Chapter 15 Busting Up Monopolies

Antitrust

By virtue of the Sherman Antitrust Act of 1890, the US government can take legal action to break up a monopoly. In 1902, President Theodore Roosevelt used the Sherman Antitrust Act as a basis for trying to break up the monopolization of railway service in the United States. The resulting legal case, known as Northern Securities Co. v. United States, involved two key elements: restraint of trade and interstate commerce.

Multiple providers of rail service had joined together in the Northern Securities Company, resulting in a restraint of trade. The Supreme Court decisionCornell University Law School, Legal Information Institute, “Northern Securities Co. v. United States (No. 277),” accessed March 14, 2011, http://www.law.cornell.edu/supct/html/historics/USSC_CR_0193_0197_ZS.html. provides direct discussion of the anticompetitive nature of the creation of this company, noting that the company had been set up to eliminate competition among the different providers. Further, this restraint of trade had interstate implications because the railway lines themselves were not contained within any one state. This was a key point of jurisdiction: the Sherman Antitrust Act spoke directly about the effects of market power on interstate trade. Under the US Constitution, interstate commerce is part of the responsibility of the federal government.

Another famous antitrust case decided by the Supreme Court (http://supreme.justia.com/us/221/1/case.html) centered around the breakup of the Standard Oil Trust in 1911.Justia.com, US Supreme Court Center, “Standard Oil Co. of New Jersey v. United States, 221 U. S. 1 (1911),” accessed March 14, 2011, http://supreme.justia.com/us/221/1/case.html. John D. Rockefeller founded Standard Oil in 1870 soon after the discovery of oil in Ohio. By the late 1870s, the company controlled nearly 90 percent of the refineries in the United States. Shortly thereafter, the Standard Oil Trust was formed with the idea that the individual shareholders in a group of companies would come together to create a single company. This company would then jointly manage all the companies that joined the trust. In doing so, a monopoly was created. (This, by the way, is where the term antitrust comes from.)

Once again, the issue of jurisdiction played a role. Standard Oil was originally deemed a monopoly by the Ohio Supreme Court, which ordered it to be broken up. Instead, the company was simply reorganized as a New Jersey–based corporation. Because the company had moved to another state, the Ohio ruling became irrelevant, and the federal government had to step in to prevent unlawful interstate trade. The court ruled that the trust be dismantled. In the modern era, similar concerns arise across national borders: if two companies merge in Europe, for example, US antitrust authorities still take an interest, and vice versa.

These cases may seem like ancient history. Yet they remain very relevant today for at least two reasons. First, the legal system relies heavily on precedent, meaning that cases decided today depend on past rulings in similar cases. The actions of the Supreme Court in these two cases created a significant precedent for the present-day antitrust actions of the federal government. Second, antitrust actions continue to this day. A recent case again alleging a violation of the Sherman Antitrust Act was brought against Microsoft Corporation by the US government in 1998. Microsoft was charged with monopolizing the software market through its Windows operating system. As stated in the case, “Virtually all major PC manufacturers find it necessary to offer Microsoft operating systems on most of their PCs. Microsoft’s monopoly power allows it to induce these manufacturers to enter into anticompetitive, long-term licenses under which they must pay royalties to Microsoft not only when they sell PCs containing Microsoft’s operating systems, but also when they sell PCs containing non-Microsoft operating systems.”A full description of this aspect of the case is contained in John Kwoka and Lawrence White, eds., The Antitrust Revolution (New York: Oxford University Press, 2004).

One noteworthy element of the case concerned the definition of the market. After all, there are competing operating systems to Microsoft, most notably Apple Computer’s operating system. If the court viewed these products as substitutes for Microsoft’s operating system, then the claim that Microsoft was a monopolist would be less clear. Interestingly, the court ruled that Apple was not in the same market as Microsoft.

Other Kinds of Price Discrimination

There are many kinds of price discrimination.

• Explicit segmentation. The monopolist may be able to divide the market into identifiable segments and charge different prices to different segments. For example, movie theaters often offer student and senior citizen discounts. As another example, businesses often charge different prices in different countries. The idea is that the monopolist can charge higher prices to segments that are less price sensitive.
• Self-selection. If a firm cannot explicitly identify different segments, it can set a menu of prices and allow customers to sort themselves into different groups. When Apple introduced the iPhone, it originally charged a high price, knowing that impatient customers would buy immediately. It then dropped the price to capture the more patient, more price-sensitive customers. Discount coupons are another example. Individuals with more time to spare will clip and redeem coupons—and these individuals will typically also be more sensitive to price.

From the perspective of a firm, the biggest danger of price discrimination is the possibility of arbitrageThe act of buying and then selling an asset to make a profit.. If you are selling your product more cheaply to some than to others, you don’t want someone to buy at the cheap price and then resell to your higher-value customers. This is a particular issue for pharmaceutical companies that charge different prices in different countries. If you need proof of this, look at the spam in your e-mail. The chances are very good that you have recently received an e-mail offering you pharmaceuticals at Canadian prices.

Competition Policy on Price Discrimination

The Robinson-Patman Act of 1936 directly forbids certain forms of price discrimination. The following is from the act, with our emphasis added: “It shall be unlawful for any person engaged in commerce…to discriminate in price between different purchasers of commodities of like grade and quality, where either or any of the purchases involved in such discrimination are in commerce…and where the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce.”Cornell University Law School, Legal Information Institute, “§ 13. Discrimination in price, services, or facilities,” accessed March 14, 2011, http://www.law.cornell.edu/uscode/15/13.html.

The type of price discrimination we have discussed certainly involves discrimination among different purchasers. What is less clear is whether such actions would “lessen competition or tend to create a monopoly.” In our earlier example, price discrimination actually makes the market more efficient, and in general, price discrimination can increase or decrease efficiency. There are no simple guidelines, and each case must be examined on an individual basis.

Price discrimination also occurs in business-to-businessA market where firms sell goods and services to other firms. transactions between firms. In a recent price-discrimination case involving Volvo, a car dealer charged that other dealers had obtained deeper discounts (price concessions) that permitted them to be more competitive.See Cornell University Law School, Legal Information Institute, “Volvo Trucks North America, Inc. v. Reeder-Simco GMC, Inc. (04-905) 546 U.S. 164 (2006) 374 F.3d 701, reversed and remanded,” accessed March 14, 2011, http://www.law.cornell.edu/supct/html/04-905.ZS.html. This seems to fit the requirements of the Robinson-Patman Act because Volvo trucks are a homogeneous good (Volvo trucks in this case), and the practice allows one dealer to undercut another.

A jury trial led to an award of $4.1 million to the injured company. Not surprisingly, Volvo appealed the decision, partly because, so it claimed, the other dealers were not in direct competition with the dealer filing the suit. Once again, you can see the critical significance of defining the market. The Supreme Court eventually decided the case in January 2006 in favor of Volvo.

The Final Stage: Competition

Once a firm’s patent expires, other firms can produce a similar or an identical product. The firm will then be operating in a competitive market and can no longer expect to gain any particular advantage from its innovation. When the patent for a pharmaceutical product expires, for example, other companies can step in and produce chemically identical copies of the product, known as generics.

In a competitive market, we expect the price of the product will decrease until it equals marginal cost. For this reason, the innovating firm cannot anticipate making very much profit at this stage. For simplicity, we can think of the firm making no profits. Although there may be some advantage in being the original producer of a product, any excess profits that remain after the patent expires are unlikely to be substantial. More precisely, the firm would earn no more than a “normal” level of profit—the same as it could earn in any other activity. Such normal profits would not provide any benefit to justify the initial innovation, so we can ignore them.

The Middle Stage: Patent Protection

If the innovating firm is going to make profits to justify the costs of developing its product, these profits must come in the middle stage when the firm has patent protection. During this period, the firm has monopoly power by virtue of the patent. We know how the firm behaves in this situation.

• It will produce a level of output such that marginal revenue equals marginal cost.
• It will set the price equal to the market’s willingness to pay for this output level.

This is exactly what we saw earlier in Figure 15.4 "Distortions Due to Market Power". The monopolist produces less output than is efficient and earns monopoly profits.

Notice that a firm’s decision about how much to produce and about what price to set does not depend on the costs that it paid for researching and developing its product. After the firm gets to this second stage of its decision, those costs are in the past. They are sunk costsA cost that, once incurred, cannot be recovered.. They have no influence on the marginal cost of production and the price/output decision of the monopolist.

To calculate the total profit that a firm earns in this monopoly stage, we must do two things: (1) calculate the firm’s profit in each year and (2) add these profits over the entire time that the firm has patent protection. The firm’s profits in any given year are given by

profits = revenues − total cost = revenues − variable cost − fixed operating cost.

The area shown as “monopoly profit” in Figure 15.4 "Distortions Due to Market Power" corresponds to revenues minus variable costs. (Businesspeople and accountants call this a firm’s profit contribution.) In any given year, the monopoly will also typically incur some costs of operation in any given year that are constant, irrespective of how much output it produces. Examples include rent on a building and other long-term contracts. These are the firm’s fixed operating costsCosts of operating a business that do not vary with the level of output., which also must be subtracted from the firm’s revenues to calculate its profits.

Because the monopolist earns profit in each year of its patent protection, we add these profits together. We do so using the tool of discounted present value. This calculation takes into account that money earned in the future is less valuable than money today whenever the rate of interest is positive. Thus the proper measure of the profits at this stage is the discounted present value of the sum of the profits made during the period of patent protection. Factors that would increase the discounted value of a firm’s profits include the following:

• Lower marginal costs
• More inelastic demand
• More years of patent protection
• Lower interest rates

The First Stage: Innovation

We are finally in a position to evaluate whether or not a firm should innovate. The gains from innovation are measured by the present discounted value of the flow of profits. A firm must compare these gains to the costs of innovation to determine whether or not the innovation is worthwhile. These innovation costs are determined by the costs of the research and development (R&D) process together with any other costs of market entry. For example, a firm must pay scientists and engineers, fund research laboratories and R&D departments, and so on. Having done the hard work of analyzing stage two, the decision for stage one is straightforward. The firm should follow this rule: “innovate if the discounted present value of profits is greater than the costs of innovation.” The firm should innovate as long as the monopoly profits it will earn in the second stage (appropriately discounted) are greater than the costs of innovation in the first stage.

Think again about a pharmaceutical firm. Such firms spend an enormous amount of money on the research and development phases of new pharmaceutical compounds. This occurs in stage one. Once the product goes to market, however, the costs of development are sunk and have no effect on the firm’s profits during stage two. The same point applies to the production of a music CD. The costs of producing, marketing, and distributing a typical CD are estimated to be around $500,000. Of this, at least $100,000 represents the costs of production.

One thing that we have neglected in our discussion is that the payoff from research and development efforts is typically uncertain. Many promising pharmaceutical compounds turn out, on further testing, to be ineffective or have unacceptable side effects. A band recording a new song cannot know for sure if it will sell hundreds of copies, thousands of copies, or millions of copies. The decision about whether or not to innovate must be based on a firm’s best estimates of the expected valueThe measure of how much you would expect to win (or lose) on average, if the situation were to be replayed a large number of times. of its profits.

An Example

Table 15.1 "Calculating the Discounted Present Value of Expected Profits" provides a numerical example of the innovation decision. The first year is the innovation stage: we suppose the cost of innovation is $150. The firm earns no revenues in that year and incurs no costs, so its first year profits are −$150. In the second year, the firm finds out if its innovation was successful. We suppose there is a 50 percent chance that it is successful, in which case the firm has monopoly power for the second and third years. It earns the revenues and incurs the costs listed in the third and fourth columns of the table. If the innovation is unsuccessful, it earns no profits. After the third year, the firm earns no profits. Suppose finally that the interest rate is 10 percent.

The profits in the second year if the innovation is successful are $132. The expected value of profits in that year is therefore given by (0.5 × $132) + (0.5 × 0) = $66. These must be discounted back one year using the 10 percent interest rate. In other words, we divide the second-year profits by 1.1. The discounted value of the second-year profits in the first year is therefore $60. The expected value of the third-year profits, by a similar calculation, is $121. These must be discounted back two years, all the way to the first year. To do so, we first divide by 1.1 to get the value of expected third-year profits in the second year ($121/1.1 = $110). We then discount this back another year by dividing by 1.1 again. The discounted value of expected third-year profits in the first year is therefore 110/1.1 = $100.

We can now legitimately add together the numbers in the last column, and we find that the discounted present value of the firm’s stream of profits is $160. This exceeds the cost of innovation ($150), so the firm should go ahead with the project. It expects to earn $10 from the project.

Commitment

Perhaps the government could both encourage innovation and avoid the monopoly distortion. Suppose that, after an innovation had been introduced, the government removed the patent protection. As an example, many individuals in the world suffer from HIV/AIDS but cannot afford medication at current prices. This problem is especially severe in much of sub-Saharan Africa. As a matter of social policy, one would like to have the drug companies first develop HIV/AIDS treatments and then sell those products at marginal cost.

Currently, some pharmaceutical companies have developed treatments that are still protected by patents. Eliminating these patents seems like it might be good social policy, given that the lives of millions of people are at stake, but it would come with a significant cost. If the government ignores patent protection for a particular product today, innovators will suspect that the government can no longer be relied on to provide patent protection in the future, which would have a huge impact on innovative activity. Thus by failing to provide patent protection for one product, the government risks destroying its reputation for patent protection in general. This is an example of a commitment problem. Prior to stage one of the innovation process, the government promises patent protection to provide an incentive for innovation. After stage one is finished and the product is introduced in the market, the government’s incentives change: it wants to remove the market distortions. Because the innovation stage is over, the government could potentially renege on its promise of patent protection.

Pharmaceutical companies like GlaxoSmithKline, which produce the antiretroviral drugs used for treating HIV, have in fact offered to make their products available more cheaply. However, pharmaceutical companies have strongly opposed suggestions that their patents not be honored. A better policy proposed by economists and others is a patent buyout. This would work as follows. The government would pay the pharmaceutical company a reasonable market price for its patent and then allow other companies to come into the market and produce generics. A related idea is that governments could offer to buy future patents if drug companies came up with treatments for particular maladies. For example, even though malaria is a major killer in the world, pharmaceutical companies have little interest in researching its treatment. The reason is one of harsh market economics: most sufferers from malaria are poor, so a malaria drug would not be very lucrative. If governments offered a substantial patent buyout for a malaria drug, however, firms might find it worth investing in this disease area after all.

International Dimension of Patents

If you have a US patent for a good that you have invented—for example, a new kind of printer—then you are protected in the United States. However, US patent law does not protect you if you are selling in other countries. A firm in another country could take your printer apart, analyze how it works (this is called reverse engineering), and then produce the good itself. If the firm tried to sell the good in the United States, you could take it to court, but if it is selling elsewhere in the world, you have very little protection.

Other countries (such as Japan) and regions (such as the European Union) have patent laws that are similar to those in the United States. (Indeed, one of the earliest biotech patents went to the famous biologist Louis Pasteur in 1873 for his method of producing yeast.) These laws are structured, as are those in the United States, to balance the gains from innovation against the costs of monopolization of the market. Thus producers who sell across the world will typically seek patent protection in many different countries.

Still, to the extent that the benefits of innovation flow to purchasers around the world, the innovation-monopoly trade-off is potentially altered. If we think about the stage two monopoly as “the price we pay for the benefits of innovation,” then residents in countries with strong patent laws are paying in part for benefits that flow to individuals and households in other economies. In some countries—China is a leading example—patents and copyrights are not very well protected. This is good news if you want to buy some cheap DVDs in China, but if you are the owner of the rights to these pirated movies, you surely wish that copyright laws were more thoroughly enforced around the world.

The Pricing Decision of a Single Firm

The first thing we need to know is the shape of the demand curve facing a firm. Let us look at the decision of one gas station (firm A). The owner of this gas station can look across the street and see the price set by its competitor (firm B) across the street. Suppose, for example, that firm B is selling gas at $2 per gallon. What does the demand curve for firm A’s gas look like?

• If firm A charges a price greater than $2 per gallon, all its potential customers will go across the street to firm B. The quantity demanded will be zero.
• If firm A charges a price less than $2 per gallon, it can capture the entire market. The demand curve faced by firm A is the same as the market demand curve.
• If firm A sets a price of exactly $2 per gallon, then the two firms will divide up the market. The simplest assumption is to suppose that they will each get half the market.

We illustrate this demand curve in Figure 15.6 "The Demand Curve Facing a Firm, Taking as Given the Price Set by a Competitor".

If firm B sets a price of $2, what should firm A do? As long as $2 is greater than the marginal cost, firm A makes the most profit if it undercuts firm B a little bit. If it sets a price of $1.99, it can capture the entire market, whereas if it sets a price of $2 it gets only half the market. But exactly the same is true of firm B. For any given price that firm A sets, firm B would do better to undercut it by a penny. Competition provides a strong incentive for firms to cut their prices.

Nash Equilibrium

The discussion so far tells us how one firm will respond to the price of the other, but we don’t yet know where the firms will end up. We do not yet know what the equilibrium will look like. Previously, we used the term equilibrium in the context of supply and demand, denoting the point where the supply and demand curves intersect. The idea of equilibrium goes beyond this, however: it denotes a situation of balance in which no one has any desire to change.

When we think about strategic situations with a small number of firms, we maintain the idea of an equilibrium where no one wants to change their decision. We use an idea of equilibrium invented by John Nash (a mathematician who won a Nobel Prize in Economics for this and other contributions). The concept of the Nash equilibriumA process for predicting outcomes in strategic situations. also expresses a sense of balance, but it is applied to strategic situations rather than markets. The key feature of the Nash equilibrium is that no one has any desire to change what is being done.

Nash Equilibrium When Firms Choose Prices (Bertrand Competition)

To see the Nash equilibrium in action, go back to our gas stations. So far, we have seen that—taking firm B’s price as given—firm A will want to set a lower price. For example, if firm B sets a price of $2, firm A will set a price of $1.99. But this is not a Nash equilibrium because firm B would then like to do something different. If firm A sets a price of $1.99, firm B will want to set a price of $1.98, and so on. This process will stop only when the firms’ prices equal marginal cost. Thus we can make an educated guess: in the Nash equilibrium, each firm sets its price equal to the marginal cost of production. In this equilibrium, both firms earn no profits.

To see that this is indeed an equilibrium, we suppose firm B sets its price equal to marginal cost and then ask if firm A would like to change its price away from marginal cost. Were firm A to set its price greater than firm B, it would get no sales and no profits. This does not make firm A better off. If firm A were to set its price below marginal cost, then it would capture the entire market, but it would make a loss on each sale. It would be selling below the costs of production and thus make negative profit. This certainly does not make it better off. So firm A has no incentive to do anything different. Obviously, the same arguments apply to firm B. If both firms set the price at the marginal cost, neither can change its price and make higher profit.

The game played between two firms producing an identical product and setting prices is called Bertrand competitionA situation in which two or more firms sell an identical product and set prices. In equilibrium, they all set price equal to marginal cost.. The remarkable implication of Bertrand competition is that the predicted outcome (price) will equal marginal cost. With only two firms, in other words, we get the same outcome as with a competitive market. Even with two firms, the market can be very competitive, and there may be no need for antitrust authorities to intervene.

The Prisoners’ Dilemma

Households certainly like Bertrand competition because they can purchase goods at marginal cost. The firms, however, clearly do not like the outcome. Both firms understand that they would both be better off if they could only charge more for a good. If they could meet together and collude, then they would want to behave like a monopoly and share the profit between them. But charging the monopoly price is not a Nash equilibrium. If one firm were to set the monopoly price, the other would have an incentive to undercut its rival by charging a slightly lower price.

Here is another way of looking at this problem. Let us again consider a situation where two firms are choosing their prices. To keep things simple, suppose now that the firms must choose between two prices: the (high) monopoly price and the (low) competitive price. We will no longer assume constant marginal cost, so the firms still earn some profit when the price equals marginal cost. There are thus four possibilities: (1) both firms set a high price, (2) both firms set a low price, (3) firm A sets a high price and firm B sets a low price, and (4) firm A sets a low price and firm B sets a high price.

Suppose that Figure 15.7 "The Payoffs (Profits) from Different Pricing Choices" shows us the profits that firm A earns in each case. (There is a similar figure for firm B.) When both firms set high prices, firm A earns profits of $100 (think of this as monopoly profits of $200 that they share.) However, if firm B sets a high price, firm A is better off setting a low price. In this case, firm A gets profits of $120. So if firm B sets a high price, firm A does best by setting a low price. What about if firm B sets a low price? Then firm A gets nothing if it sets a high price and profits of $60 if it sets a low price. Firm A is clearly better off setting a low price in this case as well. No matter what firm B does, firm A should set a low price. The same is true for firm B, so the Nash equilibrium is for both to set a low price. If they could collude, they would both agree to set a high price and earn higher profits. But collusion is not a Nash equilibrium because both firms have an incentive to cheat. This is an example of a prisoners’ dilemmaThere is a cooperative outcome that both players would prefer to the Nash equilibrium of the game. game.

Is there any way that the firms might be able to change the incentives so that they can collude? Each firm would like some means of punishing the other if it cheats. If they get to set their prices only once, there is no obvious punishment. But if these firms are competing over a long period of time—as happens in the real world—then more possibilities open up.

Think again about our two gas stations. They might both agree to set a high price for their gas. Then they keep a careful eye on each other. As long as firm B keeps its prices high, firm A is content to do the same. But if firm B ever drops its prices, then firm A can punish it by retaliating. More specifically, suppose the firms adopt the following rules:

• We both set the monopoly price (cooperation).
• However, if one of us ever fails to set the monopoly price (defection), then the other firm will set the competitive price forever after.

Figure 15.8 "The Payoffs (Profits) from Cooperating and Defecting" shows an example of firm A’s profits in this case. (This should now be understood as a discounted present value of profits because we are imagining the firms competing for months or years.)

Will the firms have an incentive to follow these rules? Put yourself in the shoes of firm A. If firm B is charging the monopoly price, then you can make a quick profit by undercutting firm B’s price. This will generate a gain in that period. But there is a cost: in future periods, firm B will charge the competitive price, and your profit will be driven to $0. For the numbers in the table, it is better for firm A to cooperate if firm B also cooperates. Thus if two firms compete with each other over and over again, they may be able to sustain collusive high prices.

There is something else striking about Figure 15.8 "The Payoffs (Profits) from Cooperating and Defecting". It is a Nash equilibrium for both to cooperate, but it is still also a Nash equilibrium for both to defect. Look at the profits of firm A if firm B defects. Firm A earns $300 if it defects but only $180 if it cooperates. If you thought the other firm was going to renege on the agreement, then you would want to do the same thing. Economists say that this game has multiple equilibria. Expectations are critical: if each firm expects the other firm to cooperate, then they will indeed both cooperate; if each firm expects the other firm to defect, then they will indeed both defect. This is called a coordination gameThere are multiple Nash equilibria, and the players all agree on the ranking of these equilibria..

One final note: we are showing how firms can (and often do) sustain high prices even in the face of competitive pressures. We are not suggesting that this is what you should do if you are ever responsible for setting prices! Conspiring to set high prices is very often a violation of antitrust laws.

Price Competition with Imperfect Substitutes

Up to this point, we supposed that the two firms were producing an identical product. If we think of two firms producing goods that are close—but not perfect—substitutes, we still reach very similar conclusions. For example, suppose there are two pizza restaurants on the same street. If one restaurant undercuts the other’s price, it would no longer expect to immediately capture the entire market, but it would still expect to gain a lot of business. In this situation, the arguments that we have just made still apply. Each restaurant would have an incentive in the short run to undercut the other’s price. If they compete repeatedly, however, they may be able to sustain high prices.

The Capacity Decision of a Firm

We again consider a situation where two firms are competing in the same market. As before, the first step is to determine the demand curve facing an individual firm. Figure 15.9 "The Demand Curve Facing One Firm Shifts to the Left as the Other Firm Increases Its Output" shows our gas station example again. In contrast to our previous analysis, firm B is choosing how much to produce (that is, how much gas to sell) rather than what price to set. As firm B increases its output, the demand curve faced by firm A shifts to the left. If firm B produces nothing, firm A faces the entire market demand curve. If firm B produces (sells) 30,000 gallons, firm A’s demand curve is shifted to the left by that amount. You can see that the demand curve faced by firm A has a familiar shape, unlike the odd demand curve in Figure 15.6 "The Demand Curve Facing a Firm, Taking as Given the Price Set by a Competitor".

When its demand curve shifts to the left, firm A’s marginal revenue curve also shifts to the left. Figure 15.10 "Firm A’s Profit-Maximizing Choice of Output as Firm B Changes Its Level of Output" shows what happens. Note that the downward-sloping curves here are now marginal revenue curves, not demand curves. We omitted the demand curves to keep the diagram from being too cluttered.

Part (a) of Figure 15.10 "Firm A’s Profit-Maximizing Choice of Output as Firm B Changes Its Level of Output" shows two marginal revenue curves for firm A associated with different levels of output for firm B. An increase in firm B’s output causes the marginal revenue curve facing firm A to shift to the left. How will firm A respond? As always, we know it will produce a level of output such that marginal revenue equals marginal cost. So as the marginal revenue curve shifts inward, firm A will produce less output. If firm B produces more output, firm A will produce less. This response of firm A to firm B is shown in part (b) of Figure 15.10 "Firm A’s Profit-Maximizing Choice of Output as Firm B Changes Its Level of Output". Here the output of firm B is on the horizontal axis, and the output of firm A is on the vertical axis. The downward sloping curve, sometimes called a reaction curveA curve that shows what happens to one player’s best strategy when the other player’s (or players’) strategy changes., shows us the output of firm A for every level of output of firm B.

Nash Equilibrium Revisited

We can now predict what will happen in this market. To simplify matters, we assume that the two firms are identical. This will make it easier to find a Nash equilibrium. In a Nash equilibrium, the following things are true.

• Firm A is choosing the level of output that maximizes its profits, which is based on its prediction of how much output firm B is producing.
• Firm A’s prediction about firm B’s level of output is correct.
• Firm B is choosing the level of output that maximizes its profits, which is based on its prediction of how much output firm A is producing.
• Firm B’s prediction about firm A’s level of output is correct.

If the two firms are identical, they will produce the same levels of output in the Nash equilibrium. Then, as shown in Figure 15.11 "Nash Equilibrium for Quantity Game", the equilibrium level of output corresponds to the intersection of the reaction curve and the 45-degree line. It is at this point, and only at this point, that all four conditions that we have listed hold. To understand this, put yourself in the position of firm A. You make a forecast about how much firm B will produce. Suppose you correctly forecast firm B’s profit-maximizing quantity. Then you will respond with your own profit-maximizing quantity. This is the point labeled as the Nash equilibrium in the figure. But why should you predict that quantity for firm B? That quantity is in fact its profit-maximizing choice, given what you are doing. The beliefs that each firm has about the other’s actions are consistent, and indeed they are self-enforcing.

Determining Prices

Now that we know how firms choose capacity (quantity), how are prices determined? The answer is easy: prices come from the demand curve. If the two firms are producing identical products, the price comes from the market demand curve, given the total output of the two firms. This is similar to what we did in the monopoly case: given the output level, we turned to the demand curve to find the price. If the two firms are producing products that are not perfect substitutes, then the analysis is similar. However, there is not one demand curve in this case; there are two. Each firm faces a demand curve that depends on the output of the other firm, as shown in Figure 15.12 "The Markets for Both Firms".

Inefficiency

When firms are choosing the price to set, it is possible for competition between two firms to drive prices all the way down to marginal cost and eliminate all monopoly inefficiency. This is what we see with Bertrand competition. When firms choose quantity rather than price, the effects of competition are much weaker. Look again at Figure 15.12 "The Markets for Both Firms". You can see that both firms are setting price in excess of marginal cost: there is still a distortion due to market power.

Competition between the firms does matter, however. Increases in firm B’s output, for example, shift firm A’s demand curve to the left. As a consequence, firm A ends up choosing a lower price than it would otherwise. Competition from firm B helps keep firm A’s prices low. By similar reasoning, competition from firm A helps keep firm B’s prices low.

In an ideal world (from their point of view), these firms would both limit their output further to get closer to the monopoly outcome. Exactly the same kind of strategies that we discussed earlier could come into play: firms that compete repeatedly over a long period of time might tacitly agree to reduce output further, punishing any defection by increasing output and cutting prices.

Collusive Behavior

The Sherman Antitrust Act is not only used against existing monopolies but is also more generally applied to groups of firms that act to jointly monopolize a market. If multiple firms act collusively to exert market power, they may fall foul of the act. Indeed, in the Standard Oil case, the key issue was the way the Standard Oil Trust brought multiple firms together into a single decision-making unit.

One form of collusive behavior occurs when firms come together to jointly decide on output levels and/or the price of the goods and services they sell. Occasionally, managers of firms are foolish enough to get caught on record making such proposals. In a famous incident a couple of decades ago, the CEO of American Airlines, Robert Crandall, proposed a 20 percent price increase to his counterpart at Braniff Airlines. Unfortunately for Crandall, the conversation was taped. (Interestingly, Crandall was not actually guilty of a violation of the Sherman Antitrust Act because no actual price-fixing took place.) More often, such agreements are likely to be tacit. One firm may try raising its prices, to see if others will follow. It can be very hard for the antitrust authorities to determine if price-fixing is actually occurring.

Other countries also have laws and agencies that seek to prevent collusion by firms. For example, in April 2008, the Office of Fair Trading in Britain charged two tobacco companies of colluding with supermarkets to set high prices for cigarettes. Specifically, the Office of Fair Trading said that the companies had set up arrangements “linking the retail price of a manufacturer’s brand to the retail price of a competing brand of another manufacturer.”Quoted in G. Wearden, “OFT Accuses Tobacco Firms and Retailers of Cigarette Price Fixing,” Guardian, April 25, 2008, accessed January 31, 2011, http://www.guardian.co.uk/business/2008/apr/25/regulators.retail.

Another form of collusive behavior occurs under the heading of “bid rigging.” Suppose only a few dairies provide milk to all schools in a region of the country. The schools set up auctions to decide which dairies will supply milk to different school districts. There is one auction for each district, and several dairies compete to provide milk to the different districts. The situation seems at first glance to be very competitive. We might expect the dairies to compete with each other in all the auctions, with this competition driving down milk prices for the schools. But imagine instead that the dairies agree ahead of time to divide up the districts. So for example, if there are three districts and three dairies, the dairies might agree that dairy 1 would win the auction in one district, dairy 2 in another, and dairy 3 in the third. They do this by each putting in very high bids (meaning they would charge a lot for a service), in the districts where they do not want to win the auction, allowing one dairy to win the auction and still charge a high price. Exactly such a scheme occurred in Texas, and an individual was charged with a felony.

Mergers and Acquisitions

Suppose you heard that Apple Computer and Microsoft Corporation were proposing a merger. You can perhaps imagine Steve Jobs and Bill Gates telling us how much we as consumers would benefit from this merger. They would say that Apple and Microsoft could combine the best features of their operating systems. The two companies could avoid costly duplication of research, so they would be able to provide goods more cheaply. They would no longer need to spend so much on advertising, again providing savings that could be passed onto the consumer, and so on. In a nutshell, Jobs and Gates might claim, the merger would bring new exciting products with lower costs of distribution and marketing.

You can be sure, however, that the Department of Justice, European antitrust authorities, and other similar bodies throughout the world would look on a proposed Microsoft-Apple merger with a highly skeptical eye. They would carry out their own studies of the costs and benefits of the merger. Even if the merger were to bring all the advertised benefits, it would also make the computer operating system market much less competitive. If there are originally two firms in a market and they then merge, they become a monopoly. The analysis in this chapter allows us to predict that output would decrease and prices would increase. Because a market with only two firms may still be very competitive, the loss in buyer surplus going from a competitive market to a monopoly can be sizable.

Stemming from the Clayton Act, the Department of Justice and the Federal Trade Commission must approve mergers and acquisitions of larger companies. The guidelines used for those decisionsUS Department of Justice, “Horizontal Merger Guidelines,” accessed March 14, 2011, http://www.usdoj.gov/atr/public/guidelines/hmg.htm. emphasize two general points: (1) the effect of the merger on efficiency and (2) the effect of the merger on market power. Efficiency here refers to cost efficiency—the extent to which a merged company will be able to reduce its costs of production. The Department of Justice puts it as follows in its merger guidelines: “Efficiencies generated through merger can enhance the merged firm’s ability and incentive to compete, which may result in lower prices, improved quality, enhanced service, or new products.”US Department of Justice and Federal Trade Commission, Revised Section 4 Horizontal Merger Guidelines, April 8, 1997, accessed March 1, 2011, http://www.ftc.gov/bc/docs/horizmer.shtm.

Efficiency may also relate to the quality of the good being produced. If a merger enables a better product to be produced at the same price as before, then the merger increases market efficiency. Put differently, if the merger increases the surplus of buyers, either because the product is improved or because the reduced costs of production lead to a lower price, then the merger has had a positive impact on efficiency. For example, one benefit from airline mergers might be the more efficient use of the information systems that handle travel reservations. Another might be more efficient use of airport landing rights.

With these guidelines in mind, how would the Department of Justice respond to a proposed merger of Microsoft and Apple? They would study the proposed merger with the goal of determining if the merger would create a more efficient market or if it would increase the market power of the sellers. Department of Justice economists would use frameworks like the ones we have presented in this chapter to help them predict the outcome of the merger.

The Department of Justice would almost certainly decide not to approve an Apple-Microsoft merger. However, it does not block most proposed mergers. There is debate among economists about whether the antitrust authorities are too lenient or too stringent. A recent study by economists Orley Ashenfelter and Daniel Hosken looked at five mergers that they suspected of being anticompetitive.Orley Ashenfelter and Daniel Hosken, “The Effect of Mergers on Consumer Prices: Evidence from Five Selected Case Studies” (National Bureau of Economic Research Working Paper 13859), February 2008. They compared prices of the goods produced by the merged firms with prices of goods that were not close substitutes yet had similar costs. They concluded that, in four of the five cases, prices increased between 3 percent and 7 percent, and “given the large amount of commerce in these industries, the implied transfer from consumers to manufacturers is substantial.” The evidence from this study thus suggests that the antitrust authorities are too permissive in allowing mergers.

The European Union also watches mergers closely, likewise balancing cost-efficiency and competitiveness considerations. Here is a statement of its policy on mergers: “If the annual turnover of the combined businesses exceeds specified thresholds in terms of global and European sales, the proposed merger must be notified to the European Commission.…These rules apply to all mergers no matter where in the world the merging companies have their registered office, headquarters, activities or production facilities. This is so because even mergers between companies based outside the European Union may affect markets in the EU.”See European Commission, “Mergers: Overview,” accessed March 14, 2011, http://ec.europa.eu/comm/competition/mergers/overview_en.html.