Oligopoly & Game Theory - Microeconomics

SECTION 1

Historical Context & Motivation

Classical economics offered elegant models for two polar extremes—perfect competition, where countless small firms are price-takers, and monopoly, where a single seller dictates terms. Yet the vast majority of real-world industries—airlines, automobiles, telecommunications, big tech—inhabit neither extreme. They feature a small number of large, interdependent firms whose strategic decisions reverberate across the entire market. Economists needed a framework that captured this mutual interdependence, a situation where each firm's optimal choice depends on what its rivals choose. That need gave rise to the modern study of oligopoly and, ultimately, to the application of game theory to industrial organization.

1838 • Cournot's Duopoly Model

Augustin Cournot published Researches into the Mathematical Principles of the Theory of Wealth, introducing the first formal model of oligopoly in which two firms simultaneously choose quantities, anticipating the logic of equilibrium in strategic settings.

1883 • Bertrand's Price Competition

Joseph Bertrand critiqued Cournot by arguing that firms compete on price rather than quantity. His model showed that with homogeneous products and simultaneous price-setting, even two competitors can drive price down to marginal cost—a striking prediction known as the Bertrand paradox.

1934 • Stackelberg's Leader–Follower Model

Heinrich von Stackelberg extended Cournot's framework by allowing one firm to move first, committing to a quantity before the rival responds. This sequential structure formalized the concept of first-mover advantage in oligopoly.

1944 • Von Neumann & Morgenstern: Game Theory Is Born

John von Neumann and Oskar Morgenstern published Theory of Games and Economic Behavior, providing the mathematical foundations for analyzing strategic interaction in economics and beyond.

1950 • Nash Equilibrium

John Nash's doctoral dissertation at Princeton introduced the concept of Nash equilibrium—a strategy profile in which no player can unilaterally improve their payoff. This concept became the central solution concept for oligopoly analysis and earned Nash the 1994 Nobel Prize in Economics.

The central question this lesson addresses is deceptively simple: How do firms behave when their profits depend not only on their own decisions but also on the decisions of a few powerful rivals? Answering that question requires a toolkit that goes beyond simple supply-and-demand analysis. Game theory provides precisely that toolkit, transforming the problem of strategic rivalry into a structured framework of players, strategies, payoffs, and equilibria.

SECTION 2

Core Principles & Definitions

An oligopoly is a market structure characterized by a small number of firms that collectively supply the majority of industry output and recognize that their pricing, output, and investment decisions are strategically interdependent. Unlike perfect competition, entry barriers—whether from economies of scale, patents, or brand loyalty—are substantial enough to limit the number of effective competitors. Unlike monopoly, the presence of even one rival fundamentally alters decision-making because each firm must anticipate the likely responses of its competitors before committing to a strategy.

Few Dominant Sellers

The market is supplied by a handful of firms—typically two to ten—each large enough that its output decisions measurably affect market price. This concentration is often measured by the Herfindahl–Hirschman Index (HHI) or the four-firm concentration ratio (CR₄).

Strategic Interdependence

Each firm's optimal strategy depends on what rivals do. A price cut by one airline triggers matching cuts by competitors, so the payoff from any action is conditional. This mutual interdependence is the defining feature of oligopoly and the reason game theory is required.

Barriers to Entry

High fixed costs, intellectual property, regulatory licenses, or network effects prevent new entrants from eroding incumbents' market power. These barriers preserve the small-numbers structure over time.

Nash Equilibrium

A set of strategies—one for each firm—is a Nash equilibrium if no single firm can increase its profit by unilaterally changing its own strategy, holding rivals' strategies fixed. It is the primary solution concept for oligopoly models.

Dominant Strategy & the Prisoner's Dilemma

A dominant strategy yields the highest payoff regardless of rivals' actions. In the classic Prisoner's Dilemma, each player's dominant strategy leads to a collectively inferior outcome—a tension that mirrors oligopolistic price wars.

✦

KEY TAKEAWAY

Think of an oligopoly as a poker table with only a few players. In a crowded tournament (perfect competition), your individual bet barely moves the pot, so you simply play the odds. At a small, high-stakes table (oligopoly), every bet you make visibly shifts the dynamics—your rivals watch, react, and adjust. Game theory is the discipline of reading that table, systematically analyzing what each player knows, what each can do, and what each will likely choose given the choices of others.

SECTION 3

Visual Explanation — The Prisoner's Dilemma Payoff Matrix

The most celebrated illustration of strategic interaction is the Prisoner's Dilemma, adapted here to an oligopoly pricing context. Consider two rival firms—Firm A and Firm B—each independently deciding whether to maintain a high price (cooperate) or cut its price (defect). The diagram below presents the resulting payoff matrix, where each cell shows the profit pair (Firm A, Firm B) in millions of dollars.

The dashed yellow box marks the Nash equilibrium at (Low Price, Low Price), where each firm earns $6M. Both would prefer the cooperative outcome of (High Price, High Price) yielding $10M each, but the temptation to defect—capturing $15M while the rival earns only $2M—makes price-cutting a dominant strategy for both players.

Notice the structural tension embedded in this matrix. If Firm A holds a high price, Firm B can increase its profit from $10M to $15M by undercutting—so B prefers to defect. If Firm A has already cut its price, B still earns more by also cutting ($6M vs. $2M)—so B defects regardless of A's choice. By symmetry, the same logic applies to Firm A. The result is a dominant-strategy equilibrium in which both firms set low prices, even though mutual cooperation would leave both better off. This tension between individual rationality and collective welfare is the hallmark of oligopolistic competition and the reason cartels, though tempting, are inherently unstable.

SECTION 4

Mathematical Framework — Cournot & Bertrand Models

Game theory provides several canonical models of oligopoly, each distinguished by the strategic variable firms control (quantity vs. price) and the timing of moves (simultaneous vs. sequential). We begin with the Cournot model, the most widely used framework in industrial organization.

The Cournot Duopoly Model

In the Cournot model, two firms simultaneously choose quantities q₁ and q₂ of a homogeneous product. Total output Q = q₁ + q₂ determines the market price through an inverse demand function. Each firm maximizes its own profit, taking the rival's quantity as given.

INVERSE DEMAND

P = a − b(q₁ + q₂)

P = market price; a = demand intercept (maximum willingness to pay); b = slope of demand; q₁, q₂ = quantities chosen by Firm 1 and Firm 2.

FIRM 1 PROFIT

π₁ = P × q₁ − c × q₁ = (a − bq₁ − bq₂ − c) × q₁

π₁ = Firm 1 profit; c = constant marginal cost (assumed identical for both firms). Revenue equals price times quantity, and cost is linear in output.

To find Firm 1's best response, we take the first-order condition ∂π₁/∂q₁ = 0, treating q₂ as a parameter. This yields the best-response function (also called the reaction function), which expresses Firm 1's profit-maximizing quantity as a function of its rival's output.

BEST-RESPONSE FUNCTION (FIRM 1)

q₁* = (a − c) / (2b) − q₂ / 2

By symmetry, Firm 2's best-response function is q₂* = (a − c)/(2b) − q₁/2. The Nash–Cournot equilibrium is found at the intersection of both reaction functions.

COURNOT–NASH EQUILIBRIUM QUANTITIES

q₁* = q₂* = (a − c) / (3b)

Each firm produces one-third of the competitive output. Total industry output Q* = 2(a − c)/(3b), which is less than the competitive quantity (a − c)/b but more than the monopoly quantity (a − c)/(2b).

Bertrand Competition

The Bertrand model assumes firms compete on price rather than quantity. With homogeneous products and equal marginal costs, the unique Nash equilibrium is for both firms to set P = c, earning zero economic profit—the same outcome as perfect competition. This dramatic result, the Bertrand paradox, arises because any firm charging even slightly above marginal cost loses its entire market share to the undercutting rival. In practice, product differentiation, capacity constraints, and switching costs soften the paradox and sustain positive margins.

💡

Cournot vs. Bertrand — Why It Matters

The choice of strategic variable is not merely a modeling convenience—it reflects industry structure. Industries where capacity is committed first (e.g., airlines choosing fleet sizes) are better modeled by Cournot quantity competition. Industries where prices can be changed instantly (e.g., online retail) more closely resemble Bertrand price competition. Understanding which model applies is critical for strategic analysis and antitrust policy.

SECTION 5

Detailed Breakdown — Cournot Reaction Functions & Market Outcomes

The diagram below plots the best-response (reaction) functions for two Cournot duopolists on a quantity–quantity plane. The intersection of the two curves identifies the Nash–Cournot equilibrium, and the diagram also marks the monopoly, competitive, and collusive outcomes for comparison.

The downward-sloping reaction function RF₁ shows Firm 1's optimal quantity for every possible q₂; the reaction function RF₂ does the same for Firm 2. Their intersection is the Cournot–Nash equilibrium. Note that total output exceeds the collusive (monopoly-sharing) level but falls short of the perfectly competitive level—reflecting the oligopolist's intermediate market power.

Market Structure	Total Output Q*	Market Price P*	Industry Profit
Monopoly	(a − c) / (2b)	(a + c) / 2	Highest
Cournot Duopoly	2(a − c) / (3b)	(a + 2c) / 3	Intermediate
Bertrand Duopoly	(a − c) / b	c	Zero
Perfect Competition	(a − c) / b	c	Zero

Comparison of equilibrium outcomes under different market structures (linear demand, constant marginal cost)

The table reveals a striking pattern: as the number of competitors increases from one (monopoly) to two (Cournot) to many (competitive), total output rises monotonically while price and industry profits fall. In fact, the general Cournot result for n identical firms yields q* = (a − c)/[(n + 1)b] per firm. As n → ∞, the Cournot outcome converges to the perfectly competitive result—a powerful theoretical bridge between market structures.

SECTION 6

Worked Example — Cournot Equilibrium with Numbers

Consider two smartphone chip manufacturers—ChipCo and SiliTech—competing as Cournot duopolists. Market demand for chips is P = 120 − 2Q (where Q is total output in millions of units), and each firm has a constant marginal cost of c = $24 per unit with no fixed costs.

Finding the Cournot–Nash Equilibrium

1
Step 1 — Write the Profit Function

Let q₁ and q₂ denote the output of ChipCo and SiliTech respectively, with Q = q₁ + q₂. ChipCo's profit is: π₁ = (P − c) × q₁ = (120 − 2q₁ − 2q₂ − 24) × q₁ = (96 − 2q₁ − 2q₂) × q₁.
2
Step 2 — Derive the Best-Response Function

Take the first-order condition: ∂π₁/∂q₁ = 96 − 4q₁ − 2q₂ = 0. Solving for q₁ yields ChipCo's best-response function:

q₁* = 24 − q₂/2
3
Step 3 — Exploit Symmetry to Find Equilibrium

By symmetry (identical costs), SiliTech's best-response function is q₂* = 24 − q₁/2. At equilibrium, q₁ = q₂, so substitute: q₁ = 24 − q₁/2, which gives 3q₁/2 = 24.

q₁* = q₂* = 16 million units each
4
Step 4 — Calculate Equilibrium Price

Total output Q* = 16 + 16 = 32 million units. Substituting into the demand function: P* = 120 − 2(32) = 120 − 64.

P* = $56 per unit
5
Step 5 — Calculate Equilibrium Profits

Each firm's profit: π* = (56 − 24) × 16 = 32 × 16.

π₁* = π₂* = $512 million each
6
Step 6 — Compare to Monopoly and Competitive Outcomes

A monopolist would produce $Q_m$ = (120 − 24)/(2 × 2) = 24 million units, charge $P_m$ = $72, and earn π_m = $1,152M total. Under perfect competition, $Q_c$ = (120 − 24)/2 = 48 million units, $P_c$ = $24, and π = $0. The Cournot duopoly output (32M) falls between these benchmarks, and combined profit ($1,024M) is less than monopoly profit but far above zero—confirming the intermediate nature of oligopoly.

SECTION 7

Strengths, Limitations & Real-World Complications

Dimension	Strengths	Limitations
Predictive Power	Cournot predictions of output and price align well with industries featuring capacity commitments (e.g., oil, steel, airlines).	Assumes firms choose output simultaneously, which may not reflect sequential decision-making or real-time price adjustments.
Strategic Insight	Nash equilibrium captures mutual best-response logic, making it applicable far beyond economics—to negotiations, R&D races, and entry deterrence.	Multiple Nash equilibria can exist in more complex games, making unique predictions difficult without refinement criteria.
Collusion Analysis	The Prisoner's Dilemma framework explains why cartels are tempting but inherently unstable, informing antitrust enforcement strategies.	One-shot games understate the scope for tacit collusion in repeated interactions where firms can punish defectors over time.
Scalability	The n-firm Cournot model smoothly bridges monopoly and perfect competition, making it a unifying theoretical tool.	Assumes symmetric firms with identical costs and homogeneous products—assumptions that rarely hold exactly in practice.
Behavioral Realism	Game-theoretic models can incorporate information asymmetries, signaling, and commitment devices for richer analysis.	Firms may not behave as perfectly rational profit-maximizers; bounded rationality and organizational inertia can alter outcomes.

Evaluating classical oligopoly models for business strategy and policy analysis

✦

KEY TAKEAWAY

Oligopoly models are like flight simulators for business strategists—they simplify the real cockpit but capture the essential dynamics of interdependence, trade-offs, and competitive tension. No single model perfectly replicates reality, but together the Cournot, Bertrand, and Stackelberg frameworks form a versatile toolkit for diagnosing competitive dynamics, predicting merger effects, and designing pricing strategies. The key is matching the right model to the right industry context.

SECTION 8

Connections to Advanced Theory — Repeated Games & Beyond

The one-shot Prisoner's Dilemma predicts that firms always defect, yet real-world oligopolists sometimes sustain cooperation—or at least tacit coordination—for extended periods. The resolution lies in repeated games. When firms interact in the same market period after period with no known endpoint, the shadow of the future enables cooperative strategies that would be impossible in a single round. The Folk Theorem formalizes this insight: in an infinitely repeated game, any outcome that gives each player at least their one-shot Nash equilibrium payoff can be sustained as a subgame-perfect equilibrium, provided players are sufficiently patient (discount factor δ close to 1).

Feature	One-Shot Game	Repeated Game (Infinite Horizon)
Cooperation	Not sustainable—defection is dominant.	Sustainable via trigger strategies (e.g., grim trigger, tit-for-tat) if δ is high enough.
Punishment	No mechanism—game ends after one round.	Future punishment (price wars, output flooding) deters defection.
Equilibria	Unique dominant-strategy equilibrium (defect, defect).	Multiple equilibria; Folk Theorem guarantees a continuum of cooperative outcomes.
Policy Implication	Less concern about collusion.	Antitrust authorities must monitor for tacit collusion in stable, transparent markets.

One-shot versus repeated game dynamics in oligopoly

Beyond repeated games, advanced industrial organization draws on mechanism design (designing auctions and contracts to align incentives), Bayesian games (modeling incomplete information about rivals' costs or demand), and evolutionary game theory (where strategies evolve through adaptive learning rather than rational calculation). For business students, the immediate next step is typically the study of entry deterrence and contestable markets—examining how incumbents use excess capacity, limit pricing, or strategic R&D to discourage potential entrants.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain why mutual interdependence distinguishes oligopoly from both perfect competition and monopoly. In your answer, describe how a firm's pricing decision in an oligopoly differs from the pricing decision of a perfectly competitive firm and a monopolist.

Show answer

In perfect competition, each firm is a price-taker—its output is too small to affect the market price, so it simply produces where P = MC without considering rivals. A monopolist faces no rivals at all and sets price by equating marginal revenue to marginal cost. In an oligopoly, however, each firm is large enough that its pricing or output decisions visibly shift total supply and therefore market price. Crucially, each firm knows that its rivals will react to those shifts, so the profit-maximizing strategy depends on anticipated competitor responses. This mutual interdependence means oligopoly pricing cannot be solved with simple optimization; it requires strategic reasoning—the domain of game theory.

Question 2: BASIC CALCULATION

Two identical firms compete as Cournot duopolists facing inverse demand P = 200 − 4Q, where Q = q₁ + q₂. Each firm has a constant marginal cost of $40 and no fixed costs. Find the Cournot–Nash equilibrium quantity for each firm, the equilibrium market price, and each firm's profit.

Show answer

Using the Cournot formula q* = (a − c)/[3b] with a = 200, c = 40, and b = 4, each firm produces q* = (200 − 40)/(3 × 4) = 160/12 ≈ 13.33 units. Total output Q* = 26.67 units, and the equilibrium price is P* = 200 − 4(26.67) ≈ $93.33. Each firm's profit is π* = (93.33 − 40) × 13.33 ≈ $711.11. Note that combined industry profit ($1,422.22) is less than what a monopolist would earn:

Q_m

= 160/8 = 20,

P_m

= $120, π_m = (120 − 40) × 20 = $1,600. The Cournot outcome splits the difference between monopoly and competitive outcomes, as theory predicts.

Question 3: INTERMEDIATE

Suppose Firm 1 becomes the Stackelberg leader in the market described in Problem 2 (P = 200 − 4Q, c = $40). The leader commits to its quantity first, and the follower (Firm 2) observes the leader's choice before responding. Derive the Stackelberg equilibrium quantities, price, and profits. Compare Firm 1's profit as the Stackelberg leader to its profit under Cournot.

Show answer

First, derive Firm 2's reaction function from the Cournot setup: q₂ = (a − c)/(2b) − q₁/2 = 20 − q₁/2. Firm 1 (the leader) substitutes this into its own profit function: π₁ = (200 − 4q₁ − 4(20 − q₁/2) − 40) × q₁ = (200 − 4q₁ − 80 + 2q₁ − 40) × q₁ = (80 − 2q₁) × q₁. Taking the first-order condition: ∂π₁/∂q₁ = 80 − 4q₁ = 0, so q₁ = 20. The follower produces q₂ = 20 − 20/2 = 10. Total Q = 30, P = 200 − 120 = $80. Firm 1's profit = (80 − 40) × 20 = $800, versus $711.11 under Cournot—a clear first-mover advantage. Firm 2's profit = (80 − 40) × 10 = $400, lower than under Cournot. The leader gains at the follower's expense by credibly committing to a larger quantity.

Question 4: APPLIED

Two competing ride-sharing platforms, RideNow and GoFast, each independently decide whether to run a promotional pricing campaign (Low Price) or maintain standard pricing (High Price). Analysts estimate the following quarterly profit payoffs (in millions): Both High → ($50M, $50M); RideNow Low/GoFast High → ($70M, $20M); RideNow High/GoFast Low → ($20M, $70M); Both Low → ($30M, $30M). (a) Identify the Nash equilibrium. (b) Is there a dominant strategy for each firm? (c) Discuss why these platforms might still attempt to maintain high prices in practice.

Show answer

(a) The Nash equilibrium is (Low Price, Low Price) with payoffs ($30M, $30M). At this outcome, neither firm can improve its payoff by unilaterally switching: if RideNow deviates to High Price, it earns $20M instead of $30M. (b) Yes, Low Price is a dominant strategy for both firms. Regardless of the rival's choice, each firm earns more by choosing Low Price ($70M > $50M if the rival goes High; $30M > $20M if the rival goes Low). (c) In practice, ride-sharing platforms interact repeatedly quarter after quarter. If both firms adopt trigger strategies—maintaining high prices until one defects, then permanently switching to low prices—cooperation can be sustained if the present value of future cooperation exceeds the one-time gain from defection. Additionally, platform differentiation, brand loyalty, and switching costs reduce the severity of the Prisoner's Dilemma by softening the demand response to price cuts.

Question 5: CRITICAL THINKING

The Bertrand model predicts that even two firms competing on price will drive the market price down to marginal cost, yet empirical evidence from many duopoly industries (e.g., Boeing vs. Airbus, Coca-Cola vs. Pepsi) shows sustained positive economic profits. Critically evaluate at least three specific assumptions of the basic Bertrand model that, when relaxed, can explain the divergence between theory and observation. For each assumption, explain the mechanism through which relaxing it restores pricing power.

Show answer

Three key assumptions and their relaxation: (1) Homogeneous products — The basic Bertrand model assumes consumers view rivals' products as perfect substitutes. In reality, product differentiation (brand identity, taste, design, network effects) means a small price increase does not cause a firm to lose all its customers. Relaxing this assumption via a differentiated Bertrand model yields equilibrium prices above marginal cost because demand is less than perfectly elastic at the firm level. (2) Unlimited capacity — Bertrand assumes each firm can serve the entire market at marginal cost. When capacity is constrained (as with aircraft manufacturing), a firm that undercuts cannot steal all demand, reducing the incentive to undercut and supporting higher prices. Kreps and Scheinkman (1983) showed that capacity-constrained price competition with prior capacity investment yields Cournot outcomes. (3) Complete information and no switching costs — Bertrand assumes consumers costlessly switch and have perfect price information. In practice, search costs, contractual lock-in, loyalty programs, and brand habit create friction that insulates incumbents from aggressive price competition. Each of these departures from the idealized model introduces a source of market power that sustains positive profits even with very few competitors.

SUMMARY

Lesson Summary

Summary

An oligopoly is a market dominated by a few interdependent firms whose pricing and output decisions are strategically linked. Because each firm's payoff depends on its rivals' choices, analyzing oligopoly requires game theory—a formal framework of players, strategies, and payoffs. The central solution concept is the Nash equilibrium, in which no firm can unilaterally improve its profit. The Prisoner's Dilemma illustrates why individually rational firms may reach collectively inferior outcomes—each has a dominant strategy to defect, even though mutual cooperation would yield higher profits for all.

The Cournot model (simultaneous quantity choice) produces equilibrium output that lies between monopoly and competitive levels, with each firm's best-response function defining optimal output as a decreasing function of rivals' quantities. The Bertrand model (simultaneous price choice) yields the competitive price with homogeneous goods—the Bertrand paradox—while the Stackelberg model introduces sequential moves and first-mover advantage. In repeated games, the shadow of future interaction can sustain cooperation, explaining tacit collusion and informing antitrust policy.