BusinessWeek magazine study of mergers and acquisitions between 1990 and 1995 found that 83 percent of these deals achieved, at best, marginal returns, and 50 percent recorded a loss. If such mergers are not especially profitable, why do they occur? U.S.

The central element of oligopoly is that there are a small number of firms in an industry so that, when making decisions, a firm must take into account the expected reaction of other firms. Oligopolistic firms are mutually interdependent and can be collusive or noncollusive.

This mutual interdependence is the big difference between monopolistic competition and oligopoly. In oligopoly, firms explicitly take other firms’ actions into account. In monopolistic competition, there are so many firms that individual firms tend not to explicitly take into account rival firms’ likely responses to their decisions. Collusion is difficult. In oligopoly there are fewer firms, and each firm is more likely to explicitly engage in strategic decision making—taking explicit account of a rival’s expected response to a decision you are making. In oligopolies all decisions, including pricing decisions, are strategic decisions. Also, in oligopolies collusion is much easier. Thus, one distinguishes between monopolistic competition and oligopoly by whether or not firms explicitly take into account competitors’ reactions to their decisions.

Why is the distinction important? Because it determines whether economists can model and predict the price and output of an industry. Nonstrategic decision making can be predicted relatively accurately if individuals behave rationally. Strategic decision making is much more difficult to predict, even if people behave rationally. What one person does depends on what he or she expects other people to do, which in turn depends on what others expect the one person to do. Consistent with this distinction, economists’ model of monopolistic competition has a definite prediction. A model of monopolistic competition will tell us: Here’s how much will be produced and here’s how much will be charged. Economists’ models of oligopoly don’t have a definite prediction. There are no unique price and output decisions at which an oligopoly will rationally arrive; there are a variety of rational oligopoly decisions, and a variety of oligopoly models.

Most industries in the United States have some oligopolistic elements. If you ask almost any businessperson whether he or she directly takes into account rivals’ likely response, the answer you’ll get is “In certain cases, yes; in others, no.”

Most retail stores that you deal with are oligopolistic in your neighborhood or town, although by national standards they may be quite competitive. For example, how many grocery stores do you shop at? Do you think they keep track of what their competitors are doing? You bet. They keep a close eye on their competitors’ prices and set their own accordingly.

Oligopolistic firms are mutually interdependent.

Oligopolies take into account the reactions of other firms; monopolistic competitors do not.

No single general model of oligopoly behavior exists. The reason is that an oligopolist can decide on pricing and output strategy in many possible ways, and there are no compelling grounds to characterize any of them as the oligopoly strategy. Although there are five or six formal models, I’ll focus on two informal models of oligopoly behavior that give you insight into real-world problems. The two models we’ll consider are the cartel model and the contestable market model. These should give you a sense of how real-world oligopolistic pricing takes place.

Why, you ask, can’t economists develop a simple formal model of oligopoly? The reason lies in the interdependence of oligopolists. Since there are few competitors, what one firm does specifically influences what other firms do, so an oligopolist’s plan must always be a contingency or strategic plan. If my competitors act one way, I’ll do X, but if they act another way, I’ll do Y. Strategic interactions have a variety of potential outcomes rather than a single outcome such as in the formal models we discussed. An oligopolist spends enormous amounts of time guessing what its competitors will do, and it develops a strategy of how it will act accordingly. As we will discuss in Chapter 20, an entire theory called game theory has developed that considers interdependent decisions. The appendix to Chapter 20 shows how game theory can be applied to oligopoly decisions.

Models of Oligopoly

If oligopolies can limit the entry of other firms and form a cartel, they increase the profits going to the combination of firms in the cartel.

A cartel is a combination of firms that acts as if it were a single firm; a cartel is a shared monopoly. If oligopolies can limit entry by other firms, they have a strong incentive to cartelize the industry and to act as a monopolist would, restricting output to a level that maximizes profit for the combination of firms. Thus, the cartel model of oligopoly is a model that assumes that oligopolies act as if they were monopolists that have assigned output quotas to individual member firms of the oligopoly so that total output is consistent with joint profit maximization. All firms follow a uniform pricing policy that serves their collective interest.

WWW

Web Note 15.1 Price-Fixing

Since a monopolist makes the most profit that can be squeezed from a market, cartelization is the best strategy for an oligopoly. It requires each oligopolist to hold its production below what would be in its own interest were it not to collude with the others. Such explicit formal collusion is against the law in the United States, but informal collusion is allowed and oligopolies have developed a variety of methods to collude implicitly. Thus, the cartel model has some relevance.

The model has some problems, however. For example, various firms’ interests often differ, so the collective interest of the firms in the industry isn’t clear. In many cases a single firm, often the largest or dominant firm, takes the lead in pricing and output decisions, and the other firms (which are often called fringe firms) follow suit, even though they might have preferred to adopt a different strategy.

This dominant-firm cartel model works only if the smaller firms face barriers to entry, or the dominant firm has significantly lower cost conditions. If that were not the case, the smaller firms would pick up an increasing share of the market, eliminating the dominant firm’s monopoly. An example of such a dominant-firm market was the copier market in the 1960s and 1970s, in which Xerox set the price and other firms followed. That copier market also shows the temporary nature of such a market. As the firms became more competitive on cost and quality, Xerox’s market share fell and the company lost its dominant position. The copier market is far more competitive today than it used to be.

In other cases the various firms meet—sometimes only by happenstance, at the golf course or at a trade association gathering—and arrive at a collective decision. In the United States, meetings for this purpose are illegal, but they do occur. In yet other cases, the firms engage in implicit collusion—multiple firms make the same pricing decisions even though they have not explicitly consulted with one another. They “just happen” to come to a collective decision.

In some cases, firms collude implicitly—they just happen to make the same pricing decisions. This is not illegal.

Implicit price collusion, in which firms just happen to charge the same price but didn’t meet to discuss price strategy, isn’t against the law. Oligopolies often operate as close to the fine edge of the law as they can. For example, many oligopolistic industries allow a price leader to set the price, and then the others follow suit. The airline and steel industries take that route. Firms just happen to charge the same price or very close to the same price.

It isn’t only in major industries that you see such implicit collusion. In small towns, you’ll notice that most independent carpenters charge the same price. There’s no explicit collusion, but were a carpenter to offer to work for less than the others, he or she would feel unwelcome at the local breakfast restaurant.

Or let’s take another example: the Miami fish market, where sport fishermen sell their catch at the dock. When I lived in Miami, I often went to the docks to buy fresh fish. There were about 20 stands, all charging the same price. Price fluctuated, but it was by subtle agreement, and close to the end of the day the word would go out that the price could be reduced.

I got to know some of the sellers and asked them why they priced like that when it would be in their individual interest to set their own price. Their answer: “We like our boat and don’t want it burned.” They may have been talking in hyperbole, but social pressures play an important role in stabilizing prices in an oligopoly.

Even if all firms in the industry cooperate, other firms, unless they are prevented from doing so, can always enter the market with a technologically superior new product at the same price or with the same good at a lower price. It is important to remember that technological changes are constantly occurring, and that a successful cartel with high profits will provide incentives for significant technological change, which can eliminate demand for its monopolized product.

Informal collusion happens all the time in U.S. businesses. One characteristic of informal collusive behavior is that prices tend to be sticky—they don’t change frequently. Informal collusion is an important reason why prices are sticky. But it’s not the only reason.

Another possible reason is that firms don’t collude, but do have certain expectations of other firms’ reactions, which changes their perceived demand curves. Specifically, they perceive that the demand curve they face is kinked. This kinked demand curve is used especially to explain why firms often do not use lower-price strategies to increase sales.

Let’s go through the reasoning behind the kinked demand curve. If a firm increases its price, and the firm believes that other firms won’t go along, its perceived demand curve for increasing price will be very elastic (D1 in Figure 15-1). It will lose lots of business to the other firms that haven’t raised their price. The relevant portions of its demand curve and its marginal revenue curve are shown in blue in Figure 15-1.

Figure 15-1 The Kinked Demand Curve

One explanation for why prices are sticky is that firms face a kinked demand curve. When we draw the relevant marginal revenue curve for the kinked demand, we see that the corresponding MR curve is discontinuous. It has a gap in it. Shifts in marginal costs between c and d will not change the price or the output that maximizes profits.

If it decreases its price, however, the firm assumes that all other firms would immediately match that decrease, so it would gain very few, if any, additional sales. A large fall in price would result in only a small increase in sales, so its demand is very inelastic (D2 in Figure 15-1). This less elastic portion of the demand curve and the corresponding marginal revenue curve are shown in orange in Figure 15-1.

When the demand curve has a kink, the marginal revenue curve must have a gap.

Notice that when you put these two curves together, you get a rather strange demand curve (it’s kinked) and an even stranger marginal revenue curve (one with a gap). I didn’t make a mistake in drawing the curves; that’s the way they come out given the assumptions. When the demand curve has a kink, the marginal revenue curve must have a gap. Shifts in marginal cost (such asMC0 to MC1) will not change the firm’s profit maximization position. A large shift in marginal cost is required before firms will change their price. Why should this be the case? The intuitive answer lies in the reason behind the kink. If the firm raises its price, other firms won’t go along, so it will lose lots of market share. However, when the firm lowers price, other firms will go along and the firm won’t gain market share. Thus, the firm has strong reasons not to change its price in either direction.

I should emphasize that the kinked demand curve is not a theory of oligopoly pricing. It does not say why the original price is what it is; the kinked demand curve is simply a theory of sticky prices.

A second model of oligopoly is the contestable market model. The contestable market model is a model of oligopoly in which barriers to entry and barriers to exit, not the structure of the market, determine a firm’s price and output decisions. Thus, it emphasizes entry and exit conditions, and says that the price that an oligopoly will charge will exceed the cost of production only if new firms cannot exit and enter the market. The higher the barriers, the more the price exceeds cost. Without barriers to entry or exit, the price an oligopolist sets will be equal to the competitive price. Thus, an industry that structurally looks like an oligopoly could set competitive prices and output levels.

In the contestable market model of oligopoly, pricing and entry decisions are based only on barriers to entry and exit, not on market structure. Thus, even if the industry contains only one firm, it could still be a competitive market if entry is open.

Because of the importance of social pressures in determining strategies of oligopolies, no one “oligopolistic model” exists. Oligopolies with a stronger ability to collude (that is, more social pressures to prevent entry) are able to get closer to a monopolist solution. Equilibrium of oligopolies with weaker social pressures and less ability to prevent new entry is closer to the perfectly competitive solution. That’s as explicit as we can be.

An oligopoly model can take two extremes: (1) the cartel model, in which an oligopoly sets a monopoly price, and (2) the contestable market model, in which an oligopoly with no barriers to entry sets a competitive price. Thus, we can say that an oligopoly’s price will be somewhere between the competitive price and the monopolistic price. Other models of oligopolies give results in between these two.

Much of what happens in oligopoly pricing is highly dependent on the specific legal structure within which firms interact. In Japan, where large firms are specifically allowed to collude, we see Japanese goods selling for a much higher price than those same Japanese goods sell for in the United States. For example, you may well pay twice as much for a Japanese television in Japan as you would in the United States. From the behavior of Japanese firms, we get a sense of what pricing strategy U.S. oligopolists would follow in the absence of the restrictions placed on them by law.

One of the things that limits oligopolies from acting as a cartel is the threat from outside competition. The threat will tend to be more effective if this outside competitor is much larger than the firms in the oligopoly.

For example, small-town banks have a tendency to collude (implicitly, of course), offering lower interest to savers and charging higher interest to borrowers than big banks charge, even though their average costs aren’t significantly higher. When I ask small-town banks why this is, they tell me that my perceptions are faulty and that I should mind my own business. But if a big bank, which couldn’t care less about increasing the wealth of a small-town banker, enters the town and establishes a branch office, interest rates to savers seem to go up and interest rates to borrowers seem to go down. The big bank can add significant competition—competition that couldn’t come from within the town.

On a national scale, the outside competition often comes from international firms. For example, implicit collusion among U.S. automobile firms led to foreign firms’ entry into the U.S. automobile market. There are many such examples of this outside competition breaking down cartels with no barriers to entry. Thus, a cartel with no barriers to entry faces a long-run demand curve that’s very elastic. This means that its price will be very close to its marginal cost and average cost. This is the same prediction that came from the contestable market theory.

On a national scale, the outside competition comes from international firms.

With oligopolies, there’s always the possibility of a price war. The reasons for such wars are varied. Since oligopolistic firms know their competitors, they can personally dislike them; sometimes a firm’s goal can be simply to drive a disliked competitor out of business, even if that process hurts the firm itself. Passion and anger play roles in oligopoly pricing because interpersonal and interfirm relations are important.

Alternatively, a firm might follow a predatory pricing strategy—a strategy of pushing the price down temporarily to drive the other firm out of business to increase long-term profits. Some argue that Microsoft followed a predatory pricing strategy by virtually giving away its Office Suite on new computer systems to make its software the industry standard. If the predatory pricing strategy is successful, the firm can charge an even higher price because potential entrants know that the existing firm will drive them out if they try to enter. It’s this continual possibility that strategies can change that makes oligopoly prices so hard to predict. (In a later chapter we will discuss game theory in detail, since it is an important tool that economists use to study strategic pricing by oligopolies.)

Since we have come to the end of our presentation of the analytics of market structure, a review is in order. The box “A Comparison of Various Market Structures” on page 321 provides that review. It lists the four central market structures, and the similarities and differences among them. It is worth a careful review.

WWW

Web Note 15.2 Porter’s Five Forces

An industry seldom fits neatly into one category or another. Inevitably, numerous arbitrary decisions must be made as to what the appropriate market is, and whether the industry comes closest to the characteristics of one or the other market structure. So to classify actual industries, a variety of procedures and measures have been developed, and in this section we review those procedures and measures.

To see the problems that arise in classifying industries, consider the banking industry. There are about 6,000 banks in the United States, and banking is considered reasonably competitive. However, a particular small town may have only one or two banks, so there will be a monopoly or oligopoly with respect to banks in that town. Is the United States or the town the relevant market? The same argument exists when we think of international competition. Many firms sell in international markets and, while a group of firms may compose an oligopoly in the United States, the international market might be more accurately characterized by monopolistic competition.

Another dimension of the classification problem concerns deciding what is to be included in an industry. If you define the industry as “the transportation industry,” there are many firms. If you define it as “the urban transit industry,” there are fewer firms; if you define it as “the commuter rail industry,” there are still fewer firms. Similarly with the geographic dimension of industry. There’s more competition in the global market than in the local market. The narrower the definition, the fewer the firms.

One of the ways in which economists classify markets in practice is by cross-price elasticities (the responsiveness of a change in the demand for a good to a change in the price of a related good). Industrial organization economist F. M. Sherer has suggested the following rule of thumb: When two goods have a cross-price elasticity greater than or equal to 3, they can be regarded as belonging to the same market.

The North American Industry Classification System (NAICS) is an industry classification that categorizes industries by type of economic activity and groups firms with like production processes. In the NAICS, all firms are placed into 20 broadly defined two-digit sectors. These two-digit sectors are further subdivided into three-digit subsectors, four-digit industry groupings, five-digit industries, and six-digit national industry groupings. Each subgrouping becomes more and more narrowly defined. Table 15-1 lists the 20 sectors and shows some subgroupings for one sector, Information, to give you an idea of what’s included in each.

When economists talk about industry structure, they generally talk about industries in the four- to six-digit subsector groupings in the United States. This is a convention. Economists are often called on to give expert testimony in court cases, and if an economist wants to argue that an industry is more competitive than its opponents say it is, he or she challenges this convention of using a four- to six-digit classification of industry, asserting that the classification is arbitrary (which it is) and that the relevant market should be the two- to three-digit classification.

To empirically measure industry structure, economists use one of two methods: the concentration ratio or the Herfindahl index.

A concentration ratio is the value of sales by the top firms of an industry stated as a percentage of total industry sales. The most commonly used concentration ratio is the four-firm concentration ratio. For example, a four-firm concentration ratio of 60 percent tells you that the top four firms in the industry produce 60 percent of the industry’s output. The higher the ratio, the closer the industry is to an oligopolistic or monopolistic type of market structure.

The Herfindahl index is a method used by economists to classify how competitive an industry is.

The Herfindahl index is an index of market concentration calculated by adding the squared value of the individual market shares of all the firms in the industry. For example, say that 10 firms in the industry each has 10 percent of the market:

The Herfindahl index weights the largest firms in the industry more heavily than the concentration ratio because it squares market shares.

Because it squares market shares, the Herfindahl index gives more weight to firms with large market shares than does the concentration ratio measure.

The two measures can differ because of their construction, but generally if the concentration ratio is high, so is the Herfindahl index. Table 15.2 presents the four-firm concentration ratio and the Herfindahl index of selected industries.

Source: Census of Manufacturers (factfinder2.census.gov).

The Herfindahl index plays an important role in government policy; it is used as a rule of thumb by the U.S. Department of Justice in determining whether an industry is sufficiently competitive to allow a merger between two large firms. If the Herfindahl index is less than 1,000, the Department of Justice generally assumes the industry is sufficiently competitive, and it doesn’t look more closely at the merger. This policy may change in the future.

Neither the four-firm concentration ratio nor the Herfindahl index gives us a picture of corporations’ bigness. That’s because many corporations are conglomerates—companies that span a variety of unrelated industries. For example, a conglomerate might produce both shoes and automobiles.

To see that concentration ratios are not an index of bigness, say the entire United States had only 11 firms, each with a 9 percent share of each industry. Both indexes would classify the U.S. economy as unconcentrated, but many people would seriously doubt whether that were the case. Little work has been done on classifying conglomerates or in determining whether they affect an industry’s performance.

To see how empirical measures of market structure relate to oligopoly models, let’s consider the cartel and contestable market models of oligopoly. The cartel model fits best with these empirical measurements because it assumes that the structure of the market (the number of firms) is directly related to the price a firm charges. It predicts that oligopolies charge higher prices than do monopolistic competitors, who in turn charge higher prices than competitive firms charge.

The contestable market model gives far less weight to the empirical estimates of market structure. According to it, markets that structurally look highly oligopolistic could actually be highly competitive—much more so than markets that structurally look less competitive. This contestable market model view of judging markets by performance, not structure, has had many reincarnations. Close relatives of it have previously been called the barriers-to-entry model, the stay-out pricing model, and thelimit-pricing model. These models provide a view of competition that doesn’t depend on market structure.

To see the implications of the contestable market approach, let’s consider an oligopoly with a four-firm concentration ratio of 60 percent and a Herfindahl index of 1,500. Using the structural approach, we would say that, because of the multiplicity of oligopoly models, we’re not quite sure what price firms in this industry would charge, but that it seems reasonable to assume that there would be some implicit collusion and that the price would be closer to a monopolist price than to a competitive price. If that same market had a four-firm concentration ratio of 30 percent and a Herfindahl index of 700, the industry would be more likely to have a competitive price.

A contestable market model advocate would disagree, arguing that barriers to entry and exit are what’s important. If no significant barriers to entry exist in the first case but significant barriers to entry exist in the second case, the second case would be more monopolistic than the first. An example is the Miami fish market mentioned earlier, where there were 20 sellers (none with a large percentage of the market) and significant barriers to entry (only fishers from the pier were allowed to sell fish there and the slots at the pier were limited). Because of those entry limitations, the pricing and output decisions would be close to the monopolistic price. If you took that same structure but had free entry, you’d get much closer to competitive decisions.

As I presented the two views, I emphasized the differences in order to make the distinction clear. However, I must also point out that there’s a similarity in the two views. Often barriers to entry are the reason there are only a few firms in an industry. And when there are many firms, that suggests that there are few barriers to entry. In such situations, which make up the majority of cases, the two approaches come to the same conclusion.

Now that we’ve gone over the four major market structures in theory, and the way in which industries are classified in practice, let’s consider government’s role in affecting market structure. That role goes under the name antitrust policy in the United States and competition policy in some other countries.

Antitrust policy is the government’s policy toward the competitive process. It’s the government’s rulebook for carrying out its role as referee. In volleyball, for instance, the rulebook would answer such questions as: When should a foul be called? When has a person caught and thrown rather than hit the ball over the net? In business a referee is needed for such questions as: When can two companies merge? What competitive practices are legal? When is a company too big? To what extent is it fair for two companies to coordinate their pricing policies? When is a market sufficiently competitive or too monopolistic?

At different times, U.S. antitrust law has been based on two competing views: judgment by performance and judgment by structure.

The United States has seen wide swings in economists’ prescriptions concerning such questions, depending on which of the two views of competition has held sway. The two competing views are:

1. Judgment by performance: We should judge the competitiveness of markets by the performance (behavior) of firms in that market.

2. Judgment by structure: We should judge the competitiveness of markets by the structure of the industry.

Two examples illustrate the difference.

In the late 1880s, a number of trusts (cartels) in the railroad, steel, tobacco, and oil industries were created by what were sometimes called robber barons (organizers of trusts who engaged in the exploitation of natural resources and other unethical behavior). The trusts were seen as making enormous profits, preventing competition, and in general bullying everyone in sight. One such cartel was the Standard Oil Trust, created by John D. Rockefeller, which used its monopoly power to close refineries, raise prices, and limit the production of oil. In response the U.S. Congress passed the Sherman Antitrust Act of 1890—a law designed to regulate the competitive process. In 1914 the Sherman Antitrust Act was clarified and strengthened with the Clayton Antitrust Act, which identified specific practices as illegal and monopolistic. The government brought a lawsuit against Standard Oil for violating the Sherman Antitrust Act.

In 1911 the U.S. Supreme Court handed down its opinion. It was determined that Standard Oil controlled 90 percent of the market and thus was definitely a monopoly. However, the Court decided that the monopolistic structure alone did not violate the Sherman Antitrust Act. To be guilty of antitrust violations there had to be evidence that the firm used its monopoly power to its benefit. In this case, the difference was academic because the court found that Standard Oil had done so and was guilty because it used “unfair business practices.” The resolution was to break up Standard Oil, which made the distinction between performance and structure academic.

The structure/performance distinction was also important in a case involving U.S. Steel. Here the Supreme Court ruled that although U.S. Steel controlled a majority of the market and was therefore structural monopoly, it was not a monopoly in performance. That is, it had not used unfair business practices to become a monopolist or once it was a monopolist, and thus it was not in violation of antitrust law. Unlike Standard Oil, U.S. Steel was not required to break up into small companies.

Judgment by performance was the primary criterion governing antitrust policy until 1945, when the U.S. courts changed their interpretation of the law with the Aluminum Company of America (ALCOA) case. ALCOA was the only producer of aluminum in the United States, a position it built by using its knowledge of the market to expand its capacity before any competitors had a chance to enter the market. Like Standard Oil, it had not used unfair business practices to become a monopolist. This time, the courts focused on the structure of the market and ruled ALCOA to be in violation of antitrust laws, even though it was not guilty of monopoly behavior.

Both judgment by structure and judgment by performance have their problems.

Both judgment by structure and judgment by performance have their problems. Judgment by structure seems unfair on a gut level. After all, in economics the purpose of competition is to motivate firms to produce better goods than their competitors are producing, and to do so at lower cost. If a firm is competing so successfully that all the other firms leave the industry, the successful firm will be a monopolist, and on the basis of judgment by structure will be guilty of antitrust violations. Under the judgment-by-structure criterion, a firm is breaking the law if it does what it’s supposed to be doing: producing the best product it can at the lowest possible cost.

An important reason supporting the structure criterion is practicality.

Supporters of the judgment-by-structure criterion recognize this problem but nonetheless favor the structure criterion. An important reason for this is practicality.

It isn’t only the federal government that has laws regulating competition. States do also. One such state law is Arkansas’s Unfair Practices Act, which prohibits selling, or advertising for sale, items below cost “for the purpose of injuring competitors and destroying competition.” In the early 1990s, three Arkansas pharmacies sued Walmart for violating this law by selling its goods at “too low” a price.

Walmart initially lost the suit in Arkansas; however, in 1995 the Arkansas Supreme Court overturned the lower court decision and held that Walmart’s pricing was not part of a strategy to price below cost over a prolonged period.

We may see more such suits, especially in the pricing of prescription pharmaceuticals, since Walmart introduced a $4 price for a 30-day prescription of a variety of generic drugs (including 14 of the top 20 best-selling prescription drugs) and other discount stores followed suit. However, these suits are likely to fail for two reasons. First, Walmart now takes into account state laws, and the $4 program is not available in states where its lawyers see the state laws as a potential problem. And second, the competition has changed. Back in the 1990s, Walmart was competing with local drug stores that had high profit margins but low volume. Now it’s competing with other chains—Kmart, Target, BJ’s, Sam’s Club—that have entered the prescription drug market and established their own low-cost generic programs.

© David McNew/Getty Images

All of these chains argue that when they charge a low price, they are not doing it to “destroy competition” or “injure competitors,” but rather to maintain low prices for consumers. They claim that their pricing policies promote, not destroy, competition.

In principle, most economists agree with Walmart and other chains; new competition, by its very nature, hurts existing businesses—that’s the way the competitive process works. Those who don’t sell for the lowest price lose, and those who do gain. But most economists also recognize that Walmart’s brand of competition can affect the social fabric of small-town economies. A new Walmart store can undermine the town centers and replace them with commercial sprawl on the outskirts of these towns. Whether these externalities are a reason to limit Walmart’s aggressive pricing policies is a debatable question.

Judgment by performance requires that each action of a firm be analyzed on a case-by-case basis. Doing that is enormously time-consuming and expensive. In some interpretations, actions of a firm might be considered appropriate competitive behavior; in other interpretations, the same actions might be considered inappropriate. For example, say that an automobile company requires that in order for its warranty to hold, owners of its warranteed vehicles must use only the company’s parts and service centers. Is this requirement of the automobile company intended to create a monopoly position for its parts and service center divisions or to ensure proper maintenance? The answer depends on the context of the action.

The problem is that judging each case contextually is beyond the courts’ capabilities. There are so many firms and so many actions that the courts can’t judge all industries on their performance. To solve this problem the courts limit the cases it looks at using market share, even though it is firms’ performance that will ultimately be judged.

Judging by market structure also has difficulties. As you saw in the discussion of monopolistic competition, it’s difficult to determine the relevant geographic market (local, national, or international) and the relevant industry (three-digit or five-digit NAICS code) necessary to identify the structural competitiveness of any industry.

The relevant industry question played a large part in the Department of Justice’s opposition to a merger between Gillette and Parker Pens in 1993. The government argued that the combined firm would control about 40 percent of the premium-fountain-pen market. The Court, however, allowed the merger, arguing that the relevant market was much larger—the market for premium writing instruments, which also included mechanical pencils, ballpoint pens, and rollerballs. The premium-writing-instruments market had many more competitors than the premium-fountain-pen market.

Similar ambiguities exist with the decision about the relevant geographic market. In the Pabst Brewing case (1966), the definition of the market played a key role. Pabst wanted to merge with the Blatz Brewing Company. On a national scale, both companies were relatively small, accounting together for about 4.5 percent of beer sales in the United States as a whole. Pabst argued that the United States was the relevant market. The Court, however, decided that Wisconsin, where Pabst had its headquarters, was the relevant market, and since the two firms held a 24 percent share of that market, the merger was not allowed.

What should one make of debates regarding relevant markets? The bottom line is that both structure and performance criteria have ambiguities, and in the real world there are no definitive criteria for judging whether a firm has violated the antitrust statutes. A firm isn’t at fault or in the clear until the courts make the call.

Both structure and performance criteria have ambiguities, and in the real world there are no definitive criteria for judging whether a firm has violated the antitrust statutes.

In recent years few major antitrust cases have been brought, in part because a century of experience has taught business what the law allows, and in part because the government has been lenient in its interpretation of the antitrust laws. That leniency has three interrelated causes. The first is a change in American ideology. Whereas in the 1950s and 1960s the prevailing ideology saw big business as “bad,” by the 1980s big business was seen as a combination of good and bad. In this new ideological framework, the political pressure to push antitrust enforcement waned. Second, as the United States became more integrated into the global economy, big business faced significant international competition and hence competition created by U.S. market structure became less important. Third, as technologies became more complicated, the issues in antitrust enforcement also became more complicated for the courts to handle. By the time the legal system had resolved a case, the technology would have changed so much that the issues in that case were no longer relevant. These issues can be seen in two recent cases—Microsoft and AT&T.

Since the 1980s the United States has been more lenient in antitrust cases because of a change in ideology, the globalization of the U.S. economy, and the increasing complexity of technology.

One of the most important antitrust cases brought in the 1990s was the Microsoft case which raises difficult questions about competition, the competitive process, and government’s role in that competitive process.

Microsoft makes computer software. From the company’s small start 40 years ago, sales of Microsoft software have grown to account for about 50 percent of the world’s software market. Its PC operating system, Windows, accounts for an even larger share—about 90 percent—of the world’s computer operating system software market.

Since all software must be compatible with an operating system, the widespread use of Windows gives Microsoft enormous power—power that competitors claim it has used to gain competitive advantage for its other divisions. Competitors’ calls for action, and reports of monopolistically abusive acts by Microsoft, led the U.S. Department of Justice in 1998 to charge Microsoft with violating antitrust laws.

The government suit against Microsoft charged the company with being a monopoly and using that monopoly power in a predatory way. Specifically, it charged Microsoft with:

Microsoft product shot(s) reprinted with permission from Microsoft Corporation

1. Possessing monopoly power in the market for personal computer operating systems.

2. Tying other Microsoft software products to its Windows operating system.

3. Entering into agreements that keep computer manufacturers that install Windows from offering competing software.

In a secretly recorded comment during a price-fixing meeting, the former president of Archer Daniels Midland (ADM), a major supplier of food and grain, stated, “Our competitors are our friends and our customers are our enemies.”

The U.S. antitrust laws concern far more than mergers and market structure; they also place legal restrictions on certain practices of businesses such as price-fixing. By law, firms are not allowed to explicitly collude in order to fix prices above the competitive level. A key aspect of the law is the explicit nature of the collusion that is disallowed. Airlines, gas stations, and firms in many other industries have prices that generally move in tandem—when one firm changes its price, others seem to follow. Such practices would suggest that these firms are implicitly colluding, but they are not violating the law unless there is explicit collusion.

To prove explicit collusion is difficult—there must be a smoking gun, and there is seldom sufficient evidence of explicit collusion to prosecute businesses. There are exceptions, however. In 1996, ADM was caught red-handed when one of its former officials gave prosecutors tapes of meetings in which price-fixing occurred. Meeting secretly around the world, in countries like Mexico, France, Canada, and Japan, ADM executives tried to fix prices of Lysine, a feed additive, and citric acid.

© AFP/Getty Images

One of ADM’s officials, working undercover for the FBI, secretly recorded these meetings. Faced with the taped evidence against them, ADM agreed to pay $100 million in fines— the largest criminal antitrust fine in history up to that year. Since that time, fines have risen to even greater sums, with LG, Sharp, and Hitachi paying a fine of $860 million for price fixing.

Microsoft had dominated the market for PC operating systems for about a decade. The U.S. Department of Justice argued that this long-standing monopoly position was the result of unfair business practices. Microsoft argued that Windows sold so well because it was a superior product. Microsoft further argued that, because it faced competition from technological change, it was not a monopolist.

Competition came from open-source (free) operating systems such as Linux and Java and more recently “cloud computing” services such as those provided by Internet browser Chrome. Cloud computing provides software services to cell phones, computers, and tablets over the Internet, which eliminates the need for compatibility between hardware and software. Each of these changes has eroded Microsoft’s monopoly advantage.

Whether one sees Microsoft as a monopolist depends in part on whether one views it in a static or dynamic framework.

With regard to the antitrust case, in 2000 a judge concluded that Microsoft violated the Sherman Antitrust Act and proposed breaking Microsoft into two companies. Microsoft appealed and eventually the Department of Justice and Microsoft reached an agreement. Microsoft would not be broken up, but its practices would be regulated to prevent predatory behavior that served to raise barriers to entry. This regulation reduced Microsoft’s dominance but the development of new technologies has played a larger role in reducing that dominance. As many of the functions of the PC have been replaced by mobile apps, many of which were not Windows-based. Microsoft’s monopoly over operating systems was reduced.

A second recent case is AT&T. The AT&T you know today isn’t the AT&T of the last century. Beginning in 1913 AT&T was what was called a regulated monopoly. It had the exclusive right to provide telephone service in the United States. AT&T controlled 90 percent of the telecommunications market: long-distance and local telephone service and the production of telephones themselves as well as other communications equipment. It was given that right because telephone service required substantial set-up costs—land lines that connect households. To have more than one company stringing competing lines made little sense, making it a natural monopoly. In exchange for exclusive rights to the telephone market, AT&T agreed to be regulated. AT&T was required to provide universal service to all Americans, even those living in rural and remote areas, where service was more costly to provide. Unregulated companies likely would have practiced cream skimming (providing service to low-cost areas and avoiding high-cost areas).

In the 1970s technological change fundamentally altered the nature of the longdistance telephone industry. Satellite transmission and fiber optics made physical line connections no longer the only option, so long-distance telephone service was no longer a natural monopoly, and slowly regulators allowed some competition to develop. In 1984, AT&T was broken up by a government antitrust case into seven Baby Bells, which continued to supply local telephone service in a regulated industry, and the parent company, AT&T, which could enter any unregulated industry it desired.

The breakup of AT&T was not the end of the changes. The seven Baby Bells merged with one another, and by 2005 only four remained: SBC Communications, Verizon, Bell South, and Qwest. In 1995, AT&T had divided itself into three companies: AT&T, Lucent Technologies, and National Cash Register. Of these, only AT&T remained in the market for communication services, expanding its offerings into wireless communications, digital cable, cable, and long distance. In 2004, however, it withdrew from the residential local and long-distance phone markets. Then in 2005 it was taken over by one of its former parts—SBC Communications, which operated Cingular Wireless—and the combined company chose to use the AT&T name. Then in 2006, the new AT&T was taken over by Bell South, another of the former Baby Bells, and the combined company again called itself at&t (initially in lowercase), making it the new, new at&t. So the name, at&t, still exists, but the company that the name is associated with is quite different from the company that was the subject of the antitrust suit, and it is a company that has twice been taken over by its former parts.

The new AT&T is still facing antitrust issues. In 2011 the Department of Justice sued AT&T and blocked its merger with T-mobile, a smaller cell phone company. The Department of Justice blocked the merger arguing that it would substantially limit competition and allow the combined company to raise prices for consumers.

What we have learned from this experience is that rapid technological change alters the nature of industries and introduces competition in ways that previously had not been possible. The slow moving antitrust laws are usually years behind.

Rapid technological change alters the nature of industries and introduces competition in ways that previously had not been possible.

Economic scholars’ overall assessment of antitrust policy is mixed. In certain cases, such as the ALCOA case, most agree that antitrust prosecution went too far. But most believe that other decisions (as in the 1911 Standard Oil case) set a healthy precedent by encouraging a more competitive U.S. business environment. Almost all agree that antitrust enforcement has not reduced the size of firms below the minimally efficient level, the level at which a firm can take full advantage of economies of scale. But they are mixed in their judgments as to whether the enforcement was needed. Performance advocates generally believe that it was not, while structural advocates generally believe that it was. They are also mixed in their judgment about whether any type of antitrust action is feasible in a technologically dynamic industry such as computers or telecommunications.

Economists’ judgment on antitrust is mixed.

We’ve come to the end of our discussion of market structure and government policy toward the competitive process. What conclusion should we reach? That’s a tough question because the problem has so many dimensions. What we can say is that market structure is important, and generally more competition is preferred to less competition. We can also say that, based on experience, government-created and protected monopolies have not been the optimal solution, especially when industries are experiencing technological change. But how government should deal with monopolies that develop as part of the competitive process is less clear. Competition has both dynamic elements and market structure elements, and often monopolies that develop as part of the competitive process are temporary—and they will be overwhelmed by other monopolies. Thus, the debate about government entering into the market to protect competition has no single answer, which makes cases like the Microsoft antitrust case difficult to resolve.

The debate about government entering into the market to protect competition has no single answer.

· The two distinguishing characteristics of an oligopolistic market are (1) there are a small number of firms and (2) firms engage in strategic decisionmaking. (LO15-1)

· A contestable market theory of oligopoly judges an industry’s competitiveness more by performance and barriers to entry than by structure. Cartel models of oligopoly concentrate on market structure. (LO15-2)

· An oligopolist’s price will be somewhere between the competitive price and the monopolistic price. (LO15-2)

· Industries are classified by economic activity in the North American Industry Classification System (NAICS). Industry structures are measured by concentration ratios and Herfindahl indexes. (LO15-3)

· A concentration ratio is the sum of the market shares of individual firms with the largest shares in an industry. (LO15-3)

· A Herfindahl index is the sum of the squares of the individual market shares of all firms in an industry. (LO15-3)

· Antitrust policy is the government’s policy toward the competitive process. (LO15-4)

· There is a debate about whether markets should be judged on the basis of structure or on the basis of performance. (LO15-4)

· Judgment by performance means judging the competitiveness of markets by the behavior of firms in that market. Judgment by structure means judging the competitiveness of markets by how many firms operate in the industry and their market shares. (LO15-4)

· In 2000 the courts found that Microsoft had a monopoly that was protected by barriers to entry and that Microsoft engaged in practices to maintain that monopoly power. Microsoft agreed to stop some practices. (LO15-4)

· The antitrust suit against AT&T ended in a settlement that required AT&T to be broken up. AT&T both divided itself and merged with other companies. (LO15-4)

· antitrust policy (324)

· cartel (316)

· cartel model of oligopoly (316)

· concentration ratio (322)

· contestable market model (318)

· Herfindahl index (322)

· implicit collusion (317)

· judgment by performance (324)

· judgment by structure (324)

· North American Industry Classification System (NAICS) (320)

· oligopoly (315)

· strategic decision making (315)

Economics, Ninth Edition

Chapter 15: Oligopoly and Antitrust Policy

ISBN: 9780078021701 Author: David C. Colander

Copyright © McGraw-Hill Company (2013)