Economics•Chapter 8•6 min read•Updated September 24, 2026

Microeconomics — Oligopoly and Monopolistic Competition

O
OiyoContributor
8/11

Microeconomics — In Oligopoly, Rivals’ Strategies Are Part of Your Demand

With only a few firms, no firm can take the price as given like a competitive firm, nor use the whole market demand alone like a monopolist. Each firm’s best choice depends on the others’ choices. This chapter keeps the same market demand, P=120−QP=120-Q, and a marginal cost of zero, and compares how outcomes change with the strategic variable firms use. The general theory of mixed strategies and repeated games is in Chapter 11.

1. Cournot: choosing quantities simultaneously

Two firms choose outputs q1q_1 and q2q_2 at the same time and the market sets the price P=120−(q1+q2)P=120-(q_1+q_2). Firm 1’s profit is π1=q1(120−q1−q2)\pi_1=q_1(120-q_1-q_2). Maximising it while taking the rival’s output as given:

∂π1∂q1=120−2q1−q2=0⇒q1=120−q22\frac{\partial \pi_1}{\partial q_1}=120-2q_1-q_2=0 \quad\Rightarrow\quad q_1=\frac{120-q_2}{2}

This is firm 1’s reaction function: the more the rival produces, the less I produce. Solving it together with firm 2’s symmetric reaction function gives q1=q2=40q_1=q_2=40, market output 80 and a price of 40 won. Each firm is best-responding to the other, so this point is a Nash equilibrium.

Cournot reaction functions
q₁ q₂ 40 40 R₁ R₂ Cournot (40, 40) Cartel Stackelberg (60, 30)

The Cournot equilibrium is where the two reaction functions cross. The cartel point is better for both firms but lies on neither reaction function.

Symmetric Cournot with n firms (demand P = a − Q, marginal cost c)
qi=a−cn+1,P=a+ncn+1q_i=\frac{a-c}{n+1},\qquad P=\frac{a+nc}{n+1}
n = 1 is monopoly; as n grows, the price converges to marginal cost c. This is the channel through which the number of firms reduces market power.

2. Stackelberg: moving first pays

If firm 1 sets its output first and firm 2 responds after observing it, firm 1 substitutes firm 2’s reaction function q2=(120−q1)/2q_2=(120-q_1)/2 into its own profit:

π1=q1(120−q1−120−q12)=q1⋅120−q12\pi_1=q_1\left(120-q_1-\frac{120-q_1}{2}\right)=q_1\cdot\frac{120-q_1}{2}

Maximising gives q1=60q_1=60, q2=30q_2=30 and a price of 30 won. The leader earns 1,800 won, more than the Cournot 1,600, and the follower’s profit falls to 900. Committing to a large output first makes the rival produce less. This is why hard-to-reverse choices such as capacity investment have strategic value.

3. Bertrand: competing on price

If two firms selling an identical product compete on price, undercutting the rival by just 1 won captures the whole market. As long as the price is above marginal cost, each has an incentive to shave it a little, so the equilibrium is P=MCP=MC. That two firms produce the competitive outcome is the Bertrand paradox.

The paradox is resolved under three conditions: capacity limits mean the cheaper firm cannot serve the whole market; product differentiation means a slightly higher price does not lose every customer; and in repeated interaction a price cut can be punished.

4. Collusion and the incentive to cheat

If the two firms agree to act like a monopolist, market output is 60, the price is 60 won and each produces 30 for a profit of 1,800 won — more than the Cournot 1,600. But if the rival produces 30, my best response is q1=(120−30)/2=45q_1=(120-30)/2=45. The price falls to 45 won, yet I earn 2,025 won and the rival’s profit drops to 1,350. In a one-shot game, the cartel does not hold.

Repetition changes this. Under a trigger strategy that reverts to Cournot forever after any deviation, the value of keeping the agreement is 1,800/(1−δ)1{,}800/(1-\delta) and the value of cheating is 2,025+1,600 δ/(1−δ)2{,}025+1{,}600\,\delta/(1-\delta).

Condition for sustaining collusion
δ≥πcheat−πcartelπcheat−πpunish=2,025−1,8002,025−1,600≈0.53\delta \ge \frac{\pi^{\text{cheat}}-\pi^{\text{cartel}}}{\pi^{\text{cheat}}-\pi^{\text{punish}}}=\frac{2{,}025-1{,}800}{2{,}025-1{,}600}\approx 0.53
If the future matters enough (δ large), the permanent loss from punishment outweighs the one-off gain from cheating and the cartel holds. Collusion is easier when interaction is frequent, cheating is detected quickly and there are few firms.

This is why competition authorities use leniency programmes, which reduce fines for the first firm to confess: they raise the payoff to defecting from within the cartel.

Comparison on the same demand (P = 120 − Q, MC = 0)
ModelMarket outputPriceProfit per firm
Monopoly / cartel6060 won1,800 (split between 2 firms)
Stackelberg9030 won1,800 / 900
Cournot8040 won1,600 / 1,600
Bertrand1200 won0 / 0

5. Monopolistic competition

Markets with many firms and free entry but slightly different products (restaurants, hair salons, clothing) are called monopolistic competition. Because of differentiation, each firm has its own downward-sloping demand curve and produces where MR=MCMR=MC. If firms make profits in the short run, similar products enter and reduce each firm’s demand. In the long run profit is zero where the demand curve is tangent to the average cost curve.

As a result, P=ATC>MCP=ATC>MC even in the long run. The price is above marginal cost and firms produce less than the output that minimises average cost, leaving excess capacity. Whether this is simply inefficient depends on how much people value variety; if they want many choices, slightly higher average cost may be a price worth paying.

Check your understanding

What is the Cournot equilibrium with demand P=100−QP=100-Q and two firms with marginal cost 10 won? From the formula qi=(100−10)/3=30q_i=(100-10)/3=30, market output 60, price 40 won and profit (40−10)×30=900(40-10)\times 30=900 won per firm. Under collusion, market output is 45 (the monopoly level), the price 55 won and each firm makes 22.5 units for a profit of 1,012.5 won, which is better for both; but the best response to a rival producing 22.5 is (90−22.5)/2=33.75(90-22.5)/2=33.75, so collusion does not hold in a one-shot game.

References

  • Hal Varian, Intermediate Microeconomics, ch. 27–28
  • Jean Tirole, The Theory of Industrial Organization, ch. 5–6
  • MIT OpenCourseWare, 14.01 Principles of Microeconomics
O

Oiyo

Editorial Desk

The OIYO editorial desk researches money, law, lifestyle, and self-understanding topics against primary sources and public statistics. Every piece carries source notes and is reviewed on a regular cycle for accuracy and usefulness.