Oligopoly

Oligopoly is a market structure in industrial organization and Business & Economics Theory in which an industry is supplied by a small number of relatively large firms—typically between two (duopoly) and a dozen major producers—protected by substantial barriers to entry such as Economies of Scale, network effects, spectrum licenses, or proprietary research and development. Oligopolistic markets account for the majority of manufacturing, telecommunications, aviation, semiconductor, banking, and digital platform output in modern industrialized economies.

The defining analytical feature that separates oligopoly from both perfect competition and pure Monopoly is strategic interdependence. In perfect competition, each atomistic firm is too small to affect the market price; in monopoly, the single firm faces no rivals at all. In an oligopoly, because each firm controls a substantial share of industry output, any change in Firm i's price p_i, physical output q_i, advertising budget, or product R&D directly alters the residual demand and profits of all rival firms j ≠ i, prompting strategic counter-moves. Consequently, oligopoly theory is analyzed primarily through the mathematical tools of non-cooperative and repeated Game Theory and Nash Equilibrium.

Because strategic interaction depends on what strategic variable firms choose (quantities vs. prices), when they move (simultaneously vs. sequentially), and whether products are homogeneous or differentiated, economists classify oligopoly equilibria across four benchmark models:

Resolving the Bertrand Paradox: Capacity Constraints and Kreps–Scheinkman

Joseph Bertrand (1883) critiqued Antoine Augustin Cournot's quantity-competition model by observing that firms in most retail markets post prices rather than quantities, and that under homogeneous price competition just two firms suffice to drive price all the way down to marginal cost (P = MC). Francis Ysidro Edgeworth (1897) noted that this result relies on the unrealistic assumption that a single firm can instantaneously supply 100% of market demand at constant marginal cost without hitting physical plant capacity limits.

In a landmark synthesis, David Kreps and José Scheinkman (1983) modeled a two-stage game: in Stage 1, firms simultaneously invest in physical production capacities (k₁, k₂); in Stage 2, after observing capacities, firms compete in Bertrand prices subject to q_i ≤ k_i. Kreps and Scheinkman proved that under efficient rationing, the unique subgame perfect equilibrium of this two-stage capacity-then-price game yields the exact Cournot equilibrium quantities and prices. Thus, Cournot Competition represents long-run capacity competition, while Bertrand describes short-run price competition when excess capacity is abundant.

Joint industry profit ∑_i=1^N Π_i is maximized if the N oligopolists form a cartel (such as OPEC in crude oil or historic international vitamin and lysine cartels), restrict aggregate output to the monopoly level Q_m, charge the monopoly price P_m, and split production quotas (q_i = Q_m ÷ N).

In a one-shot simultaneous game, however, a cartel agreement is structurally isomorphic to the Prisoner's Dilemma. Because market price exceeds marginal cost (P_m > MC), each individual cartel member has a unilateral incentive to secretly cheat on its quota by expanding output or offering secret price discounts to steal market share, driving the one-shot Nash Equilibrium back to the non-cooperative Cournot or Bertrand outcome.

Tacit Collusion and the Folk Theorem in Infinitely Repeated Games

When oligopolists interact repeatedly over an indefinite horizon (t = 0, 1, 2, ...) and discount future profits by the per-period discount factor δ = 1 ÷ (1 + r) ∈ (0, 1) (linking directly to Net Present Value), the Folk Theorem of repeated game theory demonstrates that tacit collusion can be sustained without legally enforceable contracts via grim-trigger strategies (Friedman, 1971):

• Cooperate: Charge the collusive monopoly price P_m (earning Π_m ÷ N per period) so long as no firm has ever deviated in any past period.
• Punish: If any firm deviates in period t (capturing the entire market and earning one-period deviation profit Π_m), all firms revert permanently to the competitive Bertrand equilibrium (P = c, earning Π_c = 0) from period t + 1 onward.

In a symmetric N-firm Bertrand supergame, cooperating yields the perpetual present value:

Unilaterally undercutting by an infinitesimal ε today yields V_cheat = Π_m + 0. Tacit collusion is a subgame perfect equilibrium if and only if V_coop ≥ V_cheat:

For a duopoly (N = 2), collusion is sustainable whenever δ ≥ 0.50; as the number of firms N increases, the critical discount factor threshold 1 − (1 ÷ N) rises toward 1.0, proving mathematically why collusion becomes harder to sustain as an industry becomes less concentrated.

Empirical industrial organization and antitrust agencies (assessing industry rivalry alongside Porter's Five Forces) quantify oligopolistic concentration using two primary indices for market shares s₁ ≥ s₂ ≥ ... ≥ s_N (expressed as percentages summing to 100):

1. Four-Firm Concentration Ratio (CR₄): CR₄ = s₁ + s₂ + s₃ + s₄. Industries with CR₄ > 60% are classified as tight oligopolies.
2. Herfindahl–Hirschman Index (HHI): HHI = ∑_i=1^N s_i². Because shares are squared, HHI weights large firms more heavily than small fringe firms and equals N × (100 ÷ N)² = 10,000 ÷ N for a symmetric N-firm oligopoly. Under the U.S. DOJ/FTC Merger Guidelines, a horizontal merger in a market with post-merger HHI > 1,800 that raises HHI by ΔHHI > 100 is presumed likely to enhance market power.

Furthermore, incumbent oligopolists may engage in strategic entry deterrence (analyzed by Avinash Dixit, 1980)—investing in irrevocable excess capacity, proliferating brands across product characteristic space, or using limit pricing under Information Asymmetry to signal low marginal costs and deter potential entrants.

Why are prices in oligopolistic markets often rigid ("sticky") over short horizons?

Two complementary mechanisms explain price rigidity in oligopolies. First, in Paul Sweezy's (1939) kinked demand curve heuristic, if an oligopolist believes rivals will match any price cut (making demand relatively inelastic below the current price) but will not match a price increase (making demand highly elastic above the current price), the marginal revenue curve exhibits a vertical jump discontinuity at the current quantity, so moderate shifts in marginal cost MC do not alter the profit-maximizing price. Second, in repeated-game models of tacit collusion, firms avoid frequent small price adjustments lest rivals misinterpret a cost-driven price cut as a defection and trigger a costly price war (Green and Porter, 1984).

What is the difference between explicit collusion and tacit collusion?

Explicit collusion involves direct communication—secret meetings, encrypted messages, price-fixing agreements, or market-division pacts among executives—which is a per se civil and criminal violation of Section 1 of the U.S. Sherman Antitrust Act and Article 101 of the Treaty on the Functioning of the European Union. Tacit collusion (or conscious parallelism) occurs when firms independently recognize their mutual interdependence in a repeated game and maintain supra-competitive prices through public price leadership or focal points without any direct agreement.

How does product differentiation soften price competition in an oligopoly?

In Harold Hotelling's (1929) spatial competition model, when firms locate at opposite ends of a product-attribute spectrum (maximum differentiation) rather than offering identical commodities at the same point, consumers develop brand preferences or incur mismatch/transport costs to switch suppliers. Each firm then faces a smooth downward-sloping residual demand curve rather than a discontinuous step function, allowing all firms to set prices above marginal cost (P > MC).

1. Bertrand, J. (1883). Book review of Théorie mathématique de la richesse sociale and of Recherches sur les principes mathématiques de la théorie des richesses. Journal des Savants, 67, 499–508. https://doi.org/10.1017/CBO9780511528231.006
2. Friedman, J. W. (1971). A non-cooperative equilibrium for supergames. The Review of Economic Studies, 38(1), 1–12. https://doi.org/10.2307/2296617
3. Kreps, D. M., & Scheinkman, J. A. (1983). Quantity precommitment and Bertrand competition yield Cournot outcomes. The Bell Journal of Economics, 14(2), 326–337. https://doi.org/10.2307/3003636
4. Green, E. J., & Porter, R. H. (1984). Noncooperative collusion under imperfect price information. Econometrica, 52(1), 87–100. https://doi.org/10.2307/1911462
5. Tirole, J. (1988). The Theory of Industrial Organization. MIT Press. https://mitpress.mit.edu/9780262200714/the-theory-of-industrial-organization/