• SUPPORT CATO
• FOR STUDENTS
• FOR MEDIA
• FREE NEWSLETTERS

• HOME

• ABOUT

• BLOG

• CATO STORE

• CENTERS & PROJECTS

• EVENTS

• EXPERTS

• MULTIMEDIA

• PUBLICATIONS

Books

Reviews & Journals

Public Filings

Policy Studies

Opinion & Commentary

• RESEARCH AREAS

• SPEAKERS BUREAU

• SUPPORT CATO

• CATO UNBOUND

• DOWNSIZING
  GOVERNMENT.ORG

• ELCATO.ORG

• LIBERTARIANISM.ORG

The Cato Review of Business & Government

---

Antitrust as Antimonopoly

Douglas H. Ginsburg

Douglas H. Ginsburg is a judge on the United States Court of Appeals
for the District of Columbia Circuit and Foundation
Professor of Law at George Mason University.

Antitrust policy over the past hundred years has lacked both consistency and a firm basis in economics. Antitrust has been used to block economically efficient mergers and related transactions such as joint ventures, and at least until the late 1970s, it was also used against business practices in markets that were functioning quite well without help from the government's visible hand. Antitrust enthusiasts, from academics to plaintiffs' lawyers to members of the Supreme Court, try to justify these efforts as appropriate, if perhaps unsuccessful, attempts to "promote competition:' That justification only highlights the problem. Government should not and in any event cannot "promote competition:' The only practical antitrust policy is simply to preserve competition from acts by those who would displace it with monopoly.

To promote competition is to intervene whenever it appears to the government that a market could, at reasonable cost, be made to work better. Promoting competition leads the government to investigate the practices of the most successful firm in each market in search of an anticompetitive explanation for the firm's success. This is what Donald Turner once called the "inhospitality tradition" of antitrust. Mere market power, far short of actual monopoly becomes the target of enforcement efforts. But market power is literally ubiquitous; it neither can nor should be stamped out, for it is precisely the quest for market power that attracts resources into new ways of appealing to the consumer. Market power is the reward that induces the rivalry that benefits the consumer, which is what antitrust is supposed to be all about.

Furthermore, antitrust's obsession with market power, which is typically viewed solely in terms of power over price, does not even begin to address antitrust's stated goal of promoting competition. There are almost always nonprice dimensions (service, reliability, and so on) in which firms compete. Measuring the competitiveness of a market solely by reference to price is like valuing a stock solely by its price-earnings ratio, without regard to yield, quality of earnings, growth prospects, or any other dimension of value.

Preserving competition from acts by those who, in the terms of the Sherman Act, would "combine, contract, or conspire" against it, that is, those who would replace competition with monopoly, is a more modest but also a more practical goal for antitrust. To reach this goal we need inquire only whether firms have combined, by merger or by agreement, to such an extent that monopoly has displaced competition in the market. We do not need to investigate the source of one firm's success at the expense of its rivals, nor to make fine distinctions between more and less competitive market conditions. This becomes clearer when we consider the nature of economic competition and our reasons for favoring competition over monopoly.

The Nature of Economic Competition

Let us start at the beginning, with the definition of competition. The word has Latin roots signifying a "seeking together," and Webster's adequately defines competition as "seeking or endeavoring to gain that for which another is also striving." Thus, there are only two elements essential to competition: two or more actors (competitors) and a common goal toward which they strive.

In the wild there is competition but no competition law; animals or species of animals compete fiercely for the scarce resources necessary for their survival. There are no rules. The law of the jungle, the battle to the death, sorts out the winners and losers as reflected in the life span of a particular animal or the persistence or extinction of a species.

To preserve competition from acts by those who would replace it with monopoly we need inquire only whether firms have combined, by merger or by agreement, to such an extent that monopoly has displaced competition in the market.

In any competition that takes place among people in a civilized society, however, there must be some limits or ground rules. (Unlimited competition is incompatible with social life; it exists among human societies only when there is the complete breakdown of order known as total war.) It follows that there must be an authority that can, by consent or by force, establish and maintain limits on the forms that competition takes by ruling out certain types of conduct.

Thus, in the somewhat more civilized groves of academe, where scholarship and teaching are the accepted forms of competition for tenure, sabotaging a colleague's experiment is out of bounds. In a bureaucracy, whether it is the welfare department of a large American city or the Soviet ministry of agriculture, the road to advancement may be in making one's superiors look good, perhaps by helping them to meet quotas; but altering the personnel file of a competitor for promotion is not permissible. Likewise, in the economic marketplace where all the producers of goods and services compete to acquire wealth, we allow some forms of rivalry such as cutting price and improving quality while prohibiting others, such as fraud, commercial defamation, and industrial espionage.

There are two types of prohibitions against particular forms of economic competition-rules devised with economic competition specifically in mind and rules of general application. To illustrate, the rule against price-fixing is specific to competitors. In contrast, torching the plant of one's competitor is not prohibited specifically; rather, arson is prohibited generally without regard to the relationship between arsonist and victim. Such a prohibition of general application clearly affects the terms upon which competition may take place, but it is not helpful to think of it as a part of competition policy.

General laws that only incidentally affect the forms of competition reflect broader political choices about the social order. Thus, if we do not want to live in a society in which children work instead of play we prohibit competition resulting from the use of child labor. When the laws of general application have been set in place, however, there is still a very considerable area of conduct left open for economic competition. It is here, within competition's domain, where lawful goods are offered to consumers without resort to force or fraud, that we find the laws specific to competition policy (which, for historical reasons, is called antitrust in the United States). To fulfill their purpose, these rules must be derived from and be consistent with the underlying reasons for preferring competition.

Antitrust vs. Competition

The mere existence of the antitrust laws is testimony to the favored status of competition over monopoly in our society. We favor competition despite the fact that (like democracy) it is often wasteful and encourages aggressive behavior, because (also like democracy) it produces more of what we want than does the alternative. As Professors Philip Areeda and Louis Kaplow put it in Antitrust Analysis: "Monopoly… produces economic inefficiency.... Where a single firm controls the production of [a] commodity the output will be smaller and the price higher [or, I might add, the quality lower, which is the same thing]. Consumers unable to pay the monopolist's high price will spend their funds elsewhere. . . [on] commodities that [they] would not want under competitive pricing conditions.... [This] distort[s] resource allocation away from the maximum satisfaction of consumer wants."

These conventional observations about competition and monopoly are drawn from the static model that economists use to simplify complex phenomena for purposes of analysis. In Understanding Antitrust and Its Economic Implications, two other antitrust scholars, Professors E.T. Sullivan and J.L. Harrison, describe the model as follows: "Levels of competition can be described along a continuum, extending from what is called 'perfect competition' to 'pure monopoly: In order for a market to be perfectly competitive, several conditions must hold: there are many buyers and sellers; no individual firm is large enough to affect price by individual action; products in the market are homogeneous, with each product capable of serving as a substitute for another; barriers to entry do not exist; and the ability to increase production is without restriction. In addition, producers and consumers have complete information of all relevant market factors. Perfect competition is characterized by uncoordinated, individual decisionmaking by each producer and consumer.

As Professor Harold Demsetz has pointed out, however, in this model "[c]ompetition is assumed, and not really analyzed. . . . The true function of the model is to understand decentralization, not to analyze competition processes." Furthermore, to make the model relevant to practical antitrust policymaking, it is often necessary to ignore the condition that the "products in the market are homogeneous with each product capable of serving as a substitute for another."

Because products are hardly ever truly homogeneous, it is necessary to decide which among the multitude of heterogeneous products competing for the consumer's dollar are sufficiently close substitutes for one another to be considered in "the relevant market:' Do colas constitute a separate market? Or are other carbonated soft drinks in the same market? Are fruit-flavored drinks in that market? Juices? Bottled water? Tap water? Beer? These issues were examined in excruciating detail when the FTC challenged the merger of Coca-Cola and Dr. Pepper (FTC v. Coca-Cola Co. (1986)). Because the degree to which two products compete with each other is determined by consumers, who may number in the millions and whose preferences vary by degrees, every product is in at least some degree of competition with every other. To do antitrust analysis, however, a line must be drawn around those that are in sufficiently close competition that an increase in the price of one will cause a significant number of consumers to switch to the other(s). But it would be foolish indeed to suggest that such a line can often be drawn without being arbitrary.

The problem of market definition is inherent in the antitrust enterprise: one cannot even talk of the existence of competition, much less the promotion of competition, apart from a defined market context. Thus, any improvement in antitrust analysis, including the reform advocated in this article, can be set to naught by a procrustean approach to market definition that creates the appearance that there are little monopolies everywhere. Indeed, such market gerrymandering is not merely of hypothetical concern; the Antitrust Division once sued each of the three television network companies for monopolizing the "market" for production of "television entertainment programs exhibited on [its own network] during prime evening hours:' (No kidding.) The problem of market definition, because it too implicates a search for degrees of competition among various potential substitutes, compounds the unreliability of traditional antitrust analysis by adding a second dimension to the supposed continuum of competition underlying the model of perfect competition.

The problem of market definition, because it implicates a search for degrees of competition among various potential substitutes, compounds the unreliability of traditional antitrust analysis by adding a second dimension to the supposed continuum of competition underlying the model of perfect competition.

Notwithstanding these limitations, in the world of antitrust, as Demsetz describes, "the assumptions upon which the [decentralization] model rests and the deductions made from these are the sources of most of the standards used to judge the intensity of competition in real markets:' Policy is made on this basis, in part because it is quite difficult to test the implications of the model against real-world observations about the operations of markets. In fact, rigorous empirical evidence for any antitrust policy is scarce and is likely to remain so.

The assumptions of the model have been carried into antitrust law even by the Supreme Court. In the 1960s the Court interpreted the 1950 amendment of the Clayton Act-which condemns mergers "the effect of [which] may be substantially to lessen competition, or to tend to create a monopoly" -to make the model's atomistic market of small firms, each without any market power, the competitive ideal (Brown Shoe Co. v. United States (1962). As in the model, the Court assumed that there are levels of competition and that certain transactions or practices can "lessen competition." The implicit notion that "competition can be described along a continuum" has not been much questioned, although it is surely beyond the state of the art in economics to support fine distinctions among market structures on a continuum of competitiveness (such as a market concentration index).

Undaunted by this difficulty antitrust analysis emphasizes the importance of blocking the creation or exercise of mere "market power" rather than the attainment of actual monopoly In economic theory market power is defined as a producer's ability to set price above the theoretical competitive level (equated with the marginal cost of producing the good) without losing so many sales as to lower his profits. Because market power is defined by reference to a theoretical benchmark price, it, like competition, is thought of as coming in degrees. A small degree of market power-the dry cleaner can charge more because it is two blocks to the nearest competitor- is tolerated as a regrettable but unavoidable fact of economic life.

When it comes to applying the concept of market power to identify practices or transactions that warrant antitrust objection-that is, those that entail an increase in market power worth proceeding against - antitrust analysts recognize, of course, that they cannot measure market power directly. Economic tools cannot tell them whether a particular price is above the level that would obtain absent market power, nor whether a change in market structure will actually cause a change in market prices. Lacking the means to confront market power directly antitrust analysts instead resort to identifying and trying to avoid the circumstances that are likely-on the basis of numerous assumptions of their model-to give rise to market power. One assumption is that market concentration-and particularly large market shares among the leading firms-implies market power. Antitrust thus attempts to ameliorate assumed problems in "the relevant market" by attacking the circumstances that are likewise assumed to increase the degree of market power that firms have. Although the competitive ideal entails zero market power, analysts realize that the complete elimination of market power-an infinite number of dry cleaners located cheek by jowl-is not practical, and perhaps not even desirable.

There is virtually no empirical evidence for the proposition that the intensity of or the economic benefit from competition in a market is inversely related to concentration in that market.

Undue attention to the model of perfect decentralization has, I believe, obscured a fundamental point, not about the model of "perfect competition" but about the problem with "pure monopoly:' Monopoly involves the complete (or near complete) occupation of a market by a single firm. The problem is not the lack of many, but the lack of any competitors. While our understanding is too limited to discern degrees of competitiveness, even if they exist, we can be highly confident that competition, in whatever form it takes; is preferable to monopoly. As long as there are even two significant competitors in a market, there is reason to believe that each firm will be under constant pressure to decrease price or to improve quality in an attempt to keep from losing sales to its rival.

Why, then, is antitrust so often concerned with the merger of two firms in a market in which there will still be two or more significant firms left to compete? There is virtually no empirical evidence for the proposition that the intensity of or the economic benefit from competition in a market is inversely related to concentration in that market. Almost all of the more rigorous research has been done in the past decade, long after the prevailing antitrust dogma had been worked out. Even Leonard Weiss concludes that "concentration makes little difference" when the four leading firms in a market account for less than half the sales. In general, the best data in support of the proposition that competition is less vigorous in the most concentrated markets derive from rather atypical markets that come closest to the model of perfect competition. One involves Portland cement, where the product is virtually homogeneous. Auction markets are another source of concentration-price data. There is evidence that, at least over a certain range, an increase in the number of bidders results in a higher price if the auctioneer is selling and a lower price if the auctioneer is buying. An auction, however, involves an inherently homogeneous product. All bidders seek to buy the same item or to supply an item pursuant to the same specifications. The implications of the auction data are of uncertain generality, therefore, because most product markets do not resemble auctions any more than they resemble the model of perfect competition.

There is also some anecdotal evidence in support of the proposition that concentration short of monopoly tends to dampen competition and hence market performance. Michael Porter, for example, finds it significant that Japan has had its greatest success in export markets, such as consumer electronics, in which there are a significant number of Japanese competitors. Such evidence is not very helpful, however, because there is at least as much anecdotal evidence to the contrary. For instance, the most successful export industry in the United States is civil aviation manufacturing. There the number of U.S. firms has gone from three to effectively one, Boeing, over the past decade in a global competition where the only other entrant is the European consortium, Airbus Industrie.

"But if we do merge with Amalgamated, we'll have enough resources to fight the anti-trust violation caused by the merger."

Export markets aside, there is quite a bit of anecdotal evidence for the proposition that numerous firms are not necessary for vigorous competition to take place. For instance, the market for long-distance telephone service appears to be very competitive, although there are only three significant firms in the market, and one has a market share of 70 to 75 percent. Similarly, the old three-firm market of television networks did not seem to lack competitive vigor. Are two firms enough? Surely no one believes that the Associated Press and the United Press were playing footsie when they appeared to be playing hardball, indeed, unto the death. No one would suggest that the answer to the rhetorical question, "Does Macy's tell Gimbels?" is "Yes." or that Coke and Pepsi would not still compete furiously with each other if Dr. Pepper were not in the market.

The primary basis for the antitrust agencies' concern is analytical, not empirical, and it rests upon the observation that the transaction costs of colluding should decline as the number of firms in a market declines. ft should be easier; in other words, for two firms to succeed in making and policing a price-fixing agreement than it would be for three firms; likewise, three should find it easier than four. This alone underlies our current policy against mergers that do not create a monopoly: they may facilitate price-fixing.

Price-fixing may occur with or without an actual agreement. When done by agreement, price-fixing is unlawful per se and indeed felonious. It can be penalized with fines of up to $1 million for a participating corporation and $100,000 for each individual conspirator, with three years in prison as well. In recent years criminal sentences have been quite stiff, thanks to the new Sentencing Guidelines used by the federal courts. Even more significant, perhaps, price-fixers face civil liability of three times the harm they cause to others, which is usually three times the amount they manage to overcharge their customers. So it is not really sensible to predicate a strong antimerger policy upon the presumed need to deter illegal conduct that is already subject to very significant penalties. If those penalties are not enough, they should be raised. In no event should we forego efficient transactions just because they might make it somewhat easier to commit a serious crime, so that more business people might succumb to that temptation.

The primary basis for antitrust agencies concern with concentration is analytical, not empirical, and it rests upon the observation that the transaction costs of colluding should decline as the number of firms in a market declines.

On the other hand, price-fixing without an actual agreement is not unlawful. Company X is free to announce a price increase in the hope that its principal competitor, company Y, will follow suit. If company Y does not go along, company X is free to rescind the scheduled increase and will be none the worse for having tried to accomplish in public what, if done by means of direct communication with its competitor, would be unlawful. It may still be called price-fixing, such tacit collusion is no more than a manifestation of an economic reality: in any competitive market short of the blackboard model of price-taking wheat farmers, each firm's pricing decisions are determined in part by those of its rivals.

Though tacit collusion is possible only in a highly concentrated market, we have seen that it is not to be assumed even there that competition will be displaced. At most, a tacit agreement on price would lead firms to compete on nonprice grounds, which are generally too subtle and too numerous to be constrained by an unarticulated agreement. Because nonprice competition will out anyway, firms with a cost advantage are likely to make price a ground of competition after all and will thus preclude price-fixing without an express agreement. Not surprisingly, therefore, there is virtually no empirical evidence to suggest specifically that tacit collusion occurs with any frequency, or to indicate under what market conditions, within a concentrated market, it is more likely to occur.

Although it is doubtful that blocking mergers short of monopoly produces significant benefits by preventing tacit collusion, doing so may be costly in the short run and futile in the long run. Smaller firms, prevented from growing by merger, may be incapable of minimizing the cost of production; if that is so, then the number of firms will shrink as some fail and the survivors approach efficient scale through internal growth. Blocking mergers thus will not prevent the eventual concentration of the market in the hands of a few efficiently scaled firms. In the meantime, however, output will be lower, and consumers will pay higher prices.

Whenever the government asserts that a particular transaction or business practice should be prohibited under antitrust law, it imposes its own monopoly decision on the market and thus displaces the multiple decisions that would otherwise be made by consumers.

Ironically, whenever the government asserts that a particular transaction or business practice should be prohibited under the antitrust law, it imposes its own monopoly decision on the market and thus displaces the multiple decisions that would otherwise be made by consumers, with their diverse tastes and preferences. To be sure, when the government exercises its legal monopoly on force in this way, it does so on the ground that, unless checked, the business practice or the transaction will increase the risk that firms will acquire market power. A supposedly benevolent (public) monopolist, that is, exerts its power to prevent the emergence, which it claims to foresee, of a malevolent (private) actor, a firm with market power or a monopoly.

The question, then, is how readily to accept the government's claimed foresight that some form of business conduct will have an outcome detrimental to competition and hence to consumers. That is a question both of the government's ability to foresee the nature of the outcome and of its incentives to predict that outcome accurately. I pass over here the problem of the government's incentives, particularly its incentive to overpredict threats to competition; that is a problem better left to public-choice economists, and even they have had only limited success in modelling the decisionmaking of the antitrust enforcement agencies. I shall focus instead on the anterior question of the government's ability to predict the likely competitive effect of a transaction or a business practice that does not, here and now, actually create a monopoly.

To predict whether a transaction or practice should be stopped on antitrust grounds, the government needs to know whether competition will significantly constrain firms in the relevant market if the government stays its own hand. Will the merger create a firm with significant market power or a market structure in which there is a substantially increased risk that the remaining firms will collude? Is the allegedly exclusionary practice-such as predatory pricing, a boycott, or a tie-in sale- likely to drive the firm's competitors from the field? Those are the ultimate questions, but the more immediate and operational question is how the multitude of consumers and actual and potential competitors will respond in the marketplace if the merger or the exclusionary practice is allowed to go forward.

The government's predictive power is severely limited by the unavoidable lack of information about how consumers and competitors will react under hypothetical circumstances. Market power depends on the slope of the demand curve for the product, that is, the elasticity of consumers' demand at current prices and above; it depends on the shape of the supply curve, which indicates the response of other producers if the merged firm raises its price; and it depends on the cross elasticity of demand between the product in question and the readiest substitutes to which consumers would turn in the event that the merged firm raises its price.

In most cases these data are not only unknown but unknowable, and only rarely can they be estimated with enough confidence to indicate whether a transaction or a business practice is contrary to the public interest. Therefore, it strains credulity for the government to claim that it can accurately predict the competitive implications of any practice or transaction that does not literally create a monopoly, as would the merger between the only two firms in a market or its functional equivalent, the formation of a marketwide cartel. Interestingly, the government does not study what happens in markets in which it has approved mergers that significantly increase concentration.

Lest there be any doubt about the government's ability to predict the future, consider its ability even to interpret the past or to observe the present competitive significance of market conduct. Time and again, the antitrust agencies have condemned as exercises of market power business practices for which academic commentators have later shown the existence of a procompetitive or at least an innocuous explanation-from United Shoe Machinery's lease-only and related policies, to Topco's division of territories, to the Maricopa County Medical Society's schedule of maximum fees. All too often, however, the benign or innocuous explanation did not appeal to or sometimes even occur to the government at the time it investigated and prosecuted the matter. Similarly, the government's assumption that highly concentrated markets are prone to collusion has led it time and again to block mergers in one market that would have created a level of concentration below that prevailing in other markets where competition seems to be quite vigorous.

In sum, on both analytical and empirical grounds, we have reason to be skeptical about the government's ability to identify transactions or practices that are a real threat to competition. In fact, the government does not even undertake, except in the rarest of cases, to predict whether a transaction or a practice threatens competition overall in its manifold forms. Instead, the government limits itself to the seemingly more achievable goal of predicting the effect of the transaction or practice on price competition.

The government recognizes that determining whether a merger or practice will increase competition overall by substituting nonprice for price competition is ordinarily too complicated for the analytic tools and empirical data that are available. Not surprisingly, therefore, with only a slight nod in the direction of nonprice competition, the government's Merger Guidelines expressly emphasize price effects. That is, in defining the relevant market the government looks for the group of products for which a monopolist "could profitably impose a 'small but significant and nontransitory' increase in price," defined as 5 percent for one year. The government is understandably diffident about considering the claim that a merger will enable the parties to realize efficiencies in production that would lower prices, because that claim is extremely difficult to confirm. Indeed, the Guidelines require "clear and convincing evidence that a merger will achieve such efficiencies"-which is well-nigh impossible-to forestall a challenge that would otherwise be brought. The government will certainly not go further to consider whether a merger is likely to enhance the merged firm's ability to do research or to invest in reputation as a means of competing. Something of an exception, however, has arisen in the case of vertical restraints, such as a territorial limitation, which a manufacturer or distributor might impose upon retailers to induce them to provide presale services. In Continental TV v. GTE Sylvania (1977) the Supreme Court interpreted the law to make the legality of a vertical restraint depend on balancing the gain in interbrand (largely nonprice) competition against the loss in intrabrand (largely price) competition. That can be done only rhetorically, however, not in any meaningful, let alone rigorous, way. Nonetheless, the lower courts have applied the teaching of Sylvania almost uniformly to uphold vertical restraints on the ground that they create substantial nonprice competition among dealers.

On both analytical and empirical grounds, we have reason to be skeptical about the government's ability to identify transactions or practices that are a real threat to competition.

Where analysis suggests that price will increase, antitrust enforcement will generally go forward. Apart from the vertical restraint context, enforcement is almost never overridden by a countervailing nonprice story. Indeed, firms rarely present such stories because the government will not sympathetically receive them. As a result, the government may interdict a merger or a business practice to preserve price competition-indeed, it may be correct in predicting that high prices would otherwise result-but it cannot claim to know whether it has helped or hurt overall competition. As the Supreme Court observed in the 1966 General Motors case, "[t]he protection of price competition… is an object of special solicitude under the antitrust laws:' The practical effect is that the many nonprice aspects of competition receive virtually no solicitude under those laws.

Antitrust's focus on price is like the drunk's search for his keys, not where he dropped them, but down the block where the street lamp sheds its light. Because price seems less intractable, it is virtually the exclusive focus-one might even say the price fixation-of antitrust analysis. There is simply no reason to believe, however, that what the antitrust agency perceives to be the public interest -lower nominal prices-will accord with what the diverse public itself desires. Insofar as consumers value nonprice aspects of the goods they buy, the government's attitude is, in effect, the consumer be damned!

Given that firms may compete in a variety of ways, we must ask when, if ever, there is reason to believe that government can identify a particular means of competing as being contrary to the Interests of consumers.

Preservation vs. Promotion

I now return to my earlier suggestion that the only proper goal of an antitrust law is to preserve competition from displacement by monopoly Promoting competition by regulating market structure and business practices is not a practical objective, and it therefore leads inevitably to mischief. Within competition's domain, we can makc no meaningful judgment about whether one nonmonopoly market structure is "more competitive" than another; we simply lack the theoretical and practical capabilities. So too is the case with the regulation of competitive business practices.

Moreover, centrally engineering the mix of competitive behaviors, even if government could do it, would be inconsistent with the reason for favoring competition over monopoly in the first place. Left to their own devices, firms choose to compete in a variety of different ways. Thus, one firm may be an aggressive discounter, as are MCI, K-Mart, and a seller of generic aspirin, while a second stresses its reputation for quality, as do AT&T, Nordstrom, and Bayer Aspirin (although its closest competitors follow the same chemical formula). A third firm may emphasize a policy of introducing periodic product enhancements-a strategy IBM pursued with great success for many years that is now practiced by software makers such as WordPerfect-while a fourth stresses the duration and scope of its warranty, as have Chrysler, Sears (Craftsman Tools), and L.L. Bean. Note that of these four strategies, only one emphasizes price. Consumers purchase from whichever producer offers the mix of quality, price, and ancillary features that best suits their diverse tastes.

Given that firms may compete in a variety of ways, we must ask when, if ever, there is reason to believe that the government can identify a particular means of competing as being contrary to the interests of consumers. If we believe that competition is superior to monopoly decisionmaking in determining which goods to produce to satisfy consumers, then it is difficult to imagine why we would believe that centralized decisionmaking is nonetheless superior in determining the shape of the firm by which those same goods are produced or the means by which they are marketed.

In fact, however, any practice challenged under the antitrust laws could be reconceptualized as just a way of offering a different good-or a different package of goods. That package might consist variably of the physical good, presale or aftermarket services, payment terms, warranties, advertising-inspired associations, or such elusive features as the cachet of exclusivity.

Why not also put a maintained resale price or a tied product in the package? The government would not try to prohibit a manufacturer from requiring a retailer to display its product in a certain way (as does L'eggs hosiery). Why then should the government insist that retailers rather than manufacturers have control over pricing policy? Or try to suppress tie-ins as part of an offering in the marketplace? If enough consumers prefer the retailer-discounted product or the untied product, they will induce a competitor to satisfy them. If consumers do not have that preference, surely the government has no proper role in promoting a product that consumers themselves have not demanded in sufficient number to interest a supplier.

When an allegedly exclusionary or unfair business practice is seen as merely a means of providing a new and different package, rather than as a different- and vaguely anticompetitive-method of providing an existing product, the anomaly of government intervention to ban the new package or to promote the old one in the name of competition becomes apparent. Indeed, the government has no proper concern with what a firm puts into the package- the mix of competitive tools it chooses. If one firm competes on price ("We will not be undersold:') and another trades on the reputation it has cultivated ("Use only genuine GM parts."), consumers are perfectly capable of expressing their preference for one type of competition or the other. The odds are that consumers will differ among themselves, with some preferring each form of competition. If enough consumers favor one form of competition over another, then the firm that loses sales will observe that fact and alter its competitive strategy accordingly.

Thus, it should be a matter of complete indifference to the government that a firm tries to make its way in the market by using vertical restraints. Nevertheless, present law condemns per se resale price maintenance. At the same time, it treats under the rule of reason (English translation: threatens with litigation) all other types of vertical restraints, such as customer, territorial, and locational restrictions. This is utterly irrational. All of these distribution practices should be per se lawful. It is of no moment to a law concerned with competition that one firm tries to compete in the market by using resale price maintenance (for which Cuisinart was indicted in 1980) while another competes by using customer and territorial restraints (for which the government pursued White Motor to the Supreme Court in 1963) or that a third (for example, Tandy! Radio Shack) chooses vertical integration, which is per se legal. Let the consumers in the market dictate who wins and who loses.

Even if the government were to get out of the business of promoting competition, antitrust would still play an important role: preserving competition within competition's domain. That means using antitrust only to prevent private parties from eliminating competition. It means not suppressing practices that may arguably, in some economist's theory create, enhance, or constitute an exercise of market power.

Practices and transactions that effectively displace competition come in two basic varieties. First, there are mergers to monopoly-defined as something shy of a 100 percent market share. For example, in United States v. Alcoa (1945) the Court asserted that 90 percent "is enough to constitute a monopoly; it is doubtful whether sixty or sixty-four percent would be enough." Similarly, Robert H. Bork wrote, "My guess is that... mergers up to 60 or 70 percent of the market should be permitted:' The amalgamation of all, or virtually all, of the firms in a market necessarily eliminates competition, at least for a time. It allows the unilateral exercise of economic decisionmaking power for as long as it takes for a new competitor to enter the market. Although some economists may correctly argue that potential competition may constrain even a monopolist from effectively exercising monopoly power, exotic economic theories should no more be relied on to stay the government's hand against palpable monopoly than they should be relied on to activate the government against anything short of monopoly.

An important role of antitrust is to preserve competition within competition's domain. That means using antitrust only to prevent private parties from eliminating competition. It means not suppressing practices that may arguably, in some economist's theory, create, enhance, or constitute an exercise of market power.

Antitrust's second concern should be with marketwide noncompetition agreements. These are agreements-often called horizontal agreements- among nominal competitors to eliminate competition among themselves, where all or almost all of the finns in the market are parties to the agreement. Thus, market division, bid-rigging, and agreements to eliminate price or any nonprice form of competition on a marketwide basis must be condemned.

Such agreements are condemned per se under the law as it stands today, and for good reason. An agreement among firms that, if they merged, would constitute a monopoly is the functional equivalent of a merger to monopoly Because of each individual firm's incentive to cheat, such an agreement may be fragile, and it is certainly less likely to endure than a merger to monopoly But while it endures, such an agreement is comparable to a merger; it substitutes a single decision about the terms on which a good is offered to consumers for the plural decisions that are essential, by definition, to the existence of competition.

Current law goes further, however, and condemns a horizontal agreement, even if it operates on a less than effectively marketwide basis. That seems anomalous. The argument is often made that this type of agreement is properly condemned, no matter how limited in scope, because it has the potential to suppress competition among the parties to the agreement, while it has no potential to benefit consumers or society generally. (I have made the argument myself in the past.) In the specific context of a less than marketwide agreement, however, the argument does not work. If three of the four significant firms in a particular market agree to eliminate some form of competition among themselves, they remain subject to the competition of the other firm in the market. If the agreement makes them more successful in appealing to consumers notwithstanding the competition of their rival, then there is no apparent reason to condemn the practice, even though we may not understand exactly why it appeals to consumers. But this is a small point because, as a practical matter, agreements to eliminate competition are almost always formed on a marketwide basis, precisely because they would otherwise be doomed to defeat by the competition of any firm that is not a party to the agreement.

In sum, from the limited goal of preserving rather than promoting competition, it follows that antitrust should be concerned only with preventing mergers to monopoly and marketwide noncompetition agreements; they alone are what we know to be threats to competition. In pursuing that modest but worthy goal, antitrust policy need not be inconsistent or unpredictable, as it has been in the past. It would, to be sure, be a more limited endeavor, but it would also be a more coherent one.

The Next Step

The government began to move toward a rational antitrust policy in the early 1980s. In abandoning the IBM case in 1982, the government ended a 12-year effort to find an anticompetitive explanation for that company's long-standing success in the market. The second major step came with the adoption of the Merger Guidelines in 1982 and their revision in 1984. The Guidelines continue to display an unwarranted concern with market power far short of actual monopoly, however, on the basis of the unsubstantiated fear that concentrated markets are prone to collusion, express or tacit. The Guidelines did, however, raise the threshold degree of market concentration required to induce government intervention, and that was a move in the right direction.

The Guidelines for Vertical Restraints (1985) marked yet another significant step. They formalized a policy of indifference to nonprice vertical restraints except insofar as they are used to facilitate horizontal collusion, and in fact no nonprice vertical restraint has drawn a government objection for over a decade. Thus, in the past dozen years antitrust enforcement policy has approached, but has not yet achieved, a rational and defensible position.

The remaining anomalies in government policy derive from the claim that there is a significant risk of tacit collusion in concentrated markets. There is insufficient evidence for that claim, which is the only support for current efforts to "promote competition" by blocking mergers short of monopoly and condemning supposedly anticompetitive business practices, such as resale price maintenance.

It is now time to take the next step-to recognize in enforcement policy that the only proper goal of antitrust is to preserve competition, not to promote it. Under the current statutory framework, the broad discretion available to the enforcement agencies and to the Supreme Court in private actions should be exercised by focusing solely on the proper concern of antitrust-monopoly.

Selected Readings Bork, R.H. The Antitrust Paradox. New York: Basic Books, 1978. Demsetz, H. 100 Years of Antitrust: Should We Celebrate? Arlington, Va.: George Mason     University School of Law, 1991. Porter, ML. The Competitive Advantage of Nations and Their Firms. New York: Free Press,     1989. Posner, R.A. "The Next Step in the Antitrust Treatment of Restricted Distribution: Per     Se Legality." University of Chicago Law Review, Vol. 48 (1981). Weiss, L W, ed. Concentration and Price. Cambridge: MIT Press, 1989.

		Subscribe to Regulation