CHAPTER SEVEN

OLIGOPOLY MARKET STRUCTURE

Learning objective
At the end of the chapter, you are expected to:
- Differentiate oligopoly from other forms of market
- Identify how interdependence between firms affect their optimal decision
- Explain the difference between different duopoly models
- Explain how equilibrium price and output determined in collusive and non-collusive
oligopoly market structure.
7.1 Introduction
Dear students, we have studied perfect competition and monopoly market structure. Both of
them in common assume that the decision taken by any particular firm has no diffused effect on
the environment in which other firms operate. Individual firm therefore safely neglect the
reaction or behavior of its competitor in making its decision. Especially, the issue of strategic
interaction between firms is irrelevant in perfect competition because the prevailing market price
convey all the external information that was relevant to the firm.
Is the above feature of firms can be applied to firms in oligopoly market structure? No, because
when small number of large firms dominating a particular industry producing homogenous or
differentiated product, any change in a firm’s price or output influence the sales and profit of its
competitors. As a result, each firm formulates its policies and strategies with an eye on its
competitor decision. A market model, which considers such interdependence of few firms when
making optimal decision, is called oligopoly market model/structure.
Oligopoly
As it is mentioned above, oligopoly is a market structure dominated by few sellers of
homogenous or differentiated product. As a result, the action of each firm affects the other firm’s
decision in the industry. A beer industry in Ethiopia is a good example of such type of industry.
Each of the major beer producers takes in to account the reaction of others when they formulate
their price and output policies. Bedel or Dashen in this case know that its own action will have
significant impact on the rest of the beer producers. Therefore, Bedel or any other producer
considers the possible reaction of its competitor in deciding prices, degree of product

1
differentiation to be introduced, the level of advertisement to be undertaken and the amount of
service to be provided and so on.
Automobile industry, which produces different cars and aerospace industry producing different
airplanes are also an example of oligopoly industry. Note also that, all oligopolists are not
necessary large firms as given in the above example. Two grocery stores that exist in isolated
community can be oligopolists given that their decisions are interdependent and they are the sole
supplier of a specific product. This implies that the distinguishing feature of oligopoly is the
interdependence of decision making by rivalry firms in an industry. Such interdependence
between firms is the natural result of the existence of few numbers of firms in an industry.
In summary, oligopoly market structure is a market structure which contain firms
characterized by the following features
 Few number of firms in a given industry
 Interdependence of firms in decision making
 Firms produce homogenous or differentiated product
 Firms have some power to set price

Causes of Oligopoly
There are many cause of oligopoly market. Some of them are
1. Economies of scale: low costs cannot achieved in some industries unless a few firms are
producing output that account for substantial percentages of the total market
demand .That means the average cost of production reach minimum only when the
output produced in large amount by a few firms. As result, the number of firms in such
type of industry should be reduced in order to make use of the advantage of economies of
scale in production. Economies of scale in sales promotion and advertising may also
promote oligopoly.
2. Barriers to entry: - there are varieties of barrier that did not allow the entry of some
firms in to the industry. This barrier may be technological, skill, cost, size of the market
in relation to economies of scale , patent right and different activities of government such
as licensing and marketing quota.
3. Collusion (merger of small firms): Small firms collide to get market power and
overcome their competitor’s pressure. If they gain market power, firms set higher price

2
 and restricts output supply that maximizes their profit. Such firms develop to oligopoly
while other removed from the market.
Dear students, from our pervious section discussion you have some idea about the general
features of oligopoly market structure and its cause. Now let us identify different types of
oligopoly model and see how equilibrium level of output and prices are determined in each
model given their underlying assumption. For simplicity, we consider only the case of two
firms that is termed as duopoly. In addition, we limit our self to the study of firms producing
homogenous product. This will allow us to avoid the problem related to analysis product
differentiation and focus on the study of strategic interaction between firms.
Based on the reaction pattern of firms in the industry, it is possible to classify oligopoly in to two
groups. These are collusive oligopoly and non-collusive oligopoly. Non-collusive oligopoly is a
condition in which firms operate independently to determine the optimal level of price and
output. That is firms in the industry will not go in to contractual agreement to cooperate in
making optimal decision. Under such cases, negotiation and enforcement of binding agreement is
not possible even though each firms make some expectation (assumption) about the reaction of
its rivalry in response to its action or observe the decision of its rivalry while setting profit
maximizing level of output and price. For example, if a firm in non-collusive oligopoly wants to
increase output or price to maximize its profit, it has to assume something about the possible
reaction of its rivalry and the effect of the reaction on profit maximization process.
Dear students do you think that the assumption firms make about their rival firm is the same in
all different type of non-collusive oligopoly model? The answer to this question will obtain after
discussing the different non-collusive oligopoly in subsequent subsections. Here are some of
non-collusive oligopoly model that we will consider right now after a moment.
1. Kinked demand model
2. Cournot duopoly model
3. Bertrand duopoly model
4. Stackelberg duopoly model
In collusive oligopoly however, firms get together to make open and formal agreement in setting
prices and output so as to maximizes the total profit of the industry as in the case of cartel. It can
be also implicit cooperation of firms in the industry without actually making explicitly
agreement with one another as in the case of price leader.

3
 7.2 Non collusive oligopoly
7.2.1 The kinked demand curve model
Dear students as you know from our previous discussions, Price in perfect competition,
monopoly and monopolistic competition markets adjusts rapidly to changing cost or demand
conditions. Is the adjustment of price possible to change in cost and demand conditions in the
case of oligopoly market like in other market? For instance, is the price of beer or soft drink
changes frequently as their demand and cost condition changes? You might answer these
questions based up on the assumption how oligopolist react to each other’s decisions according
to kinked demand curve model.
The kinked demand curve model, developed by Paul Sweezy in 1939, explains why prices are
rigid in some oligopoly market. The model assume that oligopolist often have strong desire to
keep stable price. Even under the condition when cost and demand changes, firms are reluctant
to change their prices. If costs fall or market demand declines, they fear that the lower prices
send the wrong message to their competitors and initiate price war among them. If costs and
demand rises, they do not increase price because they are afraid that their competitors may not
raise their price.
According to this model therefore, demand curve facing each firm in oligopoly market is kinked
at prevailing market price (Chamberlin’s intersection of individual and market demand)
reflecting the following behavioral pattern of oligopolists. Rivalry firms expected to follow
decrease in price but ignore price increase. That means if a firm increases its price, due to
increase cost of production or increase demand for its product, it would loss most of its customer
and cause total revenue to decrease. This is because other firms in the industry would not follow
the increase in price. As result, given that the product produced in the industry is homogenous,
consumer preference shifts from the other hand an oligopolist could not increase its market share
by lower intend to maintain prices constant even under the condition where their demand and
cost changes as shown in figure 7.1. The demand curve is much more elastic above point E
(kinked point) than below the kink on the assumption that the competitors will not follow price
increase but quickly follow price decrease.

4
 D
MC1
MC2
d
Price MC3

E
Po

B
D

MR
Q0
Quantity

Figure 7.1: The kinked demand curve

As shown in the above figure 7.1, an oligopolist firm faces two demand curves for different
ranges of prices. Above P0 the relevant demand curve for the firm is dE. Because if the firm
increases its price, it would lose some of its customer to firm that maintained their previous
price. The firm will then face a demand curve given by dE, which is very elastic. On the other
hand, if the firm decreases its price below P o form the intention of increasing their market share,

5
they will not able to increase their market share, since other firms also decrease their price in
order to keep up their customers. Therefore, below Po the relevant demand curve of the firm is
ED. This implies the demand curve facing oligopolies is not straight line, rather kinked at a
certain price.

Well students, we have seen that oligopolist is reluctant to price change in response to change in
their cost and demand condition. Thus, their demand curve is kinked at the intersection point of
market and individual demand curve to reflect the rigidity of prices. Now let us see how their
marginal revenue curve derived and how they undertake optimal decision. Normally, marginal
revenue curve derived from demand curve. So the marginal revenue curve associated to kinked
demand curve is discontinues at the level of output corresponds to the kinked point. It has two
segments as indicated in figure 7.1, dA and BMR. Segment dA corresponds to the upper part of
demand curve, while the segment BMR corresponds to the lower part of demand curve and the
kink point on demand curve corresponds to the discontinuous portion of marginal revenue curve.

The point of kink defines equilibrium of the firm since at any point to the left of the kink, MC is
below MR, while to the right of the kink; MC is larger than MR. Thus profit of the firm
maximized at the point of the kink. However, this equilibrium is not necessary defined by the
intersection of MC and MR curve. So long as MC passes through segment AB, the firm
maximizes its profit by producing Qo and charging po level of price. This level of price and
output is compatible with wide range of cost.
To sum up, the kinked demand curve model give us some insight about why price and output
will not change despite changes in cost and demand in oligopoly market structure. The only case
where a rise in cost results in increase in price is when the rise in cost equally affects all firms in
the industry. It is important to note here that a kinked demand curve model does not explain how
equilibrium price and output determined like other models rather, explain why price once set
remain fixed. Lastly some economists have been very critical of the model’s assumption .It
would certainly be a mistake to conclude that oligopolies in general are unresponsive to change
in cost as well as unresponsive to change in demand in the world of technological advancement
and change in taste of consumer preference. Perhaps we observe the opposite in many oligopoly
industries.

6
7.2.2 Cournot model
A French mathematician, Augustine cournot in 1938 using two firm producing homogenous
products, illustrated cournot model for the first time. He illustrates the model assuming two firms
having identical cost facing linear market demand. Given these assumptions, each firms known
that the market price will depend on the total output of both firms. Thus to maximize their profit
each of the cournot duopolist simultaneously decides, how much to produce by taking its rivals’
output constant at existing level regardless of what output it decides to produce. Thus, each firm
takes the other firms output level as given and chooses its own output level to maximize profit.
The level of output that it chooses will, of course depend on how much it thinks its rival will

produce. In other words, each firm recognizes that its own decision about output will affect its
revenue through affecting market price but any one firm has output decisions do not affect those

of any other firm. That is, each firm recognizes that but assume that , for

.
Dear students now let us see cournot simplified illustration of the model. He started his
illustration by assuming that there are two firms (firm-A and firm-B), each producing mineral
water at zero cost and face linear demand curve DD as shown in figure 7.2. Each firm also acts
on the naïve assumption that its competitor will not change its output when deciding its profit
maximizing level of output.
Assume that firm A is the first to start producing and selling mineral water assuming that firm B
produce nothing. Firm-A therefore thinks that its effective demand curve is the market demand
because the firm thinks that it will be the sole producer of mineral water. So to find the profit
maximizing level of output and price, we use the marginal principle (MR=MC). Since it is
assumed that cost of production equals to zero MC is also zero. Therefore profit maximization
condition of the firm reduced to MR= 0. This point corresponds to Q 1 level of output that is half
of the total market demand.

Price
D

C
P
E 7
P1
P2
Now firm-B assume that firm-A will keep its output fixed at Q 1 and define CD’ as its relevant
demand curve to determine its profit maximizing level of output. Applying the same marginal
rule to the perceived demand section, firm-B produce half of the market which is not supplied by
firm-A.
7.2.2.1 Reaction curve approach
Under this approach, we try to see how cournot duopolist choice output level that maximizes
their profit after relaxing the identical cost and demand assumption. Here we use the reaction
curves of the two firms to determine the optimal level of output of the duopolist under the basic
cournot behavioral assumption. So before using the reaction curves as a tool for determination of
the optimal choose of the firms, it is helpful first if we understand what a reaction curve is and
how one can derive a reaction curve.

QB

Reaction curve of A

Isoprofit curve of the firm-A

QA
Figure 7.3.: (a): Isoprofit of firm-A

A reaction curve is a graphic representation of reaction function. A reaction function is a

function that shows the functional relationship between the optimal output level of one firm
and its beliefs about other firm optimal choice. For instance, reaction function of firm-A

8
 shows, how firm-A will react in making its output decision to its perception of how much it
thinks firm-B will produce and sell. That is for firm-A to decide to produce let say Y 1 level of
output, it has to forecast first the amount of output firm-B produce such as Y 2. A function
which shows the amount of Y1 as a function of Y2 gives the reaction of firm-A. Similarly,
reaction function of firm-B defined in the same fashion.
Graphically the reaction curve of cournot duopolists derived from their respective isoprofit
curve. A curve contains a locus of point that yields the same level of profit for the firm.
Isoprofit of firm-A is the locus of points defined by different level of output of firm-A and its
rival B which yields the same level of profit to firm-A. Similarly, isoprofit curve of firm-B is
the locus of point defined by different level of output of firm-B and its rival firm-A which
yields the same level of profit to firm B. The isoprofit curves of the two firms given in figure
7.3 (a) & (b). From the definition, it should be clear that the isoprofit curves are a type of
indifference curve.

QB

Isoprofit curve of the firm-A

Reaction curve of B

QA
Figure 7.3.: (b): isoprofit of firm- B
Properties of isoprofit curve
1. Isoprofit curves for substitute commodities are concave to the axis along which we measure
the output of the rival as indicated in figure 7.3.a and b. Therefore, isoprofit of firm-A is concave
to axis that measures the output of A and isoprofit curve of firm-B is concave to the quantity axis
of firm-B. The shape of isoprofit curve of firm-A shows how A react to B’s output decision so as

9
 to retain a given level of profit. For example, consider the isoprofit curve of firm-A, and firm-

B, as in the figure- 7.4.below. It is concave to A’s quantity axis.

Let us suppose that B decide to produces B1 amount of output, firm-A will realizes the same
level of profit by producing Ah or Ag amount of output. Assume firm-A decide to produce A g
amount of output. If firm-B increases its output to B 2, firm-A must decreases its output to A f to
maintain the same level of profit. If firm-A continue producing A g amount while firm-B increase
its output level to B2 the total amount of output supplied to the market increase and result in
decrease in prices and hence profit of the firms. If B further increases its output to point B 3, firm-
A should decrease its output to Ae, otherwise it result in reduction of price and then
profit .Beyond point E, increase in output by firm-B result in fall in market price and increase
cost of production which result in decrease in profit of the firm. This is indicated by out ward

shift in isoprofit function to ,representing lower level of profit .This implies the isoprofit line
which is found far from quantity axis represent lower level of profit while the isoprofit curve
found nearer to quantity axis represent higher level of profit

QB

B3
k f B4
B2

h g B3 ¶A2
B1
¶A1
¶A1

Ah Ak Ae Af Ag QA
2.

Figure 7.4.: isoprofit curves

2. For firm-A the highest points of successive isoprofit curves lies to the left of each as we
go further away from A’s quantity axis and lies to the right of each other as we go

10
 towards to quantity axis. On the other hand the highest point of the isoprofit curves of
firm-B lies to the right of each other as we move further away from the quantity axis) Q A
and lies to the left as we move toward the quantity axis. Joining these respective highest
points of isoprofit curve of the two firms gives their reaction curve. Hence, the reaction
curve of firm-A is a locus of highest point of Isoprofit that firm-A can attain, given the
level of output of rival B. The reaction curve of firm-B also represent the locus of highest
isoprofit point that firm-B can attain as indicated in figure 7.3. (a) and (b).
Dear students we have seen what a reaction curve is and how to derive it form isoprofit curve.
Now let us see how reaction curves used to determine the equilibrium level of output for cournot
duopolist. Each firms reaction curve tell us how much to produce given the output of its
competitor. Therefore, at equilibrium the belief that each firm make about other’s firm output
level during setting their output is equal to the actual amount of output produced by its rival. This
point corresponds to the intersection of two reaction curves, point e as shown in figure 7.4.

QB
QB ¶B3 ¶B41
¶B2

¶B1

¶A4
e
¶A3
¶A2
¶A1

QA

Figure 7.4.: cournot’s duopoly model equilibrium

Note that at point e, each firm maximizes its own profit but the industry’s (joint profit) profit is
not maximized. This can easily seen by referring to the contract curve. It is the curve formed by
joining the tangency point between the two firms’ isoprofit curves. Points on this curve

11
represents optimal point (maximize industry’s profit). Points off contract curve however
represent lower level profit to one firm or both profits. Therefore, since point e found off the
contract curve it cannot maximizes industry’s profit and represents suboptimal level of output.
Such suboptimal resource allocation is the result of naïve cournot behavioral assumption of firms
cannot learn from their experience. Had they learn from experience, they would show a different
strategic behavior. For instance, they would have come together and agree to share the market in
such a way that would raise the profit of the two firms or the industry as the whole.

7.2.2.2 Mathematical version of cournot duopoly model

So far, we have seen graphical presentation of cournot model. Now let us see mathematical
treatment of the model. Determining equilibrium output level of the cournot duopolists
mathematically given their cost and demand function. Here we begin from the assumption of the
model, each firm decide how much to produce by treating its competitor level of output fixed at
existing level. For example, if the units of output that firm-2 produce currently is Y 2, firm 1
assume firm 2 continue producing Y 2 amount of output while deciding to produce Y 1 unit of
output. So the total output supplied to the market equals Y=Y 2 +Y1 and the market price of the
product, P(y) = P (Y1+Y2). Given the cost function of firm 1, C 1 (Y1) and firm 2, C2 (Y2), we can
define the profit maximization problem of the firms as follows.
For firm 1: Max P (Y1+Y2) Y1-C1 (Y1)
For firm 2: Max P (Y1+Y2) Y1- C2 (Y2)

Following the usually optimization procedure, take the first order condition of the profit function
of each firm with respect to choice variable (their output level). The result of first order condition
gives the reaction function of each firm. Then solve the two-reaction curves to gather
simultaneously to get the equilibrium output level of the two firms.
Example: Suppose that two firms producing identical product and supply to a market with a

linear demand function, . Where and represent unit of output

produce by firms-1 and represent unit of output produce by firm two. The cost of production
2
of firm 1 and firm-2 is given by = y 1 and 4 y 2 respectively .Find the output level

that the two firms should produce to maximizes their profit under cournot behavioral assumption.

12
 Solution: starts solving the problem by defining the profit maximization problem of the two
firms.

Firm-1

Π 1=[ 60−2 ( Y 1+Y 2 ) ] Y 1−Y 2=60Y 1−2Y 12−2Y 1 Yalignl ¿ 2 ¿ ¿−Y 21 ¿

1
Π 1 =60 Y 1−3Y 2 −2Y 1 Y 2
1

F.O.C:

∂
( 60 y 1−3 y12−2 y 1 y 2 ) =0
∂ y1

60−6 y 1 2 y 2=0
1
y 1 =10− y 2 . . .. .. . .. .. ( 1 )
3
Equation (1) represents the reaction curve of firm –1

Firm 2 Max Π 2 =max [ p ( y 1 + y 2 ) ] y 2 −c 2 ( y 2 )

Π 2 =[ 60−2 ( y 1 + y 2 ) ] y 2 −4 y 2

Π 2 =60 y 2 −2 y 1 y 2 −2 y 2 −4 y 2
2

Π 2 =56 y 2 −2 y 1 y 2 −2 y
22

∂ Π2
=56−24 1−4 y 2 =0
F.O.C: ∂ y2

y 2 =14−0 .5 y 1 . .. . .. .. .. . .. .. . .. . ( 2 )
Equation (2) represents the reaction function of firm 2. Solve simultaneously the two reaction
curves together to get the cournot equilibrium level of output.
1
y 1 =10− y 2 . . .. .. .(1)−
3 Reaction curve of firms -1
y 2 =14−0 .5 y 1 . .. . .. .(2) -Reaction curve of firms –2

Substituting equation (1) in equation (2)

1 1
y 2 =14−0 .5 (10− y 2 )=14−5+ y 2
3 6

13
 1 5 54
y 2 =9+ y 2 → y 2 =9 y 2 = =10 . 8
6 6 5 (Equilibrium level of output of firm –2)
1 1
y 1 =10− y 2 =10− ( 10 .8 )=10−3 . 6
3 3
The total output of the industry at equilibrium is the sum of Y1 and

Y2, ( y=6 . 4 +10 . 8=17 . 2 )

The above result can represented graphically as shown in figure 7.5 below

Y2

30 Reaction curve for firm 1

14 Cournot’s equilibrium

10.8 Reaction curve for firm 2

Y1
6.4 10 28

Figure 7.5 Cournot equilibrium

The market price of the product
p( y∗)=60−2 ( y∗) =60-2(17.2) =60-34.4 P(Y*) =25.6
The profit of each duopolistic is
Π 1 = py 1 −c ( y 1 )
¿ ¿ Π 1 =163 .84−40 .96
Π 1 =163 .84 …….. Profit of firm-1
Π 2 = py 2 −c ( y 2 ) = (25.6) (10.8)–4(10.8) = 276.48-43.2
¿ ¿

Π 2 =233 .28 …..Profit of firm–2

Total industry’s profit, Π=Π 1 + Π 2

= 163.84+233.28 Π=397 . 12

14
____________________________________________________________________________
Activity
1. Given the following demand and cost function of cournot duopolist
Y= 40-0.2p where Y=Y1+Y2
C1 =50+2y1 C2 =100 +10y2
A. Find the profit function of each firm and their reaction function
B. Find cournot equilibrium level of output and price
C. Calculate the profit of each firm
D. Show graphically the equilibrium level of output of the two firms.

2. Suppose that we have two firms that face a linear demand curve and have
constant marginal costs, c, for each firm. Find the equilibrium level of output for the two firms in
terms of a and b assuming that are act according to cournot model.

7.2.3 The Bertrand duopoly model

In cournot model, we have seen that firms are choosing quantities of output produced and letting
the market to determine price. However, in case of Bertrand model, firms set their price and
letting the market to determine the amount of output sold. This implies the strategic variable up
on which firms are competing to maximize their profit is price rather than output for Bertrand
duopolist. Similar to cournot model however, each firm makes decision about the level of price
that maximizes their profit simultaneously by assuming their competitor’s price fixed at existing
level.

The model also assumes that firms operating in the industry produce homogenous product with
identical cost. This implies that each firm faces the same demand curve and consumer will prefer
to purchase from lower price seller or firm. Thus if the two firm charge different price, lower
price firm will supply the entire market while the firm which charges higher price sell nothing. If
both firms charge the same price, the consumers are indifferent between the two firms’ product.
This leads to a return to firm-1 of the form:

15
Equation (1) to (3) represent the profit earned by firm-1, when it set price less than, equal to and
greater than firm-2 respectively.
Therefore, Bertrand model is an oligopoly model in which firms producing homogenous product
set price simultaneously that maximize their profit by assuming their competitors price fixed at a
certain level.
It is known that price cannot be set less than marginal cost since each firm has the incentive to
decrease their price to increases its profit by reducing production level. So let us see how
Bertrand duopolist set price under the case where price is greater than marginal cost for the two
competing firms.
Suppose that both firms sell their product at some price greater than marginal cost. If firm-1
lower its price by small amount while the other firm keeps its price fixed, the entire consumer
will prefer to purchase from firm-1 and the other firm sell nothing as stated above. However, the
other firm also acts in the same way (have the incentive to reduce its product) if its price is
greater than marginal cost. Therefore, price greater than marginal cost cannot be stable
equilibrium because each firm has an incentive to cut price as long as production remain
profitable. The only equilibrium point where firms have no incentive to change their decision is
where price equal to marginal cost. This is the same as competitive market equilibrium
condition. Both firms set price equals to marginal cost in the short run and both will set price
equal to average cost in the long run because of constant cost assumption, which leads to earning
of zero profit.

The equilibrium price level of Bertrand model can be also determined through reaction curve
approach. As we have seen under cournot model, reaction curves are derived from Isoprofit
maps. However, Bertrand Isoprofit curve represent different thing from cournot’s isoprofit curve.
Bertrand model isoprofit curves contain locus of point that represents a combination of prices of

16
 a firm and its competitor that yields the same level of profit to the firm. Unlike cournot isoprofit
curve , the shape of Bertrand duopolist isoprofit curve is convex to the price axis of the firms.

Dear students, what does the shape of Bertrand duopolist isoprofit imply? The shape of Isoprofit
of Bertrand duopolist show how a firm reacts to price cut by its competitor. For example if the
competitor of a firm cut price, the firm also adjust its price to maintain its profit at the same
level. Such process continues up to the minimum point of the isoprofit curve, which represents
lower level of profit. If the competitor cut price beyond the minimum point, the firm cannot
adjust its price to keep the same level of profit. The profit of the firm decreases due to fall in
price and increase in output level. This can be indicated by moving to the lower level of Isoprofit
curve.

PA

PA 4B

4A
3B
P2A 2B
3A

2A 1B

1A

PB
P2B

PB
Figure: 7.6. Reaction and Isoprofit curves of Bertrand duopolist

17
 For example, firm-A in figure 7.6 moves from Isoprofit curve to Isoprofit curve when
firm-B cut its price below P 2B. This implies Isoprofit curve found nearer to the price axis of the
duopolist represent lower level of profit. If we join the minimum point of successive isoprofit
curves of firm-A, it result in reaction curve of A. They are locus of point firm-A can attain by
charging a certain price, given the price of its rival. The reaction curve for firm-B also derived in
the similar way by joining the lowest point of isoprofit curves given the price level of firm-A.
Given the two reaction curves, Bertrand equilibrium defined by the intersection of the reaction
curves of the firms as indicated by figure 7.7.

PA

B’s reaction curve

PA* e A’s reaction curve

Bertrand’s equilibrium

Price level (PA*, PB*) PisB* the equilibrium level of

PB Bertrand duopoly model. Also like cournot
model, Bertrand equilibrium does not lead to maximization of industry (joint) profit, due to the
Figure: 7.7. Bertrand equilibrium
fact that firms behave naively (firms set their price by assuming its rival keep its price fixed and
they never learn from past experience) to decide the price that maximizes their profit. The
industry profit could be increased if firms recognized their past mistakes and abandoned the
Bertrand pattern of behavior.
Although the Bertrand model used to understand the strategic interaction of oligopolist on price
setting, it has plenty of shortcomings for various reasons. For one thing, firms, which produce
exactly the same product, seem to compete more by focusing on non-price competition than on
price competition. Moreover, if they focus on price competition, they set the same price in
accordance with the model; there is no real assurance that they split the market equally. Also like
cournot model, Bertrand model is criticized for its naïve assumption of firms.

7.2.4 Stackelberg Model

18
Dear students, in the previous duopoly model we have seen that duopolists make simultaneous
optimal decision about level of output produced and price charged. However, according to
Stackelberg model firms make decision about output that maximizes their profit sequentially.
That is, there is a firm known as a dominant (stackelberg leader) which knows the other firm
behaves naively in cournot fashion (i.e. known the reaction function of naïve firm). The firm,
which behaves in cournot, fashion (take competitor’s output as given) and make decision after
observing leader’s decision is called the follower. The leader has extra information and potential
than the follower firm to make decision before the follower firm. In choosing its own output,
therefore, the leader could account the effect of its output on follower behavior while the
follower naively took leader’s output as given.
For instance, IBM is often considered as a dominant firm in the computer industry. Small firms
in the industry wait for the decision made by IBM in order to make decision how much to
produce and the type of product they supply to the market. In general, the leader firm (the first
mover) decides to produces certain amount of output, which maximizes its profit; it will take into
consideration the impact of follower firm. The follower firm, after identifying the level of output
produced by the leader firm, responds by producing certain amount of output to maximize its
profit. Each firm act in the stated way while making its output decision because each of them
know that the total output produced determined the market price and then profit they earn.

Given the structure of the model, what output should the leader choose to maximize its profits?
Leader firm recognizes its influence on the action of followers firm and total level of output
when making decision on the level of output to maximize profit. This relationship between
follower and leader optimal choice can be summarized with the help of reaction function of
follower firm. So the leaders first determine the reaction function of its follower and then
incorporate it to its own profit function. Then it maximizes the newly formed profit function like
monopoly firm by setting marginal revenue equals to marginal cost. On the other hand follower
firm react to the optimal choice of the leader according to its reaction function to come up with
its profit maximizing level of output.
The stackleberg solution can also be illustrated graphically using isoprofit curves and reaction
curves. Both firms have the same shaped isoprofit curve as in the case of cournot. The highest
profit level is represented by the isoprofit curve found away from the quantity axis of each firm

19
and the lowest level of profit is represented by isoprofit curve, which found near to the quantity
axis. Given such properties of isoprofit curves of the firms, they are acting in the following ways
to determine their optimal output level.

Y2 Reaction curve of
firm-1

Cournot
equilibrium

Reaction
curve of
firm-2 Stackelberg
equilibrium

Isoprofit curve of
e firm-1
Y*2

Firm 2 as a follower will choose an output

Y2 * along its reaction curve,
Y1 while firm-1 (the leader)
choose the output level on the reaction curve of firm-2 that gives him the highest maximum
Figure: 7.8. Stackelberg equilibrium
profit. As a result, the equilibrium point of the stackelberg duopolists is not defined by the
intersection of their reaction curve. This is because the leader firms no more take the follower’s
output as given. It knows the follower output would depend on its own output level in
accordance with its reaction function. This implies the equilibrium point of the Stackelberg
model defined by the tangency of the isoprofit of the leader with the reaction curve of the
follower as indicated by figure 7.8 above.

Since the leader, firm-1 makes choose on the reaction curve of follower, firm 2, point e
represents stable stackleberg’s equilibrium. At point-e, the leader get higher profit and the
follower firm get lower profit compared to cournot equilibrium.
In short if one firm is sophisticated, it will emerge as a leader and stable equilibrium will
established since the naïve firm act as a follower. However if both firms are sophisticated, both
wants to be a leader to get higher profit. In this case, market situation becomes unstable. Such

20
situation is known as Stackleberg disequilibrium. The effect of such situation will be either a
price war until one of them surrender and agree to act as follower or collusion of both firms
abandoning their reaction function and move to a point close to edge worth contract curve where
both of them get higher profit

Numerical illustration of Stackleberg model

Consider the demand and cost function used for cournot model
P= 60-2(y1 + y2)
C1= y12, C2= 4y2
Find
a) Stackleberg equilibrium assuming firm – 1 is a leader
b) Stackleberg equilibrium assuming firm – 2 is a leader.
Solution
a) If firm-1 is a leader, it is the one which chooses output level that maximizes
its profit first by incorporating reaction function of its follower in to its own
isoprofit function as follows:
Follower’s profit function,
╥2 = [60-2(y1+y2)] y2 -4y2
╥2 = 60y2 - 2y1 y2 - 2y22 - 4y2
╥2 = 560y2 - 2y1 y2 - 2y22
Take first order partial derivative of the profit function with respect to y 2 to determine the
reaction function of follower firm;
( )
2
The2 reaction function of follower
∂ π 2 ∂ 56 y 2 −2 y 1 y 2−2 y
= firm=0
∂ y2 ∂ y2

56− y 1 −4 y 2 =0
⇒ y 2=14−0 . 5 y 1

Leader’s profit function can define as:

π 1 =[ 60−2 ( y 1 + y 2 ) ] y 1− y 12 … ( 2 )

21
To determine its optimal level of output the leader incorporate follower’s reaction function into
its own profit function:

Then take first order condition to determine the level of output that maximizes the profit of the
leader.

The leader produces 8 unit of output to maximize its profit.

To determine follower’s optimal choice substitute leader’s output level in the reaction function
of the follower firm.

Profit of the leader

Profit of follower

b) Assume firm-2 is a leader and firm-1 is a follower

Solve backward starting from the follower profit maximization

22
To determine the optimal level of output for the follower, substitute the optimal output level of

the leader (firm-2) in the reaction function of the follower:

Leader profit’s follower’s profit

23
7.3 Collusive oligopoly
All oligopoly models described up until now are an example of non-collusive oligopoly. In such
type of industry each firm make independent decision to maximize its profit even though each of
them take other’s likely behavior in to account during the process of decision making. Now we
consider other possibility in which firms found in a given industry make collusive agreement
implicitly or explicitly to some degree in setting price and output. Firm enter in to such collusive
agreement in order to cultivate the advantage of increasing profit, decreasing uncertainties and to
create better opportunity to prevent other’s entry to the industry.

7.3.1 Cartel
A cartel is a formal organization of firms producing the same product. It is formed to coordinate
the policies of member firm so as to increase their joint profit by limiting the scope of
competition among them. It also reduces uncertainty arising from their mutual interdependence.
Here we will study two typical forms of cartels.
a) Cartels aiming at joint profit maximization
b) Market sharing cartel
7.3.1.1 Cartel aiming at joint profit maximization.
Cartel aiming at joint profit maximization is formed as its name indicates with the objective of
maximizing the total industry’s profit. In order to achieve this objective member firm appoint a
central agency to which they delegate the authority to decide not only the total quantity and price
at which the industry’s profit maximized but also allocation of output and profit among the
member of cartel.
To make such decisions, it is assumed that central agency have access to information about the
cost of individual firms and know market demand. Given the information about cost and
demand, central agency determine price and output that maximizes industry’s profit defined by
the intersection of marginal revenue and marginal cost curve. Marginal revenue curve can be

24
derived from market demand and the marginal cost used for decision can be obtained by
summing up individual marginal cost. In this case, the cartel act as multi-plant monopoly and

chooses to maximize the total industry profits:

Take first order condition of industry profit function to determine profit maximization level of
output and price:

For simplicity let us assumes there are only two firms in the cartel with marginal cost MC 1and
MC2 as indicated in figure 7.9. So the industry’s marginal cost (MC) obtained by summing up
marginal costs of the two firms. From the market demand DD, we can derive industry’s marginal
revenue. The intersection point of MC and MR gives profit-maximizing level of output and
price.

. MC2 AC2
MC1
AC1 MC=MC1 +MC2
D

P*

D
Q*1 Q*2 Q*=Q*1 + Q*2

Figure: 7.9 Joint profit maximizing cartel

At the same time central agency, allocate production between the two firms similar to Multi-
plant monopoly by equating the MR to individual marginal cost. Mathematically the profit
maximization of cartel can be derived as

25
 Max π = R1+R2-C1-C2
Y1, Y2

π= p (y1+y2) (y1+y2) – C1 (y1) –C2 (y2)

Taking F.O.C for joint profit maximization
∂ π1 ∂[ ( y 1 + y 2 )( y 1 + y 2 )−c1 ( y 1 ) −c 2 ( y 2 ) ]
=
1. ∂ y 1 ∂ y1 =0

2.
∂ y1 [
∂ π 1 ∂ ( y 1 + y 2 )( y 1 + y 2 ) −C2 ( y 1 ) −C2 ( y 2 )
=
∂ y2 ] =0

Rearranging equation (1) and (2) we can get

∂ R1 ∂ R 1 ∂ R
= =
MR= MC1 = MC2 Since ∂ y 1 ∂ y2 ∂ y

Cartels aiming at joint profit maximization determine the amount of output produced by each
member firm using the above condition (MR= MC1 = MC2 )
Numerical illustration
Assume that the Market demand facing the members of a cartel is P = 60-2 (y 1+y2) and their cost
of production is given by C1= Y12, C2= 4y2. Find the level of output and price that maximizes the
industry’s profit.
Solution
We derive profit maximization condition for cartel aiming at joint profit maximization as MR=
MC1 = MC2
∂TR ∂ ( 60−2 y ) . y
MR= = =60−4 y
∂y ∂y
2
∂ c1 ∂ y 1
MC 1 = = =2 y
∂ y1 ∂ y1
∂c ∂ y
MC 2 = 1 = 4 2 =4
∂ y2 ∂ y2
1) MR = MC1
2) MR = MC2
 60-4y = 2y1 Where y = y1+y2
60 – 4(y1+y2) = 2y1
60 – 4y1-4y2 = 2y1

26
 60 – 6y1- 4y2 = 0
From Equation (3)
60 – 4y = 4
60 – 4y1- 4y2 = 4
56 – 4y1- 4y2 = 0… (4)
Solve simultaneously equation 3 and 4
- 56 - 4y1- 4y2 = 0
60 - 6y1- 4y2 = 0
4-2y1 = 0
 Y1 = 2
Substitute in equation (1)
60-6y1 - 4y2 = 0
60 - 12- 4y2 = 0
48-4y2 = 0
 Y2 = 2
Total output that maximizes industry’s profit is the sum of individual firm’s output
Y= y1 + y2 = 12 + 2 = 14
Y = 14
P = 60 -2 (y)
= 60 -2 (14) = 22
P= 22

Joint profit of the industry π (y1 + y2) = [p (y1 + y2)] [y1 + y2]-
(y1 + y2) = 60y1 + 56y2 – 3y12 – 2y22 -4y1y2
π (y1 + y2) = 60y1 + 56y2 – 3y12 – 2y22 -4y1y2
= 60(2) + 56 (12) – 3 (2)2 – 2 (12)2- 4 (2) (12)
= 120 + 672 – 12 – 288 -96
= 792 – 396 = 396
In Practice however, cartels rarely achieve maximum joint profit. There are several reasons why
cartel’s industry profit cannot be maximized. Some of them are
1. Mistakes in the estimation of market demand. Mistake in estimation of market demand leads to
mistake in the derivation of the MR and hence price and output which maximizes profit.

27
2. Mistakes in estimation of MC due to incomplete knowledge of the individual firm’s costs at
different level of output. Such error leads to equilibrium level of output and price, different from
the expected cartel solution.
3. Slow process of cartel negotiation. Since member firms have different cost and market share, to
bring them to common level of price it takes time. When the time taken for making decision is
longer, the price of goods and serves changes for what it is expected to be. So the information used
by central agency for decision making becomes out dated and correct level of price that maximizes
the industry profit cannot be set.
4. Stickiness of the negotiated price. Even when market condition changes, once price set, it need
long time for negotiation and set profit maximizing price.
5. Cheating by member firm during bargaining process through giving biased information in order to
achieve large market share.
6. Existence of high cost firms. If a firm operates with higher than equilibrium MC, it reduces the
total joint profit. As a result, it has to close down to maximize joint industry’s profit. Nevertheless,
there is no such thing with firm aiming of joining industry maximization.
7. Fear of government interference to change the level of price which maximizes cartels’ industry
profit
8. The wish to have good public image cartel agency will not increase price to profit maximization
level of the industry.

_________________________________________________________________________

Self-test exercise
Suppose that the market demand facing two firms producing the same product is

given by p=300−3 y where y= y 1 + y 2 , total output. If the production cost of the

2
two firms is equal and given by C ( y )=30 y +1 .5 y and then the agree to form cartel
aiming at joint profit maximization, find:
a) The amount of output that maximizes their profit.
b) The industry’s price.
c) The output and the profit of each firm.
______________________________________________________________________

28
7.3.1.2 Market Sharing Cartel
In market sharing cartel the member firms agree on how to share the market but they keep a
considerable degree of freedom concerning the style of their output, their selling activities and
other decisions. Each firm therefore operates in one area or region agreed upon without
encroaching on others’ territories. This form of cartel is more common in real world and more
popular than cartel aiming at maximization of joint industry profit. There are two basic methods
for sharing market by the member firm of cartel: non-price competition and quota.

1. Non- price competition

In such form of cartel the member firm agree on a common price at which they sell their output.
The price up on which they agreed set by the process of bargaining. In the bargaining process,
low cost firm pressing for lower price and high cost firm press for higher price. At the end
common price which allow all members certain amount of profit, is set. At such price firms
compete to maximizing their profit by increasing their sell volume through different way of non-
price competition like quality, style, selling activities and advertising. Such form of cartel is
unstable most of the time compared to cartel aiming at joint profit maximization. Because,
whenever there is cost and liquidity difference, low cost firms have the incentive to cheat by
lowering their price and initiates price war among member firms, which result in instability.
2. Sharing of the market by agreement on quotas
In this case, member firms agree to supply a certain quantity of output at agreed price. The quota
allocation made based on the cost structure of the firms. If they have identical costs, the
monopoly solution will emerge by sharing the market equally. However if costs are different, the
quota of market share differs between the firms. Under such cases quota sharing depend on the
bargaining power of the firms. During bargaining process past levels sales and or the basis for
productive capacity of the firms are considered for decision.
Most of the above forms of cartels are unstable. The firms will not act according to output and
price level upon which they agreed due to a number of reasons. Some of the reasons includes: the
member firms have different costs, different assessments of the market demand and even have
different objective. So most of the time they wants to set price at different level. Furthermore,
each member of the cartel tempted to cheat by lowering price slightly to capture larger share of

29
market than allocated for the firm. This implies there should be some sort of enforcement
mechanism for a cartel to be successful. Threat in the long run for the member who breaks the
agreement should be there. Also gaining certain monopoly power to set price by cartel is other
factor for the success of cartel i.e., if the potential gain form cooperation are large ,cartel member
have more incentive to solve their organizational problem and hence profit for cartelization is
large enough to give incentive for the members to act according to the contractual agreement.

7.3.2 Price leadership

Dear students, unlike firms in cartel, which agree explicitly to cooperate in setting price and
output, firms in price leadership collusive oligopoly, agree to cooperate implicitly in making
decisions about price and output without any formal discussion. They enter in to such agreement
voluntarily to avoid any uncertainty about the competitor reaction. In such type of model, one
firm implicitly recognized as the leader and set price. The other remaining firm, the follower take
price as given and adopt the price set by the leader firm even though its profit did not
maximized.
This follow from the assumption that the two firm selling identical products. If one charged a
different price form the other, all of the consumers would prefer the producers with the lower
price. In addition, if two firms do not agree implicitly on common price, there is possibility to
enter in to a price war. Price war through reduction of prices cause low profit level and even
causes destruction of the firm. These conditions may force oligopoly firms to cooperate without
actually making explicitly agreements with one other.
There are various form of price leadership, the most common ones are:
1. Price leadership by low cost firm
2. Price leadership by large (dominant) firm
7.3.2.1 Low cost price leadership
This model assumes that there are two firms in the industry producing homogenous product at
different cost. One firm produce at low cost compared to its competitor. Moreover, firms may
have equal or unequal market share. Given firms with the stated features, low cost firm
becomes the leader and set price, which maximizes its profit. Follower firm by scarifying some
of its profit take the price set by the leader. This is to avoid a price war, which would eliminate
the firm from the industry if price set lower than its LAC.

30
However, the price set by the leader firm using marginal principle (MC=MR) would remain at
the stated position through maintaining output constant. Deviation of output from the point
where MR=MC due to over or under supply of output by follower firm will change the price
and then the profit of the leader will not maximized. This implies that the follower must supply
a quantity sufficient to maintain the price set by the leader. So at the optimal price level, the
firms must also enter agreement on the share of the market formally or informally. Otherwise
even though the follower adopt leaders price, producing higher or lower level of output
required to maintain the price (set by the leader) in the market push the leader to non-profit
maximizing position. In this respect, the follower is not completely passive; it can affect the
market price and then the profit of the leader unless they enter in to formal or informal
agreement to supply certain proportion of the total output.

Numerical illustration
Recall that the examples we use for others duopoly model with:
Market demand: P = 60-2(y1+y2)
Cost C (y1) =y12 C (y2) = 4Y2
When we see their cost structure, firm two is a low cost firm so that it can be taken as the
leader. So, it is the one who set price using the marginal principle but the two firms go into
agreement about the share of the market at leader’s profit maximizing level of price. Let
assume they agree to share equally. With this assumption, the demand curve relevant to
leader’s decision would be modified as:
P= 60-2(Y1+Y2) , Y1=Y2
P= 60-2(Y2+Y2) = P=60-4Y2
Π 2 =PY 2 −C 2 , Π 2 =(60−4 Y 2 )−4 Y 2=56Y 2−4 Y 22

dΠ 2
=0
Take F.C.O dY 2 to determine equilibrium level of output that maximizes leader’s profit.
dΠ 2 (56 Y 2 −4 Y 22 )
=d =0
dY 2 dY 2 , 56-8Y2=0 Y2=Y1=7
P=60-4Y2 =60-4(7) = 60-28=32 P*=32

31
 However, p=32 and output level Y1=7 would not maximizes the profit of the follower firm.
The price that maximizes the profit of the follower firm can be determined through
maximizing its profit function as follows;
Π 1 =R1 −C1 =(60−4 Y 1 )Y 1 −Y 1 , Π 1 =60 Y 1−5Y 12

dΠ 1
=60−10 Y 1 =0
Take F.O.C dY 1
Y 1 =6 P=60-4Y 1 , P=60-4(6) =60-24=36, P*=36
The output level which maximizes the profit of firm one (follower) would be Y 1 =6 units and
charge a price P*=36 if it did not recognizes the leader ness of firm two.

7.3.2.2 Dominant firm price leadership

In such oligopoly industry there is one large firm having large share of the total market with a
group of smaller firms supply the rest. The larger firm, which is also known as, the dominant
firm become the leader and set price that maximize its profit. The other smaller firms, which
could have little influence over price, would be a price taker like firms in perfect competitive
firm. That is each of the small firms produces an amount determined by setting marginal cost
equals to price announced by the leader and the leader supply the residual.
By assumption, the dominant firm knows the market demand and the marginal cost of small
firms. This assumption enables the dominant (the leader) firm to derive its own demand function
form the market demand and marginal costs of smaller firms. That is since small firms in the
industry behave like prefect competitive firm, the dominant firm equate price with the sum of

their marginal costs to get their supply function ( ). Then the leader firm’s demand
function, which is also known as, the residual demand curve, can be obtained by subtracting the
supply function of the smaller firms from the market demand.
Once the leader identify its demand function, it act like a monopoly to set price and output level
which maximizes its profit by equating marginal revenue with its marginal cost . Here the
marginal revenue used for decision is derived from residual demand that actually measures how
much output it will be able to sell at each given price.

Small firms supply 32

Leader's MC
 Output

As indicated in figure 7.10. for example at P 1 the demand for the product of the leader will be
zero, because the total quantity D1 is supplied by smaller firm. As price fall below p 1 the demand
for the leader’s product increases. At p the total demand is D 2 out of which PA is supplied by
small firms and the remaining AD2 is supplied by the leader.
Having derived the demand curve of the leader as in figure 7.10(b) and given its MC, the
dominant firm will set price at which his MR=MC. However smaller firms may or may not
maximizes it profit depending up on its cost structure at the set level of price.

33