Can Economic Theory Give Guidance to Antitrust Policy?

Presidential Address delivered at the annual meeting of The Western Economic Association
Los Angeles, California

Introduction

I intend to focus my remarks today on the interplay between economic thought and antitrust policy. In recent years, a number of thoughtful commentators have raised questions about contemporary antitrust policy. I choose this occasion to argue that the old economic foundations of antitrust policy—rather than antitrust policy itself—are flawed.

Some noteworthy advances in microeconomic analysis have occurred in the past 25 years. Indeed, many would agree that micro is where the action, excitement, and progress has been most impressive. The economics of contracts [2], information [3], property rights [4], agency relationships [5], vertical integration [6] warranties [7], advertising [8], entry [9], transactions costs and monopolistic competition [10] have each in their own way enriched our understanding of how markets are formed and how the exchange process works.[11] Law and Economics has come of age, with a half dozen or more journals and a dozen centers. Law schools teach economics, and economics departments increasingly have staff with sufficient legal training to teach law and economics courses. Finally, and by no means the least important, our growing appreciation of competition as a dynamic process as opposed to a static structure has enormous potential for influencing and guiding public policy in the area of antitrust.

Although I would not care to argue that the decisions and the approaches of the courts and commissions have faithfully followed contemporaneous economic theory for the past 90 years, the case can be made that there is a correlation. The recent change in attitude toward vertical activities as represented in the GTE Sylvania decision [12] is a case in point. The termination of the IBM and AT&T cases along with Professor Baxter’s positions on mergers and vertical activities could, I am confident, be traced to academic research and criticism. The appointments of Bark and Posner to the Federal Bench cannot fail to have an impact. My position is that the time is ripe to relate recent advances in economic theory to antitrust policy.

The Relevant Antitrust Law

The original draft of Senator Sherman’s bill contained language intending to make illegal business arrangements that tended to “advance the cost to the consumer.”[13] Unfortunately, the Senate did not adopt that language in the final version of the bill. Instead, the Sherman Act focused on classes of acts without regard to their consequences. Section 1 reads:

Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several states, or with foreign nations, is hereby declared to be illegal . . .

Section 2 extends and complements section 1:

Every person who shall monopolize, or attempt to monopolize, or combine or conspire with any other person or persons, to monopolize any part of the trade or commerce among the several states or with foreign nations, shall be deemed guilty . . .[14]

The court’s interpretation of Section 1 has come to mean that only certain kinds of restrictive contracts and activities would be regarded as unreasonable restraints of trade as a matter of law. In these cases, violations would be deemed illegal per se and there would be no further occasion to examine the circumstances of the case. Thus the Sherman Act prohibited only “unreasonable” restraints of trade and it was up to the courts to decide what was lawful and what was criminal under that statute. In time, price fixing and group boycotts became per se violations of the antitrust law. Once the government proves such conduct, the court turns a deaf ear to the defendant’s wishes to provide a defense based upon reasonable conduct, economic necessity, or absence of anticompetitive effect.

The Addyston Pipe Case of 1898 resulted in the establishment of a per se prohibition of all price fixing agreements among firms. Judge Taft of the sixth circuit court read the common law as voiding all such agreements unless they were contingent to some legitimate cause. Consequently, any arguments by the defense that the conspiracy was reasonable became irrelevant.[15]

The next significant development subsequent to the passage of the Sherman Act occurred in 1911 when the Supreme Court decreed the dissolution of the Standard Oil Company [16] and the American Tobacco Company [17], finding both to be monopolies and combinations in restraint of trade. More important than the specifics of the several cases was Chief Justice White’s enunciation of the “Rule of Reason” which, as a practical matter, inserted the word “unreasonable” as an adjective before “restraint of trade” in Section 1 of the Sherman Act. Henceforth, only unreasonable restraints of trade were illegal at common law.

The court recognized that in an ex post sense every transaction is, by definition, an act of monopoly and restraint of trade, albeit frequently trivial and irrelevant. Indeed, competition takes place prior to a sale but the ultimate winner is a monopolist on that particular transaction. And a trade once concluded can also be thought of as a refusal to deal.

The Trenton Potterys Case in 1927 strengthened the earlier per se interpretations. Justice Stone reaffirmed the trial court’s finding of a price-fixing agreement and ruled that the existence of any price-fixing agreement, reasonable or unreasonable, constituted a violation of the Sherman Antitrust Act. A literal interpretation of Justice Stone’s opinion means the prosecution need demonstrate the mere existence of a price-fixing agreement.[18]

Section II of the Sherman Act has had a life of its own. The language indicates it is illegal to monopolize or attempt to monopolize: however, it is silent on monopoly itself. Hence, it is presumably legal to be a monopolist but it is illegal to try to become one. For example, in the 1945 Alcoa case, the court said:

Persons may unwittingly find themselves in possession of a monopoly, automatically so to say: that is, without having intending either to put an end to existing competition or to prevent competition from arising when none had existed: they may become monopolist by force of accident. Since the Act makes “monopolizing” a crime, as well as a civil wrong, it would be not only unfair, but presumably contrary to the intent of Congress, to include such instances. . . . A single producer may be the survivor out of group of active competitors, merely by virtue of his superior skill, foresight, and industry. . . . The successful competitor, having been urged to compete, must not be turned upon when he wins.[19]

Having advanced an economically sound opinion, the court preceded to rule against Alcoa with the following explanation:

It was not inevitable that (Alcoa) should always anticipate increases in the demand for ingot and be prepared to supply them. Nothing compelled it to keep doubling and redoubling its capacity before others entered the field. It insists that it never excluded competitors: but we can think of no more effective exclusion than progressively to embrace each new opportunity as it opened, and to face every newcomer with new capacity already geared into a great organization, having the advantage of experience, trade connections and the elite personnel.[20]

The message seems to be that “superior skill, foresight and industry”—the very hallmarks of competitive behavior are to be encouraged only up to a point. If the firm is too successful, it must be monopolizing because it is not inevitable that they succeed.

This interpretation was elaborated in the United Shoe case.[21] Referring to United’s machinery leasing system, the lower court ruled “they are not practices which can be properly described as the inevitable consequences of ability, natural forces, or law.”[22] The court further found that the leasing practices were unnatural barriers that unnecessarily excluded actual and potential competition and restricted a free market, instead of encouraging competition based on pure merit. Several passages are worth citing because of the repeated use of the word “inevitable”:

While the law allows many enterprises to use such practices, the Sherman Act is now construed by superior courts to forbid the continuance of effective market control based in part upon such practices. The courts hold that market control is inherently evil and constitutes a violation of section II unless economically inevitable, or specifically authorized and regulated by law . . .

The violation with which United is now charged depends not on moral considerations, but on solely economic considerations. United is denied the right to exercise effective control of the market by business policies that are not the inevitable consequences of its capacities or its natural advantages. That those policies are not immoral is irrelevant . . .

. . . the law does not allow an enterprise that maintains control of a market through practice is not economically inevitable, to justify that control because of its supposed social advantage.[23]

The “economically inevitable” test for conduct under Section II poses interesting and serious policy questions. What kind of conduct ought to be permitted and encouraged if we want to promote superior skill, foresight, and industry yet require the conduct be economically inevitable?

Who really knows what conduct is economically inevitable? By their very essence, markets are the settings in which firms continually improvise new types of contracts, property rights, and agreements in the quest for efficiency and profits. Each new innovation of this nature is automatically suspect if the test it must meet is one of “economic inevitability.”

Franklin Fisher has written: “the crucial difference between monopoly and competition is the compulsion which market forces place on the competitor and the lack of it on the monopolist.”[24] Fisher proposes two principles for assessing conduct: (1) conduct, to be suspect, ought at least to be more restrictive than necessary. (2) conduct should not be condemned if it is precisely the conduct which competition would lead us to expect. Elaborating on these points, he states:

One has to be careful to distinguish between cases in which competition is forcing firms to react and cases in which firms are taking unnecessary action to forestall competition. The competitive model itself points to situations in which firms, faced with competition, will be forced to do certain things or lose business. Firms observed to be doing those things in those situations should not be regarded as monopolizing. They are engaging in conduct which competition makes “economically inevitable.” [25]

Under this approach, the firm is passive, reacting to “market forces” and there seems to be a little role for foresight, anticipation, and innovation. Does this mean, for example, that a unilateral price reduction by a firm is illegal because it is not economically inevitable? Should the firm have waited until a rival cut a price before it initiated its own cut? A retaliatory price reduction may well be the result of economic inevitability but it is then difficult to make a case that the behavior of the initiator of the price cut was economically inevitable. “Least restrictive” or “inevitability” tests are even more troublesome when the issue is one of changes in product design, contracts, advertising, and similar activities. Unfortunately, the competitive model of economic theory not only offers little guidance, but actually points us in the wrong direction.

The confusion arises because many courts and many economists fail to realize that the “competitive model” is silent on the subject of competition. Perceptive as they are, the courts sense the truth of the matter, but in the absence of a robust theory of competition, must forge their own theories on a case by case basis. Economists of the new school have assisted by developing “special” rather than general theories of competition. Hence, we have a mosaic of conflicting and ambiguous opinions.

The Competitive Model as a Standard for Business Behavior

The traditional model of pure competition is a marvel of elegant simplicity but it is not concerned with competition. Each firm is vanishingly small and sells a homogeneous product at a price over which it has no control. If sellers and buyers have full information, the market—open to any and all at no cost—is in equilibrium with price equal to average cost and so there are no profits in the economic sense of the term.

The model is silent on the dynamics of price changes in the event the market equilibrium or balance condition is not satisfied for one reason or another. Who knows whose decision it is to change prices; if each individual transactor is a price taker, who has the ability or motive to change price?[26]

The question of price adjustment aside, the traditional competitive model logically precludes any act that economists, businessmen, and laymen alike would judge to be competitive acts. Such common business practices as advertising, special promotions, product variation, price cutting, research and development, innovation, and information acquisition have no place in the traditional competitive model. On the contrary, a strict interpretation of the model and its complement—the monopoly model—associates every competitive act by a business firm with some degree of monopoly power. Yet, by definition, the formal monopoly model also excludes the possibility of competitive behavior.[27] It might be useful to coin a new phrase or name for the “competitive model.” In the interest of clarity, we could call it “the costless information, no communication, zero transaction costs, well-defined property rights, perfect markets idealization;” or, to have a somewhat more memorable designation suitable for the classroom, we could borrow from Lanier’s advertising copy and simply call it the “no problem” model.

The shortcomings of the traditional models have been apparent for some time. Yet these models continue to serve as the theoretical foundation of antitrust policy in the view of a good many economists. The “old school” of industrial organization elaborates the competitive/monopoly paradigms into the well known structure-conduct-performance approach. The number of firms in an industry, along with various measures of concentration, are taken as given and are thought to cause conduct and performance. Preoccupation with structural measures leads to highly suspect evaluations of conduct (e.g., who gave and who heard the speech on “The Future of Our Industry” at the latest trade association meeting? How did the sales force discover a rival was offering price protection over the next six months?) In turn, performance evaluation usually focuses on the relationship of price to some measure of marginal cost and on whether or not profits are “excessive.” These “tests” clearly relate to the norm of the competitive model.

In contrast to this traditional approach, the emerging view of industrial organization holds structure to be endogenous—a result of a process influenced by property rights, contracts, institutions, information systems, and communication. In the view of the emerging new school of industrial organization of economics, perfect markets are neither feasible nor ideal. Information is costly to acquire and verify and is even more difficult to interpret.

When the time comes for some obscure business historian to write the History of the Apple Computer Corporation, I am confident that the author will not offer the following narrative: Founders Jobs and Wozinack designed and assembled the first Apple computer in the family garage in 1977 knowing full well that they had no unique information and knowledge at their disposal, that their computers would be sold at marginal cost, that they would be price takers, and that their expected rate of return on investment would equal their next best alternative, risk free government bonds. Maybe the monopoly model would provide a better explanation: Perhaps the two entrepreneurs cornered the market for microprocessor chips and stored them in their garage. Perhaps the Apple trademark itself conferred monopoly power (a la the Real-lemon case) since its obvious that rivals would scarcely fair as well with a Cumquat II Plus or a Banana III. Perhaps Apple achieved success because they sold a complete microcomputer instead of dozens of separately priced computer components separately in plastic bags. Such unbundling would have certainly promoted “competition” according to the old school of industrial organization.

If you expect me to offer a coherent model to explain why and how Apple succeeded, and at that particular time, I shall disappoint you. I can do no more than to observe there was nothing “economically inevitable” in Apple’s history and to refer you to Armen Alchian’s 1950 article on “Uncertainty, Evolution and Economic Theory.”[28] But at the risk of jumping to conclusions, I am willing to venture that the founders of Apple Computer neither expected nor intended to become theoretically indistinguishable from that Kansas wheat farmer who produces to the point where the marginal cost of production equals the price dictated by “the market.”

Some Conjectures from the Business World

There is an old cliche to the effect that “if you haven’t met a payroll, you have no business theorizing about business behavior.” Suffice it to say that I’m less inclined to scoff at that attitude as I mature. I keep in touch with a number of former students who have achieved unambiguous success in entrepreneurial and managerial business positions. Our roles are now reversed: ignorant of recent advances in analysis, they tell me that formal economic theory—to the extent they recall it—is wrong, incomplete, or irrelevant. They instruct me on how firms behave in the real world which, as I recall, is what positive economics is all about.

I must add that I have been impressed with the economic intuition and intelligence of the attorneys and judges I have had the occasion to work with in the course of my consulting and expert testimony. Unencumbered by and uninterested in the formal niceties of mathematical economics and econometrics, they have an unerring sense for homing in on the truly important aspects of business behavior.

I would like to offer a list of conjectures by my sources in the real world of commerce and competition. These are offered in the spirit of challenging us to model and test the intentions of real world business people.

Conjecture I: The average cost of production falls over the relevant range for a typical firm. In other words, economies of scale in production are ubiquitous.

The cost curve in this conjecture is the ex ante long run cost curve where the firm can fully adjust to different production techniques and take full advantage of economies in materials acquisition. This is not the tradeoff between volume of production and rate of production for a given production function in a given time interval [29] Whether the product is a microprocessor, a barrel of cement, a half-ton truck, or a crap table, businessmen believe long run marginal cost is falling over the volume they can sell. Unfortunately, this state of affairs is not consistent with the second order condition for profit maximization by a firm in the traditional competitive model: since marginal cost does not cut the constant average and marginal revenue from below, no profit maximizing equilibrium exists for this firm.

Instead of speculating on whether it is economically inevitable that the firm will exploit the divergence between price and marginal cost and become a monopolist, I move to the second conjecture.

Conjecture II: The marginal cost of selling output (including advertising, distribution, inventories, service, warranties, special terms and considerations, etc.) limits the size of the firm that, under Conjecture I, threatened to gobble up the entire market. In more colorful language, “any fool can produce a widget, it takes a genius to sell one.” Moreover, the sales function is fraught with far more risk and uncertainty than is the production activity.

I venture to say that after a review of the advertising and marketing literature, no general theorems, principles, or even rules-of-thumb emerge. At least, there are none that my sources would dignify by including them in this list of conjectures. However, I would observe that the current state of the art as practiced by the best minds of Madison Avenue is represented by the T.V. spot commercials sandwiched every evening between segments of Laverne and Shirley.

The enormous variety of sales techniques and distribution channels actually observed defies generalization. Apple sells its personal computers to independent retailers while Tandy sells through its company owned or franchised Radio Shack stores. IBM markets its micros through Sears & Roebuck but has its own sales force for its minis, its large frame computers, and its office equipment. SCM and European manufacturers distribute their typewriters through independent retailers. Xerox sells copiers with its own integrated sales force while Canon and Savin distribute through independent middlemen.

Madison Avenue is the home of creative artists, not scientific technicians. John Wanamaker, the Philadelphia department store magnate, once confessed that half of his advertising budget was wasted. When asked why he didn’t reduce his advertising by half, he responded he would be delighted to except he didn’t know which half to cut.

For the most part, selling effort has been ignored by economists in the traditional theory of the firm. The majority regard sale expenditures with some suspicion—a waste of resources leading to or resulting from product differentiation and a symptom of market power. But note the context in which such behavior is judged: traditional theory superimposes an average and marginal revenue curve of one variety or another on a set of cost curves reflecting output production exclusively.

It is as if the consumers whose tastes, etc. are represented by the average revenue curve line up at the back door of the factory to the purchase onesy-twosy the instantaneous or daily output from the production line. Of course, that is not the case. Both before and after production, the firm must arrange to discover its customers, inform and persuade them, and deal with the myriad of other concerns associated with the activity of selling.

Conjecture III: It is very difficult to beat a rival or a competitor by lower production costs for a standardized commodity. A firm’s most serious competitors—actual or potential—are well informed about state of the art production technologies. Markets screen personnel so that the capabilities of one’s rivals are fairly uniform in the mastery of science and technology. Access to current technology is a necessary but not a sufficient condition for success. That is, it is very difficult to beat a rival with lower production costs alone: product design, quality, service, warranties, and marketing is where the battle is won or lost.

With tongue even further in cheek, I suggest the successful firm follows what I call the double thirty/twenty rule. Whether defined by importance, concern, worry, or activity—none subject to cardinal measure, of course—the manager/entrepreneur assigns the following weights: thirty percent to product design, twenty percent to production, thirty percent to sales activities, and twenty percent to collecting the money.

Of course it is arbitrary and even frivolous to assign such weights, let alone to generalize them into a rule. Nevertheless, the successful firm must excel in all four areas or its very survival is at risk. No renaissance man or team running a business firm can safely afford not to put forth their best efforts in pricing, product design, selling effort, production, and finance—exactly the opposite of the price-taking passivity of the “competitive model” ideal. Note also that the traditional competitive model focuses almost exclusively on the narrowly defined production activities of the firm, assuming that sales and distribution along with finance fall into place costlessly and effortlessly. Furthermore, old school industrial organization theory has little to say about product design except in the negative as it is condemned as wasteful duplication via brand proliferation and product differentiation.

Conjecture IV: Although firms trade in markets, it is more accurate to state individuals representing firms trade with other individuals (both consumers and representatives of other firms). Clearly, there are times when it is handy to talk about “the market” but in so doing we ignore the role played by personal contact, relationships, trust, reputation, etc.

Each vertical chain of firms involving suppliers, intermediate producers, assemblers, jobbers, distributors, and retailers has an important history of accumulated dealings at any moment of time. These chains are created at some considerable time, difficulty, and expense, and if they work for every participant, they tend to be stable and enduring in spite of occasional offers of “better deals” from potential substitutes in the chain. It is common for participants in such an established chain to act as a team and, because of the special, usually satisfactory, past relationships, to provide other team members with financing, advice on product design, cooperation on special price promotions, advertising allowances, assistance on trouble shooting customer problems, and, of course, information and intelligence about what “the market,” i.e., what competitors are doing or are attempting to do. Such joint effort could easily be attacked as a Sherman 1 violation under a superficial interpretation.

Conjecture V: The collection of firms in any vertical chain (of suppliers, assemblers, processors, distributors, retailers) is itself an efficient institution for acquiring and signaling information and for distributing risks. The conduct of the various participants is shaped by the methods of communication improvised and by the negotiated rules and contracts which coordinate the activities of the team. The relevant information that flows up and down the chain contains far more than mere price signals. I say “mere price signals” because marketing information is equally important. There are multitudes of products that, while technically producible, are commercially infeasible: the proposition can be put a number of ways but perhaps the simplest is to say that consumers will not buy sufficient quantities at a price sufficient to earn even a traditional “normal” rate of return. Similarly, there are hosts of products (perhaps best described by bundles of characteristics) that consumers would be willing to buy if only firms knew how to design and produce them. Interaction among members of the vertical chain of firms is a process by which the team members grope toward a set of products that are technologically and economically feasible.

Conjecture VI: Entrepreneurs price their products—especially new products—according to their estimate of consumers’ perceived value of the product, not its cost of production. Prices determine costs rather than the converse. The entrepreneur attempts to forecast what the consumer wants and what he is willing to pay: the task then becomes to see if it is possible to design, produce, and market a product at a cost that makes the effort worthwhile. Incidentally, I take it as axiomatic that every intelligent entrepreneur knows it is easier to sell a product that a consumer wants at a price he is willing to pay than it is to sell him something he does not want at a price above his perceived value. My prototypical entrepreneurs from Apple Computer forecast that sufficient numbers of consumers would want to buy a personal computer and would be willing to pay several thousand dollars. According to anecdotal evidence, IBM couldn’t understand why ordinary people would want to own a computer.[30]

Conclusion

If businessmen believe these conjectures and act as if they were true, perhaps economists should attempt to devise appropriate test. We have the mathematical and statistical skills required. The insights gained could be of considerable value since antitrust policy wrestles with a variety of practical questions. Perhaps the most important is: which behavior is competitive and which is not. To the extent that the “competitive model” serves as the economic foundation of antitrust policy, we can continue to expect confusion since it does not deal with asymmetric information, or with transaction costs. That model—elegant as it is—does not deal with competition nor with its absence. Instead, it assumes away the interesting behavioral and institutional questions because of the way it treats information. Until we address that issue and the business community’s conjectures squarely, perhaps the best advice we can offer the businessman is to keep a low profile and stay out of trouble by following the counsel of Francis Bacon:

It were good therefore, that men in their innovations would follow the example of time itself, which indeed innovateth greatly, but quietly and by degrees scarce to be perceived.[31]

Notes

1. Armentano, Bark, Brozen, Posner.
2. Steven Chung, Arthur DeVany, Mitchel Polinsky, Steven Shavell.
3. Akerlof, Hirschleifer, Spence.
4. Alchian, Coase, DeAlessi, Demsetz, Frech.
5. Jensen & Meckling.
6. Alchian, Klein, Crawford, McGee, Pashigan, Warren-Bouton.
7. (7)
8. Nelson, Schmalensee.
9. Baumol, Panzar, Willig.
10. Archibald, Lancaster, Leland, Spence, Stiglitz.
11. Alchian, Barzel, Leland, Ostroy, Leland.
12. GTE—Sylvania.
13. Hans Thorelli. The Federal Antitrust Policy (Baltimore: Johns Hopkins Press, 1955) p. 169.
14. Antitrust Laws with Amendments, 1890–1964, United States Government Printing Office, Wash. (1964) 1.
15. United States v. Addyston Pipe and Steel Company. Sixth circuit, 1898. 29 c.c.a. 141,85f.271.
16. Standard Oil Company of New Jersey v. U.S., 221 U.S. (1911).
17. U.S. v. American Tobacco Company. 221 U.S. 106 (1911).
18. United States v. Trenton Potterys Company. Supreme Court of the United States, 1927. 273 U.S. 392,47s.ct.377,711.ed.700.
19. United States v. Aluminum Company of America, 148 f.2d 416 (2d CIR. 1945) 429,431.
20. Ibid.
21. United States v. United Shoe Machinery Corporation., 110 F. Supp. 295 (d.) Mass. 1953 (, aff’d, 347 u.S. 521 1954).
22. Op. cit. 344.
23. Op. cit. 345.
24. Franklin M. Fisher, “Diagnosing a Monopoly,” The Quarterly Review of Economics and Business, vol. 19, no. 2 (summer, 1979) 12.
25. Ibid. 28, 29.
26. Kenneth J. Arrow, “Toward a Theory of Price Adjustment.” The Allocation of Economic Resources by Moses Abramovitz et al., Stanford University Press, 1959.
27. Paul J. McNulty, “Economic Theory and the Meaning of Competition,” The Quarterly Journal of Economics, Volume 82 (1968).
28. Armen A. Alchian, “Uncertainty, Evolution and Economic Theory,” Journal of Political Economy, vol. 58, no. 3 (June 1950) pp. 211–21
29. A. Alchian and W. Allen, Exchange and Production: Competition, Coordination, and Control, second edition. Wadsworth Publishing Co., Inc. (1977) pp.250–255.
30. The Jeffries Report. Vol. 1, No. 1 (July, 1982) 6.
31. Francis Bacon, Of Innovations.