263

At the time of that elaborately bugged meeting, no one company controlled the world lysine industry, but there were only a few major producers. An industry with only a few sellers is known as an oligopoly; a firm in such an industry is known as an oligopolist.

Oligopolists obviously compete with one another for sales. But neither ADM nor Ajinomoto were like a firm in a perfectly competitive industry, which takes the price at which it can sell its product as given. In the case of ADM and Ajinomoto, each firm knew that its decision about how much to produce would affect the market price. That is, like monopolists, each of the firms had some market power. So the competition in this industry wasn’t “perfect.”

Economists refer to a situation in which firms compete but also possess market power—which enables them to affect market prices—as imperfect competition. As we saw in Chapter 8, there are actually two important forms of imperfect competition: oligopoly and monopolistic competition. Of these, oligopoly is probably the more important in practice.

Although lysine is a multibillion-dollar business, it is not exactly a product familiar to most consumers. However, many familiar goods and services are supplied by only a few competing sellers, which means the industries in question are oligopolies. For example, most air routes are served by only two or three airlines: regularly scheduled shuttle service between New York and either Boston or Washington, D.C., has been provided only by Delta and US Airways. Four firms—Kellogg, General Mills, Post, and Quaker—control 85% of the breakfast cereal market. Two firms—Apple and Samsung—control about 70% of the smartphone market. Most cola beverages are sold by Coke and Pepsi. This list could go on for many pages.

It’s important to realize that an oligopoly isn’t necessarily made up of large firms. What matters isn’t size per se; the question is how many competitors there are. When a small town has only two grocery stores, grocery service there is just as much an oligopoly as air shuttle service between New York and Washington.

Why are oligopolies so prevalent? Essentially, oligopoly is the result of the same factors that sometimes produce monopoly, but in somewhat weaker form. Probably the most important source of oligopoly is the existence of increasing returns to scale, which give bigger producers a cost advantage over smaller ones. When these effects are very strong, they lead to monopoly; when they are not that strong, they lead to an industry with a small number of firms.

For example, larger grocery stores typically have lower costs than smaller ones. But the advantages of large scale taper off once grocery stores are reasonably large, which is why two or three stores often survive in small towns.

If oligopoly is so common, why has most of this book focused on competition in industries where the number of sellers is very large? And why did we study monopoly, which is relatively uncommon, first, in Chapter 8? The answer has two parts.

First, much of what we learn from the study of perfectly competitive markets—about costs, entry and exit, and efficiency—remains valid despite the fact that many industries are not perfectly competitive. Second, the analysis of oligopoly turns out to present some puzzles for which there are no easy solutions. It is almost always a good idea first to deal with the questions you can answer, then to puzzle over the harder ones. We have simply followed that strategy, developing the relatively clear-cut theories of perfect competition and monopoly first, and only then turning to the puzzles presented by oligopoly.

How much will a firm produce? Up to this point, we have always answered: the quantity that maximizes its profit. Together with its cost curves, the assumption that a firm maximizes profit is enough to determine its output when it is a perfect competitor or a monopolist.

264

When it comes to oligopoly, however, we run into some difficulties. Indeed, economists often describe the behavior of oligopolistic firms as a “puzzle.”

A Duopoly Example Let’s begin looking at the puzzle of oligopoly with the simplest version, an industry in which there are only two producing firms—a duopoly—and each is known as a duopolist.

Going back to our opening story, imagine that ADM and Ajinomoto are the only two producers of lysine. To make things even simpler, suppose that once a company has incurred the fixed cost needed to produce lysine, the marginal cost of producing another pound is zero. So the companies are concerned only with the revenue they receive from sales.

Table 9-1 shows a hypothetical demand schedule for lysine and the total revenue of the industry at each price–quantity combination.

If this were a perfectly competitive industry, each firm would have an incentive to produce more as long as the market price was above marginal cost. Since the marginal cost is assumed to be zero, this would mean that at equilibrium lysine would be provided free. Firms would produce until price equals zero, yielding a total output of 120 million pounds and zero revenue for both firms.

However, surely the firms would not be that stupid. With only two firms in the industry, each would realize that by producing more, it drives down the market price. So each firm would, like a monopolist, realize that profits would be higher if it and its rival limited their production.

So how much will the two firms produce?

One possibility is that the two companies will engage in collusion—they will cooperate to raise their joint profits. The strongest form of collusion is a cartel, an arrangement between producers that determines how much each is allowed to produce. The world’s most famous cartel is the Organization of Petroleum Exporting Countries (OPEC).

As its name indicates, OPEC is actually an agreement among governments rather than firms. There’s a reason this most famous of cartels is an agreement among governments: cartels among firms are illegal in the United States and many other jurisdictions. But let’s ignore the law for a moment (which is, of course, what ADM and Ajinomoto did in real life—to their own detriment).

So suppose that ADM and Ajinomoto were to form a cartel and that this cartel decided to act as if it were a monopolist, maximizing total industry profits. It’s obvious from Table 9-1 that in order to maximize the combined profits of the firms, this cartel should set total industry output at 60 million pounds of lysine, which would sell at a price of $6 per pound, leading to revenue of $360 million, the maximum possible.

Then the only question would be how much of that 60 million pounds each firm gets to produce. A “fair” solution might be for each firm to produce 30 million pounds with revenues for each firm of $180 million.

But even if the two firms agreed on such a deal, they might have a problem: each of the firms would have an incentive to break its word and produce more than the agreed-upon quantity.

Collusion and Competition Suppose that the presidents of ADM and Ajinomoto were to agree that each would produce 30 million pounds of lysine over the next year. Both would understand that this plan maximizes their combined profits. And both would have an incentive to cheat.

265

To see why, consider what would happen if Ajinomoto honored its agreement, producing only 30 million pounds, but ADM ignored its promise and produced 40 million pounds. This increase in total output would drive the price down from $6 to $5 per pound, the price at which 70 million pounds are demanded. The industry’s total revenue would fall from $360 million ($6 × 60 million pounds) to $350 million ($5 × 70 million pounds). However, ADM’s revenue would rise, from $180 million to $200 million. Since we are assuming a marginal cost of zero, this would mean a $20 million increase in ADM’s profits.

But Ajinomoto’s president might make exactly the same calculation. And if both firms were to produce 40 million pounds of lysine, the price would drop to $4 per pound. So each firm’s profits would fall, from $180 million to $160 million.

Why do individual firms have an incentive to produce more than the quantity that maximizes their joint profits? Because neither firm has as strong an incentive to limit its output as a true monopolist would.

Let’s go back for a minute to the theory of monopoly. We know that a profit-maximizing monopolist sets marginal cost (which in this case is zero) equal to marginal revenue. But what is marginal revenue? Recall that producing an additional unit of a good has two effects:

The negative price effect is the reason marginal revenue for a monopolist is less than the market price. In the case of oligopoly, when considering the effect of increasing production, a firm is concerned only with the price effect on its own units of output, not those of its fellow oligopolists. Both ADM and Ajinomoto suffer a negative price effect if ADM decides to produce extra lysine and so drives down the price. But ADM cares only about the negative price effect on the units it produces, not about the loss to Ajinomoto.

This tells us that an individual firm in an oligopolistic industry faces a smaller price effect from an additional unit of output than does a monopolist; therefore, the marginal revenue that such a firm calculates is higher. So it will seem to be profitable for any one company in an oligopoly to increase production, even if that increase reduces the profits of the industry as a whole. But if everyone thinks that way, the result is that everyone earns a lower profit!

Until now, we have been able to analyze producer behavior by asking what a producer should do to maximize profits. But even if ADM and Ajinomoto are both trying to maximize profits, what does this predict about their behavior? Will they engage in collusion, reaching and holding to an agreement that maximizes their combined profits? Or will they engage in noncooperative behavior, with each firm acting in its own self-interest, even though this has the effect of driving down everyone’s profits? Both strategies sound like profit maximization. Which will actually describe their behavior?

Now you see why oligopoly presents a puzzle: there are only a small number of players, making collusion a real possibility. If there were dozens or hundreds of firms, it would be safe to assume they would behave noncooperatively. Yet when there are only a handful of firms in an industry, it’s hard to determine whether collusion will actually materialize.

Since collusion is ultimately more profitable than noncooperative behavior, firms have an incentive to collude if they can. One way to do so is to formalize it—sign an agreement (maybe even make a legal contract) or establish some financial incentives for the companies to set their prices high. But in the United States and many other nations, you can’t do that—at least not legally. Companies cannot make a legal contract to keep prices high: not only is the contract unenforceable, but writing it is a one-way ticket to jail. Neither can they sign an informal “gentlemen’s agreement,” which lacks the force of law but perhaps rests on threats of retaliation—that’s illegal, too.

266

In fact, executives from rival companies rarely meet without lawyers present, who make sure that the conversation does not stray into inappropriate territory. Even hinting at how nice it would be if prices were higher can bring you an unwelcome interview with the Justice Department or the Federal Trade Commission.

For example, in 2003 the Justice Department launched a price-fixing case against Monsanto and other large producers of genetically modified seed. The Justice Department was alerted by a series of meetings held between Monsanto and Pioneer Hi-Bred International, two companies that account for 60% of the U.S. market in maize and soybean seed. The two companies, parties to a licensing agreement involving genetically modified seed, claimed that no illegal discussions of price-fixing occurred in those meetings. But the fact that the two firms discussed prices as part of the licensing agreement was enough to trigger action by the Justice Department.

Sometimes, as we’ve seen, oligopolistic firms just ignore the rules. But more often they find ways to achieve collusion without a formal agreement, as we’ll soon see.

How oligopolies usually work in practice depends on the legal framework that limits what firms can do and on the underlying ability of firms in a given industry to cooperate without formal agreements.

The Legal Framework To understand oligopoly pricing in practice, we need to understand the legal constraints under which oligopolistic firms operate. In the United States, oligopoly first became an issue during the second half of the nineteenth century, when the growth of railroads—themselves an oligopolistic industry—created a national market for many goods.

Large firms producing oil, steel, and many other products soon emerged. The industrialists quickly realized that profits would be higher if they could limit price competition. So, many industries formed cartels—that is, they signed formal agreements to limit production and raise prices. Until 1890, when the first federal legislation against such cartels was passed, this was perfectly legal.

However, although these cartels were legal, they weren’t legally enforceable—members of a cartel couldn’t ask the courts to force a firm that was violating its agreement to reduce its production. And firms often did violate their agreements, for the reason already suggested by our duopoly example: there is always a temptation for each firm in a cartel to produce more than it is supposed to.

In 1881 clever lawyers at John D. Rockefeller’s Standard Oil Company came up with a solution—the so-called trust. In a trust, shareholders of all the major companies in an industry placed their shares in the hands of a board of trustees who controlled the companies. This, in effect, merged the companies into a single firm that could then engage in monopoly pricing. In this way, the Standard Oil Trust established what was essentially a monopoly of the oil industry, and it was soon followed by trusts in sugar, whiskey, lead, cottonseed oil, and linseed oil.

Eventually there was a public backlash, driven partly by concern about the economic effects of the trust movement, partly by fear that the owners of the trusts were simply becoming too powerful. The result was the Sherman Antitrust Act of 1890, which was intended both to prevent the creation of more monopolies and to break up existing ones. At first this law went largely unenforced. But over the decades that followed, the federal government became increasingly committed to making it difficult for oligopolistic industries either to become monopolies or to behave like them. Such efforts are known to this day as antitrust policy.

267

Among advanced countries, the United States is unique in its long tradition of antitrust policy. Until recently, other advanced countries did not have policies against price-fixing, and some had even supported the creation of cartels, believing that it would help their own firms against foreign rivals. But the situation has changed radically as the European Union (EU)—a supranational body tasked with enforcing antitrust policy for its member countries—has moved toward U.S. practices. Today, EU and U.S. regulators often target the same firms because price-fixing has “gone global” as international trade has expanded.

During the early 1990s, the United States instituted an amnesty program in which a price-fixer receives a much-reduced penalty if it informs on its co-conspirators. In addition, Congress increased the maximum fines levied upon conviction. These two new policies clearly made informing on your cartel partners a dominant strategy, and it has paid off as executives from Belgium, Britain, Canada, France, Germany, Italy, Mexico, the Netherlands, South Korea, and Switzerland, as well as from the United States, have been convicted in U.S. courts of cartel crimes.

Life has gotten much tougher for those who want to operate a cartel.

Tacit Collusion and Price Wars If a real industry were as simple as our lysine example, it probably wouldn’t be necessary for the company presidents to meet or do anything that could land them in jail. Both firms would realize that it was in their mutual interest to restrict output to 30 million pounds each and that any short-term gains to either firm from producing more would be much less than the later losses as the other firm retaliated. So even without any explicit agreement, the firms would probably achieve the tacit collusion needed to maximize their combined profits.

Firms are said to be engaged in tacit collusion when, as in our example, they restrict output in a way that raises the profits of another firm, and expect the favor to be returned even without an enforceable agreement—although they act “as if” they had such an agreement and are in legal jeopardy if they even discuss prices.

Real industries are nowhere near this simple. Nonetheless, in most oligopolistic industries, most of the time, the sellers do appear to succeed in keeping prices above their noncooperative level. Tacit collusion, in other words, is the normal state of oligopoly.

Although tacit collusion is common, it rarely allows an industry to push prices all the way up to their monopoly level; collusion is usually far from perfect. As we discuss next, there are four factors that make it hard for an industry to coordinate on high prices.

Less Concentration In a less concentrated industry, the typical firm will have a smaller market share than in a more concentrated industry. This tilts firms toward noncooperative behavior because when a smaller firm cheats and increases its output, it gains for itself all of the profit from the higher output. And if its rivals retaliate by increasing their output, the firm’s losses are limited because of its relatively modest market share. A less concentrated industry is often an indication that there are low barriers to entry.

Complex Products and Pricing Schemes In our lysine example the two firms produce only one product. In reality, however, oligopolists often sell thousands or even tens of thousands of different products. Under these circumstances, keeping track of what other firms are producing and the prices they are charging is difficult. This makes it hard to determine whether a firm is cheating on the tacit agreement.

Differences in Interests In the lysine example, a tacit agreement for the firms to split the market equally is a natural outcome, probably acceptable to both firms. In real industries, however, firms often differ both in their perceptions about what is fair and in their real interests.

268

For example, suppose that Ajinomoto was a long-established lysine producer and ADM a more recent entrant to the industry. Ajinomoto might feel that it deserved to continue producing more than ADM, but ADM might feel that it was entitled to 50% of the business. (A disagreement along these lines was one of the contentious issues in those meetings the FBI was filming.)

Alternatively, suppose that ADM’s marginal costs were lower than Ajinomoto’s. Even if they could agree on market shares, they would then disagree about the profit-maximizing level of output.

Bargaining Power of Buyers Often oligopolists sell not to individual consumers but to large buyers—other industrial enterprises, nationwide chains of stores, and so on. These large buyers are in a position to bargain for lower prices from the oligopolists: they can ask for a discount from an oligopolist and warn that they will go to a competitor if they don’t get it. An important reason large retailers like Walmart are able to offer lower prices to customers than small retailers is precisely their ability to use their size to extract lower prices from their suppliers.

These difficulties in enforcing tacit collusion have sometimes led companies to defy the law and create illegal cartels. We’ve already examined the case of the lysine industry and in the upcoming Economics in Action, we’ll look at the chocolate industry.

Because tacit collusion is often hard to achieve, most oligopolies charge prices that are well below what the same industry would charge if it were controlled by a monopolist—or what they would charge if they were able to collude explicitly. In addition, sometimes collusion breaks down and there is a price war, which sometimes involves a collapse of prices to their noncooperative level. Sometimes they even go below that level, as sellers try to put each other out of business or at least punish what they regard as cheating.