In this article, we continue by reviewing studies these four completed, under Director and Levi’s direction, of business practices that courts had long viewed as suspect under the antitrust laws: vertical integration, predatory pricing, price discrimination, resale price maintenance, and tying. We will then review an article Director and Levi wrote near the end of the project laying out their agenda for further research, which others characterized as the “manifesto” of the Chicago School. This will set the stage for our next series of articles on the Chicago School Reformation. In those articles, we will first examine the continuing work done by other scholars at Chicago, who built on the foundation the Antitrust Project laid, and will then show how Robert Bork, Ward Bowman, and a new generation of Chicagoans began spreading the teachings of Aaron Director and his disciples to the broader antitrust community.
The Business Practices the Antitrust Project Studied
In our last article, we discussed William Letwin’s articles on the early history of the Sherman Act. We will now review the articles Bork, Bowman, and McGee wrote reporting on the results of their studies into five business practices (or, as Levi had called them, “abuses”) which had long been viewed as suspect under the antitrust laws.
Robert Bork, Vertical Integration and the Sherman Act: A Legal History of an Economic Misconception
When the Antitrust Project began in 1953, there were already a number of recent studies on vertical integration. Bork acknowledged that these earlier studies had already discussed “the economics of the subject . . . rather fully,” so that he felt it necessary only to provide a short summary of what they had covered.
Before getting to the economics of vertical integration, Bork reviewed a long list of cases decided before World War II, seeking to show that judicial hostility to vertical integration dated back to the earliest days of the Sherman Act. Herbert Hovenkamp, after reviewing these decisions more recently, concluded that Bork had overstated the degree of hostility these earlier cases had shown toward vertical integration.
Bork turned next to four more recent cases involving vertically integrated firms: A&P, Yellow Cab, Paramount, and Columbia Steel. Again, Bork seems to overstate the degree of judicial hostility to vertical integration in these decisions. While the Department of Justice had mounted a very aggressive campaign against vertical integration, arguing that it should be per se illegal, the Supreme Court never accepted that position. Instead, in Paramount and Columbia Steel, the Supreme Court held that vertical integration required rule of reason analysis.
Applying the rule of reason, the Supreme Court in its final decision in Yellow Cab affirmed a district court decision holding that Yellow Cab had legitimate business reasons for requiring its operating companies to use taxis it manufactured. Similarly, in Columbia Steel, the Supreme Court held that U.S. Steel’s acquisition of Consolidated Steel did not violate Section 1 even though U.S. Steel was the largest producer of rolled steel sheet on the West Coast and Consolidated was the largest independent fabricator of steel sheet in the same region. The Court held that despite their leading positions in these two vertically related markets, the shares of the two merging firms would be too small to enable them to restrain trade even if U.S. Steel required Consolidated to buy all of its steel requirements from U.S. Steel.
Despite shooting largely at straw men, Bork’s article advanced two novel theories based on the economic literature to show why vertical integration posed less of a threat to competition than many had supposed. Both have since become cornerstones of the Chicago School approach to vertical restraints generally: the one-monopoly profit principle and the elimination of double marginalization.
The one-monopoly profit principle. Bork argued that a vertically integrated monopoly “can take but one monopoly profit” and, therefore, “the second monopoly adds no power the first did not confer.” As he explained,
If, for example, a firm operates at both the manufacturing and retail levels, it maximizes over-all profit by setting the output at each level as though the levels were independent. Where both levels are competitive, the firm maximizes by equating marginal cost and price at each level; each level makes the competitive return. Where the firm has a monopoly at the manufacturing level but is competitive in retailing, it will, of course, exact a monopoly profit at the first level. And the manufacturing level will sell to the retail level at the same price as it sells to outside retailers.
Bork does not cite any sources for this one-monopoly profit principle other than an unpublished paper by Bowman and McGee explaining that the principle applies only when the two products are used in fixed proportions. McGee, in an article he wrote a quarter-century later, traces the original source of the one-monopoly profit principle back to Alfred Marshall’s theory of derived demand in his classic treatise, Industry and Trade, first published in 1919.
(2) Elimination of double marginalization. Bork argued further that vertical integration may actually lower downstream prices by eliminating what has come to be called “double marginalization.” As he explained in another footnote:
An interesting sidelight on vertically related monopolies is that it may be better from the consumers’ point of view that such monopolies be integrated rather than in separate hands. Where the monopolies are integrated consumers will pay the monopoly price. But where the monopolies are in separate hands, the higher monopolist may charge the full monopoly price to the lower who will in turn accept that price as a cost and further restrict output and raise price. . . . Therefore, if both monopolies are legal . . . vertical integration . . . is to be desired rather than attacked.
As with the one-monopoly profit principle, Bork does not identify where he first learned about this potential benefit to downstream customers resulting from vertical integration. The original source of the concept, however, appears to be Joseph Spengler, a professor of economics at Duke University who spent 1952 as a visiting professor at Chicago. In 1950, Spengler published an article entitled Integration and Antitrust Policy in the Journal of Political Economy. In it, he presented an economic model to show that vertical integration may enable a producer or distributor at one level to evade monopolistic surcharges imposed by sellers at earlier stages of production. Spengler concluded that, by allowing a producer or distributor “to evade imposts generated by horizontal integration and similar arrangements” at another level, vertical integration may enable him “to reduce his selling prices below the level that would obtain in the absence of vertical integration.”
John McGee, Predatory Pricing Cutting: Standard Oil (N.J.) Case
As Bork wrote in The Antitrust Paradox, “There was a time not long ago when everybody knew that the great American trusts had established and maintained monopoly by the ruthless extermination of smaller rivals.” Bork credits McGee with putting “a very large crack” in this story with an article McGee wrote as part of the Antitrust Project examining the record of the Standard Oil case and coming to the “startling” conclusion that Standard Oil “had not used predatory price cutting in its march to monopoly.”
In his article, McGee begins by acknowledging that he was “profoundly indebted to Aaron Director . . . who in 1953 suggested that this study be undertaken.” He explained that:
Professor Director, without investigating the facts, developed a logical framework by which he predicted that Standard Oil had not gotten or maintained its monopoly position by using predatory price cutting. In truth, he predicted, on purely logical grounds, that they never systematically used the technique at all.
McGee continued: “I was astounded by these hypotheses, and doubtful of their validity, but was also impressed by the logic which produced them.” McGee determined, therefore, to “investigate the matter, admittedly against my better judgment; for, like everyone else, I knew full well what Standard had really done.”
McGee goes on to describe Director’s logic more fully. It started, McGee writes, with the fact that when Rockefeller founded Standard in the 1860s, the oil industry was highly competitive with hundreds of small refiners. Director reasoned that, as a fledgling company, Standard Oil would not have had the large war chest needed to finance a predatory pricing campaign. He therefore concluded that it would have made more sense for Standard to use its own stock to acquire other refiners, thereby agreeing to share the profits of his growing company with them. Director surmised that even when Standard later reached a size that might have enabled it to engage in predatory pricing, it still would have been cheaper to acquire its competitors rather than try to run them into the ground. Rockefeller simply would have had to pay the owners of the acquired firm an amount up to the discounted value of the expected monopoly profits resulting from that acquisition, instead of incurring substantial losses trying to drive the target from the market without any assurance of success.
To test Director’s hypothesis, McGee examined the massive trial record in the Standard Oil case. What he found convinced him that Director had been right in surmising that Standard Oil had acquired and maintained its monopoly position through acquisitions and discriminatory rebates from railroads, not through predatory pricing. McGee reported that he had not found “a single instance in which Standard used predatory price cutting to force a rival refiner to sell out, to reduce asset values for purchase, or to drive a competitor out of business.” “I do not believe,” he concluded, “that Standard even tried to do it; if it tried, it did not work.”
Aided by Bork’s praise, McGee’s study of the Standard Oil case has become perhaps the best-known study funded by the Antitrust Project. One critic of the Chicago School, Christopher Leslie, describes McGee’s article as having been “wildly successful” in persuading other scholars and, ultimately, the Supreme Court in Matsushita, that predatory pricing is “inherently irrational” and, therefore, “does not occur.” Leslie goes on to try to refute McGee’s “revisionist history” of Standard Oil based on his own study of the record in that case.
As Joshua Wright has shown, Leslie exaggerates the influence of McGee’s article. The article most directly responsible for persuading the courts to reevaluate their approach to predatory pricing was not McGee’s article, but rather a 1975 article on predatory pricing in the Harvard Law Review by Phillip Areeda and Donald Turner. In their article, Areeda and Turner discounted McGee’s study, noting that, during the period when Standard Oil had acquired its monopoly position in the oil industry, no laws prohibited mergers that might substantially lessen competition. Not surprisingly, therefore, the Supreme Court in Matsushita cited a number of other, more recent studies in addition to McGee’s article to support its assertion that there was “a consensus among commentators that predatory pricing schemes are rarely tried, and even more rarely successful.”
John McGee, Price Discrimination and Competitive Effects: The Standard Oil of Indiana Case
Although not nearly as famous as his study on predatory pricing, McGee did a similar study three years earlier about a more recent case against one of Standard Oil’s offspring, Standard Oil of Indiana, known better as Amoco. In deciding this case in 1951, the Supreme Court reversed the Seventh Circuit’s decision upholding an FTC decision finding Amoco liable for price discrimination. The FTC had found that Amoco had violated the Robinson Patman Act by selling gasoline in Detroit between 1936 and 1940 to four large “jobbers” at 1 1/2 cents per gallon less than the price at which it sold gasoline to small service stations. In reversing, the Court held that the FTC had failed to consider whether Amoco gave those price reductions to large jobbers “in good faith” to meet an equally low price of a competitor, which under the Robinson Patman Act would be an “absolute defense” to a claim of illegal price discrimination.
As he would do in his later article on predatory pricing, McGee first set out the reasons he and his colleagues on the project were skeptical of the Robinson Patman Act’s prohibition on price discrimination. Noting that “[m]any economists claim that price discrimination is both a mark of monopoly and an effective method by which competition may be crushed out,” McGee argued to the contrary that price discrimination could often be procompetitive. As he explained, in a highly concentrated market with few competitors, a seller who “finds it advantageous to cut prices . . . will usually find it preferable to do so in a selective or discriminatory fashion” because that will make it more difficult for his rivals to detect and punish his price cutting.
McGee proceeded to “test this logic by examining an actual fact situation”—in this case, “the Detroit gasoline market from about 1936 to 1940”—based on the record in the Amoco case. McGee found that prices in the Detroit market had declined significantly during the period when Amoco had allegedly engaged in price discrimination, just as his logic had predicted.
McGee admitted that there was “no doubt that the declining price level injured some gasoline retailers as well as refiners,” and that “price discrimination must carry some of the blame for these injuries.” McGee argued, however, that these “were injuries sustained in a competitive struggle.” Recognizing that it “is sometimes said that competition tends to destroy itself through the accumulation of individual injuries and the eventual extinction of rivals,” McGee nevertheless asserted that the record in the Amoco case “lends no support to the notion.” Rather than facing extinction, “[T]he number of retail filling stations increased considerably between 1933 and 1939, a period that embraces the principal events studied in this essay.”
Ward Bowman, Prerequisites and Effects of Resale Price Maintenance
More than two decades after the Supreme Court held vertical resale price maintenance per se unlawful in its 1911 decision in Dr. Miles, Congress passed the Miller-Tydings Fair Trade Act, which allowed states to pass “fair trade laws” legalizing resale price maintenance on intrastate transactions. By 1951, 45 states had passed fair trade laws.
Director and Levi had difficulty understanding why some manufacturers supported fair trade laws and would choose to use resale price maintenance. To Director and Levi, maintaining resale prices seemed to be against the manufacturers’ self-interest because it was likely to raise distribution costs by eliminating competition among dealers. They asked Ward Bowman to find an answer to this conundrum.
In searching for an answer, Bowman reviewed what the manufacturers themselves had said while lobbying for fair trade laws, what other economists had written on the subject, and a large number of court decisions involving resale price maintenance. He found that while many manufacturers engaged in resale price maintenance in states that allowed it, most did not. Those that did most often used it only for differentiated products, claiming it arose from “the necessity of creating and maintaining product differentiation.”
Bowman identified seven explanations these manufacturers presented for their use of resale price maintenance. These included protecting their good will, enhancing product differentiation, protecting against loss leaders, and assuring dealer services needed to market a product. Of these explanations, the only one Bowman found convincing was the last: assuring the dealer services needed to market a manufacturer’s products. Bowman explained that many products, “especially new ones,” may require “costly demonstrations or services by the dealer,” for which neither the dealer nor the manufacturer can charge the customer directly. In these cases, where a customer can get his service from a service dealer and a cut price from a non-service dealer, “the manufacturer may suffer because of the elimination of service outlets.” In essence, Bowman was identifying what we now know as the free-rider problem, although he did not use that terminology.
Bowman recognized that this dealer service rationale might not apply in every case. He argued, therefore, that whether any “particular use of resale price maintenance” was lawful or unlawful would depend on knowing whether it had “come about through coercion by dealers, as a result of bargaining between organized dealers and organized manufacturers, or as a means of assuring performance of essential dealer services.”
In the second part of his study, Bowman reviewed a number of prior price studies undertaken to determine the price effects of resale price maintenance. He found that some had found higher prices and some lower prices. This led Bowman to do his own study comparing the prices of toothpaste in six states with fair trade laws with those in four states without these laws. Bowman found widespread lack of adherence to the specified minimum prices in both types of states, but a somewhat better record in fair trade states. Perhaps for this reason, Bowman found that toothpaste prices were significantly higher in free trade states than in non-free trade states.
Ward Bowman, Tying Arrangements and the Leverage Problem
Ward Bowman wrote the final article growing out of the Antitrust Project, entitled Tying Arrangements and the Leverage Problem, which was published in the Yale Law Journal in 1958. In it, Bowman acknowledged his indebtedness to Director both “for encouraging my interest in the tie-in problem” and for having formulated two explanations for tying other than as a mean to gain a second monopoly. The first was that tying could be a way to evade price regulation; the second was that tying could serve as a counting device for price discrimination. Both later became cornerstones of the Chicago School argument against treating tying as per se illegal.
When Bowman published his article on tying in 1958, the courts had been treating tying the sale of a second non-patented product to a patented product as a type of patent misuse for over 40 years. In 1947, the Supreme Court, in International Salt, further held tying per se illegal under the antitrust laws when its effect was “to foreclose competitors from any substantial market” for the tied product. In his opinion for a unanimous Court, Justice Robert Jackson declared flatly that “the tendency of the arrangement to accomplishment of monopoly seems obvious.”
The point of Bowman’s article was to “say it ain’t so.” Bowman begins by explaining that “a monopolist cannot necessarily improve his position—increase his monopoly revenue—by imposing restrictions on his customers.” If a monopolist is already charging a profit-maximizing monopoly price, “imposing an additional restriction on his customers [is] the equivalent” of charging them a higher price. A tie-in, therefore, will necessarily force him to lower his price on the tying product. Thus, for a tie-in to make economic sense, what a monopolist “sacrifice[s] in the way of return from the sale or lease of the tying product must be more than compensated by increased return from the tied product.”
This argument is an extension of the “one-monopoly profit” principle that Bork had used to explain why a firm with a monopoly in an upstream product market that is already charging a profit-maximizing monopoly price for that product will not be able earn a second monopoly profit in a downstream distribution market. With this one-monopoly principle as a predicate, Bowman identified five plausible explanations for tying arrangements, concluding that for only one of them was the purpose the acquisition of a second monopoly.
(1) Complementary products used in fixed proportions. Where two complements are used in fixed proportions—such as left and right shoes—Bowman explained that “from the buyer’s point of view the two together might as well be a single product.” That being the case, “The price of the combination is the only matter of interest,” and a firm could not raise the price of one product (e.g., a left shoe) without having to lower the price of the other (e.g., a right shoe). Therefore, the likely explanation for tying them together (pardon the pun) is that it is more efficient to sell them in a single package.
(2) Tying of complementary products used in variable proportions. Like Bork, Bowman recognized that the one-monopoly principle applies only when two products are complements used in fixed proportions, like left and right shoes. Where the two products instead are used in variable proportions, there may be some substitution—and hence competition—between them. In that case, Bowman acknowledged that a firm might have an incentive to use a tie-in to gain a second monopoly and thereby increase its monopoly profits.
(3) Evasion of price regulation. Bowman next explained that another way a seller might be able to increase its profits through a tie-in might be to tie the sale of a price regulated product to the purchase of an unregulated product, allowing him to charge a higher price for the unregulated tied product while complying with the price regulations on the tying product. Bowman explained that, while evading rate regulation in this manner might increase a firm’s monopoly profits, it did not require a second monopoly to do so.
(3) Price discrimination. Bowman explained that tying could also serve as a counting device to discriminate in price between more intensive and less intensive users of the tying product. Bowman gave three examples drawn from tying cases: (1) tying ink to mimeographs, (2) tying punch cards to tabulating machines, and (3) tying rivets to riveting guns. Again, however, while this use of a tying arrangement could increase a firm’s monopoly profits, Bowman argued that it was not anticompetitive and might actually promote consumer welfare by increasing output.
(4) Technological Interdependence. Another non-anticompetitive explanation Bowman gave of tying arrangements is that a manufacturer might use tying to assure that its product performs as expected. When the exact source of poor performance is difficult to trace, Bowman argued that the supplier could use a tie-in “to assure that no foreign elements are used with his machines” at much less cost than other methods of preventing their use.
(5) Economies of Joint Production or Sale. Finally, Bowman explained that tying often may be due to economies of production or sale when “the cost of producing and selling the combination is less than the cost of producing and selling the parts separately,” citing an automobile as an example. In these cases, Bowman argued that “[n]o tie-in can be said to exist” because “[n]o coercion is required when a cost advantage exists, for these lower costs will be reflected in lower prices.”
(6) Delaying entry. Bowman last addressed the claim that tying can delay entry into the market for the tying product by making it necessary for a prospective entrant to enter both the market for the tying product and for the tied product. Bowman sought to minimize this potential effect by arguing that it would be valid only if the time needed to enter the market for the tied product is greater than that needed to enter the market for the tying product. He argued that this was unlikely where the market for the tied product was competitive before the tie was in place.
After reviewing all six explanations for tying, Bowman concluded that tying was more likely to result from the promotion of legitimate business purposes rather than efforts at acquiring a second monopoly. Bowman thus made a strong case for evaluating tying arrangements under the rule of reason.
Levi and Director’s “Manifesto”
In 1956, as the Antitrust Project was nearing its end, Director and Levi published an article in the Northwestern Law Review entitled Law and the Future: Trade Regulation, in which they sought to summarize some of what they had learned from their project. Although some have called this article their “manifesto,” it was far from the kind of strident policy statement often associated with that term.
Director and Levi began by raising their concern over the Supreme Court’s increased use of the per se doctrine to find business practices illegal. “Throughout its history,” they wrote, “the Sherman Act has exhibited the twin tendencies of flexibility and ambiguity, on the one hand, and a drive for certainty and automaticity, on the other.” They worried that a desire for “certainty and automaticity” seemed “paramount,” and that it rested “not so much on the concept of fair warning, which is inherent in any idea of the rule of law, but rather more on the belief that new and automatic applications of the laws will catch objectionable conduct and effects in their incipiency.”
Director and Levi were further concerned that the Court’s application of this incipiency doctrine seemed “to rest on economic doctrines” they questioned as “substitutes for an observed effect.” They argued that in many cases this reliance on economic doctrine seemed unjustified. “We believe,” they wrote, “that the conclusions of economics do not justify the application of the antitrust laws in many situations in which the laws are now being applied.”
Director and Levi acknowledged that, even if they were right, it was “possible, perhaps probable, that the law will continue on its own, for the law is not economics.” The reason, they suggested, was that there was still “uncertainty” about whether the antitrust laws were intended as a code of “fair conduct, which may have nothing whatever to do with economics,” or instead as “minimal rules protecting competition or prohibiting monopoly or monopolizing in an economic sense.” In either case, they argued, the laws’ application would benefit from sharpening application of the economic doctrines on which the courts relied.
Director and Levi next identified three areas that they believed would be “the central field of antitrust enforcement” over the coming years. These three issues were “size, the concept of abuse, and . . . the idea of collusion.”
Size. Director and Levi saw monopoly as less of a problem in mid-century America than it had been at the turn of the century. They argued that size was now a problem mostly in industries where three or four large firms controlled most industry capacity. They maintained, however, that while mergers had been responsible for much of the economic concentration early in the 20th century, more recent increases in concentration no longer seemed to be “due in any widespread way to recent mergers or acquisitions.”
Collusion. Director and Levi viewed the problems of size and collusion as closely related. Because of the growing number of industries dominated by three or four large firms, Director and Levi believed that the problem of collusion, which had “always been central to the antitrust laws,” had become even more important. Without using the term, they seemed to be suggesting that interdependent behavior by firms in these highly concentrated industries was likely to become a major focus of attention under the antitrust laws in the future. They cautioned, however, that it “would appear to be extremely difficult and unwise for the law to assume that action taken on general knowledge implies a concert of action equivalent to collusion, conspiracy or agreement,” even though the result might be the same.
Abuse. On the concept of abuse, Director and Levi argued that “economic teaching gives little support to the idea that the abuses create or extend monopoly.” Echoing Bork and Bowman, they maintained that large firms lack the power to impose coercive restrictions on their suppliers or their customers as a means of obtaining a monopoly. While “firms that have some monopoly power over prices and output can impose coercive restrictions on suppliers and customers,” they argued that those firms would likely lose revenue if they imposed such restrictions “because they cannot both obtain the advantage of the original power and impose additional coercive restrictions so as to increase their monopoly power.”
Having identified size, collusion, and abuse as the issues most likely to be the “central field of antitrust enforcement” in the future, Director and Levi closed by reiterating that they did not intend “to suggest that the law must conform to the prescriptions of economic theory.” They repeated again, however, that there should be “a recognition of the instability of the assumed foundation for some major antitrust doctrines” that might require “a re-evaluation of the scope and function of the antitrust laws.”
The Initial Reception of the Antitrust Project
The initial impact of the Chicago Antitrust Project was akin to throwing several pebbles into the ocean. Based on our own search on Lexis for the period from 1954 to 1975, we were able to find only a handful of citations to the articles written as part of the Antitrust Project in any law reviews or any antitrust cases decided before 1975.
One reason for this lack of attention may have been an overlapping project around the same time at Harvard on “Competition and Monopoly in American Industry.” The Harvard project, led by Dean Edward Mason of the Harvard Graduate School of Public Administration, included several highly respected antitrust law professors and industrial organization economists, whose views were more in the antitrust mainstream in the 1950s. The Harvard project resulted in six book-length studies: five of major U.S. industries, and one by the economist Joe Bain examining entry conditions in 20 industries.
The Harvard project also held “a continuing seminar of lawyers and economists who met together regularly over a period of seven years” to discuss antitrust policy. The attendees included such influential antitrust law professors and economists as Carl Kaysen and Donald Turner at Harvard, Kingman Brewster at Yale, and Morris Adelman at MIT. At the end of the project, Kaysen and Turner wrote a seventh book, entitled Antitrust Policy: An Economic and Legal Analysis, which Dean Mason described as “the product of the discussion at the series of seminars the project had sponsored.” Published in 1959, it quickly came to be viewed as “the classic statement of the Harvard school” during the Kennedy-Johnson years when Harvard was at the height of its influence in Washington.
Overshadowed by the Harvard project, none of the Antitrust Project’s studies had any immediate impact on antitrust policy. The Warren Court continued to treat tying and resale price maintenance as per se illegal, and further extended per se treatment beyond minimum resale price maintenance to other vertical restraints, such as maximum resale price maintenance and territorial restraints. The Warren Court also subjected vertical mergers to harsh treatment under the newly amended Section 7 of the Clayton Act, holding vertical mergers presumptively unlawful under Section 7 at much lower market shares than the Vinson Court had under the Sherman Act in Columbia Steel. The Court also continued to apply the Robinson-Patman Act broadly, applying its terms literally to find discriminatory price cuts in local markets illegal in Utah Pie even though they had served to increase competition in the market for frozen pies in Salt Lake City.
The Coming Chicago Reformation
The end of the Antitrust Project in 1957 was only the end of the beginning of the Chicago School antitrust movement. In a future series of articles on the Chicago Reformation, we will study how Aaron Director and economists he helped bring to Chicago continued to build the intellectual framework of the Chicago School over the next decade and beyond. We will also study how Bork and Bowman became the first missionaries of that movement while teaching together at Yale in the 1960s. We will then see how the Chicago Reformation gained strength in the 1970s, as Richard Nixon remade the Supreme Court during his one-and-a-half terms in office, and how Chicago emerged triumphant during the Reagan administration in the 1980s. Finally, we will conclude by evaluating the criticisms leveled at the Chicago School over the last several years by the New Brandeisians and others.
