CHAPTER 9 – MATHEMATICAL TUTORIAL

Cournot and Stackelberg

Oligopoly Models

A Step-by-Step Tutorial in Quantity Competition

Reaction Functions • Nash Equilibrium • Backward Induction

Prepared by Mr. Farouk Saleh

Economics • Intermediate Microeconomics

Mr. Farouk Saleh | +201062415129 | Page 1

Table of Contents

1. Introduction to Chapter 9 ........................................................................................................ 3

2. Learning Objectives................................................................................................................. 3
3. Mathematical Foundations and Notation .............................................................................. 3
4. Section 9.1 – Cournot Duopoly: Reaction Functions & Equilibrium .................................. 6
4.1 Assumptions of the Cournot Model ..................................................................................... 6
4.2–4.4 Revenue, Marginal Revenue & Reaction Function of Firm A ........................................ 6
4.5 Firm B’s Reaction Function ................................................................................................. 7
4.6–4.9 Cournot–Nash Equilibrium, Output, Price & Profit ......................................................... 8
4.10 Graphical Interpretation .................................................................................................... 9
4.11 Worked Numerical Cournot Example .............................................................................. 10
5. Section 9.2 – The Stackelberg Model .................................................................................. 11
5.1 Assumptions..................................................................................................................... 11
5.2 The Follower’s Problem .................................................................................................... 11
5.3 The Leader’s Problem ...................................................................................................... 12
5.4–5.5 Follower Output, Total Output, Price & Profits ............................................................ 12
5.6 Symmetric-Cost Special Case .......................................................................................... 13
5.7 Worked Numerical Stackelberg Example .......................................................................... 14
6. Cournot and Stackelberg Comparison ................................................................................ 15
7. Common Mathematical Mistakes ......................................................................................... 17
8. Chapter Summary .................................................................................................................. 19
9. Key Formula Sheet ................................................................................................................ 20
10. Practice Problems................................................................................................................ 21
11. Model Answers ..................................................................................................................... 22

Mr. Farouk Saleh | +201062415129 | Page 2

1. Introduction to Chapter 9

An oligopoly is a market dominated by a small number of firms, each large enough that its
decisions noticeably affect the whole market. Because there are only a few sellers, the firms are
interdependent: the profit each firm earns depends not only on its own choices but also on the
choices of its rivals. This interdependence is what makes oligopoly the most strategically
interesting of all market structures, and it is why we need mathematics to analyse it.
In a competitive market a single firm is too small to influence price, so it simply takes the price as
given. A monopolist, by contrast, faces no rivals at all. An oligopolist sits between these two
extremes: it has some power over price, but it must constantly anticipate how its competitors will
react. A price cut that would be safe for a monopolist might trigger a damaging price war in an
oligopoly. For this reason, every oligopolist must reason not just about demand and cost, but about
the behaviour of the other firms.
Economists study this behaviour using two broad families of models. In quantity competition,
firms decide how much to produce, and the market price then adjusts so that total output is sold.
In price competition, firms set prices directly and let quantity demanded follow. This chapter
concentrates on quantity competition, which is described by the two most important models in the
theory of oligopoly:
• The Cournot model – two firms choose their output levels simultaneously, each treating
the rival’s output as fixed.
• The Stackelberg model – one firm (the leader) chooses output first, and the other (the
follower) responds afterwards.
The two models share the same demand and cost structure but differ in one crucial respect: the
timing of decisions. That single difference produces strikingly different equilibrium outputs, prices
and profits, and it reveals a genuine first-mover advantage for the firm that commits early.
Mathematical models let us pin down these differences exactly – not with vague intuition, but with
precise reaction functions, equilibrium formulas and comparative results that can be checked and
reused. That precision is the whole purpose of this tutorial.

2. Learning Objectives

By the end of this tutorial you should be able to:

• State the assumptions of the Cournot duopoly model.
• Construct the total revenue and profit functions of both firms.
• Derive marginal revenue by differentiating total revenue.
• Derive the reaction (best-response) function of each firm and explain what it means.
• Solve the two reaction functions simultaneously to find the Cournot–Nash equilibrium.
• Calculate equilibrium output, market price and profit under Cournot.
• State the assumptions of the Stackelberg model and solve it by backward induction.
• Derive the follower’s reaction function and the leader’s optimal output.
• Compare the Cournot and Stackelberg equilibrium outcomes and explain the first-mover
advantage.

Mr. Farouk Saleh | +201062415129 | Page 3

3. Mathematical Foundations and Notation

Before building the models we fix the notation that will be used throughout the chapter. Defining
every symbol now means that later derivations can move quickly without ambiguity.

Symbol Meaning

𝑃 Market price of the (single, homogeneous) product

𝑄 Total market output

𝑄 Output produced by Firm A

𝑄 Output produced by Firm B

𝑎 Vertical intercept of the inverse demand function (the choke price)

𝑏 Slope parameter of inverse demand (how fast price falls as output rises)

𝑐, 𝑐 , 𝑐 Constant marginal (and average) costs

𝑇𝑅 , 𝑇𝐶 Total revenue and total cost of a firm

𝜋 Profit of a firm

𝑀𝑅 , 𝑀𝐶 Marginal revenue and marginal cost of a firm

The inverse market demand function

Demand tells us how much consumers buy at each price. The inverse demand function turns this
around and expresses price as a function of total quantity:
𝑃 = 𝑎 − 𝑏𝑄
Here a is the price at which quantity demanded falls to zero (the highest price the market will bear),
and b measures how steeply the price must fall to sell additional units. We require 𝑏 > 0, so that
demand slopes downward: selling more output can only be achieved at a lower price.
In a duopoly the total quantity supplied is the sum of the two firms’ outputs:
𝑄 = 𝑄 + 𝑄
Substituting this into the inverse demand function gives the form we use for the rest of the chapter:
𝑃 = 𝑎 − 𝑏(𝑄 + 𝑄 )

KEY POINT — Why price depends on both firms

The single most important feature of this equation is that the price received by either firm
depends on the combined output of both firms. If Firm B floods the market, the price falls for
Firm A as well. This shared price is the mathematical source of all strategic interaction in
oligopoly.

Mr. Farouk Saleh | +201062415129 | Page 4

The profit-maximisation condition
Every firm in this chapter chooses output to maximise profit, defined as total revenue minus total
cost:
𝜋 = 𝑇𝑅 − 𝑇𝐶
There are two equivalent ways to find the profit-maximising output. The first is the familiar rule MR
= MC. The second is to differentiate the profit function directly and set the derivative to zero. These
are the same statement, because:
𝑑𝜋 𝑑𝑇𝑅 𝑑𝑇𝐶
= − = 𝑀𝑅 − 𝑀𝐶
𝑑𝑄 𝑑𝑄 𝑑𝑄
Setting this derivative equal to zero gives 𝑀𝑅 − 𝑀𝐶 = 0, i.e. 𝑀𝑅 = 𝑀𝐶 . Throughout the
tutorial we use whichever form is more convenient, but they always give the same answer.

Mr. Farouk Saleh | +201062415129 | Page 5

4. Section 9.1 – Cournot Duopoly: Reaction Functions and
Equilibrium

4.1 Assumptions of the Cournot Model

The Cournot model rests on a specific set of assumptions. Keeping them clearly in mind prevents
most of the errors students make later.

Assumptions of the basic Cournot model

1. There are two firms, A and B (a duopoly).
2. Both firms produce the same homogeneous product, so consumers see one market
price.
3. The firms compete by choosing quantities, not prices.
4. Each firm chooses its output simultaneously, without observing the rival’s choice.
5. Each firm treats the rival’s output as a fixed number when choosing its own output.
6. The market price depends on the combined output of both firms.
7. Both firms know the market demand and cost conditions (complete information).
8. In the basic model the two firms have identical constant marginal costs.

We use the inverse demand function derived above and the simplest possible cost functions, in
which each unit costs a constant 𝑐 to produce:
𝑃 = 𝑎 − 𝑏(𝑄 + 𝑄 ) 𝑇𝐶 = 𝑐𝑄 𝑇𝐶 = 𝑐𝑄
Because total cost is exactly 𝑐 multiplied by output, the extra cost of one more unit – the marginal
cost – is simply the constant 𝑐 for each firm:
𝑑𝑇𝐶
𝑀𝐶 = = 𝑐 𝑀𝐶 = 𝑐
𝑑𝑄

4.2 Total Revenue of Firm A

Total revenue is price multiplied by the firm’s own quantity sold. For Firm A:
𝑇𝑅 = 𝑃 × 𝑄
The price 𝑃 is common to both firms, so we substitute the inverse demand function directly:
𝑇𝑅 = [ 𝑎 − 𝑏(𝑄 + 𝑄 ) ] 𝑄
Now expand the bracket, multiplying every term inside by 𝑄 one step at a time:
𝑇𝑅 = 𝑎𝑄 − 𝑏𝑄 − 𝑏𝑄 𝑄

KEY POINT — Interpretation

Firm A’s revenue contains the term −𝑏𝑄 𝑄 . This is the mathematical fingerprint of oligopoly:
Firm A’s revenue falls when Firm B produces more, even though Firm A has changed nothing.
The two firms are linked through the shared market price.

Mr. Farouk Saleh | +201062415129 | Page 6

4.3 Marginal Revenue of Firm A
Marginal revenue is the rate at which total revenue changes as Firm A produces one more unit.
We obtain it by differentiating 𝑇𝑅 with respect to 𝑄 . The essential point is that Firm A treats 𝑄
as a constant, because under the Cournot assumptions it cannot influence Firm B’s choice.
𝜕𝑇𝑅 𝜕(𝑎𝑄 − 𝑏𝑄 − 𝑏𝑄 𝑄 )
𝑀𝑅 = =
𝜕𝑄 𝜕𝑄
Differentiate term by term:
• The derivative of 𝑎𝑄 with respect to 𝑄 is 𝑎 (a constant times 𝑄 ).
• The derivative of −𝑏𝑄 is −2𝑏𝑄 (bring down the power 2).
• The derivative of −𝑏𝑄 𝑄 is −𝑏𝑄 , because 𝑄 is held constant, so −𝑏𝑄 behaves like a fixed
coefficient on 𝑄 .
Collecting the three results gives Firm A’s marginal revenue:

𝑀𝑅 = 𝑎 − 2𝑏𝑄 − 𝑏𝑄

4.4 Reaction Function of Firm A

At the profit-maximising output, marginal revenue equals marginal cost. Setting 𝑀𝑅 = 𝑀𝐶 with
𝑀𝐶 = 𝑐:
𝑎 − 2𝑏𝑄 − 𝑏𝑄 = 𝑐
We now make 𝑄 the subject, moving one term at a time. First take the 𝑄 term to one side and
everything else to the other:
2𝑏𝑄 = 𝑎 − 𝑐 − 𝑏𝑄
Then divide every term on both sides by 2𝑏:

𝑎 − 𝑐 𝑄
𝑄 = −
2𝑏 2

DEFINITION — Reaction (best-response) function

A reaction function gives a firm’s profit-maximising output for every possible output the rival
might choose. The equation above is Firm A’s reaction function: hand it any value of 𝑄 and it
returns Firm A’s best reply.

Notice the coefficient of 𝑄 is −½, which is negative. Economically, if Firm B raises its output, total
market supply rises and the price falls; Firm A responds by cutting its own output to avoid
depressing the price further. Every extra unit Firm B produces induces Firm A to withdraw half a
unit. The two outputs are strategic substitutes.

Mr. Farouk Saleh | +201062415129 | Page 7

4.5 Total Revenue, Marginal Revenue and Reaction Function of Firm B
We now repeat the derivation in full for Firm B. Begin with its total revenue:
𝑇𝑅 = 𝑃 × 𝑄 = [ 𝑎 − 𝑏(𝑄 + 𝑄 ) ] 𝑄
Expanding the bracket:
𝑇𝑅 = 𝑎𝑄 − 𝑏𝑄 𝑄 − 𝑏𝑄
Differentiate with respect to 𝑄 , treating 𝑄 as constant:
𝑀𝑅 = 𝑎 − 𝑏𝑄 − 2𝑏𝑄
Set marginal revenue equal to marginal cost, 𝑀𝑅 = 𝑐:
𝑎 − 𝑏𝑄 − 2𝑏𝑄 = 𝑐
Rearrange to isolate 𝑄 : move the 𝑄 term across, then divide by 2𝑏:
2𝑏𝑄 = 𝑎 − 𝑐 − 𝑏𝑄

𝑎 − 𝑐 𝑄
𝑄 = −
2𝑏 2
Firm B’s reaction function has exactly the same shape as Firm A’s, with the roles of the two firms
swapped. It tells us Firm B’s best output for any level of 𝑄 , and it too slopes downward: the more
Firm A produces, the less Firm B wishes to produce.

4.6 The Cournot–Nash Equilibrium

The market settles at a point where both firms are simultaneously playing their best response. At
such a point:
• Firm A is choosing its best reply to Firm B’s output; and
• Firm B is choosing its best reply to Firm A’s output; so
• neither firm can raise its own profit by unilaterally changing output.
This is the Cournot–Nash equilibrium, and it occurs where the two reaction functions intersect.
We solve them together:
𝑎 − 𝑐 𝑄
𝑄 = − (𝐴)
2𝑏 2
𝑎 − 𝑐 𝑄
𝑄 = − (𝐵)
2𝑏 2
Substitute equation (B) into equation (A) wherever 𝑄 appears:
𝑎 − 𝑐 1 𝑎 − 𝑐 𝑄
𝑄 = − ( − )
2𝑏 2 2𝑏 2
Distribute the across the bracket:

𝑎 − 𝑐 𝑎 − 𝑐 𝑄
𝑄 = − +
2𝑏 4𝑏 4
( )
Combine the two constant terms. Since = , their difference is . Then collect the 𝑄
terms on the left:
𝑄 𝑎 − 𝑐 3 𝑎 − 𝑐
𝑄 − = ⇒ 𝑄 =
4 4𝑏 4 4𝑏

Mr. Farouk Saleh | +201062415129 | Page 8

Finally multiply both sides by . The 4 cancels and we obtain Firm A’s equilibrium output:

𝑎 − 𝑐
𝑄∗ =
3𝑏
By the symmetry of the two reaction functions, Firm B’s equilibrium output is identical:

𝑎 − 𝑐
𝑄∗ =
3𝑏
The equilibrium quantities are equal because the firms are mirror images: same demand, same
constant cost, same simultaneous timing. With nothing to distinguish them, they must produce the
same amount.

4.7 Total Cournot Output

Adding the two equilibrium quantities gives total market output:
𝑎 − 𝑐 𝑎 − 𝑐
𝑄∗ = 𝑄 ∗ + 𝑄 ∗ = +
3𝑏 3𝑏

2(𝑎 − 𝑐)
𝑄∗ =
3𝑏

4.8 Cournot Equilibrium Price

Substitute total output into the inverse demand function 𝑃 = 𝑎 − 𝑏𝑄:
2(𝑎 − 𝑐)
𝑃∗ = 𝑎 − 𝑏 ·
3𝑏
The 𝑏 in the numerator cancels the 𝑏 in the denominator:
2(𝑎 − 𝑐) 3𝑎 − 2(𝑎 − 𝑐) 3𝑎 − 2𝑎 + 2𝑐
𝑃∗ = 𝑎 − = =
3 3 3

𝑎 + 2𝑐
𝑃∗ =
3

KEY POINT — Price lies above marginal cost

Because 𝑎 > 𝑐 in any sensible market, 𝑃∗ = exceeds 𝑐. The two firms therefore earn
positive profit: unlike perfect competition, Cournot price is above marginal cost. But it is lower
than the monopoly price, so competition between the two firms benefits consumers relative to
monopoly.

Mr. Farouk Saleh | +201062415129 | Page 9

4.9 Cournot Equilibrium Profit
Each firm’s profit is price minus unit cost, multiplied by its output. Using 𝜋 = (𝑃∗ − 𝑐)𝑄 ∗ :
𝑎 + 2𝑐 𝑎 − 𝑐
𝜋 = ( − 𝑐) ·
3 3𝑏
Simplify the bracket first: − 𝑐 = = . Substituting back:

𝑎 − 𝑐 𝑎 − 𝑐 (𝑎 − 𝑐)
𝜋 = · =
3 3𝑏 9𝑏

(𝑎 − 𝑐)
𝜋 = 𝜋 =
9𝑏
Both firms earn the same profit, again because the model is symmetric. The profit rises with the
size of the market gap (𝑎 − 𝑐) and falls as demand becomes more price-sensitive (larger 𝑏).

4.10 Graphical Interpretation of the Reaction Functions

The two reaction functions can be drawn on a single diagram, with Firm A’s output on the horizontal
axis and Firm B’s output on the vertical axis. Each line shows one firm’s best response to the other.
The point where they cross is the Cournot–Nash equilibrium 𝐸 – the only pair of outputs at which
both firms are simultaneously satisfied.

Figure 4.1 — Cournot reaction functions and the Nash equilibrium E.

Reading the diagram:

• Firm A’s reaction line meets the 𝑄 -axis at : this is Firm A’s output when Firm B produces
nothing (the monopoly output).
• It meets the 𝑄 -axis at : the level of Firm B’s output at which Firm A is driven out of the
market entirely.
• Both lines slope downward, capturing the fact that more output by one firm calls for less by
the other.
• They intersect at 𝐸 = ( , ), the symmetric Cournot–Nash equilibrium.

Mr. Farouk Saleh | +201062415129 | Page 10

4.11 Fully Worked Numerical Cournot Example

WORKED EXAMPLE — Cournot with a = 100, b = 2, c = 20

Suppose the inverse demand is 𝑃 = 100 − 2𝑄 (so 𝑎 = 100, 𝑏 = 2) and each firm has
constant marginal cost 𝑐 = 20.

Step 1 – Reaction functions. Substitute the numbers into the general reaction functions 𝑄 =
− :

100 − 20 𝑄
𝑄 = − = 20 − 0.5 𝑄
2 × 2 2
𝑄 = 20 − 0.5 𝑄
Step 2 – Solve simultaneously. Substitute the expression for 𝑄 into the equation for 𝑄 :
𝑄 = 20 − 0.5(20 − 0.5 𝑄 ) = 20 − 10 + 0.25 𝑄
𝑄 − 0.25 𝑄 = 10 ⇒ 0.75 𝑄 = 10 ⇒ 𝑄 ∗ = 13.33

By symmetry 𝑄 ∗ = 13.33. (Check against the formula: = = 13.33. ✓)

Step 3 – Total output, price and profit.

𝑄 ∗ = 13.33 + 13.33 = 26.67
𝑃∗ = 100 − 2(26.67) = 100 − 53.33 = 46.67
𝜋 = (𝑃∗ − 𝑐) 𝑄∗ = (46.67 − 20)(13.33) = 26.67 × 13.33 = 355.56

𝑄 ∗ = 𝑄 ∗ = 13.33, 𝑃∗ = 46.67, 𝜋 = 𝜋 = 355.56

Interpretation. The two identical firms split the market equally, producing about 13.3 units each
for a total of 26.7 units. The resulting price of 46.67 sits well above the marginal cost of 20, so each
firm earns a healthy profit of about 355.6. Neither firm can do better by deviating: given the rival’s
output, 13.3 units is each firm’s best possible reply.

Mr. Farouk Saleh | +201062415129 | Page 11

5. Section 9.2 – The Stackelberg Model

5.1 Introduction and Main Assumptions

The Stackelberg model keeps the same demand and cost structure as Cournot and again features
quantity competition, but it changes one thing: the timing. Instead of moving at the same time, the
firms move sequentially. We call Firm A the leader and Firm B the follower.

Assumptions of the Stackelberg model

1. The leader (Firm A) chooses its output first and commits to it.
2. The follower (Firm B) observes the leader’s output before choosing its own.
3. The follower then picks its profit-maximising output given what the leader has done.
4. The leader anticipates the follower’s reaction and takes it into account before choosing.
5. The model is solved by backward induction: solve the follower’s problem first, then the
leader’s.

To make the model slightly more general we now allow the firms to have different constant
marginal costs, 𝑐 for the leader and 𝑐 for the follower:
𝑃 = 𝑎 − 𝑏(𝑄 + 𝑄 ) 𝑇𝐶 = 𝑐 𝑄 𝑇𝐶 = 𝑐 𝑄
𝑀𝐶 = 𝑐 𝑀𝐶 = 𝑐
Allowing different costs is realistic: the two firms may use different technology or face different
input prices. When we later set 𝑐 = 𝑐 = 𝑐 we recover the symmetric-cost special case.

KEY POINT — Why we solve the follower first

The leader can only choose wisely if it knows how the follower will react. So we must first work
out the follower’s response as a formula, then feed that formula into the leader’s decision.
Solving the last mover first and working backwards is called backward induction.

5.2 Solving the Follower’s Problem

The follower behaves exactly like a Cournot firm: it takes the leader’s output 𝑄 as a fixed number
and maximises its own profit. Its total revenue is:
𝑇𝑅 = 𝑃𝑄 = [ 𝑎 − 𝑏(𝑄 + 𝑄 ) ] 𝑄
Expanding the bracket:
𝑇𝑅 = 𝑎𝑄 − 𝑏𝑄 𝑄 − 𝑏𝑄
Differentiate with respect to 𝑄 (holding 𝑄 constant) to get marginal revenue:
𝑀𝑅 = 𝑎 − 𝑏𝑄 − 2𝑏𝑄
Set 𝑀𝑅 = 𝑀𝐶 = 𝑐 and solve for 𝑄 :
𝑎 − 𝑏𝑄 − 2𝑏𝑄 = 𝑐 ⇒ 2𝑏𝑄 = 𝑎 − 𝑐 − 𝑏𝑄

𝑎 − 𝑐 𝑄
𝑄 = −
2𝑏 2

Mr. Farouk Saleh | +201062415129 | Page 12

This is the follower’s reaction function. It is the key object in the whole model: it tells the leader
precisely how Firm B will respond to any output the leader might commit to. Everything else follows
from substituting it into the leader’s problem.

5.3 Solving the Leader’s Problem

The leader knows the follower’s reaction function, so it can predict 𝑄 as a function of its own 𝑄 .
It therefore substitutes that reaction function into the market price before choosing output. Start
from the demand function and replace 𝑄 :
𝑎 − 𝑐 𝑄
𝑃 = 𝑎 − 𝑏𝑄 − 𝑏𝑄 = 𝑎 − 𝑏𝑄 − 𝑏( − )
2𝑏 2
The 𝑏 multiplies into the bracket. The 𝑏 and partially cancel:

𝑎 − 𝑐 𝑏𝑄
𝑃 = 𝑎 − 𝑏𝑄 − +
2 2
Group the constant terms and the 𝑄 terms separately. The constants give 𝑎 − = , and
the 𝑄 terms give −𝑏𝑄 + = − . Hence the leader faces the simplified price:

𝑎 + 𝑐 𝑏𝑄
𝑃 = −
2 2
Now build the leader’s total revenue, 𝑇𝑅 = 𝑃 𝑄 :
𝑎 + 𝑐 𝑏𝑄
𝑇𝑅 = 𝑄 −
2 2
Differentiate with respect to 𝑄 to obtain the leader’s marginal revenue:
𝑎 + 𝑐
𝑀𝑅 = − 𝑏𝑄
2

KEY POINT — The leader’s marginal revenue is steeper

Compare this with a Cournot firm’s 𝑀𝑅 = 𝑎 − 2𝑏𝑄 − 𝑏𝑄 . Because the leader has already
folded the follower’s response into the price, its marginal revenue falls twice as fast in its own
output. This is exactly why the leader ends up producing more than a Cournot firm would.

Set 𝑀𝑅 = 𝑀𝐶 = 𝑐 and solve for the leader’s optimal output:

𝑎 + 𝑐 𝑎 + 𝑐
− 𝑏𝑄 = 𝑐 ⇒ 𝑏𝑄 = − 𝑐
2 2
Write the right-hand side over a common denominator, , then divide by 𝑏:

𝑎 + 𝑐 − 2𝑐
𝑄∗ =
2𝑏

Mr. Farouk Saleh | +201062415129 | Page 13

5.4 Finding the Follower’s Output
To find how much the follower actually produces, substitute the leader’s optimal output back into
the follower’s reaction function:
𝑎 − 𝑐 1 𝑎 + 𝑐 − 2𝑐
𝑄∗ = − ·
2𝑏 2 2𝑏
( )
Put both terms over the common denominator 4𝑏. The first term becomes ; the second
becomes . Subtracting:

2(𝑎 − 𝑐 ) − (𝑎 + 𝑐 − 2𝑐 )
𝑄∗ =
4𝑏
Expand the numerator carefully: 2𝑎 − 2𝑐 − 𝑎 − 𝑐 + 2𝑐 = 𝑎 − 3𝑐 + 2𝑐 . Therefore:

𝑎 − 3𝑐 + 2𝑐
𝑄∗ =
4𝑏
Both outputs are positive only if the numerators are positive. The leader produces a positive
amount when 𝑎 + 𝑐 > 2𝑐 (the leader is not too high-cost), and the follower survives when 𝑎 +
2𝑐 > 3𝑐 (the follower is not too high-cost relative to the leader). If a firm’s formula turns negative,
that firm would optimally produce zero and the market becomes a monopoly.

5.5 Total Output, Price and Profits (General Costs)

Adding the two outputs and simplifying over the denominator 4𝑏 gives total output:
3𝑎 − 2𝑐 − 𝑐
𝑄∗ = 𝑄∗ + 𝑄∗ =
4𝑏
Substituting total output into 𝑃 = 𝑎 − 𝑏𝑄 and simplifying yields the market price:

𝑎 + 2𝑐 + 𝑐
𝑃∗ =
4
The leader’s profit is 𝜋 = (𝑃∗ − 𝑐 )𝑄∗ . Using 𝑃 ∗ − 𝑐 = , which is exactly half of 𝑏𝑄 ∗ ,
the profit simplifies neatly:

(𝑎 + 𝑐 − 2𝑐 )
𝜋 =
8𝑏
The follower’s profit is 𝜋 = (𝑃∗ − 𝑐 )𝑄∗ . Following the same steps:

(𝑎 − 3𝑐 + 2𝑐 )
𝜋 =
16𝑏
The leader receives a first-mover advantage. By committing to a large output first, it forces the
follower to accommodate by producing less. The leader cannot be undercut on this commitment,
so it captures the larger share of the market and the larger share of profit – an advantage that
comes purely from moving first, not from any cost or demand difference.

Mr. Farouk Saleh | +201062415129 | Page 14

5.6 Symmetric-Cost Stackelberg Model
The most-quoted version of the Stackelberg model sets both marginal costs equal, 𝑐 = 𝑐 = 𝑐.
Substituting into the general results collapses them to very clean forms. For the leader:
𝑎 + 𝑐 − 2𝑐 𝑎 − 𝑐
𝑄∗ = =
2𝑏 2𝑏
and for the follower:
𝑎 − 3𝑐 + 2𝑐 𝑎 − 𝑐
𝑄∗ = =
4𝑏 4𝑏
𝑎 − 𝑐 𝑎 − 𝑐
𝑄∗ = 𝑄∗ =
2𝑏 4𝑏
Total output, price and profits then follow. Adding the outputs:
𝑎 − 𝑐 𝑎 − 𝑐 3(𝑎 − 𝑐)
𝑄∗ = + =
2𝑏 4𝑏 4𝑏
3(𝑎 − 𝑐) 𝑎 + 3𝑐
𝑃∗ = 𝑎 − 𝑏 · =
4𝑏 4
(𝑎 − 𝑐) (𝑎 − 𝑐)
𝜋 = (𝑃∗ − 𝑐) 𝑄∗ = 𝜋 =
8𝑏 16𝑏

KEY POINT — Why the leader produces more than the follower
The leader’s output is exactly twice the follower’s . By moving first the leader behaves
like a monopolist on the residual demand it leaves after anticipating the follower’s cut-back.
The follower, forced to react to an aggressive leader, scales down to half the leader’s output –
( ) ( )
and earns only a quarter of the leader’s profit, since is one-quarter of .

Mr. Farouk Saleh | +201062415129 | Page 15

5.7 Fully Worked Numerical Stackelberg Example

WORKED EXAMPLE — Stackelberg with a = 100, b = 2, cₐ = cᵦ = 20

We reuse the demand and costs from the Cournot example – 𝑃 = 100 − 2𝑄 with 𝑐 = 𝑐 =
20 – so the two models can be compared directly. Firm A leads; Firm B follows.

Step 1 – Follower’s reaction function. Using 𝑄 = − :

100 − 20 𝑄
𝑄 = − = 20 − 0.5 𝑄
2 × 2 2
Step 2 – Leader’s optimal output. Using 𝑄 ∗ = :

100 + 20 − 2(20) 100 + 20 − 40 80

𝑄∗ = = = = 20
2 × 2 4 4
Step 3 – Follower’s output. Substitute 𝑄 ∗ = 20 into the follower’s reaction function:
𝑄 ∗ = 20 − 0.5(20) = 20 − 10 = 10
Step 4 – Total output, price and profits.
𝑄 ∗ = 20 + 10 = 30
𝑃 ∗ = 100 − 2(30) = 40
𝜋 = (40 − 20)(20) = 20 × 20 = 400
𝜋 = (40 − 20)(10) = 20 × 10 = 200

𝑄∗ = 20, 𝑄 ∗ = 10, 𝑃∗ = 40, 𝜋 = 400, 𝜋 = 200

Interpretation. The leader produces twice as much as the follower (20 versus 10). Because it
commits first, it seizes the larger share of the market and earns double the follower’s profit.
Compared with the Cournot outcome (13.3 each), total output has risen from 26.7 to 30, so the
market price has fallen from 46.67 to 40 – good news for consumers. The leader is clearly better
off (400 versus 355.6 under Cournot), while the follower is worse off (200 versus 355.6). This is
the first-mover advantage in action.

Mr. Farouk Saleh | +201062415129 | Page 16

6. Cournot and Stackelberg Comparison

The two models share identical demand and cost assumptions and differ only in timing. That single
difference – simultaneous versus sequential moves – drives every difference in outcomes below.

Feature Cournot Stackelberg

Timing of decisions Simultaneous Sequential (leader, then follower)
Strategic roles Both firms equal Leader and follower
Method of solution Solve reaction functions together Backward induction
𝑎 − 𝑐 𝑎 − 𝑐
Firm A output
3𝑏 2𝑏
𝑎 − 𝑐 𝑎 − 𝑐
Firm B output
3𝑏 4𝑏
2(𝑎 − 𝑐) 3(𝑎 − 𝑐)
Total output
3𝑏 4𝑏
𝑎 + 2𝑐 𝑎 + 3𝑐
Market price
3 4
First-mover advantage None (symmetric) Yes, to the leader
Equilibrium concept Nash equilibrium Subgame-perfect equilibrium

In words: under Cournot the two firms are perfectly symmetric and split the market evenly. Under
Stackelberg the leader exploits its ability to commit first. It produces the larger quantity , which
forces the follower down to – half the leader’s output. The follower is not naive; it is playing its
genuine best response, but that best response is to accommodate the leader’s aggressive output.

Numerical comparison (a = 100, b = 2, c = 20)

Quantity Cournot Stackelberg
Output per firm A / leader 13.33 20.00
Output firm B / follower 13.33 10.00
Total market output 26.67 30.00
Market price 46.67 40.00
Profit, firm A / leader 355.56 400.00
Profit, firm B / follower 355.56 200.00
Industry profit 711.11 600.00

Three comparisons stand out. First, total output is higher under Stackelberg (30 > 26.67), so the
price is lower (40 < 46.67). Consumers are better off when one firm leads. Second, the leader
gains and the follower loses: moving first is worth 44.4 in extra profit to the leader, while the
follower gives up 155.6. Third, industry profit falls (600 < 711.11), because the leader’s
aggressive expansion makes the market more competitive overall. The gain to the leader is smaller
than the loss to the follower, so the firms jointly earn less than under Cournot even though one of
them does better.
The economic intuition is commitment. In Cournot, neither firm can commit to a large output,
because a large output would be undercut in the simultaneous game. In Stackelberg the leader

Mr. Farouk Saleh | +201062415129 | Page 17

can commit, and a credible commitment to produce a lot shifts the outcome in the leader’s favour.
This is why the ability to move first – to build the factory, sign the supply contract, or announce
capacity before rivals – is genuinely valuable.

7. Common Mathematical Mistakes

These are the errors that most often cost marks. Read each one and note the correction.

1. Forgetting that total output is the sum of both firms.

Incorrect: writing 𝑃 = 𝑎 − 𝑏𝑄 in Firm A’s revenue, ignoring Firm B.
Correct: always use 𝑃 = 𝑎 − 𝑏(𝑄 + 𝑄 ) – price depends on both outputs.

2. Multiplying demand by the wrong firm’s output.

Incorrect: computing 𝑇𝑅 = 𝑃𝑄 .
Correct: a firm’s revenue uses its own quantity: 𝑇𝑅 = 𝑃𝑄 .

3. Differentiating with respect to the wrong variable.

Incorrect: finding 𝑀𝑅 by differentiating 𝑇𝑅 with respect to 𝑄 .
Correct: differentiate with respect to the firm’s own choice variable: 𝑀𝑅 = 𝜕𝑇𝑅 /𝜕𝑄 .

4. Not treating the rival’s output as constant.

Incorrect: differentiating −𝑏𝑄 𝑄 and getting a term in 𝑄 ’s derivative.
Correct: under Cournot 𝑄 is fixed, so 𝜕(−𝑏𝑄 𝑄 )/𝜕𝑄 = −𝑏𝑄 .

5. Losing the factor of 2 when differentiating a squared term.

Incorrect: writing 𝜕(−𝑏𝑄 )/𝜕𝑄 = −𝑏𝑄 .
Correct: the power comes down: 𝜕(−𝑏𝑄 )/𝜕𝑄 = −2𝑏𝑄 .

6. Confusing total revenue with profit.

Incorrect: maximising 𝑇𝑅 and reporting that as profit.
Correct: profit subtracts cost: 𝜋 = 𝑇𝑅 − 𝑇𝐶 . Always deduct total cost.

7. Mixing up the two models’ timing.

Incorrect: solving both reaction functions simultaneously in a Stackelberg question.
Correct: Cournot is simultaneous (solve together); Stackelberg is sequential (backward
induction).

Mr. Farouk Saleh | +201062415129 | Page 18

8. Substituting the leader’s output too early.
Incorrect: plugging a number for 𝑄 before deriving the follower’s reaction function.
Correct: first derive 𝑄 = 𝑓(𝑄 ) as a formula, then substitute it into the leader’s problem.

9. Not checking that quantities are positive.

Incorrect: reporting a negative 𝑄 ∗ as a valid answer.
Correct: if a formula gives a negative output, that firm produces zero and the market is a
monopoly – check the sign.

Mr. Farouk Saleh | +201062415129 | Page 19

8. Chapter Summary

This chapter analysed quantity competition between two firms using two models that differ only in
the timing of decisions.
• The Cournot model. Two firms choose output simultaneously, each treating the rival’s output
as fixed. Maximising profit gives each firm a reaction function – a downward-sloping best-
response rule – and the two reaction functions intersect at the Cournot–Nash equilibrium.
With identical costs the firms produce equally, 𝑄 ∗ = 𝑄∗ = , at a price above
marginal cost.
• Reaction functions. A reaction function shows a firm’s profit-maximising output for every
rival output. Its negative slope means the outputs are strategic substitutes: more from one
firm calls for less from the other.
• The Stackelberg model. One firm leads and the other follows. The model is solved by
backward induction: first derive the follower’s reaction function, then substitute it into the
leader’s problem so the leader chooses output anticipating the follower’s reply.
• Leader and follower outputs. With identical costs the leader produces and the follower
half of that, . The leader enjoys a first-mover advantage and earns more; the follower
earns less.
• The key difference. Because the leader can commit first, Stackelberg produces more total
output and a lower price than Cournot. Consumers gain, the leader gains, the follower loses,
and combined industry profit falls.

Mr. Farouk Saleh | +201062415129 | Page 20

9. Key Formula Sheet

All results assume inverse demand 𝑃 = 𝑎 − 𝑏(𝑄 + 𝑄 ) with 𝑏 > 0.

COURNOT — identical marginal cost c

𝑎 − 𝑐 2(𝑎 − 𝑐)
𝑄∗ = 𝑄∗ = 𝑄∗ =
3𝑏 3𝑏
𝑎 + 2𝑐 (𝑎 − 𝑐)
𝑃∗ = 𝜋 = 𝜋 =
3 9𝑏

STACKELBERG — different marginal costs (leader cₐ, follower cᵦ)

𝑎 + 𝑐 − 2𝑐 𝑎 − 3𝑐 + 2𝑐
𝑄∗ = 𝑄∗ =
2𝑏 4𝑏
3𝑎 − 2𝑐 − 𝑐 𝑎 + 2𝑐 + 𝑐
𝑄∗ = 𝑃∗ =
4𝑏 4
(𝑎 + 𝑐 − 2𝑐 ) (𝑎 − 3𝑐 + 2𝑐 )
𝜋 = 𝜋 =
8𝑏 16𝑏

STACKELBERG — identical marginal cost c

𝑎 − 𝑐 𝑎 − 𝑐 3(𝑎 − 𝑐)
𝑄∗ = 𝑄∗ = 𝑄∗ =
2𝑏 4𝑏 4𝑏
𝑎 + 3𝑐 (𝑎 − 𝑐) (𝑎 − 𝑐)
𝑃∗ = 𝜋 = 𝜋 =
4 8𝑏 16𝑏

Mr. Farouk Saleh | +201062415129 | Page 21

10. Practice Problems

Attempt all six problems before looking at the model answers. They progress from straightforward
derivations to a full comparison.
Problem 1
The inverse demand facing two firms is 𝑃 = 200 − 4𝑄 where 𝑄 = 𝑄 + 𝑄 . Derive Firm A’s
total revenue and marginal revenue as functions of 𝑄 and 𝑄 , treating 𝑄 as given.
Problem 2
Market demand is 𝑃 = 90 − 𝑄, with 𝑄 = 𝑄 + 𝑄 . Both firms have constant marginal cost 𝑐 =
18. Derive the two Cournot reaction functions.
Problem 3
Using the demand and cost from Problem 2 (𝑃 = 90 − 𝑄, 𝑐 = 18), calculate the complete
Cournot equilibrium: each firm’s output, total output, market price, and each firm’s profit.
Problem 4
Two firms face 𝑃 = 120 − 2𝑄 and both have marginal cost 𝑐 = 24. Firm A is the Stackelberg
leader. Calculate the leader’s output, the follower’s output, total output, market price, and both
profits.
Problem 5
Two firms face 𝑃 = 100 − 𝑄. The Stackelberg leader (Firm A) has marginal cost 𝑐 = 10; the
follower (Firm B) has 𝑐 = 20. Calculate both outputs, total output, market price, and both profits.
Problem 6
For the market 𝑃 = 90 − 𝑄 with common marginal cost 𝑐 = 18, compare the Cournot outcome
from Problem 3 with the Stackelberg outcome in which Firm A leads. Compute the Stackelberg
outputs, price and industry profit, and explain which model gives the lower price and what happens
to the follower’s profit relative to Cournot.

Mr. Farouk Saleh | +201062415129 | Page 22

11. Model Answers

Model Answer 1
Total revenue is price times own output, 𝑇𝑅 = 𝑃 𝑄 . Substitute 𝑃 = 200 − 4(𝑄 + 𝑄 ):
𝑇𝑅 = [200 − 4(𝑄 + 𝑄 )] 𝑄 = 200𝑄 − 4𝑄 − 4𝑄 𝑄
Differentiate with respect to 𝑄 , treating 𝑄 as constant:

𝑀𝑅 = 200 − 8𝑄 − 4𝑄
Interpretation: marginal revenue starts at 200 and falls as Firm A expands; it also falls when Firm
B produces more, because extra output from either firm depresses the shared price.

Model Answer 2
Here 𝑎 = 90, 𝑏 = 1, 𝑐 = 18. Firm A’s revenue is 𝑇𝑅 = [90 − (𝑄 + 𝑄 )]𝑄 = 90𝑄 − 𝑄 −
𝑄 𝑄 . Differentiating, 𝑀𝑅 = 90 − 2𝑄 − 𝑄 . Set 𝑀𝑅 = 18:
90 − 2𝑄 − 𝑄 = 18 ⇒ 2𝑄 = 72 − 𝑄

𝑄 𝑄
𝑄 = 36 − 𝑄 = 36 −
2 2
The second reaction function follows by the identical argument for Firm B. Each firm’s best output
falls by half a unit for every extra unit produced by the rival.

Model Answer 3
Substitute Firm B’s reaction function into Firm A’s:
1 𝑄 𝑄 𝑄
𝑄 = 36 − (36 − ) = 36 − 18 + = 18 +
2 2 4 4
3
𝑄 = 18 ⇒ 𝑄 ∗ = 24 ⇒ 𝑄 ∗ = 24
4
Total output, price and profit:
𝑄 ∗ = 48, 𝑃 ∗ = 90 − 48 = 42, 𝜋 = (42 − 18)(24) = 576

𝑄 ∗ = 𝑄 ∗ = 24, 𝑃 ∗ = 42, 𝜋 = 𝜋 = 576

Interpretation: the symmetric firms split the market at 24 units each; price (42) exceeds marginal
( ) ²
cost (18), giving each firm a profit of 576. (Formula check: = = 576.)

Mr. Farouk Saleh | +201062415129 | Page 23

Model Answer 4
Identical-cost Stackelberg with 𝑎 = 120, 𝑏 = 2, 𝑐 = 24. Use the symmetric formulas 𝑄 ∗ =
and 𝑄 ∗ = :

120 − 24 96 96
𝑄∗ = = = 24 𝑄∗ = = 12
2 × 2 4 8
𝑄∗ = 36, 𝑃∗ = 120 − 2(36) = 48
𝜋 = (48 − 24)(24) = 576, 𝜋 = (48 − 24)(12) = 288

𝑄∗ = 24, 𝑄 ∗ = 12, 𝑃∗ = 48, 𝜋 = 576, 𝜋 = 288

Interpretation: the leader produces twice the follower’s output and earns twice its profit – the first-
mover advantage that comes from committing first.

Model Answer 5
Different-cost Stackelberg with 𝑎 = 100, 𝑏 = 1, 𝑐 = 10, 𝑐 = 20. Leader output:
𝑎 + 𝑐 − 2𝑐 100 + 20 − 20 100
𝑄∗ = = = = 50
2𝑏 2 2
Follower output:
𝑎 − 3𝑐 + 2𝑐 100 − 60 + 20 60
𝑄∗ = = = = 15
4𝑏 4 4
Total output, price and profits:
𝑄 ∗ = 65, 𝑃 ∗ = 100 − 65 = 35
𝜋 = (35 − 10)(50) = 1250, 𝜋 = (35 − 20)(15) = 225

𝑄 ∗ = 50, 𝑄∗ = 15, 𝑃 ∗ = 35, 𝜋 = 1250, 𝜋 = 225

Interpretation: the leader’s lower cost reinforces its first-mover advantage, so it produces far more
than the follower and earns roughly five-and-a-half times the follower’s profit. (Formula checks:
² ²
𝜋 = = 1250 and 𝜋 = = 225.)

Model Answer 6
The Cournot outcome (Problem 3) was total output 48, price 42, and profit 576 for each firm, so
industry profit 1152. Now solve the identical-cost Stackelberg case with 𝑎 = 90, 𝑏 = 1, 𝑐 = 18:
90 − 18 90 − 18
𝑄∗ = = 36, 𝑄∗ = = 18
2 4
𝑄∗ = 54, 𝑃∗ = 90 − 54 = 36
𝜋 = (36 − 18)(36) = 648, 𝜋 = (36 − 18)(18) = 324

𝑆𝑡𝑎𝑐𝑘𝑒𝑙𝑏𝑒𝑟𝑔: 𝑄 ∗ = 54, 𝑃 ∗ = 36, 𝜋 = 648, 𝜋 = 324, 𝑖𝑛𝑑𝑢𝑠𝑡𝑟𝑦 = 972

Comparison and interpretation: Stackelberg total output (54) exceeds Cournot (48), so the
Stackelberg price (36) is lower than the Cournot price (42) – consumers are better off when a
firm leads. The leader’s profit rises from 576 to 648, but the follower’s profit falls from 576 to
324: it must accommodate the leader’s larger output. Industry profit falls from 1152 to 972, because
the leader’s aggressive commitment makes the market more competitive. The lower price is a
direct consequence of the leader’s first-mover commitment to a larger quantity.

Mr. Farouk Saleh | +201062415129 | Page 24