The managers and financiers who built the systems that came to domi- nate American railroad transportation also collaborated in devising the structures to manage them. The speculators, smaller investors, and larger capitalists contributed little. In the 1880s railroad men employed two al- ternative structures for the management of the huge new consolidated megacorps. One, which was entirely the creation of the most able senior career managers, was strikingly similar to those adopted by the largest industrial corporations in the mid-twentieth century. However, it was the other, the one favored by the financiers and the specialized operating executives, which became by 1900 the standard for large American railroad systems.
A memorandum Charles E. Perkins wrote his managers in May 1883 outlining a proposed organizational structure for managing the many properties he had recently obtained for the Chicago, Burlington & Quincy outlined these two alternatives: “There are essentially two different meth- ods practiced by large railroad systems. One method is to spread the working organization so to speak, over the entire system; the other makes a number of different working organizations, or units of management, each complete in itself.”77 Perkins preferred the latter. “It involves a somewhat more expensive management; but I believe this is far more than made up by the greater efficiency and economy in details.”
This second form, invented by the Pennsylvania and enthusiastically endorsed by Perkins, had proved a brilliant success. One British railroad expert writing in 1893 stressed that the Pennsylvania’s administration was the best in the country, and indeed in the world. “The Pennsylvania is in every respect the standard railway of America,” he wrote. “Its rails and rolling stock, its ballast and bridges, its stations and service are regarded as embodying a state of perfection to equal which should be the highest ambition of every railroad company in the country.”78 On this point few railroad men disagreed. Yet despite the success and the convincing argu- ments made by its advocates, relatively few systems adopted this “de- centralized” type of government. Instead they spread their existing cen- tralized structure over their greatly enlarged domains.
The Pennsylvania began to plan a new administrative structure for its system, as it was still carrying out its strategy of expansion. The initial legal changes, which have already been described, placed the control of the system in three interlocking corporations—the Pittsburgh, Cincinnati & St. Louis Railroad known as the Panhandle Company, the Pennsylvania Company, and the Pennsylvania Railroad Company. These three legal entities became the basis for three self-contained administrative networks. The Panhandle or “southern system,” which operated 1,150 miles of road in 1873, included the lines legally held by the Panhandle. The “northern system” of 1,564 miles took the lines controlled by the Pennsylvania Company. The third, the “eastern or Pennsylvania” system, totaling 2,408 miles in 1873, was administered directly by the Pennsylvania Railroad Company.79 As Thomson told his stockholders early in 1873, the object of the new administrative and legal changes was: to secure, by a single
management of these works, harmonious action throughtout the entire system of railways that we control, and at the same time to obtain the best results from the large amount of rolling stock upon them, by transferring, as occasions may require, portions of that on one line to another, where the demand for its use was more urgent and important to the interest of the Company and the public.80
The administration of each of these three systems (each much larger than the Pennsylvania Railroad had itself been in 1870) was placed under a general manager, who had full responsibility and authority for the “ safe and economical operation of the Roads committed to his charge.” He directly controlled the transportation, traffic, and purchasing department of his territorial unit, and was responsible, with the assent of the president, for the hiring, firing, and promotion of all administrative personnel.81 The general managers of the two western units reported to the same set of senior executives, since the Pennsylvania Company and the Panhandle had identical top management.82 One man was the first vice president of both enterprises, watching over traffic and transportation, another was the second vice president of both, responsible for finance, and a third was the third vice president and comptroller of both. The president of the Pennsylvania Railroad was also the president of these two companies.
The internal organization of the three subsystems was similar. The largest, the eastern system, was divided into three large administrative subdivisions and two smaller ones.83 All five were built around what had been independent railroad managements before 1870, the three major units being the Philadelphia and Erie, the United Railroads of New Jersey, and the original line between Philadelphia and Pittsburgh. Their boundaries were now reshaped to meet more satisfactorily the needs of traffic and administrative oversight. So too were their internal subdivisions. The Philadelphia and Erie with relatively little traffic had two such divisions. The United Railroads of New Jersey had three, while the old Pennsylvania Division reached seven.
The general managers, then, supervised, appraised, and coordinated the daily operations of the major subunits within their large territorial ad- ministration. They took the initiative, working closely with each other, on ratemaking within the framework set at the regional interfirm con- ferences.84 They also determined capital requirements for their divisions and appointed managerial personnel. In all three operating systems, the general managers, who had a great deal of freedom, remained responsible for financial performance. They operated, however, within a set of general policies and procedures in whose definition they often played a role. The duties of the general superintendents who reported to the general managers involved, in the words of a contemporary, “constant supervision rather than independent direction.”85 Finally, the division superintendents at the fourth level of management were involved completely in the routine, day-to-day movement of trains and traffic. At all levels, the line and staff distinction prevailed.88
The system’s top managers had their offices in the company’s head- quarters in Philadelphia. The president and the three (soon four) vice presidents were responsible for coordinating and evaluating the perform- ance of the three autonomous subsystems and for planning and allocating resources for the system as a whole. Although these vice presidents had some supervision of operating activities on the lines east of Pittsburgh, they were expected to concentrate their attention on the larger system. When the new structure was first installed, the first vice president handled external strategy for the system as a whole and the relations with all con- necting roads.87 The second vice president, in addition to maintaining an oversight of the traffic and the comptroller’s departments on the lines east of Pittsburgh, was to advise on and review the recruitment and selection of executive personnel throughout all three systems. In addition, the second vice president was particularly charged with assisting the first vice presi- dent “in all matters relating to connecting railroads west of Pittsburgh.”
The third general officer had supervision over construction and acted as a consulting engineer for the three autonomous systems. He also was as- signed the task of keeping a close watch on the “financial condition” and performance of the parent company and its many subsidiaries, including steamship, express, and coal companies. He was to “obtain from the books and accounts in the general offices of such companies periodical state- ments of their business operations, and report them quarterly in clear and concise form to the President.” In 1882, a number of the duties of the third vice president were given to a fourth vice president.88 On the whole, however, the duties of the general officers in the Philadelphia headquarters and of the general managers and the middle managers in the operating units remained relatively unchanged from the early 1870s until after World War I.
The general officers who determined the strategies of expansion and competition and who appraised and coordinated the work of their major units of management did so by constant consultation and correspondence with general managers and department heads of the three primary oper- ating units. They also relied heavily on accounting and statistical data provided by the comptroller’s department. In addition the general office had a staff, including a legal department and a testing and standards laboratory. Since the general executives and staff officers were housed in the same building on South Fourth Street in Philadelphia as were the
senior operating officers of the Pennsylvania Railroad Company, they consulted one another with little difficulty when the occasion arose. However, they undoubtedly did have regularly scheduled meetings to consider the allocation of resources, promotion of personnel, and so on.
On the other hand, their oversight of the two western subsystems fol- lowed carefully planned agenda. On the first Tuesday of each month the president and vice presidents met in Pittsburgh with general officers of the western companies as the Finance Committee of both the Pennsylvania and the Panhandle Companies to review their financial policies and per- formance and to approve or disapprove of expenditures for capital equipment. On the following day they met, this time as the Executive Committee of both enterprises, to review “all matters relating to the busi- ness (except the matter of rates), police, and working of the railways or lines of traffic, owned or controlled by the Company.”80
This structure, with its autonomous subsystem responsible for day-today operations and its general office to handle long-term supervision and planning, was as sophisticated as any modern giant industrial enterprise. It was not, it must be stressed, the result of an evolutionary process. It was instead an almost immediate response to a totally new managerial chal- lenge. Contemporaries credited the innovation to one man, J. Edgar Thomson. As a stockholders’ investigating report noted in 1874: “Your corporation has grown to its present status under the inspiration and guidance of one mastermind—a man of honest intentions and remarkable ability.”90 And in the words of one Pennsylvania executive: “We are specialists, that is, pygmies. Thomson was great in everything—operating, traffic, motive power, finance; but most important of all in organization.” Thomson was indeed one of the most brilliant organizational innovators in American history.
In adopting Thomson’s decentralized structure at the Burlington, Perkins had much to say about the advantages of this type of organization. His road, though somewhat smaller than the Pennsylvania, had four au- tonomous operating divisions including “lines east of the Missouri,” “lines west of the Missouri,” the Hannibal and St. Joseph Railroad, and the “Kansas City lines.” Each had its own transportation, traffic, legal, accounting, and purchasing departments. Only the accounting and pur- chasing departments had direct contact with the general office—the first to provide effective financial controls through uniform accounting and reporting, and the second to take advantage of the economies of large- scale purchasing. The other three units reported directly to the general manager in charge of the subsystem.
These general managers, Perkins stressed, must be generalists rather than specialists. Such an executive should not be simply a “train and track” man, wrote the Burlington’s president, rather “he ought to be more of a man of business experience who can come into contact with businessmen of the community.”91 Perkins considered “a sound head and good judge- ment” more necessary than engineering and technical skills.92 Such man- agers must avoid becoming involved, Perkins repeatedly pointed out, in operating details.93
For Perkins the most important duties of the top managers in the gen- eral offices were strategic planning and recruitment of senior managerial personnel. “In the administration of so large a property as we now have, the chief business of a President and the Vice President must be with questions of policy and in selecting and keeping the good men in important places.”94 Perkins himself concentrated on the second, for, in his opinion, “nothing is more important in the management of our large railroad properties than to make and keep good men.”95 The president and the first and second vice presidents were to maintain a watch on policy and strategy, while the second vice president was also to specialize in the coordination and appraisal of the operating units. In the early 188os strategy was critical. Perkins reminded his managers that:
Every mile of railroad added to the system anywhere is just so much more property exposed to the attacks of our enemies; the country we now serve is so large that we are exposed to attacks in a.great many directions. All this wants careful watching, so that we may provide against such attacks, where it is possible to do so. Then, too, the country is growing; and the opportunities for building profitable lines in connection with those which we now have, has to be watched. This particular branch of our business, taking care of our geographical relations, is, in itself, of so much consequence, and involves so much study, and so much going on and about from one place to another, that it should be the duty of one man, acting under the Second Vice-President, and also coming in more or less direct contact with the President, when necessary, to look after it.96
The second vice president was also to keep in touch with all “pooling ar- rangements, especially the important pools of through business.” In co- ordinating and appraising the activities of the different units of manage- ment, he and the president were not only to review regularly the accounts and statistics of the different units but also to spend “a certain number of days every month or two with each General Manager, on the ground, for the purpose of observing him and his methods of dealing with questions that come before him.”
To Perkins an organization of regionally autonomous “systems” had obvious advantages over the centralized functionally departmentalized structure. It “made possible obtaining the advantages of the large property and organizations, without losing the advantages of the small property and the small organization.”97 It brought responsible senior management closer to the firing line. In addition “the local population in the country or towns through which the road passes can more readily know and often more readily see in person the General Manager.”98 Such an organization encouraged initiative and independent thought. “Men’s minds and abilities grow and expand with use and responsibility.”99 Finally the decentralized structure aided in “preparing and educating men” for top managerial positions. Much the same arguments would be made again in the mid- twentieth century by advocates of comparable decentralized structures in large multiunit industrial enterprises.
The decentralized structure with its autonomous operating divisions and its policy making, evaluating, and coordinating in the general office was adopted by a few large roads whose managers paid close attention to organization matters. In the 1880s the Baltimore & Ohio, the Rock Island, the Sante Fe, the Union Pacific (under Adams), the St. Louis & South- western (before Gould took it over), and the Plant lines were using this type of organization.100 On the other hand, in the same decade those roads where financiers had a strong influence on top management turned to another model. They looked instead to the New York Central, the Penn- sylvania’s major rival in trunk line territory. One reason was that J. P. Morgan, the nation’s most powerful investment banker and foremost railroad reorganizer, received his practical knowledge of railroading as a director with many years of service on the New York Central’s board.
In May 1883 William H. Vanderbilt, on deciding to retire from active business, brought forward to the Central’s board of directors a plan of government for the properties it had recently obtained.101 Each of the roads that Vanderbilt and his associates had acquired remained administratively as well as legally independent entities. The operating heads, normally their presidents, were carefully selected career managers. The roads were unified by means of interlocking directorates and a common financial office in New York City. In the memorandum to the central, board outlining his plan Vanderbilt noted: “Under the reorganization, each of them [the roads controlled by the Central] will elect a Chairman of the Board, who in connection with the Executive and Finance Committee, will have immediate and constant supervision of all the affairs of the companies, and bring to the support of the officers, the active assistance of the Directors.”102 The executive and finance committee of the New York Central referred to here was a single committee and acted as the central office of the system. But unlike that of the Pennsylvania it consisted not of salaried managers but part-time representatives of investors with other business activities of their own. Vanderbilt’s two sons then became chairmen of the boards (or in the case of smaller companies, presidents) of the several roads. Cornelius took the chairmanship of the New York Central (which also operated the Harlem) and the Michigan Central (which also operated the Canada Southern). William K. became the chairman of the Lake Shore (which also operated the Nickel Plate). E. D. Worcester, secretary of the New York Central, became treasurer of the Michigan Central (and the Canada Southern) and the Lake Shore (which operated the Nickel Plate). On the other hand, Vanderbilt did not create similar arrangements for those roads in which the Central had large blocks of stock but did not fully control. On the Chicago & Northwestern, the Bee Line, the Boston & Albany, and later the Cleveland, Cincinnati, Chicago & St. Louis, members of the Vanderbilt family and their associates did no more than sit on their boards, usually as members of their finance committees.
As a result, the New York Central system had no general office or general command comparable to that of the Pennsylvania or the Burlington. The third vice president of the New York Central had the responsibility for calling the meetings of the presidents of the roads in the system to consider rates and connections, but he did so only occasionally. The chairmen of the boards appeared to have met on a somewhat regular basis. But no full-time executive or set of executives had the responsibility for planning and coordinating the system as a whole.103 The one group entrusted with this function, the members of the Central’s executive and finance committee, were all active businessmen in their own right and could devote only part of their time to the affairs of the system. Even the younger Vanderbilts were part-time executives, spending much more of their time on leisure and social affairs than on railroading.
One result of this loose organization was that the New York Central was unable to obtain the economies of scale provided by the staff units in the general office. There were no standardization or testing laboratories for the system as a whole comparable to those set up on the Pennsylvania in 1875 and on the Burlington in 1876.104 Nor could the Vanderbilt system benefit from the advantages derived from centralized purchasing, a centralized legal staff, or a centralized management of insurance and pension funds for workers.
More serious was the lack of a central office to evaluate the performance of the operating units and to plan and allocate resources for the system as a whole. The statistical data reviewed by the board and its committees were financial rather than operating. The finance and executive committee looked at the balance sheets and operating ratios provided by Worcester’s office but not at the operating figures or cost accounting data that flowed into the office of the different presidents and on which evaluation of managerial performance had to be based.
In allocating the funds for several roads, the Central’s board appears to have acted in an ad hoc manner. As renewal and repairs were considered operating expenses, capital expenditures for such items remained com- pletely under the control of the operating managers. But all expenditures for new equipment and construction required the approval of the local boards and apparently the Central’s executive and finance committee. There is no evidence that that committee developed any systematic pro- cedures to review carefully the financial needs of the system as a whole. It merely responded to individual requests from the career managers. Thus Cornelius Vanderbilt replied to a proposal by John Newall of the Michigan Central with a brief note saying: “Newport, R.I., 31 Aug. 92: You can proceed with freight house, Cleveland: also the grading for second track Pettisville to Stryker and Kennelsville to Goshen.”105 The financiers on the board had a powerful veto power over the proposal of the managers to improve or expand facilities, but they had neither the time nor the information to make their own constructive suggestions about capital investment.
This division of labor in top management in which the professional managers supervised operations but the financiers controlled financial policy became standard on American railroads. For those roads controlled by speculators like Gould, Sage, Brice, Clyde, and the Moore brothers, the gap between operations and finance was greater than on the Vanderbilt roads. The speculators paid almost no attention at all to operating needs, nor were they particularly concerned about the caliber of the managers operating their lines. Not surprisingly the Gould roads became, in Robert Riegel’s words, “a synonym for bad management and poor equipment.”106
On those roads financed or refinanced by the investment bankers (and these included most of the major systems in the country), the relations between the boards and the operating managers came to be similar to those on the Vanderbilt roads. Morgan, trained in the Vanderbilt school, carefully picked experienced, tested career managers as presidents of the roads he reorganized. He gave them almost complete autonomy in op- erating matters, while having the board retain a close oversight of financial affairs including dividend policy and the allocation of financial resources. Members of the Morgan firm chaired the boards and sat on their executive and finance committees. (On most roads these became separate committees.) Kuhn, Loeb; Lee, Higginson; Kidder, Peabody, Belmont; and Speyer all acted in much the same manner. So too did such financiers as Harriman and Hill, although because both had long experience in rail- roading they paid closer attention to operating data than did the others. No financier, not even Harriman who did build an abbreviated superstructure to oversee the Union Pacific and Southern Pacific, created a structure comparable to the Pennsylvania tó administer the systems they financially controlled.107
In their railroad reorganizations Morgan and the other financiers did much more than merely appoint presidents and members of boards of directors. They instituted financial and administrative reforms within the systems they refinanced.108 On the financial side they lowered the fixed charges on the bonded debt by converting bonds into preferred stock. Common stock issues were reduced through exchanging four, five, or more shares of old for one of new and even then assessing the stockholders to provide new capital. In issuing new securities the amounts were based on the earning power of a road as indicated by its operating ratio. Bonds to be used for new capital equipment were to be expended in specified amounts over a specified period of time. In most cases the bankers insisted on setting up a voting trust which gave them the power to vote the majority of the stock for a period of normally five years or up to the time when the preferred stock began to pay its 4 or 5 percent dividend regularly. This last provision was adopted as much to prevent speculators from obtaining control of reorganized roads, as those companies became once again financially viable, as it was to assure the bankers of a continuing oversight of the road’s finances.
In their administrative reorganizations the bankers adopted the cen- tralized operating structure rather than the decentralized one used on the Pennsylvania and the Burlington. In making this move they often had the support of the more specialized operating managers. The experience of the Illinois Central indicates why both financiers and middle managers favored the centralized structure.
In the mid-1880s, the managers and investors of the Illinois Central who went east to find funds to cover the costs of system-building, obtained the support of a group of conservative and respected New York bankers including August Belmont, Robert Goelet, Sidney Webster, and young Edward H. Harriman.109 In 1887 these financiers appointed as president Stuyvesant Fish, who had for the previous ten years worked in the road’s financial department, and they appointed Harriman to Fish’s former posi- tion of vice president in charge of finance. The executive committee then set up a subcommittee to outline a “plan adequate for conducting the present and prospective business of the Company.”110 In the resulting dis- cussions the financiers relied heavily on the operating men for suggestions. The acting general manager favored a scheme of autonomous territorial units similar to that of the Burlington.111 The traffic manager, however, argued strongly that he should have full control over all traffic activities of all the lines incorporated into the system.112 He wanted to report directly to the president instead of to the general manager. By his plan, the president would coordinate and decide disagreements between traffic and transportation departments. The executive who had worked under Fish in the financial office wanted similar centralized control over the road’s accounting, auditing, and purchasing officers, and strongly supported the traffic manager’s proposal. So the centralized structure was adopted.
The new organization thus concentrated all decisions regarding traffic, transportation, and finance in Chicago.113 The three major functional de- partments remained quite independent. Even their regional subdivisions did not cover the same geographical areas. In addition, the central office at Chicago housed the chief engineer who was in charge of new construction and acted as a staff engineer to the transportation department, and the smaller legal, secretary’s, and land offices, as well as the relief (employee benefits) department. Only the president residing in Chicago coordinated all these activities. Since nearly all the board members lived in New York and were involved in other tasks, they had little time to review past operations or plan for future ones.
The New York financiers preferred this plan for several reasons. By having fewer managers, administrative costs were reduced. By having all the senior executives housed in one Chicago office, these managers were able to consult with one another and to be easily reached by the New York directors. Finally, the traffic department’s autonomy permitted it to adjust its schedules swiftly to meet continuing rate changes. To many managers, as well as to many bankers, these considerations outweighed the advantages that Perkins had outlined for the decentralized structure with its possibilities for increased managerial efficiency and better training.
By the beginning of the new century, nearly all American railroad systems were using this type of internal organization structure. Those roads that had adopted the Pennsylvania’s decentralized form reverted, usually during financial reorganizations, to the centralized form. These first modern megacorps thus came to be administered by career managers who used operating structures similar to those devised by McCallum and Thomson in the 1850s, structures which were, in Perkins’ phrase, “spread . . . over the entire system.” Because of the increased size these organizations had at least two levels of middle management between the division superintendent and the president. Some roads even moved away from the divisional form with its line and staff differentiation to the departmental one. Most, however, continued to use the line and staff device to help assure effective coordination of movement of trains and traffic. Other matters requiring coordination between the transportation, traffic, and financial departments had to be decided by the president.
Source: Chandler Alfred D. Jr. (1977), The Visible Hand: The Managerial Revolution in American Business, Harvard University Press.