ECO1 The Firm and Market Structures
Every firm operates in one of two broad environments, and the difference shows up in the demand curve the individual firm faces. A perfectly competitive firm must take the market price as given, so its demand curve is a horizontal line at that price: price, marginal revenue (MR), and average revenue (AR) are all the same number. A firm facing a downward sloping demand curve must cut its price to sell one more unit, so its MR always sits below price.
For an imperfectly competitive firm, price is a decision variable, so revenue depends on the price chosen at each quantity. Take a monopolist facing a linear market demand curve. As price falls, TR first rises, because the volume gained outweighs the price given up, and then falls once the price cut dominates. TR peaks where MR reaches zero, which is also where demand switches from elastic to inelastic.
The profit-maximizing rule
The rule is universal; what differs across structures is whether MR equals price (perfect competition only) or sits below it (everywhere else). A firm can earn positive economic profit in the short run whenever price exceeds average total cost (ATC). Under free entry that profit is a signal: competitors arrive, supply expands, and the market price gets pushed down toward each firm’s ATC.
Breakeven
A firm breaks even when total revenue equals total cost, which is the same as saying price equals ATC. The costs here are economic costs: accounting costs plus implicit opportunity costs. A firm that exactly covers its economic costs earns a normal profit, meaning its capital earns just what an equally risky alternative investment would pay. Zero economic profit therefore does not mean zero accounting profit.
The shutdown decision
Some fixed costs, such as a lease that must be paid until it expires, are sunk: they are owed whether the firm operates or not, so they must be ignored in the short-run decision. As long as revenue covers variable cost, operating beats closing, because every sale also contributes something toward the unavoidable fixed costs. Once price falls below minimum average variable cost (AVC), even the variable costs go uncovered, and shutting down loses less. The minimum AVC point is called the shutdown point, and the minimum ATC point is the breakeven point.
| Revenue and cost relationship | Short-run decision | Long-run decision |
|---|---|---|
| TR = TC | Stay in market | Stay in market |
| TR = TVC but < TC | Stay in market | Exit market |
| TR < TVC | Shut down production | Exit market |
TR: total revenue. TC: total cost. TVC: total variable cost.
When revenue covers total variable cost but not all fixed cost, the firm survives the short run but cannot stay solvent in the long run. For start-ups and turnaround candidates, identifying the sales levels at which the firm enters or leaves profitability, and the level below which it stops being viable at all, is one of the more useful habits an analyst can build.
WR International is a newly formed maker of low-cost prefabricated housing units for urban markets in emerging economies, operating with few competitors under imperfect competition. Its projected revenue and cost data, including all opportunity costs, are below. Quantity moves in blocks of 10 units for each 250 step down in price.
| Quantity (Q) | Price (P) | Total revenue (TR) | Total cost (TC) | Profit |
|---|---|---|---|---|
| 0 | 10,000 | 0 | 100,000 | −100,000 |
| 10 | 9,750 | 97,500 | 170,000 | −72,500 |
| 20 | 9,500 | 190,000 | 240,000 | −50,000 |
| 30 | 9,250 | 277,500 | 300,000 | −22,500 |
| 40 | 9,000 | 360,000 | 360,000 | 0 |
| 50 | 8,750 | 437,500 | 420,000 | 17,500 |
| 60 | 8,500 | 510,000 | 480,000 | 30,000 |
| 70 | 8,250 | 577,500 | 550,000 | 27,500 |
| 80 | 8,000 | 640,000 | 640,000 | 0 |
| 90 | 7,750 | 697,500 | 710,000 | −12,500 |
| 100 | 7,500 | 750,000 | 800,000 | −50,000 |
Total cost includes all opportunity costs.
A London-based business reports revenue of GBP 2 million for the latest period against total costs of GBP 2.5 million, made up of GBP 1 million in fixed costs and GBP 1.5 million in variable costs, a net loss of GBP 500,000 before tax effects. The firm was profitable in earlier periods.
The short run is the period during which at least one factor of production, such as plant size or technology, is fixed, so output can only be varied through the flexible inputs, typically labor. The long run is the period over which every factor is variable and firms can enter or exit the market. It is often called the planning horizon: the firm always operates in the short run but plans in the long run. How long that takes depends on the industry, from under a year for a small low-capital business to more than a decade for a capital-intensive one.
From short-run costs to the long-run curve
Each possible plant size has its own short-run total cost curve. A small plant has low fixed cost, but its costs climb steeply as output presses against capacity; a large plant carries higher fixed cost and a flatter slope. In the long run the firm is free to pick whichever plant is cheapest for its planned output, so long-run cost traces the lowest edge of all the short-run curves, an envelope curve. Applied to average cost, the envelope of the short-run average total cost (SATC) curves is the long-run average cost curve (LRAC).
Why unit costs fall, and why they rise again
Economies of scale exist when cost per unit falls as output rises, pulling the LRAC downward. They come from increasing returns to scale (output growing proportionately more than inputs), division of labor and specialization, better and more efficient equipment, reduced waste, sharper market information, and volume discounts on inputs. Electric utilities grew from local to multiregional firms on exactly this logic, and Walmart’s bulk purchasing power is a textbook case.
Diseconomies of scale exist when cost per unit rises with output, turning the LRAC upward. They come from decreasing returns to scale, organizations too large to manage, overlapping functions and duplicated product lines, and input prices that rise when a very large buyer strains supply. Before its restructuring, General Motors showed several of these at once: duplicated models, worldwide coordination problems, and labor costs above its competitors.
Both forces can operate at the same time; the LRAC shape shows which dominates. Commonly it falls over one range, stays flat over another, and then rises. The lowest point on the LRAC is the minimum efficient scale, the optimal firm size under perfect competition, because the long-run market price settles at that cost level. A firm not producing at the least-cost point has a long-run viability problem.
Market structure determines the choices open to a firm’s decision makers, and through them its pricing and long-run profitability. In a highly competitive market, long-run profits are competed away; in less competitive markets, large profits can persist, and in the short run anything is possible. An analyst forecasting a firm’s profit stream therefore starts by asking what kind of market the firm lives in.
A market is a group of buyers and sellers who are aware of each other and can agree on a price. Some markets are worldwide, such as internet search; some are strictly local, such as premixed cement, which travels only as far as a truck can carry it before it sets. Any service that can be digitized, such as a radiographic scan taken in one country and read in another, tends toward a worldwide market. Concentration varies just as widely: as of January 2018, three firms held 98.9 percent of the US internet search market (Google 63.5 percent, Microsoft 24 percent, and Oath, formerly Yahoo!, 11.4 percent), while automobile repair remains fragmented among small independent shops.
The four structures
Perfect competition is not just a theoretical ideal; several commodity markets match it, with a strictly homogeneous product and no producer large enough to move the price. Profits are driven to the normal level, the required return on capital, yet millions of businesses do perfectly well under exactly this pressure.
Monopolistic competition, the hybrid identified by Edward Chamberlin and Joan Robinson, combines a large number of sellers with differentiated products. A seller that convinces buyers its product is genuinely different gains some pricing power; brand loyalty in soft drinks, fashion, and cosmetics is the classic illustration.
Oligopoly rests on a small number of firms, so every pricing decision is made with an eye on the expected retaliation of rivals. Airline pricing on a given route, where only a few carriers compete, is the standard example.
Monopoly is a single seller of a product with no close substitutes. Left unconstrained it holds considerable power over price and output, which is why pure monopolies, such as a local electricity provider, are usually regulated to earn a normal return at prices the authority sets.
What determines the structure
- The number and relative size of the firms supplying the product.
- The degree of product differentiation.
- The seller’s power over pricing decisions.
- The strength of barriers to entry and exit. High exit costs, such as plants that cannot be redeployed, deter entry too, so they count as entry barriers.
- The degree of non-price competition, which dominates wherever differentiation is critical or firms fear retaliatory price changes.
| Structure | Sellers | Differentiation | Entry barriers | Pricing power | Non-price competition |
|---|---|---|---|---|---|
| Perfect competition | Many | Homogeneous | Very low | None | None |
| Monopolistic competition | Many | Differentiated | Low | Some | Advertising, differentiation |
| Oligopoly | Few | Homogeneous or standardized | High | Some to considerable | Advertising, differentiation |
| Monopoly | One | Unique product | Very high | Considerable | Advertising |
Owners prefer the structures with the most control over price, since control can mean large profits; consumers prefer the most competitive structures, since prices are generally lower there. There is a trade-off, though: quantities are largest and prices lowest under perfect competition, but less competitive structures can fund innovation that competitive margins never could. Clinical trials for new medicines would bankrupt perfectly competitive firms, yet an oligopoly can carry them, so consumers can benefit from imperfect competition too.
Michael Porter organized these questions into five forces: the threat of entry, the power of suppliers, the power of buyers, the threat of substitutes, and rivalry among existing competitors. Four of the five map directly onto the structure factors above; supplier power is the addition that matters more in practice than in economic theory. Related is the idea of an economic moat: the deeper the moat, the smaller the threat of entry and the higher the switching costs that keep customers locked in.
The single most distinctive feature of monopolistic competition is product differentiation. Its characteristics:
- A large number of potential buyers and sellers.
- Products that are close substitutes for one another, with each firm working to make its own look different.
- Entry and exit at fairly low cost.
- Some pricing power for each firm.
- Differentiation pursued through advertising and other non-price strategies.
Differentiation can run through styles, colors, networks, bundled features, contract terms, and above all advertising; successful branding builds the kind of loyalty motorcycle riders famously attach to Harley-Davidson. The more successful the differentiation, the closer the firm’s position moves to a single-seller market. Entry costs are low, though, so in the long run competition drags prices and revenues back toward the perfectly competitive outcome. The structure genuinely is a hybrid of the two.
Short-run equilibrium
Because each product is somewhat distinct, each firm faces its own downward sloping demand curve, elastic over higher price ranges and inelastic at lower ones. The firm produces the quantity at which MR equals MC, then charges the highest price its demand curve supports at that quantity.
One subtlety: a monopolistically competitive firm has no well-defined supply function. Its output level comes from the intersection of MR and MC, while its price comes from the demand schedule, so no single curve links price to quantity supplied.
Long-run equilibrium
Economic profit is a signal, and with entry cheap the signal gets answered. New firms arrive, lure customers away, and push each incumbent’s demand curve down until economic profit reaches zero.
The zero-profit end point resembles perfect competition, with two differences. Output Q₁ sits below Q₂, the minimum-cost scale a perfectly competitive market would enforce. And economic cost now includes differentiation spending such as advertising, which homogeneous products never need. Prices end up somewhat higher; in exchange, consumers get variety.
Prices are higher and quantities lower than under perfect competition, yet monopolistic competition is the most common structure in the real world. Part of the answer goes back to Schumpeter: differentiation rewards innovation and experimentation with new products. Part is that tastes differ, so slight variations of a good each capture their own niche, the way a candy shelf carries chocolate, licorice, mint, and fruit. And part is that people simply value variety: Germany and Japan sell each other cars in both directions, which classical trade theory says should not happen, and variety is the explanation.
An oligopoly has only a few firms in the relevant market, selling products that are close substitutes, branded (breakfast cereals, canned beverages) or homogeneous (petroleum, cement). Its characteristics:
- A small number of potential sellers.
- Products that are close substitutes, differentiated by brand or unbranded.
- Difficult entry, with high costs and significant barriers.
- Substantial pricing power for the typical firm.
- Heavy use of marketing, features, and other non-price strategies.
The defining trait is interdependence: with so few players, every pricing decision is really a forecast of how rivals will retaliate. Few firms plus pricing power also make collusion attractive; the best-known colluding oligopoly is the OPEC cartel in crude oil. Three basic pricing frameworks organize the analysis: pricing interdependence, the Cournot assumption, and the Nash equilibrium.
Pricing interdependence: the kinked demand curve
In markets prone to price wars, the standard assumption is that rivals match a price cut but ignore a price rise. Raise your price and customers defect to rivals who stayed put, so demand is highly elastic above the prevailing price. Cut it and everyone follows, so volume barely moves and demand is much less elastic below. The firm’s demand curve therefore bends at the prevailing price.
This is why oligopoly prices are so often sticky: marginal cost can move over a wide range without changing the optimal price. The model’s weakness is that it cannot say where the prevailing price came from in the first place, so it is a useful but incomplete theory of pricing.
The Cournot assumption
Augustin Cournot’s model from 1838 has each firm choose its profit-maximizing output on the assumption that the other firm’s output will not change, which removes the guesswork about retaliation. In a two-firm duopoly, equilibrium is the pair of outputs at which neither firm wants to move, given the other’s choice; from there, output and price are stable.
Aggregate demand in a duopoly market has been estimated as
with supply represented by a constant marginal cost of 30 for both firms. Writing total output as the sum of the two firms’ outputs and rearranging demand in terms of price:
The Nash equilibrium
John Nash, a 1994 Nobel laureate, generalized the idea. In game theory, a Nash equilibrium holds when no participant can improve its outcome by changing strategy alone, after anticipating the rational choices of its rivals. In an oligopoly it means no firm can raise profit by unilaterally changing its price; actions are interdependent but non-cooperative, each firm maximizing its own profit.
Take two firms, ArcCo and BatCo, each able to price high or low.
| BatCo: low price | BatCo: high price | |
|---|---|---|
| ArcCo: low price | ArcCo 50, BatCo 70 | ArcCo 80, BatCo 0 |
| ArcCo: high price | ArcCo 300, BatCo 350 | ArcCo 500, BatCo 300 |
The lower-left cell, ArcCo high and BatCo low, is the Nash equilibrium.
Joint profit peaks at 800 when both price high. But BatCo earns more by pricing low whatever ArcCo does: 70 against 0 when ArcCo is low, and 350 against 300 when ArcCo is high. Given that, ArcCo’s best reply is to price high and take 300. The equilibrium is therefore ArcCo high and BatCo low: not the jointly best cell, only the individually stable one. ArcCo could keep BatCo pricing high by sharing at least 51 of its 500, which is collusion, tempting and, in most countries, unlawful.
When collusive agreements are made openly and formally, the firms form a cartel. Six conditions make collusion more likely to hold:
- Few sellers, or one dominant seller. Similar market shares actually make collusion harder, because competitive instincts take over.
- Homogeneous products.
- Similar cost structures.
- Small, frequent, regular orders, which shrink the payoff to cheating on the agreement.
- Severe and credible retaliation against members who cheat.
- Limited external competition. In 2016, extracting a barrel of crude cost roughly 9 dollars in Saudi Arabia, about 23.50 in US shale fields, and about 27 in Canadian tar sands. Gulf producers collude relatively easily on their similar costs, but by holding prices well above 30 dollars OPEC made shale and tar sands viable, inviting the very competition the cartel now faces.
Other strategic games apply too. In the Stackelberg model decisions are sequential rather than simultaneous: the leader commits to its output first and can overproduce aggressively, a top dog strategy, forcing the follower to scale back and earning more than in Cournot’s simultaneous game. The general lesson is that a firm’s optimal strategy depends on what its adversary does.
Price leadership and the long run
Where one firm holds roughly 40 percent or more of the market, through greater capacity, lower costs, first-mover advantage, or customer loyalty, it tends to act as price leader, setting price like a monopolist over its own segment of demand while the rest follow. Followers rarely undercut, because leadership usually rests on lower production cost, and a price war against a cheaper rival threatens their survival; letting one firm make the pricing decision is also simply convenient. The leader picks the output where its own MR equals its own MC, prices from its segment of demand, and the followers supply the remainder of market demand at that price. Neither side has a single functional supply relationship.
Long-run economic profits are possible in oligopoly, but dominance erodes: profits attract entrants, and entrants’ costs fall as they adopt more efficient techniques. US Steel controlled 66 percent of its market when formed in 1901, 46 percent by 1920, and 42 percent by 1925. History also argues against price wars: share gains are temporary and total revenue falls for every participant. Innovation, even when uneconomical, is one way leaders defend their position.
A few firms selling an easily replicated product, with limited scale economies and no brand or patent protection, cannot charge high prices: the credible threat of entry makes them behave much like perfect competitors. At the opposite extreme, a very effective cartel behaves like a single firm, reaching the cooperative outcome instead of the Nash one. Cartels can be explicit contracts or implicit signals, for instance one firm declining to match a rival’s price cut, which the cutter may read as a suggestion that both should move prices higher.
Pricing power lets producers restrict output and lift prices, which is inefficient for the market as a whole, so many countries regulate competition. The hard part is measurement. Regulators have prosecuted price fixing (the Archer Daniels Midland lysine cartel case ended in fines above 100 million US dollars), broken up monopolies (the 1970s AT&T divestiture, which brought competition, innovation, and lower prices to US telephony), and blocked mergers (the European Commission stopped General Electric and Honeywell on concentration grounds, having earlier acted in the vitamin price-fixing cases). An analyst hearing merger news should always ask whether competition law might block the deal. Sometimes the question is prospective: would a merger that has not yet happened create excessive market power?
The econometric route
In principle, market power is measured through the elasticity of demand and supply: very elastic demand points to something close to perfect competition, while rigid demand means the firms may hold power. In practice the estimation is awkward. Observed prices and quantities are equilibrium values, jointly determined by supply and demand (the endogeneity problem), so a proper estimate needs a two-equation model. Time-series data may span decades across which the structure itself changed, and a merger can shift the supply curve so that past data say little about the market afterward. Cross-sectional work demands heavy data gathering and is sensitive to specification choices. Analysts therefore lean on simpler measures.
The concentration ratio
The concentration ratio is the combined market share of the largest N firms, running from 0 (perfect competition) to 100 percent (monopoly). Its virtue is that it is trivially easy to compute. It does not directly measure market power: a sole incumbent facing low entry barriers may price as if it were in perfect competition, the way a country’s only sugar wholesaler is disciplined by large food wholesalers who could start importing sugar at any time. And it barely reacts to mergers among the leaders.
A market has eight producers with shares of 35, 25, 20, and 10 percent, plus four firms at 2.5 percent each.
Real markets do concentrate. The top four firms in US retail gasoline held 33 percent of the market in 1992 and 78 percent by 2001 (Exxon Mobil 24, Shell 20, BP/Amoco/Arco 18, and Chevron/Texaco 16 percent).
The Herfindahl-Hirschman index
To fix the merger blindness, economists O. C. Herfindahl and A. O. Hirschman proposed squaring the top N shares before adding them.
In the merger example above, the three-firm HHI is 0.352 + 0.252 + 0.202 = 0.225 before the merger and 0.602 + 0.202 + 0.102 = 0.410 after, a jump the concentration ratio missed entirely. Competition regulators use the HHI widely. Its limits are the same in kind as the ratio’s: it ignores the threat of entry and says nothing about demand elasticity, so it is of limited use to an analyst trying to forecast profitability.
A market has 10 suppliers, each holding 10 percent.