PLANNING, REGULATION, AND COMPETITION 53 (1) the court shall not approve a plan involving division of the assets of a single plant; (2) in determining the feasibility of division or divestiture of assets, the court shall take into account any probable permanent loss of substantial economies intrinsic to the defendant company or companies as currently constituted; (3) the court shall not order division or divestiture of assets where defendants show that such relief would not materially improve the competitive conditions which other relief, proposed by defendants, would achieve; and (4) the court shall not approve a proposed plan of divestiture or division of assets where defendants show that one or more companies resulting from the plan would lack reasonable prospects for survival under the competitive conditions likely to prevail. Comments. Structural reorganization of one or more firms, and creation of new independent companies, would be the usual and normal remedy for unreasonable market power, rather than a last resort. Defendants have the burden of showing that structural reorganization is inappropriate, or that any proposed plan is not feasible, except in the "single plant" case. Whenever it is apparent from evidence received on market power that some reorganization is feasible-which is likely to be the usual case-the court should not allow defendants to pursue the purely negative role of objecting to specific plans proposed by the enforcement agency, but should direct defendants to submit a specific plan or plans of their own. APPENDIX C (Material received from Dr. Adams subsequent to the hearings) (Walter Adams and Joel B. Dirlam) Introduction: the "Schumpeterian" hypothesis, 167.-I. Oxygen The view attributed to Schumpeter, that large firms with substantial market power have both greater incentives and more ample resources for research and innovation, has become part of popular mythology and an article of faith among many economists as well. Ostensibly, Schumpeter felt "that firms had to be protected by some degree of monopoly-to have some room to maneuver * * *” in order to bring about massive innovations. Presumably, he implied "that more concentration would increase innovation and progress." Though Schumpeter never stated it without careful qualification, this idea has been widely used to explain why some industries, like textiles, are "backward," and others, like petroleum, are not. Galbraith, for example, argues that "a benign Providence *** has made the modern industry of a few large firms an almost perfect instrument for inducing technical change. It is admirably equipped for financing technical development. Its organization provides strong incentives for undertaking development and for putting it into use. The competition of a competitive model, by contrast, almost completely precludes technical 1 Richard Caves, American Industry: Structure, Conduct, Performance (New York: Prentice Hall, 1964), p. 98. 2 Schumpeter qualified his hypothesis more carefully than did his disciples. To be sure, he argued that "・・・ largest-scale plans could in many cases not materialize at all if it were not known from the outset that competition will be discouraged by heavy capital requirements or lack of experience, or that means are available to discourage or checkmate it so as to gain the time and space for further developments"; but he also observed that "it is certainly as conceivable that an all-pervading cartel system might sabotage all progress as it is that it might realize, with smaller social and private costs, all that perfect competition is supposed to realize." Capitalism, Socialism, and Democracy (New York: Harper, 1942), pp. 89-91. For a balanced restatement of the Schumpeter hypothesis, see Edward S. Mason, Economic Concentration and the Monopoly Problem (Cambridge: Harvard University Press, 1957), pp. 91-101, and Jesse W. Markham, "Market Structure, Business Conduct, and Innovation," American Economic Review, Papers and Proceedings, LV (May 1965), 323-32. 54 PLANNING, REGULATION, AND COMPETITION development."' In a whimsical vein he adds that "The foreign visitor, brought to the United States to study American production methods and associated marvels, visits the same firms as do the attorneys of the Department of Justice in their search for monopoly." Similarly, Lilienthal argues that firms that are small and competitive do not have the profits to finance research: "Only large enterprises are able to sink the formidable sums of money required to develop basic new departures." Villard points out that the financing of research is less strategic than the assurance that, after an innovation is introduced, the firm will have a sufficient share of the market to recoup its outlays. And, he holds, only oligopolists in fact can enjoy such assurance." This hypothesis has not remained unchallenged.' Moreover, there has been a recent flurry of empirical studies, replete with regression analyses, designed to test the relationship between concentration and innovation. Unfortunately, these studies have yielded inconclusive results. Therefore, an unhurried exploration, in some depth, of a single, revolutionary invention and its introduction into a major oligopolized industry may provide some rewarding insights. For testing the "Schumpeterian" hypothesis, we have selected the oxygen steelmaking process-the circumstances surrounding its invention, its delayed adoption by the dominant firms in the United States steel industry, and the cost of this delay in terms of the industry's social performance. I "In my opinion," Avery C. Adams, chairman of the board and president of Jones & Laughlin, told his stockholders in 1959, "the basic oxygen process represents the only major technological breakthrough at the ingot level in the steel industry since before the turn of the century. With the exception of what we in the industry call trick heats, i.e., one heat made under ideal conditions, the best open-hearth practice today results in a production rate of 39 to 40 tons per hour. Our basic oxygen furnaces have produced at the rate of 106 tons per hour to date this month. On a trick heat basis, we have hit 160 tons per hour." By 1965 this opinion had become virtually unanimous in the industry. Indeed, most steel experts were willing to predict that no new open-hearths would ever again be built in the United States. Nevertheless, Mr. Adams' 1959 pronouncement came some ten years after the potentials of the new process should have been a matter of course to every steelman in the United States. Despite its revolutionary character, the basic oxygen process employs a relatively simple principle. It refines pig iron into steel by jetting oxygen vertically downward into a molten bath of pig iron. The conversion is accomplished in a pear-shaped vessel that looks something like a cocktail shaker or water carafe John K. Galbraith, American Captalism (Rev. ed.; Boston: Houghton Mifflin, 1956), p. 86. Ibid., p. 91. David E. Lilienthal, Big Business: A New Era (New York: Harper, 1953), p. 69. Henry H. Villard, "Competition, Oligopoly, and Research," Journal of Political Economy, LXVI (Dec. 1958). 483. John Jewkes, David Sawers, and Richard Stillerman, The Sources of Invention (London: Macmillan, 1958). Jacob Schmookler. "Bigness. Fewness, and Research," Journal of Political Economy, LXVII (Dec. 1959), 628-35. And esp. Daniel Hamberg, "Size of Firm, Monopoly, and Economic Growth," Employment, Growth, and Price Levels, Part 7, Hearings before the Joint Economic Committee, 86th Congress, 1st Session, 1959, pp. 2337-53; "Invention in the Industrial Research Laboratory," Journal of Political Economy, LXXI (April 1963), 95–115; and "Size of Firm, Oligopoly, and Research: The Evidence," Canadian Journal of Economics and Political Science, XXX (Feb. 1964), 62-75. * Edwin A. Mansfield, for example, has conducted some highly useful statistical research into the relation between size of firm and both the importance and adoption speed of innovations. ("Size of Firm, Market Structure and Innovation," Journal of Political Economy, LXXI (Dec. 1963), 556-76, and "The Speed of Response of Firms to New Techniques," this Journal LXXVII (May 1963), 290-311.) Ils conclusions, however, as he would be the first to concede, do not permit assured generalizations with regard to the central hypothesis. For instance, he found some evidence that the length of time a firm waits before using a new technique tends to be inversely related to the size of the innovator ("The Speed of Response of Firms to New Techniques," op. cit.) On the other hand, the steel industry remains an unexplained exception to his conclusion that the larger firms were more likely to innovate than the smaller. ("Size of Firm, Market Structure and Innovation, op. cit.). As we see it. the major weakness of the Mansfield approach is that it drowns in aggregate generalization what must be qualitatively evaluated in a careful case-by-case analysis. After a comprehensive review of the recent literature, Jesse Markham concludes that "The difficulty with such regression analyses as these is not so much their statistical as their conceptual inconclusiveness." (Op. cit., p. 331.) PLANNING, REGULATION, AND COMPETITION 55 bellied at its central portion and having a restricted mouth. Not only does it produce top-grade, “open-hearth" quality steels more quickly and efficiently than older methods, but it entails lower investment (as well as operation) costs. Finally, and ironically, the process was foreseen by Sir Henry Bessemer almost a century ago. History of the Invention Bessemer ushered in the steel age with his principle of pneumatic conversion, patented in 1855. This consisted of passing a "gaseous fluid containing oxygen" through molten pig iron. The Bessemer converter, equipped with an acid refractory lining, was charged with molten pig iron through a top opening. Atmospheric air would then be forced through a number of pipes (tuyeres) in the bottom of the converter and forced upward through the bath of molten metal. No extraneous source of fuel was necessary, because the oxygen in the air blast reacted exothermically with the impurities in the iron which were burned off as a gas or carried off into the slag. This, the so-called “acid” Bessemer process, could be used only to refine lowphosphorus ores but was not adapted to refining the immense deposits of highphophorus ores in Lorraine and Sweden. With a view to using these phosphoric ores, S. G. Thomas invented and patented in 1876 a process which differed from Bessemer's principally in the use of a basic converter lining (dolomite bound with tar) instead of the acid lining employed by Bessemer. It was this Thomas converter (or basic Bessemer process, as it was known in the United States) on which the great development of steelmaking in Europe was based. The Thomas process was uniquely adapted to the use of Europe's large phosphoric ore deposits. Bessemer recognized that the air blast used in his process posed a major problem. Since air is composed of 80 per cent nitrogen and 20 per cent oxygen; since nitrogen is bad for steel (making it brittle and less malleable); and since there was no way of preventing the injection of nitrogen into the Bessemer steels through the use of atmospheric air, Bessemer stated as early as 1856: "And here I would observe, that although I have mentioned air and steam because they contain, or are capable of evolving, oxygen at a cheap rate, it will nevertheless be understood that pure oxygen gas or a mixture thereof with air or steam may be used." Indeed, Bessemer not only entertained the possibility of using "pure oxygen gas" in the converter, but also of introducing it through the top instead of the bottom of the vessel. In spite of Bessemer's insights, early atempts to apply his teachings failed. Two major problems bedeviled steel technology: (1) pure oxygen was not available in commercial quantities and was prohibitively expensive; and (2) an increase in the oxygen content of the air-blast used by Bessemer would reduce the nitrogen content of the refined steel, but would also cause serious damage to the converter's tuyeres and refractory lining. European steelmakers using the Thomas converter faced the additional problem of producing steels with an excessive phosphorus content and hence inferior quality. In view of these problems, it is not surprising that the basic open-hearth furnace, the so-called Siemens-Martin process, was almost an immediate success after its introduction in 1880. While slower and more expensive than pneumatic methods of steelmaking-requiring about eight hours for a batch of steel as compared with one hour in a Bessember converter-the open-hearth had two signal advantages. It produced steel almost free of nitgrogen, and hence of far greater quality in terms of malleability, and it could use a relatively high percentage of scrap in lieu of pig iron. In the United States, therefore, blessed as it was with plentiful scrap supplies, the Siemens-Martin furnace provided an excellent solution to the quality problems of the Bessemer and Thomas conversion processes. Indeed, by 1909, the open-hearth had outstripped the Bessemer converter as the workhorse of the American steel industry, and by 1953, about 89 per cent of the steel produced in the United States was of the basic open-hearth variety. But pneumatic conversion remained the quickest and cheapest way of refining steel. Hence experiments continued, especially in Europe, to solve the problems of the oxygen supply and the longevity of the refractory lining. One breakthrough occurred in 1929, when the Gesellschaft fur Linde's Eismaschinen AG in Germany perfected a method (the Linde-Fränkl process) of producing bulk oxygen of 99 1 British Patent No. 2768 of 1855. British Patent No. 1292 of 1856. 56 PLANNING, REGULATION, AND COMPETITION per cent purity at very low cost. From then on, except for the actual building of the needed oxygen plants, the technical and economic problem of an adequate oxygen supply for steelmaking was of no further concern." The problem of the tuyere and lining longevity, however, was more stubborn and vexing. Attempts to use high purity oxygen in botton-blown (Bessemer or Thomas) converters resulted in the rapid deterioration of the converter bottomsometimes within the short time of one heat. Other attempts, i.e., to use.oxygen in side-blown converters, encountered similar difficulties. Here the oxygen jet directed at the melt surface caused excessively high temperatures on the side of the vessel facing the oxygen inlet and resulted in serious damage to the refractory lining. Still other attempts, i.e., to use lower concentration of oxygen or oxygen-steam combination-in order to conserve the refractory bottom or the sidewalls of the converter-suffered from the inherent liabilities of the conventional Bessemer method: an excessive nitrogen content of the refined steel and failure to take full advantage of the exothermic role of oxygen as a converter fuel. The final breakthrough in the development of the oxygen process was based on the work of Schwarz, Miles, and Durrer. In an application filed in 1939 and issued as German Patent No. 735,196, on July 3, 1943, Professor C. V. Schwarz of Berlin-Charlottenburg stated: "The object of the present invention is a method of bringing gases into particularly intimate contact with liquid baths, for instance metal baths, by providing the jet of gas directed onto the surface of the bath with such a high kinetic energy that it is capable of penetrating in the manner of a solid body deep into the bath by the use of extremely high velocities lying preferably above the speed of sound. In this way it is possible, without any additional means such as for instance a pipe or the like which is subject to rapid wear, to cause the jet of gas to act within the liquid baths so that the reaction takes place extremely rapidly and completely." In this top-blown pure-oxygen process, Schwarz observed, "the danger of rapid wear of the container liner is By 1948 A. B. Robiette could report that "Developments in the production of cheap oxygen by the Linde-Fränkl and other systems have so reduced the cost of oxygen that it can now be seriously considered both for combustion systems and for the refining of pig iron and the production of steel." "Use of Oxygen for Steelmaking," The Iron and Coal Trades Review. May 28, 1948. p. 1103. Between 1936 and 1940, for example, O. Lellep conducted experiments at Oberhausen, Germany, with the use of pure oxygen in a bottom-blown converter. While he succeeded in producing high quality steel at low cost, he found no way of preserving the service life of the converter bottom, and hence failed to come up with a commercially feasible process. "Versuche zur Stahlherstellung im Herdofen und Konverter unter Benutzung von konzentriertem Sauerstoff, ausgeführt in der Gutehoffnungshütte A.-G., Oberhausen, in der Zeitperiode von 1936 bis 1940," published in Mexico City in 1941; cited in Stahl und Eisen, Vol. 71 (Dec. 20, 1951), p. 1442. By 1945 the Russians had built a special converter plant at their Kuznetsk Steel Works to study the production of Bessemer steel by use of an oxygen-enriched or pure-oxygen blast in a bottom-blown converter. They too failed to develop a method for preserving the service life of the tuyeres when using 100 per cent concentration of oxygen. See the article by V. V. Konjakov in Engineers' Digest, Nov. 1947, p. 522, cited in Iron Age, Feb. 19, 1948. p. 70. The Germans conducted successful experiments in bottom-blown converters by use of 64 per cent pure oxygen at Haspe and 73 per cent pure oxygen at Oberhausen (Stahl und Eisen, Vol. 70 (Apr. 13, 1950), pp. 303-21 and Vol. 71 (Nov. 8, 1951), pp. 1189-99), but failed in further efforts to increase the oxygen concentration in the blast without excessive wear and tear of the converter bottom. As early as 1904 Herman A. Brassert described a side-blown converter using "dry air," oxygen, or oxygen-enriched air. He suggested that a suitable number of tuyeres be posttioned around the converter vessel above the metal line "So as to direct the air issuing from them downwardly onto the surface of the metal in the bath whereby a whirling or rotary motion will be given to the metal." (U.S. Patent No. 1,032,653 applied for on November 11, 1904, and issued on July 16, 1912). Notable among the experiments and pilot projects subsequently undertaken were those of Jones & Laughlin (started in 1942) and Canegle-Illinois (started in 1946). By 1949 both companies had concluded that their side-blown converter (turbo-hearth) process was "fundamentally sound" and that it could be made to yield low-nitrogen, low-phosphorous steels in commercial quantities, if certain operating problems were solved and the equipment design modified. See E. C. Bain (vicepresident, Carnegie-Illinois) and H. W. Graham (vice-president, Jones & Laughlin). "The Turbo-Hearth A New Steelmaking Technique." Iron Age, Apr. 21, 1949, pp. 62-65. For a discussion of other side-blown converter experiments, see Stahl und Eisen, Vol. 62 (Sept. 3. 1942) pp. 749-56 and Vol. 64 (June 1, 1944) pp. 349-58. Both of these volumes of Stahl und Eisen were reproduced and distributed to scientific centers in the United States during World War II by the Allen Poperty Custodian. •Extensive experiments with oxygen-steam combinations were conducted by Coheur Marbais. Daubersy et al. at the Belgian Centre National de Recherches Metallurgiques in Liège. For accounts of these experiments, see Stahl und Eisen, Vol. 70 (Oct. 26. 1950), pp. 1015-17. (Nov. 9, 1950), pp. 1077-79, and Revue Universelle de Mines, Vol. 93 (1950) pp. 104-8, 401-2, 402-7, 408-17, 418-23 and 423-30. 92-578 O 73 66 (2B) PLANNING, REGULATION, AND COMPETITION 57 eliminated since the reaction between oxygen and iron ・・・ takes place in the center of the steel bath and therefore the walls of the vessel are not substantially attacked." In Belgian Patent No. 468,316 applied for on October 4, 1946, granted on November 30, 1946, and opened for public inspection on March 1, 1947-John Miles offered some refinements on the art taught by Schwarz. He, too, worked with a top-blown converter and emphasized the importance of keeping the source of the chemical oxidation reactions within the bath "at a good distance from the refractory lining of the furnace." Finally, Robert Durrer, a Swiss professor who had begun his experimentations at the Institut fur Eisenhüttenkunde of the Technische Hochschule at Berlin-Charlottenburg as early as 1938 and continued them at the Louis von Roll Eisenwerke in Gerlafingen, Switzerland, after the war, succeeded in producing steel with a top-blown, pure-oxygen process in a 2.5 ton experimental converter. On March 21, 1948, as Durrer's associate later reported, he proved that “it is possible to refine pig-iron of varying composition with pure oxygen. There are no difficulties with respect to the durability of the nozzle or the converter lining The qualities of the steel correspond to those of normal open-hearth steel." These experiments by Durrer and Hellbrügge provided the last crucial link in the process of technology diffusion, because it was Durrer who transmitted the Schwarz and Miles teaching (and his experimental findings based thereon) to the eventual patentees the Austrian steel firm VOEST. The sequence of events was as follows: In 1948 VOEST was contemplating an expansion of its steel plants at Linz and was actively considering all available steelmaking processes. Aware of the Durrer-Hellbrügge experiments at Gerlafingen, VOEST dispatched its Works Manager, Dr. Trenkler, to Gerlafingen on May 12, 1949, to inspect the equipment and examine the techniques which had there been employed to produce steel in a top-blown oxygen converter. Encouraged by Trenkler's favorable report, VOEST immediately initiated a test series in a two-ton modified converter which on June 25, 1949, yielded further refinements of the Schwarz-Miles-Durrer art: "first, the blowing of pure oxygen from above onto ⚫ a highly reactive zone in the upper region of the melt, which zone is spaced from the refractory lining of the vessel. Second, the avoiding of deep penetration of the oxygen jet into the bath [again to avoid damage to the converter lining]. Third, the avoiding of material agitation of the bath by the stirring effect of the oxygen jet. Fourth, the creation of a circulatory movement of the bath, not by mechanical action of the jet, but by the chemical reactions."" These refinements of the process solved not only the problem of safeguarding the converter lining, but also the need for dephosphorization through a proper slag composition. In any event, by mid-August of 1949, VOEST was convinced of the soundness of the process and initiated the final experiments to test the process operationally and practically. These were concluded successfully by November 1950, and a new metallurgy had been born.10 VOEST then constructed its first L-D plant which went into large-scale, commercial production in 1952. It is noteworthy that the three major revolutions in steelmaking-the Bessemer, Siemens-Martin (open-hearth), and basic oxygen processes-were not the products of American inventive genius nor the output of giant corporate research laboratories. The oxygen process was developed in continental Europe and perfected by the employees of a nationalized enterprise, in a war-ravaged country, with a total steel ingot capacity of about 1 million tons-by a firm that was less than one-third the size of a single plant of the United States Steel Corporation. History of the Innovation In innovation, as in invention, the giants of the United States steel industry lagged, not led. The first large-scale commercial use of the oxygen process was 7R. Durrer, "Sauerstoff-Frischen in Gerlafingen," von Roll Werkzeitung, Vol. 19 (May 1948), pp. 73-74. H. Hellbrügge, "Die Umwandlung von Roheisen in Stahl in Konverter bei Verwendung von reinem Sauerstoff." Stahl und Eisen, Vol. 70 (Dec. 21, 1950), p. 1211 (freely translated from the original German). Testimony of Dr. Hauttmann, one of the co-inventors of record in Kaiser v. McLouth, Civil Action No. 16,900, U.S. District Court (E.D. Mich.), 1946, Record p. 2754. 10 The Austrians refer to the process as LD which either stands for Linz Düsenverfahren or for Linz-Donawitz (the location of the patente's steel plants). In the United States, it is variously referred to as the Oxygen Converter Process, Basic Oxygen Furnace Process, BOP, or OSM. |