing quantity: 50 - 10Q
2
= 12, or
Q
2
= 3.8. P = 50 5*3.8 = $31
The profits for Firm 2 will be: t
2
= (31)(3.8) - (10 + (12)(3.8)) = $62.20

b. What is each firms equilibrium output and profit if they behave noncooperatively?
Use the Cournot model. Draw the firms reaction curves and show the equilibrium.

In the Cournot model, Firm 1 takes Firm 2s output as given and maximizes
profits. The profit function derived in 2.a becomes
t
1
= (50 - 5Q
1
- 5Q
2
)Q
1
- (20 + 10Q
1
), or

t = 40Q
1
5Q
1
2
5Q
1
Q
2
20.
Firm 1s reaction function:

ct
c
1
Q
= 40 10
1
Q - 5
2
Q =0, or
1
Q = 4 -
Q
2
2
|
\
|
.
.
7

Similarly, Firm 2s reaction function is

Q
2
= 3.8
Q
1
2
|
\
|
.
.
Solving for the Cournot equilibrium,

Q
1
= 4
1
2
|
\
|
.
3.8
Q
1
2
|
\
|
.
, or Q
1
= 2.8.and
Q
2
= 2.4.
P = 50 5(2.8+2.4) = $24.

The profits for Firms 1 and 2 are equal to
t
1
= (24)(2.8) - (20 + (10)(2.8)) = 19.20
t
2
= (24)(2.4) - (10 + (12)(2.4)) = 18.80

c. How much should Firm 1 be willing to pay to purchase Firm 2 if collusion is illegal
but the takeover is not?

In order to determine how much Firm 1 will be willing to pay to purchase Firm
2, we must compare Firm 1s profits in the monopoly situation versus those in
an oligopoly. The difference between the two will be what Firm 1 is willing to
pay for Firm 2. From part a, profit of firm 1 when it set marginal revenue equal
to its marginal cost was $60. This is what the firm would earn if it was a
monopolist. From part b, profit was $19.20 for firm 1. Firm 1 would therefore
be willing to pay up to $40.80 for firm 2.
8

Chapter 12: Problem 6
Suppose that two identical firms produce widgets and that they are the only firms in the
market. Their costs are given by C
1
= 60Q
1
and C
2
= 60Q
2
, where Q
1
is the output of Firm 1
and Q
2
the output of Firm 2. Price is determined by the following demand curve:
P = 300 Q where Q = Q
1
+ Q
2
.
a. Find the Cournot-Nash equilibrium. Calculate the profit of each firm at this
equilibrium.

Firm 1s profit function:

t
1
= 300Q
1
Q
1
2
Q
1
Q
2
60Q
1
= 240Q
1
Q
1
2
Q
1
Q
2
.


c
1 t
c
1
Q
= 240 2
1
Q
2
Q .= 0
Firm 1s reaction function: Q
1
= 120 - 0.5Q
2
.
Firm 2s reaction function: Q
2
= 120 - 0.5Q
1
.

Solving for the Cournot equilibrium, Q
1
= 120 - (0.5)(120 - 0.5Q
1
), or Q
1
= 80.
Q
2
= 80. P = 300 - 80 - 80 = $140.

t
1
= (140)(80) - (60)(80) = $6,400 and t
2
= (140)(80) - (60)(80) = $6,400.

b. Suppose the two firms form a cartel to maximize joint profits. How many widgets will
be produced? Calculate each firms profit.

Given the demand curve P = 300-Q MR=300-2Q
MR = 300 2Q = 60 = MC
Q = 120, P = 180
Each firm produces 60
Profit for each firm is: t = 180(60)-60(60)=$7,200.

c. Suppose Firm 1 were the only firm in the industry. How would the market output and
Firm 1s profit differ from that found in part (b) above?

MR = 300 2Q = 60 = MC
Q = 120, P = 180
Profit = $14,400.

d. Returning to the duopoly of part (b), suppose Firm 1 abides by the agreement, but Firm
2 cheats by increasing production. How many widgets will Firm 2 produce? What will
be each firms profits?

Assuming their agreement is to split the market equally, Firm 1 produces 60
widgets. Firm 2 cheats by producing its profit-maximizing level, given Q
1
= 60.
Given Q
1
= 60 into Firm 2s reaction function:

Q
2
= 120
60
2
= 90.
Total industry output, Q
T
, is equal to Q
1
plus Q
2
: Q
T
= 60 + 90 = 150.
P = 300 - 150 = $150.

t
1
= (150)(60) - (60)(60) = $5,400 and t
2
= (150)(90) - (60)(90) = $8,100.