The Modem Corporation: Railroad Organization

The 1840s mark the beginning of a great wave of organizational change that has evolved into the modem corporation (Chandler, 1977). According to Stuart Bruchey, the fifteenth-century merchant of Venice would have understood the form of organization and methods of managing men, records, and investment used by Baltimore merchants in 1790 (1956, pp. 370-71). Those practices evidently remained quite serviceable until after the 1840s. The two most significant developments were the appearance of the railroads and, in response, forward integration by manufacturers into distribution. Selective forward integration is described and interpreted in Chapter 5. The experience of the, railroads has yet to be addressed.

Although a number of technological developments—including the telegraph (Chandler, 1977, p. 189), the development of continuous process machinery (pp. 252-53), the refinement of interchangeable parts manufacture (pp. 75-77), and related mass manufacturing techniques (chap. 8)—contributed to organizational changes in the second half of the nineteenth century, none was more important than the railroads (Porter and Livesay, 1971, p. 55). Not only did the railroads pose distinctive organizational problems of their own, but the incentive to integrate forward from manufacturing into distribution would have been much less without the low-cost, reliable, all-weather transportation afforded by the railroads.

The appearance and purported importance of the railroads have been matters of great interest to economic historians. But with very few exceptions the organizational—as opposed to the technological—significance of the railroads has been neglected. Thus Robert Fogel (1964) and Albert Fishlow (1965) “investigated the railroad as a construction activity and as a means of transport, but not as an organizational form. As with most economists, the internal workings of the railroad organizations were ignored. This appears to be the result of an implicit assumption that the organization form used to accomplish an objective does not matter” (Temin, 1981, p. 3).

The economic success of the railroads entailed more, however, than the substitution of one technology (rails) for another (canals). Rather, organizational aspects also required attention. As Chandler puts it:

[The] safe, regular reliable movement of goods and passengers, as well as the continuing maintenance and repair of locomotives, rolling stock, and track, roadbed, stations, roundhouses, and other equipment, required the creation of a sizable administrative organization. It meant the employment of a set of managers to supervise these functional activities over as extensive geographical area; and the appointment of an administrative command of middle and top executives to monitor, evaluate, and coordinate the work of managers responsible for the day-to-day operations. It meant, too, the Simulation of brand new types of internal administrative procedures and accounting and statistical controls. Hence, the operational requirements of the railroads demanded the creation of the first administrative hierarchies in American business. (1977, p. 87]

To be sure, that can be disputed. Markets, after all, can and do perform many of those functions. What is it about the “operational requirements” of the railroads that was responsible for the displacement of markets by hierarchies? Does similar reasoning apply to other transportation systems, such as trucking?

The natural railroad units, as they first evolved, were lines of about fifty miles in length. Those roads employed about fifty workers and were administered by a superintendent and several managers of functional activities (Chandler, 1977, p. 96). That was adequate as long as traffic flows were uncomplicated and short hauls prevailed. The full promise’of the railroads could be realized, however, only if traffic densities were increased and longer hauls introduced. How was that to be effected?

In principle, successive end-to-end systems could be joined by contract. The resulting contracts would be tightly bilateral, however, since investments in site-specific assets by each party were considerable. Contracting difficulties of two kinds would have to be faced. Not only would the railroads need to reach agreement on how to deal with a series of complex operating matters— equipment utilization, costing, and maintenance; adapting cooperatively to unanticipated disturbances; assigning responsibility for customer complaints, breakdown, and so on—but problems of customers contracting with a set of autonomous end-to-end suppliers would have to be worked out.

There were several possibilities. One would be to be patient: The marvel of the market would work things out. A second would be to move to the opposite extreme and coordinate through comprehensive planning. A third would be to evolve organizational innovations located in between.

David Evans and Sanford Grossman interpret the railroad response in market terms. They note that “market systems in which property ownership is dispersed among numerous self-interested businesses and individuals have demonstrated a remarkable ability to coordinate the provision of goods and services” (1983, p. 96), and they specifically apply this argument to the railroads;

The experience of the railroads in the 19th century . . . demonstrates how the market system encourages physical coordination. Many separate companies built segments of our railroad system during the mid-19th century. By interconnecting with each other and enabling produce and passengers to transfer easily between railroads, these companies were able to increase revenues and profits. [Evans and Grossman, 1983, p. 103]

Evans and Grossman support this by reference to the study by George Taylor and Irene Neu, who reported that traffic between New York City and Boston moved easily over tracks owned by four different companies in 1861 (Taylor and Neu, 1956, p. 19). They also cite Chandler in support of their argument that “the market provides strong incentives for physical coordination without common ownership” (Evans and Grossman, 1983, p. 104, n. 22).

Evidently there is more to railroad organization than “physical coordina- tion,” however. Otherwise the natural railroad units of fifty miles in length would have remained intact. And there is also more to railroad organization than unified ownership. Thus the Western and Albany road, which was just over 150 miles in length and was built in three sections, each operated as a separate division with its own set of functional managers, experienced severe problems (Chandler, 1977, pp. 96-97). As a consequence a new organizational form was fashioned whereby the first “formal administrative structure manned by full-time salaried managers” in the United States appeared (pp. 97-98).

That structure was progressively perfected. The organizational innovation that the railroads eventually evolved is characterized by Chandler as the “decentralized line-and-staff concept of organization.” It provided that “the managers on the line of authority were responsible for ordering men involved with the basic function of the enterprise, and other functional managers (the staff executives) were responsible for setting standards” (Chandler, 1977, p. 106). Geographic divisions were defined, and the superintendents in charge were held responsible for the “day-to-day movement of trains and traffic by an express delegation of authority” (p. 102). The division superintendents were on the “direct line of authority from the president through the general superintendent” (p. 106), and the functional managers within the geographic divisions—who dealt with transportation, motive power, maintenance of way, passenger, freight, and accounting—reported to them rather than to their functional superiors at the central office (pp. 106-7). That administrative apparatus permitted individual railroads to operate thousands of miles of track by 1893.

To be sure, that falls well short of central planning. Moreover, the contractual difficulties posed by coordination and efficient utilization to which Chandler referred were not the only factors that elicited the large system response. Chandler also gives great weight to the strategic putposes served (1977, chap. 5). The làttçr, however, also has contractual origins: Unable to control prices and allocate traffic thrtpgh interfirm organization, thé railroads were driven to merger. Thus v/hereÆtitere is a widely held view that express or tacit collusion is easy to effecMte—John Kenneth Galbraith opines that “the firm; in tacit collaboration wth other firms in the industry, has wholly sufficient power to set and maintain prices” (1967, p. 200)—that is repeatedly refuted by the evidence. The history of cartel failures among the railroads is especially instructive. The early railroads evolved a series of progressively more elaborate interfirm structures in an effort to curb competitive pricing. The first involved informal alliances, which worked well until the volume of through traffic began to fall off and competitive pressures increased. With the onset of the depression in 1873 there began an increasingly desperate search for traffic … Secret rebating intensified. Soon roads were openly reducing rates.” The railroads thereupon decided to “transform weak, tenuous alliances into strong, carefully organized, well- managed federations” (Chandler, 1977, pp. 134, 137). The membership of the federations was expanded, and other federations in other geographic regions appeared. As Albert Fink, who headed up the largest such federation realized, however, “the only bond which holds this government together is the intelligence and good faith of the parties composing it” (Chandler, 1977, p. 140). To rectify that weakness, Fink urged the railroads to seek legislation that would give the actions of the federation legal standing.

Lack of legal sanctions means that loyal members of the cartel must exact penalities against deviants in the market place. Unless such disciplinary actions (mainly price cuts) can be localized, every member of the cartel, loyalist and defector alike, suffers. That is a very severe (if little remarked) limitation on the efficacy of cartels. Inasmuch as national legislation was not forthcoming, Fink and his associates “found to their sorrow that they could not rely on the intelligence and good faith of railroad executives” (Chandler, 1977, p. 141). In the end, the railroads turned to merger. The high-powered incentives of autonomous ownership evidently presented too strong a temptation for cheating in an industry where sunk costs were substantial.

The railroad industry thus progressed from small, end-to-end units with fifty miles of track to systems of several hundred and, eventually, to several thousand miles of track. Market coordination was thus supplanted by admin- istrative organization in substantial degree:

The fast-freight lines, the cooperatives, and finally the traffic departments of the larger roads had completed the transformation from market coordination to administrative coordination in American overland transportation. A multitude of commission agents, freight forwarders, and express companies, as well as stage and wagon companies, and canal, river, lake, and coastal shipping lines disappeared. In their place stood a small number of large multi-unit railroad enterprises… By the 1880s the transformation begun in the 1840s was virtually completed. [Chandler, 1977, p. 130]

To be sure, transaction cost economics does not predict the final configu- ration in detail. It is nevertheless noteworthy that (1) efficient technological units were very small in relation to efficient economic units in this industry, which is to say that organizational rather than technological factors were responsible for the creation of large systems; (2) transaction cost economics predicts that severe problems will arise in attempting to coordinate autonomous end-to-end systems that are characterized by site specificity by contract;2 and (3) the limits of cartels also have organizational origins and are evident upon posing problems of interfirm organization in contracting terms. That the trucking industry, which does not have the same site specificity (mainly roadbed) features, should differ from the railroad industry in significant respects is furthermore predicted by transaction cost reasoning. (Indeed, the absence of site-specific investments in the trucking industry make it a much better candidate than the railroads to illustrate the Evans and Grossman argument that market coordination is a marvel.3 If Chandler’s account is accurate, the railroad industry illustrates the importance of hierarchy.)

Operating the large railroad systems was possible only upon solving administrative complexities which greatly exceeded those faced by earlier business enterprises. As discussed below, the hierarchical structure that the railroad managers crafted was broadly consistent with the principles of efficient hierarchical decomposition stated by Simon. Thus, support activities (lower- frequency dynamics) were split off from operations (higher-frequency dynamics), and the linkages within each of those classes of activity were stronger than the linkages between. That organizational innovation, in Chandler’s judgment, paved the way for modem business enterprise.

Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.

Leave a Reply

Your email address will not be published. Required fields are marked *