INDUSTRIAL ORGANIZATION

PRICE-BASED
COMPETITION
BERTRAND AND DOMINANT FIRM

Truong Dang Thuy

[email protected]
BERTRAND COMPETITION
SIMULTANEOUS PRICE SETTING
 BERTRAND
COMPETITION ▪ Homogeneous goods/services
▪ Price competition, and consumers buy from cheaper firms
▪ No capacity constraints in production
▪ No product differentiation

Basic assumptions ▪ Firms simultaneously decide price

▪ No strategic collusion
 BERTRAND ▪ Consider a duopoly, where Firm 1 and Firm 2 produce
identical products and compete by choosing price
COMPETITION simultaneously.
▪ The market demand
𝑄=𝑞 𝑃
where 𝑃 is price and 𝑄 the total quantity supplied.

Bertrand competition with ▪ Identical cost function of two firms

identical cost 𝐶 = 𝑐𝑞𝑖
where 𝑞𝑖 is the quantity of Firm 𝑖, (𝑖 = 1,2)
▪ Note the marginal cost is constant: 𝑚𝑐1 = 𝑚𝑐2 = 𝑐.
 ▪ Consumers will buy from the cheaper firm.
BERTRAND ▪ If Firm 1 sets 𝑝1 > 𝑐, Firm 2 can set 𝑝2 = 𝑝1 − 𝜖 and capture
the entire market
COMPETITION
▪ If Firm 1 will undercut in response.
▪ The process continues until 𝑝1 = 𝑝2 = 𝑐.
▪ Meaning zero profit and the market outcome is identical to
perfect competition. This is referred to as the Bertrand paradox.

▪ The two firms split the market equally, but no profit.

Bertrand competition with
identical cost ▪ The market demand can be introduced, which helps identify
the quantity at price 𝑐: 𝑄 = 𝑞 𝑐
▪ Two firms split the market equally
𝑞 𝑐
𝑞1 = 𝑞2 =
2
 ▪ Assume the market demand
𝑄 = 500 − 20𝑃
BERTRAND
where 𝑃 is price and 𝑄 the total quantity supplied.
COMPETITION
▪ The cost function of two firms
𝐶 = 10𝑞𝑖
where 𝑞𝑖 is the quantity of Firm 𝑖, (𝑖 = 1,2)
▪ The marginal cost is constant: 𝑚𝑐1 = 𝑚𝑐2 = 10.
Bertrand competition with Consider the following:
specific functional form
▪ What if the two firms set the same price 𝑝1 = 𝑝2 = 12?
▪ What if Firm 2 cuts its price to 𝑝2 = 11 while Firm 1 remains
its price?
▪ What if Firm 1 retaliates by cutting its price to 10.5?
 BERTRAND
▪ In Bertrand competition, firms always have the incentive to
COMPETITION slightly undercut its rivals
▪ As a result, long-run equilibrium price is equal to the marginal
cost, and firms equally split the market with zero profit.
▪ The above is true even in the case of N firms.
▪ Note the above may be no longer true in case of
Summary ▪ product differentiation
▪ different marginal costs
▪ capacity constraints
BERTRAND-EDGEWORTH MODEL
SIMULTANEOUS PRICE SETTING WITH CAPACITY CONSTRAINTS
 ▪ Bertrand predicts price = marginal cost and zero profits, even
BERTRAND-EDGEWORTH with just two firms
MODEL ▪ Real-world firms often face capacity constraints
▪ Edgeworth (1897) extended the model: firms cannot serve
unlimited demand
▪ Bertrand-Edgeworth model take all the assumptions of the
Bertrand competition, except:
▪ each firm can only sell up to its capacity
Motivation and assumptions ▪ consumers buy from the cheapest firm first
▪ if one firm’s capacity is exhausted, consumers buy from the next
cheapest firm
▪ if both prices are equal, they split demand, constrained by
capacity
 ▪ Consider a duopoly, where Firm 1 and Firm 2 produce
BERTRAND-EDGEWORTH identical products and compete by choosing price
simultaneously.
MODEL
▪ The market demand
𝑄=𝑞 𝑃
where 𝑃 is price and 𝑄 the total quantity supplied.
▪ Firms decide 𝑝1 , 𝑝2 simultaneously
Bertrand competition with ▪ Identical cost function of two firms
capacity constraint 𝐶 = 𝑐𝑞𝑖 𝑖 = 1,2
▪ Capacity constraints
▪ Firm 1 can produce at most 𝐾1
▪ Firm 2 can produce at most 𝐾2
 If 𝑝1 < 𝑝2 , then
▪ Firm 1 sells
𝑞1 = min 𝐾1 , 𝑞 𝑝1
BERTRAND-EDGEWORTH
▪ Firms 2 sells
MODEL 𝑞2 = min 𝐾2 , max 𝑞 𝑝2 − 𝑞1 , 0
▪ If 𝑞 𝑝1 < 𝐾1 , the constraint is not binding, and
▪ Firm 1 sells 𝑞1 = 𝑞 𝑝1 , captures the whole market,
▪ Firm 2 sells nothing.
▪ If 𝑞 𝑝1 > 𝐾1 , the constraint is binding, and
▪ Firm 1 sells 𝑞1 = 𝐾1 , leaving some potential residual demand for
Market allocation if 𝑝1 < 𝑝2 Firm 2
▪ If Firm 2 sets 𝑝2 such that 𝑞 𝑝2 − 𝑞1 < 0, it sells nothing
▪ Otherwise if 𝑞 𝑝2 − 𝑞1 > 0, it sells
▪ 𝑞2 = 𝑞 𝑝2 − 𝑞1 if the residual demand is within capacity
constraint, or 𝑞 𝑝2 − 𝑞1 < 𝐾2
▪ 𝑞2 = 𝐾2 if this residual demand is greater than the capacity
constraint 𝑞 𝑝2 − 𝑞1 > 𝐾2
 If 𝑝1 > 𝑝2 , the reverse logic applies
▪ Now Firm 2 sells first
𝑞2 = min 𝐾2 , 𝑞 𝑝2
BERTRAND-EDGEWORTH
▪ Firms 1 sells the residual
MODEL 𝑞1 = min 𝐾1 , max 𝑞 𝑝1 − 𝑞2 , 0
▪ If 𝑞 𝑝2 < 𝐾2 , the constraint is not binding, and
▪ Firm 2 sells 𝑞2 = 𝑞 𝑝2 , captures the whole market,
▪ Firm 1 sells nothing.
▪ If 𝑞 𝑝2 > 𝐾2 , the constraint is binding, and
▪ Firm 2 sells 𝑞2 = 𝐾2 , leaving some potential residual demand for
Market allocation if 𝑝1 > 𝑝2 Firm 1
▪ If Firm 1 sets 𝑝1 such that 𝑞 𝑝1 − 𝑞2 < 0, it sells nothing
▪ Otherwise if 𝑞 𝑝1 − 𝑞2 > 0, it sells
▪ 𝑞1 = 𝑞 𝑝1 − 𝑞2 if the residual demand is within capacity
constraint, or 𝑞 𝑝1 − 𝑞2 < 𝐾1
▪ 𝑞1 = 𝐾1 if this residual demand is greater than the capacity
constraint 𝑞 𝑝1 − 𝑞2 > 𝐾1
BERTRAND-EDGEWORTH If 𝑝1 = 𝑝2 = 𝑝, there are one million possibilities ☺
▪ Case A: If 𝑞 𝑝 > 𝐾1 + 𝐾2
MODEL 𝑞1 = 𝐾1 𝑞2 = 𝐾2
▪ Case B: If 𝑞 𝑝 < 𝐾1 + 𝐾2 , there are 3 possibilities
▪ B1: No constraint for both, then equally splitting the demand
𝑞 𝑝 𝑞 𝑝
≤ min 𝐾1 , 𝐾2 , then 𝑞1 = 𝑞2 =
2 2
▪ B2: Constrant for Firm 1
Market allocation if 𝑝1 = 𝑝2 𝑞 𝑝 𝑞 𝑝
> 𝐾1 and < 𝐾2 , then 𝑞1 = 𝐾1 and 𝑞2 = 𝑞 𝑝 − 𝐾1
2 2
▪ B2: Constraint on Firm 2
𝑞 𝑝 𝑞 𝑝
< 𝐾1 and > 𝐾2 , then 𝑞2 = 𝐾2 and 𝑞1 = 𝑞 𝑝 − 𝐾2
2 2
 BERTRAND-EDGEWORTH

MODEL ▪ Each firm profit is

𝜋𝑖 = 𝑝𝑖 − 𝑐 𝑞𝑖
▪ This means that firms may choose higher prices if:
▪ they expect the rival’s capacity to be exhausted, and
▪ they can still serve residual demand at a higher price
Profit ▪ In some cases, the firm with higher price has lower demanded
quantity but earns higher profit.
 ▪ Assume the market demand
𝑄 = 500 − 20𝑃
BERTRAND-EDGEWORTH where 𝑃 is price and 𝑄 the total quantity supplied.
MODEL ▪ Firms decide 𝑝1 , 𝑝2 simultaneously
▪ Identical cost function of two firms
𝐶 = 10𝑞𝑖 𝑖 = 1,2
▪ Identical capacity constraints
▪ Firm 1 can produce at most 150
▪ Firm 2 can produce at most 150
Example Consider the following:
▪ What if both firms set prices at 𝑝1 = 𝑝2 = 12?
▪ What happen if Firm 1 cut its price to 11.5?
▪ Does Firm 1 really need to cut to 11.5, or should it just set at,
say, 11.9?
THE DOMINANT FIRM MODEL
PRICE LEADERSHIP
 ▪ Brief history
▪ Developed in early 20th century to explain real-world
DOMINANT FIRM oligopolies
MODEL ▪ Formalized in the 1930s–1950s by economists like Heinrich von
Stackelberg and George Stigler

▪ Market structure: two firms, one acts as the price leader, the
other follows (follower).
▪ Key assumptions:
▪ Firms aim to maximize profit
Basic assumptions ▪ Leader assumes followers will match its price
▪ Followers are price takers within the leader’s price decision
▪ No retaliation from followers (no price wars)
▪ Demand and cost structures are known or estimable by the
leader
 ▪ Two firms:
DOMINANT FIRM ▪ Firm 1 (leader): large, sets price to maximize profit
▪ Firm 2 (follower): small, takes price as given and maximize
MODEL profit

▪ The follower must take price as given:

▪ If it sets higher price, it sells nothing
▪ If it sets lower price, it sells up to its capacity constraint
▪ Given that the follower is small, the market price should not be
affected even if the follower set lower price
Model setup ▪ If the follower set lower price and can affect the market price, it
become the leader. Or this somehow becomes the Bertrand
competition

▪ So we go with the assumption that the follower is the price

taker.
 ▪ The cost functions of the two firms
DOMINANT FIRM 𝑐𝐿 = 𝑐𝐿 𝑞𝐿
MODEL 𝑐𝐹 = 𝑐𝐹 (𝑞𝐹 )
▪ The inverse demand function
𝑃 = 𝐷 −1 𝑄
where 𝑄 is the total supply of the leader (𝑞𝐿 ) and follower (𝑞𝐹 ),
or 𝑄 = 𝑞𝐿 + 𝑞𝐹
▪ The corresponding demand is 𝑄 = 𝐷 𝑃
Model setup
▪ Example: if the demand function is 𝑄 = 𝐷 𝑃 = 𝑎 − 𝑏𝑃, the
inverse demand function is
−1
𝑎−𝑄
𝑃=𝐷 𝑄 =
𝑏
 ▪ Suppose the Leader set the price 𝑃
DOMINANT FIRM ▪ As a price taker, the follower chooses the output level that
MODEL maximize profit
max 𝜋𝐹 𝑞𝐹 = 𝑃𝑞𝐹 − 𝑐𝐹 𝑞𝐹
𝑞𝐹

▪ The FOC is familiar:

𝑃 = 𝑚𝑐𝐹 𝑞𝐹
▪ We can obtain the follower’s output supply function in the
form
The Follower’s problem 𝑞𝐹 = 𝑆𝐹 𝑃
▪ Note: the follower supplies only if 𝑃 ≥ 𝑚𝑐𝐹
▪ Exercise: Assume a linear demand function 𝑄 = 𝑞 − 𝑏𝑃 and
2
𝑞𝐹
𝑐𝐹 = . Solve for the follower’s output supply function.
2
 ▪ The residual demand for the Leader is
𝑞𝐿 = 𝐷 𝑃 − 𝑆𝐹 𝑃
DOMINANT FIRM ▪ With a specific functional form, we can invert the above
function to get the corresponding inverse residual demand
MODEL 𝑃 = 𝑃 𝑞𝐿
▪ The Leader’s profit function
𝜋𝐿 = 𝑃 𝑞𝐿 𝑞𝐿 − 𝑐𝐿 𝑞𝐿
▪ The FOC:
𝑀𝑅𝐿 = 𝑀𝐶𝐿
The Leader’s problem ▪ The Leader takes the residual demand and acts as a
monopoly.
▪ Once it chooses optimal output, it also chooses optimal
price
▪ Exercise: assume 𝑐𝐿 = 𝑐𝑞𝐿 , solve for the optimal price 𝑃 and
𝑞𝐿∗ .
 DOMINANT FIRM ▪ Leader sets price 𝑃 and choose 𝑞𝐿∗ that maximize profit, after
MODEL considering the response from Follower
▪ The leader internalizes the follower’s response and prices
strategically

▪ Follower takes price 𝑃 and choose 𝑞𝐹∗ that maximize profit.

▪ The follower is a price taker, supplying only if price covers
marginal cost

Equilibrium ▪ At the equilibrium

▪ price is above competitive level, but below monopoly level
▪ there is asymmetric market power with one firm leading and
others following
 DOMINANT FIRM
MODEL

Example with specific

functional form
 DOMINANT FIRM ▪ Assume a linear demand function
MODEL 𝑄 = 𝑎 − 𝑏𝑃
▪ The inverse demand is thus
𝑎−𝑄
𝑃=
𝑏
▪ And the cost function
▪ Leader:
Example with specific 𝑐𝐿 = 𝑐𝑞𝐿
functional form ▪ Follower:
𝑐𝐹 = 𝑑𝑞𝐹2
▪ Note that the Follower is less efficient.
 DOMINANT FIRM ▪ Suppose the Leader set price 𝑃, the follower’s problem is
MODEL max 𝜋𝐹 𝑞𝐹 = 𝑃𝑞𝐹 − 𝑑𝑞𝐹2
𝑞𝐹

▪ The FOC:
𝑃 = 2𝑑𝑞𝐹
▪ So the Follower’s ouput supply function is
𝑃
𝑞𝐹 = 𝑆𝐹 𝑃 =
Example with specific 2𝑑
functional form ▪ The profit
𝑃2
𝜋𝐹 =
4𝑑
 ▪ The residual demand for the Leader:
𝑞𝐿 = 𝐷 𝑃 − 𝑆𝐹 𝑃
𝑃 1
𝑞𝐿 = 𝑎 − 𝑏𝑃 − =𝑎− 𝑏+ 𝑃
2𝑑 2𝑑
DOMINANT FIRM 1
▪ The demand facing Leader is 𝑞𝐿 = 𝑎 − 𝑏 + 2𝑑 𝑃. The inverse demand is
MODEL 𝑎 − 𝑞𝐿
𝑃=
1
𝑏+
2𝑑
▪ The Leader’s profit function is then
𝑎 − 𝑞𝐿
𝜋𝐿 = 𝑞 − 𝑐𝑞𝐿
1 𝐿
𝑏+
2𝑑
Example with specific ▪ The FOC is
functional form 𝑎 − 2𝑞𝐿
=𝑐
1
𝑏+
2𝑑
▪ So the Leader’s ouput supply function is
1
𝑎−𝑐 𝑏+
𝑞𝐿 = 2𝑑
2
 DOMINANT FIRM ▪ Once Leader decides output, it also decides the price because
MODEL 𝑎 − 𝑞𝐿
𝑃=
1
𝑏+
2𝑑
1
𝑎−𝑐 𝑏+
▪ Substitute 𝑞𝐿 = 2𝑑
into the above equation to yield the
2
market price
Example with specific
functional form ▪ The Leader’s profit
2
1 1
𝜋𝐿 = 𝑎−𝑐 𝑏+
1 2𝑑
4 𝑏+
2𝑑
 DOMINANT FIRM
MODEL ▪ Assume a linear demand function
𝑄 = 180 − 1.75𝑃
▪ The inverse demand is thus
180 − 𝑄
𝑃=
1.75
Example with specific ▪ And the cost function
functional form 𝑐𝐿 = 10𝑞𝐿
𝑐𝐹 = 2𝑞𝐹2
 DOMINANT FIRM
MODEL ▪ The follower’s MC is
𝑀𝐶𝐹 = 4𝑞𝐹
▪ So when ever the Leader set the price 𝑃, it knows that the
Follower’s ouput supply is
𝑃
𝑞𝐹 =
4
Example with specific ▪ The follower’s profit
functional form 𝑃2
𝜋𝐹 =
8
 ▪ The residual demand for the Leader:
𝑃
DOMINANT FIRM 𝑞𝐿 = 180 − 1.75𝑃 − = 180 − 2𝑃
4
MODEL ▪ The inverse demand is
180 − 𝑞𝐿
𝑃=
2
▪ The Leader’s profit function is then
180 − 𝑞𝐿 𝑞𝐿
𝜋𝐿 = − 10𝑞𝐿
2
Example with specific ▪ The FOC is
functional form 180 − 2𝑞𝐿
= 10
2
▪ So the Leader’s ouput supply function is 𝑞𝐿 = 80
180−𝑞𝐿 180−80
▪ The price is 𝑃 = = = 50
2 2
 DOMINANT FIRM
MODEL
▪ Given that 𝑃 = 50, 𝑞𝐿 = 55 and 𝑞𝐹 = 12.5 (why?), the profits
are:
▪ Leader: 𝜋𝐿 = 3,200
▪ Follower: 𝜋𝐹 = 312.5

▪ What if at 𝑃 = 12.5, the follower choose an output level

Example with specific different from 12.5?
functional form
▪ What if the Leader choose a price level different from 50?