As Latrobe, McCallum, and Thomson so clearly understood, a constant flow of information was essential to the efficient operation of these new large business domains. For the middle and top managers, control through statistics quickly became both a science and an art. This need for accurate information led to the divising of improved methods for collecting, col- lating, and analyzing a wide variety of data generated by the day-to-day operations of the enterprise. Of even more importance it brought a revolution in accounting; more precisely, it contributed substantially to the emergence of accounting out of bookkeeping. The techniques of Italian double-entry bookkeeping generated the data needed, but these data, required in far larger quantities and in more systematic form, were then subjected to types of analysis that were new. In sum, to meet the needs of managing the first modern business enterprise, managers of large American railroads during the 1850s and 1860s invented nearly all of the basic techniques of modern accounting.
Of all the organizational innovators, J. Edgar Thomson and his associates on the Pennsylvania Railroad made the most significant contributions to accounting. Their work and that of other managers received much public attention. Investors, shippers, and railroad directors were as much concerned about the accuracy and value of the new procedures as were the managers themselves. Railroad trade journals, particularly Henry Varnum Poor’s American Railroad Journal, and the new financial journals (first the Banker’s Magazine, and then the Commercial and Financial Chronicle) carried articles, editorials, and letters about the subject. Comparable public discussion of accounting methods had never occurred before in the United States; and it would be another thirty or forty years before similar accounting discussions took place in manufacturing and marketing.
The new accounting practices fell into three categories: financial, capital, and cost accounting. Financial accounting involved the recording, compiling, collating, and auditing of the hundreds of financial transactions carried out daily on the large roads. It also required the synthesizing of these data to provide the information needed for compiling the roads’ balance sheets and for evaluating the company’s financial performance. Where the largest of the textile mills had four or five sets of accounts to process and review, the Pennsylvania Railroad had, by 1857 (the year Thomson reorganized his accounting office), 144 basic sets of accounting records.50 Of these accounts, the passenger department had 33, the freight department 25, motive power 26, maintenance of cars 9, and maintenance of way 22. Eight more were listed under general expenses, while construc- tion and equipment had 21. Moreover, where the textile company’s accounts were compiled only semiannually, those of the Pennsylvania were summarized and tabulated monthly, and were forwarded in printed form by the comptroller to the board of directors by the fifteenth of the following month. The totals of the monthly reports were then consolidated in the road’s annual report.
In the preparation of these reports the accounting office collected, summarized, and printed detailed operating as well as financial data. As early as 1851 the Pennsylvania’s annual report showed for each month the number of passengers entered at each station, as well as the tonnage on local and through freight to Pittsburgh and Philadelphia and from each of the way stations. By 1855 traffic data of over two hundred major products were listed.51 This mass of printed information on expenses and receipts, and on passengers and products moved, remains a magnificent and little- used source for the flows and costs of American transportation at mid- century.
The processing and analyzing of these data required the Pennsylvania and other large railroads to build extensive comptrollers’ departments and to hire full-time internal auditors. By i860 the railroads probably employed more accountants and auditors than the federal or any state government. In any case, after 1850 the railroad was central in the development of the accounting profession in the United States.
In reviewing the balance sheets and other condensed information pro- vided by the new comptrollers’ department, railroad managers, directors, and investors quickly employed these data to evaluate and compare the performance of the different roads. In addition to the balance sheets them- selves, they began in the late 1850s to use the “operating ratio” as a standard way to judge a road’s financial results. Profit and loss were not enough. Earnings had to be related to the volume of business. A better test was the ratio between a road’s operating revenues and its expenditures or, more precisely, the percentage of gross revenue that had been needed to meet operating costs.52 Such ratios had never before been used by American businessmen. They remain today a basic standard for judging the performance of American business enterprise.
In drawing up their balance sheets, the railroads were the first American businesses to pay close and systematic attention to capital accounting. Again the problem was unprecedented. No other type of private business enterprise had ever made such huge investments in capital, plant, and equipment. In discussing capital accounting in the 1850s, railroad man- agers, stockholders, and journalists at first gave the most attention to defining clearly the distinction between the construction or capital account and the operating account.53 On the one hand, by charging operating expenses to construction accounts, promoters and managers could give the appearance of making profits that were not really earned. This they did to improve their chances of raising funds for completing or continuing construction.
On the other hand, by charging construction costs to operating costs, the investors in the road benefited at the expense of its users. Railroad reformers, such as Henry Varnum Poor in the 1850s and Charles Francis Adams, the chairman of the Massachusetts Railroad Commission, in the 1860s and 1870s, repeatedly urged the railroad officials to delineate clearly these two sets of accounts. To see that they were properly differentiated, the reformers proposed that outsiders—either groups of investors or rail- road or legislative commissions—have the opportunity to review a rail- road’s books.
Once a road was completed and the construction account closed, its total amount was recorded on the asset side of the consolidated balance sheet as a capital or property account. The problem then arose as to how to account for depreciation and even obsolescence of the road’s capital assets. For not only were such capital assets of far greater value than those of the factory, but they depreciated at a more rapid rate. The early roads, such as the Boston & Worcester, began by following the textile mill procedures. They put money aside in contingency funds or in their profit and loss or their surplus accounts, in order to have it available for expensive repairs or the purchase of new equipment. Every now and then, usually in good years, the financial officers wrote down the value of their plant and equipment. During the 1850s, however, the managers on the new large roads began to find it easier to consider depreciation as an operating cost and did so by charging repairs and renewals to the operating accounts.
The directors of the Pennsylvania Railroad explained these new con- cepts of renewal accounting in their annual report of 1855. By charging repairs and renewals to operating expenses, the property accounts would continue to reflect the true value of the capital assets. “The practice of the Company in relation to its running equipment is to preserve the number of cars and locomotives charged to construction account, in complete efficiency; thus, if a car or locomotive is destroyed, or has become old and worthless, a new one is substituted in its place, and its cost charged to the expense account.”54 The same was true for rails, cross ties, and bridges.
Such a procedure neatly avoided the complex problem of determining depreciation, but it did not assure the availability of funds for extensive renewal and repairs. The company estimated that the charge for “the annual decay” of the roadbed was $110,000 and the “depreciation” on “running machinery” was $40,000. “If the Company had been declaring dividends from its profits, it would be prudent to carry a portion of the year to a reserved fund.” After balancing receipts with expenditures, the company deducted for taxes, interest, and other expenses; then it set dividends at 6 percent. The balance or surplus went into a “contingent fund,” part of which was used to invest in bonds of connecting roads.55 The funds in these contingency accounts, as those in sinking funds set up for the payment of bonds, were to be placed in “safe” investments. These accounts, however, quickly became mere bookkeeping devices with funds “loaned” out to other accounts of the road itself. After the Civil War, even the Pennsylvania dropped the use of separate contingency accounts, and merely kept the surplus account high enough to meet anticipated demands for repair and renewal of rails and equipment.
By the 1870s this type of renewal accounting had become the standard form of capital accounting used by American railroads. Repair and renewals were charged to operating expenses and not to the capital or the property accounts. These two accounts—one for construction and the other for equipment—were to be altered only when new facilities were added or existing ones dropped. A convention of state railroad commis- sioners meeting in June 1879 to set up uniform accounting methods for American railroads defined the procedure in this manner: “No expenditure shall be charged to the property accounts, except it be for actual increase in construction, equipment, and property, unless it be made on old work in such a way as to clearly increase the value of the property over and above the cost of renewing the original structures, etc. In such cases only the amount of increased cost shall be charged, and the amount allowed on account of old work shall be stated.”56 In the model financial statement proposed by the commissioners (table 1) such additions (or subtractions) were to be listed under a separate heading “Charges and Credits to Property During the Year.” Under that heading was also listed changes in the value of real estate and other property held by the company.
By charging repairs and renewals to operating expenses, the value of the property was theoretically maintained at its original value. The method of renewal accounting meant the profit would continue to be considered, as it always had been in American business, as the difference between operating income and expenses but not as the rate of return on investment on actual capital assets. In fact, the use of renewal accounting made it impossible to know how much capital had been invested in road- bed, plant, and equipment since so much of the cost of capital equipment had been absorbed as operating expense. Such accounting methods thus, of necessity, made the operating ratio, rather than the rate of return, the basic tool for analyzing the financial performance of railroad enterprises. Finally, this method of defining depreciation also meant that American railroad accounting overstated operating costs and understated capital consumption.
Table 1. Form of accounts recommended by the convention of railroad commissioners held at Saratoga Springs, New York, June 10, 1879
The basic innovations in financial and capital accounting appeared in the 1850s in response to specific needs and were perfected in the years after the Civil War. Innovations in a third type of accounting—cost accounting—came more slowly. In making his recommendations for detailed divisional accounts, McCallum had emphasized the need to develop comparative cost data for each of the operating divisions on a large road. “This comparison [of division accounts] will show,” McCallum wrote in 1855, “the officers who conduct their business with the greatest economy, and will indicate, in a manner not to be mistaken, the relative ability and fitness of each for the position he occupies. It will be valuable in pointing out the particulars of excess in the cost of management of one Division with another, by comparison of details; will direct atten- tion to those matters in which sufficient economy is not practiced; and it is believed, will have the effect of exciting an honorable spirit of emulation to excell.”58 Not until the late 186os, however, did cost accounting become a basic tool for railroad management.
The railroad manager who most effectively developed McCallum’s proposals for cost accounting and control was Albert Fink, a civil engineer and bridge builder. Fink, after receiving his training on the Baltimore & Ohio, joined the managerial staff of the Louisville & Nashville, becoming its general superintendent in 1865 and the senior vice president in 1869.59 Fink’s aim was to determine with much more precision the basic measure of unit cost, the ton mile. His first step in obtaining accurate cost of carry- ing one ton for one mile in each of his divisions was to reorder the financial and statistical data compiled by his accounting and transportation depart- ments.00 He consolidated some of the existing accounts and subdivided others. Most important of all, he recategorized existing accounts according to the nature of their costs rather than according to the departments in which the functions were being carried out.
Table 2 shows how Fink reordered his accounts into four fundamental categories. One included those costs which, within limits, did not vary with the volume of traffic. Here he placed twenty-seven accounts involving primarily the maintenance of roadway and buildings and “general superintendence” or overhead. A second category included nine sets of accounts that varied with the volume of freight but not with the length of road or train-miles run. These were largely station expenses “incurred at stations in keeping up an organized force of agents, laborers, etc. for the purposes of receiving and delivering freight, selling tickets, etc.” A third class of thirty-two sets of items, “movement expenses,” varied with the number of trains run. But, as Fink pointed out, since the trains rarely ran fully loaded, the expenditures did not vary precisely with the volume of business. The accounts in these categories were determined for each division on a per-train-mile run basis. In addition to these operating ex- penses Fink had a fourth category, the interest charges that, of course, had no relation to traffic carried or trains run. Interest charges increased only when expanding business called for new construction and an enlarged debt. Table 2 gives the complex formula Fink used to convert these sixty- eight sets of accounts into costs per ton-mile. A comparison of these internal accounts (and the methods devised to use them to ascertain and control costs) with those employed in the textile mills, armories, shipping, and merchant enterprises, emphasized dramatically how much more com- plex railroads were to manage than any other contemporary business enterprise.
Fink stressed how costs varied on the different divisions or “branches,” as they were then called, on the Louisville & Nashville. Movement ex- penses, for example, went from a high of 41.3 percent of total expenses on the main stem to a low of only 17.6 percent on the less-traveled Richmond branch. Station expenses ran from only 4.3 percent of all expenses on the Knoxville branch to 18.1 percent on the main stem, maintenance of road from 9.3 percent on the Glascow branch to 22.5 percent on the Bardstown branch, and the interest account from 26 percent on the main stem to 59.2 percent on the Richmond branch. By developing a time series on the costs of the different divisions and by knowing the division’s physical and economic characteristics, the general superintendent was able to identify with some precision the reasons for the differences in costs. Such historical data and constant reviewing of current financial and operating data permitted him to evaluate performance of different divisions and their operating executives.
In addition, Fink emphasized that such cost analysis was fundamental to ratemaking. The “mere knowledge of average costs per ton mile of all expenditures” was of “no value,” for “no freight is ever transported under the average condition.” If rates are to be based on costs, then “we must classify freight according to conditions affecting the cost of transportation, and ascertain the cost of each class separately.”61 And Fink knew, as did every railroad manager, that costs were only one factor in the complex calculus that determined rates.
Cost per ton-mile rather than earnings, net income, or the operating ratio thus became the criterion by which the railroad managers controlled and judged the work of their subordinates. One reason was that revenues, particularly those from through traffic, could not be easily allocated to separate divisions. Also, many factors completely out of the division superintendent’s control affected the amount of revenues his jurisdiction produced. Thus while financial and capital accounts remained primarily the concern of the financial officers, cost accounting became increasingly the province of the transportation department and came to be used as an operational rather than a financial control.
The volume of financial transactions handled by a large railroad, as well as the volume of traffic and passengers carried, encouraged, indeed forced, railroad managers to pioneer a modern business accounting. This sharp increase in the business activity of the firm thus revolutionized accounting practices. The new methods, devised in the 1850s and perfected in the following years, were quickly adopted by the first large industrial enter- prises when they appeared in the 1880s. They remained the basic account- ing techniques used by American business enterprise until well into the twentieth century. Only in cost accounting did the large industrial enter-prises modify and adjust the methods initially devised by the railroads in the mid-nineteenth century, and this because the operations being costed were so different from those in transportation.
Table 2. Albert Fink: classification of operating expenses and computation of unit costs
In order to make use of this formula it is necessary to know . . . fifty-eight items of expense [above], all of which vary on different roads, and enter into different com- binations with each other. Some of the items of movement expenses (41 to 74) change with the weight of trains, and have to be ascertained in each individual case. The average cost for the year can be made the basis of the estimate. Besides the items shown [above], the following other items enter into the calculation: the average number of tons of freight in train per mile of the round trip of the train, the average length of haul, the number of miles run over the road with freight and passenger-trains per annum, the cost of the road, the rate of interest, and the total number of tons of freight carried during a year over one mile of road. Without these data it is impossible to make a correct estimate of the cost of transportation on railroads.
Source: Chandler Alfred D. Jr. (1977), The Visible Hand: The Managerial Revolution in American Business, Harvard University Press.