Chapter 1: Market Power Measurement and Market Definition#

Why can a drug company charge hundreds of dollars for a life-saving pill that costs a few dollars to make? The answer: market power—the ability to set prices far above cost without losing all your customers. This chapter shows how economists measure that power and how they decide which products actually compete with each other.

The Big Picture#

In this chapter, we ask two key questions. How much market power does a firm (or a group of firms) really have? And how do we draw the boundaries of a market so we can measure that power properly? You’ll learn about simple tools like the Lerner index and the Herfindahl-Hirschman Index. You’ll see the logic behind the SSNIP test, used by competition authorities worldwide. And you’ll explore more advanced ideas such as the dominant-firm model and the conduct parameter. By the end, you’ll be able to read an antitrust case and understand the numbers behind the arguments.

What Is Market Power?#

A firm has market power if it can profitably charge a price above its marginal cost (MC)—the cost of making one more unit—and still keep a meaningful number of customers. In a perfectly competitive market, every firm is a tiny price-taker. If any of them raise the price by even one cent, all customers leave. A firm with market power, on the other hand, faces a downward-sloping demand curve. It can raise the price without losing every customer, but it will sell less. So it faces a trade-off: a higher price brings more profit on each unit kept, but fewer sales overall.

We assume firms are profit maximizers. That means they choose the quantity (or price) that makes the gap between total revenue and total cost as large as possible. For any firm, the profit-maximizing rule is to keep producing as long as the extra revenue from one more unit—marginal revenue (MR)—is above the extra cost—marginal cost (MC). The sweet spot is where .

For a price-taking firm, simply equals the market price , so it stops at . But a firm with market power faces a downward-sloping demand. Selling one more unit forces it to lower the price not just on that unit, but on all earlier units. So is less than . The result: the firm chooses output where .

Here is a quick monopoly example. Suppose demand is a straight line: , and marginal cost is constant at . Total revenue is , so . Set : . The monopoly quantity is , and the monopoly price is . The markup over marginal cost is .

This markup is not harmless. Because the monopolist cuts output to raise the price, some customers who value the good more than its cost are priced out. The lost welfare is called deadweight loss. To see it, think about consumer surplus—the gap between what consumers are willing to pay (shown by the demand curve) and what they actually pay. In a competitive market, consumer surplus is the whole area under the demand curve above . Under monopoly, the higher price shrinks that surplus, and part of it moves to the firm as profit. But the triangle of lost sales between the demand curve and marginal cost is pure waste—deadweight loss. This is why market power matters for public policy.

📝 Section Recap: Market power is the ability to set price above marginal cost without losing all customers. It comes from downward-sloping demand and creates deadweight loss. Profit maximization means , and for a firm with market power that gives .

The Lerner Index: A Direct Measure of Markups#

If market power means charging above marginal cost, a handy yardstick is the Lerner index, named after economist Abba Lerner. It captures the price–cost margin as a share of price:

The index goes from 0 (perfect competition, where ) up to—in theory—1 (if were zero and the firm could still charge a positive price). In most real markets, sits somewhere between 0 and, say, 0.5. A higher Lerner index means more market power.

There is a helpful link between the Lerner index and the price elasticity of demand. Recall that a profit-maximizing firm sets . For any firm facing a demand curve, marginal revenue can be written as

where is the price elasticity of demand (the percentage change in quantity demanded for a 1% change in price). Setting and rearranging gives

So the Lerner index is simply the inverse of the demand elasticity. A firm facing very elastic demand (customers are price-sensitive) has little market power. A firm facing inelastic demand can charge a high markup. For example, if , the Lerner index is 0.5: price is twice marginal cost. If , the markup is only 20%.

This relationship is very useful. In principle, if we can estimate the demand elasticity a firm faces, we can figure out its market power without ever seeing its marginal cost directly. In practice, measuring marginal cost is hard, so the Lerner index is more a conceptual guide than a simple accounting number. But variants of it—using average variable cost as a rough stand-in—are common in competition cases.

📝 Section Recap: The Lerner index directly measures the price–cost margin. Under profit maximization it equals , so market power is higher when demand is less elastic.

Concentration and the Herfindahl Index#

While the Lerner index looks at one firm’s pricing, market structure matters too. In many industries, market power is shared among a few firms. A simple way to sum up competition is to measure concentration—how much of the market the largest firms control.

The most widely used concentration measure is the Herfindahl-Hirschman Index (HHI). You take each firm’s market share (as a percentage), square it, and add them up:

Here, is firm ’s market share in percent. If one firm has 100% of the market, . If an industry has many tiny firms, each with a share near 0%, the HHI is close to 0. Antitrust agencies in the US and Europe use the HHI as a screening tool. Markets with an HHI below 1,500 are considered “unconcentrated,” 1,500–2,500 “moderately concentrated,” and above 2,500 “highly concentrated.”

Why squares? Squaring gives more weight to larger firms. A market with two firms each holding 50% has , while a market with ten equal firms has . The first is far more concentrated.

The HHI is not just a number. Under the Cournot model (where firms compete by choosing quantities), there is a direct link between concentration and market power. In a Cournot equilibrium with firms, the industry-average Lerner index (weighted by market share) is

where is on the 0–10,000 scale and is the market demand elasticity. So higher concentration leads to higher average markups, all else equal. This gives the HHI real economic meaning: it tells us something about pricing power, not just structure.

📝 Section Recap: The HHI sums squared market shares and captures concentration. Higher HHI values mean fewer, larger firms. In the Cournot model, the average Lerner index is proportional to the HHI.

Defining the Relevant Market: The SSNIP Test#

Before you can measure market power with a Lerner index or HHI, you must answer a question that sounds simple but is actually tricky: which products belong in the same market? If you define the market too narrowly, you might think a firm has huge power when in fact customers can easily switch to a close substitute. Define it too broadly, and real market power gets hidden.

The main tool for market definition is the hypothetical monopolist test, often called the SSNIP test (Small but Significant Non-transitory Increase in Price). The idea is straightforward. Start with a narrow candidate market—say, all brands of cola. Then ask: if a single firm owned every cola brand, could it profitably impose a small, permanent price increase (usually 5–10%)? If yes, then cola is a relevant market by itself—customers don’t switch to other drinks enough to make the price rise unprofitable. If no, the hypothetical monopolist would lose too many sales to, say, lemon-lime sodas. In that case, you expand the candidate market to include those close substitutes. Repeat until you find the smallest set of products for which a 5–10% price increase would be profitable.

The SSNIP test is the basic idea behind market definition in competition policy. To apply it, economists often use a critical loss analysis. The critical loss is the percentage drop in sales that would just make the price increase unprofitable. If the actual sales loss from a 5% price rise turns out to be smaller than the critical loss, the price increase is profitable and the candidate market stands. The critical loss for a given price increase and a given margin is roughly

For example, if the margin is 40% and we test a 5% price rise, . If the estimated actual loss is, say, 8%, then the price increase is profitable and the market is defined.

The SSNIP test is not a mechanical formula. It needs good data on substitution patterns and demand elasticities. But it gives a clear, economic logic for drawing market boundaries.

📝 Section Recap: The SSNIP test defines a market by asking whether a hypothetical monopolist could profitably raise price by 5–10%. Critical loss analysis turns that question into a workable empirical test.

Dominant Firm with a Competitive Fringe#

Not every market fits the neat boxes of monopoly or symmetric oligopoly. Sometimes one large firm—the dominant firm—faces a competitive fringe of many small, price-taking rivals. Think of a national brand competing against generic store brands, or a legacy airline facing low-cost carriers on some routes.

Here, the dominant firm does not control the whole market. The fringe firms take the market price as given and supply whatever quantity makes their marginal cost equal to that price. The dominant firm, however, knows that its own output choice affects the market price, after accounting for the fringe’s response. It faces a residual demand curve: total market demand minus what the fringe is willing to supply at each price.

The dominant firm’s profit-maximizing condition is still , but now its marginal revenue depends on both the slope of market demand and how easily the fringe can expand. This leads to a modified Lerner index for the dominant firm:

where is the dominant firm’s market share, is the absolute value of the market demand elasticity, and is the elasticity of fringe supply (how much the fringe raises output when price rises). The denominator shows that a more elastic fringe supply (larger ) cuts the dominant firm’s market power. Any attempt to raise price is met by a quick expansion of fringe output. If the fringe can easily ramp up production, the dominant firm’s markup gets squeezed.

Suppose the dominant firm has 60% of the market, market demand elasticity is 1.5, and fringe supply elasticity is 2. Then

The dominant firm can sustain a 26% markup over its marginal cost—substantial, but far less than if there were no fringe at all.

📝 Section Recap: In a dominant-firm model, the large firm’s market power is limited by a competitive fringe. Its Lerner index depends on its market share, market demand elasticity, and how easily the fringe can expand supply.

Estimating Market Power: The Conduct Parameter#

The Lerner index and concentration measures give us snapshots of market power under specific assumptions (like Cournot behavior). But what if we don’t know which game firms are playing? The conduct parameter approach lets us estimate the degree of market power directly from data, without committing to a particular oligopoly model up front.

Recall that a firm’s profit-maximizing condition can be written as

Here, captures the firm’s conduct—how much it takes into account the effect of its output on the market price. If , the firm behaves as a price-taker (perfect competition). If , it acts as a Cournot competitor, taking rivals’ outputs as given. If firms perfectly collude to maximize joint profits, each symmetric firm effectively acts as if (since its share , this means ). So the conduct parameter nests a wide range of competitive behaviors.

In elasticity terms, the firm-level Lerner index becomes

If we estimate an industry-average conduct parameter (assuming symmetry for simplicity), the average Lerner index is

For Cournot, ; for perfect collusion, the joint monopoly outcome gives an industry Lerner index of , which requires . (In a symmetric market, , so , matching the firm-level logic.) So matches Cournot, and signals conduct more collusive than Cournot. A value of close to 0 suggests highly competitive pricing.

In practice, economists estimate by observing prices, market shares, and demand elasticities, then solving backward for the implied conduct. This usually means estimating a demand system and a set of supply (pricing) equations at the same time. The estimated can then be compared to benchmarks to infer whether the market behaves like Cournot, Bertrand, or something in between.

The conduct parameter method is powerful but debated. It relies on strong assumptions about demand and cost, and small mistakes in those assumptions can lead to misleading estimates. Still, it offers a flexible way to measure market power without assuming a specific game in advance.

📝 Section Recap: The conduct parameter generalizes the Lerner index to allow for different competitive regimes. Estimating from data lets us infer the intensity of competition without pre-judging the market structure.

Summary#

We’ve covered the key tools for measuring market power and defining markets. Market power is the ability to price above marginal cost, and it creates deadweight loss. The Lerner index captures that markup directly and ties it to demand elasticity. The HHI measures concentration and, under Cournot competition, links structure to pricing power. The SSNIP test gives us a reliable way to figure out which products belong in the same market. For markets with a dominant firm and a fringe, the markup is held down by the fringe’s ability to expand. Finally, the conduct parameter offers a flexible, data-driven way to estimate market power without assuming a specific game. Together, these concepts form the analytical backbone of modern competition economics.