The Tycoons: How Andrew Carnegie, John D. Rockefeller, Jay Gould, and J. P. Morgan Invented the American Supercompany

by Charles R. Morris

  Attempts to stamp the Taylorist label on the multi-division corporate management systems worked out early in the century, by Gerard Swope at General Electric and Hamilton Barksdale at DuPont, are similarly unconvincing. DuPont’s paint, varnish, dye, and dynamite lines were all flow-process businesses on the model Rockefeller had pioneered at Standard Oil, while Barksdale’s ingenious blend of centralized control systems and decentralized product divisions was wholly outside Taylor’s experience. General Electric, whose major product early in the century was consumer lightbulbs, had much more to learn from Ford and the sewing machine industry than from Taylor, while Swope’s control systems, like Barksdale’s, were of a scale and breadth Taylor had never dealt with. It seems that only Taylor and his acolytes, and, inexplicably, several generations of business historians, could nurture the delusion that he was the discoverer of planning and schedules. Taylor implicitly undercut his own claims when he cautioned Gantt and others against accepting engagements from well-managed companies, where they were not likely to make much difference.

  Railroads were naturally incensed at the notion that Taylorism was the solution to their budget problems. A riposte to Emerson’s claims at the Eastern Rate hearings was published the following year in the Quarterly Journal of Economics. While accepting all Scientific Management claims at face value, the author, an economist named William Cunningham, quite correctly pointed out that its methodologies were designed for job shops or manual laborers, and had limited applicability to railroads. Railroads, in any case, had been the pioneers of the scheduling and control systems at the heart of Scientific Management. Emerson’s specific claim that he had reduced Santa Fe locomotive repair times from sixty to thirty days actually came off badly in comparison with the Harriman lines’ average of fifteen days. The Santa Fe, in fact, was below average on virtually all cost and performance indicators, and so was a poor exemplar for global industry cost reductions.

  It was still true, however, that run-of-the-mill American enterprises, in contrast to the railroads and other big companies, were typically badly managed. With the sea change in the scale and tempo of American business since the 1880s, many companies were struggling to cope and sorely needed help installing basic management processes. Business consultancy was becoming an industry in its own right about the time that Taylor died. As the former band of Taylorists built profitable businesses, their practical eclecticism caused much friction with Taylor, especially between Taylor and Gantt, his very first disciple. Preaching a “Taylor system” was great for marketing, but in their day-to-day work the disciples quietly dropped “system” pretenses in favor of selling solutions that fit their clients’ circumstances.*

  The phenomenon of Taylorism, in any case, had little to do with bread-and-butter management consulting, for practical businessmen quickly learned to pick and choose among consultants’ offerings without falling victim to -isms. The true audience for Taylorism was rather America’s new intelligentsia—journalists, professors, and pundits. The essence of Scientific Management, Taylor preached, was to remove “all possible brain work” from the practical men in the shops, whose sole job would be to see that “operations planned and directed from the planning room are promptly carried out.” Now that was a sentiment intellectuals could rally behind. For the “chattering classes,” as we now call them, Scientific Management seemed to open a path to the “social control” they so ardently desired.

  . . . And There Were Consequences

  Taylorism says more about its devotees than about Taylor. Taylor himself was a narrow-minded obsessive, a hard-driving plant manager, and a snob. While he made important contributions to machining management and technology, his conviction that there was a science that would allow planners to determine the “one best way” for every task was statistically dubious and wrong in practice. But it had immense, if short-lived, resonance for opinion leaders.

  The first rush of enthusiasm for Taylorism arose because it offered a path for coming to terms with the awesome new corporate power centers. Instead of impotently ruing the might of the trusts, intellectuals could celebrate big companies as triumphs of American enterprise, with the comfort of knowing there was a technology for controlling them. As the faith in technocratic Progressivism began to wane in the 1920s, however, pundits lost interest in Taylorism, although its appeal to European and Soviet planners* lasted some years longer. In the public mind, the name of Taylor came to be associated with the semi-comic figure of the time-and-motion man—the 1954 musical The Pajama Game is a typical spoof.

  Aspects of Taylorism, however, especially the attitudes of Taylorism, took root in the business schools and continued to exert a subtle, but deep, influence on the intellectual assumptions of an increasingly professionalized cadre of managers. Three path-breaking books by Alfred Chandler—Strategy and Structure (1962), The Visible Hand (1977), and Scale and Scope (1990)—offer the canonical account of the rise of the American managerial tradition. Chandler’s approach to history has been called “teutonic”; his account, indeed, is of a dialectical progression, a kind of triumphal Hegelian unfolding of management consciousness, from the small, single-unit enterprise of the era of family capitalism through “administrative coordination” of the multi-unit business, to the large, vertically integrated company and the discovery of “the economies of speed,” and, finally, in the highest realization of the industrial Spirit, the vertically integrated, hierarchical, multi-divisional, broadly owned and professionally managed “modern industrial enterprise.” Those “Center” companies, as Chandler called them, produced many different products, controlled almost all their essential resources, and were marked by long-run, high-speed production lines and clear managerial stratification. “[M]anaging and coordinating the processes of production and distribution” was the job of middle managers, while “the top managers concentrated on evaluating, planning, and allocating resources for the enterprise as a whole.” Chandler’s top management was the central intelligence that energized the whole organization, the hot star at the heart of the corporate galaxy. It is, in fact, just a colossal version of Taylor’s Planning Room, the place where the “brainwork” happened.

  The new business schools evangelized a highly intellectualized “professional” version of management, which is faithfully reflected in Chandler. It is striking, for example, how little attention Chandler pays to production technology, or to technology in general; instead, it appears in his story almost as an exogenous variable. The companies he admires all have research laboratories, but they are black boxes on a chart, intermittently producing recommendations to be mulled by the philosopher-kings at the top. In the Harvard Business School’s maiden (1908) curriculum, of fifteen course choices only two—railroad operations and municipal administration—dealt explicitly with management. The remainder were spread among accounting, commercial law, investment, organization, economics, insurance, and the like, as if the school’s management hatchlings would skip right by gritty plant-level problems into the top management orbit. In fact, from the beginning, the school’s graduates have overwhelmingly gravitated to finance and consultancy.

  The business school’s founders had actually considered whether to include a manufacturing curriculum, but decided against it, since it appeared that Taylor and his disciples had exhausted the topic; so as a compromise they engaged Taylor as a guest lecturer. There was evidently little further thought on the subject, for a half century later, Chandler noted that post–World War II plant managers “had plenty of information to go on, for it was on this lowest administrative level that Frederick W. Taylor, Frank Gilbreth . . . and other advocates and practitioners of ‘scientific management’ concentrated their energies.”

  The remoteness of the business schools is unwittingly demonstrated by Chandler’s Strategy and Structure. It is a fine book, the first to explore the evolution and implications of the centralized/decentralized organization blend pioneered at companies like Dupont, and it triggered a vast amount of spin-off rese
arch, if mostly of the anecdotal variety. What is striking, however, is that Chandler’s book appeared forty years after the multidivisional organization became widely adopted in business—Chandler’s case studies are mostly from the 1920s. Imagine if engineering professors at MIT first noticed the integrated circuit only in 2002 or so.

  The disconnect had real consequences. Consider just one arcane, but important, episode—the search for the EOQ, or the Economic Order Quantity, the mathematically optimized solution to the very practical, and very challenging, business problem of setting the “right” amount of inventory. The research, which consumed quite formidable amounts of brainpower throughout the post–World War II era, proceeded from three apparently incontrovertible axioms:

  An abandoned steel mill, about 1980, detritus from the cataclysmic failure of a complacent, “professionalized” American management cadre. The jury is still out on whether America has regained its competitiveness in basic industry.

  a) At some point, the cost of eliminating additional defects starts rising, and becomes prohibitive.

  b) Unplanned stops in a high-speed production line are catastrophically expensive.

  c) Unit production costs fall with longer production runs, because setup and changeover costs are spread over the larger volume.

  The EOQ, or right inventory level, is the one that optimizes within those constraints—that is, you need just enough inventory to replace the economically optimum level of defective parts; enough to cover for a botched assembly operation to avoid stopping the line; plus the temporary overages that result from optimum-length production runs. EOQ-style problem solving was a deeply engrained characteristic of the American managerial psyche, a milestone toward the ideal of a professionalized Scientific Management that inspired Taylor and was celebrated by Chandler.

  The 1970s/1980s sweep-the-board triumph by Japanese companies in nearly every important mass-production industry shocked and demoralized American executives. It was not just the humiliation of catastrophic defeat but the stomach-wrenching discovery that fundamental compass points, like the EOQ axioms, were utterly, and disastrously, wrong.* The most famous of the Japanese paradigms, the “Toyota system,” developed over more than twenty years under the leadership of Taiichi Ohno, was a direct refutation of the EOQ logic. Costs fell with zero defects. The right amount of inventory was none at all. One always stopped the production line to prevent a defect (or else it would always recur). Long production runs always produced wasteful amounts of inventory. (The solution was to reduce the cost and time of changeovers to near zero.) Ohno’s system emphasized close contact between top management and the plant floor, and a deep respect for workers—in contrast to the quite open disdain that pervaded Taylorism. The American “pursuit of quantity and speed,” Ohno suggested, produced only “unnecessary losses.” Vertical integration was usually wasteful; it was more efficient to develop stable supply relationships with specialist contractors (in contrast to the adversarial American contracting culture).

  American managers naturally came in for widespread and well-deserved criticism. One of the earliest, and most scathing, “Managing Our Way to Economic Decline,” was published in 1980 by two Harvard Business School professors, Robert Hayes and William Abernathy. It attacked American executives for “priz[ing] analytical detachment and methodological elegance over insight based on experience” and “the false and shallow concept of the professional manager, a ‘pseudo-professional’ really.” The professors’ attack is right on target, but comes with rather poor grace, for there is no wisp of recognition that their own institution had spent three quarters of a century drilling precisely those values into American business elites. Robert McNamara, after all, was the paragon of the business school’s generalist manager tradition, and his meaningless body-count database in Vietnam the perfect expression of its Taylorist genes.

  At the conclusion of The Visible Hand, Chandler remarks that “the businessman of today [the 1970s] would find himself at home in the business world of 1910.” Indeed he would have, which is testimony not only to the greatness of the original tycoons but to the inordinate length of time we lived off their capital.

  *Immigrants tended to come off badly in these exercises. True to the Pearsonian ideal of cold adherence to (apparent) facts, the scientists tracked immigrants’ assimilation, measured their skulls, and generally pronounced them inferior. Race protection was one of the era’s important subthemes. Pearson himself was a strong advocate for race-based eugenics.

  *In a steam- or water-driven factory, power was transmitted through shaft-and-belt systems. Loose or lightweight belts could have a drastic effect on performance, so consistent belt maintenance paid large productivity dividends. Belt-driven machines disappeared with the spread of electric motors toward the end of the century.

  *A Gantt chart is a visual representation of project tasks as a series of time bars; they are still a standard output of project planning software packages.

  *His address was followed by a metallurgist’s presentation that specified the detailed composition of the new steels. Taylor, whose science was in the older empirical tradition, pooh-poohed the metallurgical results; but his failure to specify the metallurgy caused him to lose his patents in 1909. The patents were difficult to defend under any circumstances, and Taylor went to ingenious lengths to conceal the underlying (and relatively simple) processes that they entailed. The small community of cutting toolmakers, still concentrated in Sheffield, had duplicated Taylor’s results by the middle of the decade.

  *Gilbreth was the model for a novel by two of his children that later became the movie Cheaper by the Dozen (1950). He is, albeit affectionately, portrayed as something of a fool. His wife, Lillian, was actively involved in his consulting work, and they both constantly motion-studied their kids.

  *In private, Emerson was refreshingly frank about his work. When he had arrived at the Santa Fe, he said he was “entirely ignorant as to workings of such a plant . . . and had to be cautious about opening my mouth, lest I put my foot in it. By saying nothing, I was credited with deep knowledge.” What was his method? “Every employee is strictly instructed to heed my requests . . . and then I sail around and find all the fault I can and propose various improvements.”

  †Even James Dodge, his good friend and disciple who chaired the ASME editorial committee, balked at Taylor’s claims, pointing to the unscientific designation of “first-class” workers, and the arbitrary P factors in Taylor’s time studies.

  *Taylor was not directly involved in the engagement, but had been its impresario, and had picked the consultants. The machine shop engagement went smoothly, as usual, but foundry workers objected to stopwatch timing. Later, they secretly shadow-timed operations being tracked by a time-study “expert,” a Taylor associate named David Merrick, not realizing that Merrick counted only “productive” time instead of the total elapsed time. When he came up with only twenty-four minutes for a task they timed out at fifty, they assumed he was lying. But when Merrick was later pressed on the point, he conceded that he knew nothing about foundry work, and was “very confident that I didn’t get a true observation. . . . I felt that 30 minutes was too long a time for it, so I made a guess at 24 minutes.” Taylor himself had shocked the head of the arsenals, William Crozier, a true-believing Taylorite, when he suggested that they could install piece rates just by making “a rough guess” at the times.

  *Gantt had been Taylor’s first choice for the Watertown Arsenal assignment, but he scandalized his old mentor by pointing out that the client had asked only for a foundry piece rate system, which he could install in a couple of months. Taylor insisted on a three-year Scientific Management engagement that began in the machine shop. Had he taken Gantt’s advice, he might have saved himself much grief in his last years. Gantt was the only member of the Taylor coterie with experience in foundries, and was especially good at avoiding the kind of worker disaffection that led to the Watertown job action.

  *Time-and-motion studi
es were enthusiastically incorporated into the Soviet Union’s Five-Year plans. By 1930, the Rates and Norms Bureaus had set some 232,000 separate norms, derived from the “science of biomechanics” and covering about 70 percent of all workers. Metals trades were reclassified from twelve categories into 176. The blizzard of statistics perfectly suited careerist managers and bureaucrats in a system in which most factories were idle much of the time for a lack of equipment and supplies.

  *As part of my business activities, through most of the 1980s I spent a substantial amount of time with manufacturing companies. It was the period of a determined, almost frenzied, top-to-bottom effort to rethink every basic plant assumption. A senior manufacturing manager at Cummins Engine, one of the most effective of the early responders, repeated to me several times, almost in awe, that “all the textbooks were wrong.”

  APPENDIX I

  The Carnegie Company’s 1900 Earnings

  The price set by Andrew Carnegie for Carnegie Co. in the run-up to the U. S. Steel merger included $80 million for “Profit of past year and estimated profit of coming year.” Early in 1900, the company had confidently expected to earn $40–50 million for the year. James Bridge, who had good inside sources, stated in his 1903 Inside History of Carnegie Steel that the company earned $40 million in 1900, and Carnegie repeated the claim in his Autobiography. Bridge is the source for Carnegie Co. background in the Stanley Committee’s 1911 U.S. Steel investigation, and the $40 million number is cited by virtually all subsequent historians. While the records are insufficient to pin down a precise number, the $40 million is far too high.* (Since total second half earnings were only about $6 million, Carnegie’s implicit forecast of $50 million for 1901 seems clearly to be a misrepresentation.)

  The Carnegie Co. was a holding company formed in April 1900, comprising: