David M. Porter
PEAT, MARWICK, MITCHELL & CO.

THE WALTHAM SYSTEM AND EARLY AMERICAN TEXTILE COST ACCOUNTING 1813-1848

Abstract: This study of the original accounting records of a pioneering American industrial enterprise narrows by one half the time lag between the earliest known English and American applications of industrial cost accounting. The research in-dicates that the precursors of the costing systems now considered essential tools of management were in use virtually from the beginning of large scale industry in America.

The transition from mercantile to industrial accounting remains one of the greatest expansions and refinements of accounting thought since the formulation of double-entry accounting itself in the 15th century. In the words of accounting historian A. C. Little¬ton, “this transition signified the expansion of bookkeeping (a record) into accounting (a managerial instrument of precision).”1 Yet information as to the emergence of industrial cost accounting in Europe and particularly in America remains sketchy. Most his¬torians have relied on published sources such as management pub¬lications and textbooks in establishing the chronology of cost accounting’s development. Because of the dearth of cost account¬ing material published between 1820 and 1885, it is generally held among accounting historians that little of significance in the field of cost accounting occurred between these dates.2 Recently, how¬ever, several accounting historians confirmed through research on original accounting records that industrial cost accounting was being developed with surprising sophistication in the textile mills of England and New England much earlier than previously supposed

In 1972 H. Thomas Johnson published the results of his study of the accounting records of the Boston based Lyman Mills. Johnson challenged the conclusions of Garner and other leading business historians by demonstrating the application of a “sophisticated cost

For assistance in the preparation of this study, I wish to express my debt to the late William Holmes, to Peat, Marwick, Mitchell & Co. and to Robert Lovett of the Manuscripts Division, Baker Library, Harvard Graduate School of Business Ad¬ministration.

In 1973, the frontier of known factory cost accounting systems was further pushed back nearly 50 years with the publishing of Williard Stone’s study of an English mill. Stone demonstrated that a “sophisticated system of cost finding” was in use at the Charlton Mills in Manchester England as early as 1810.

Johnson’s article describes the Lyman mills as “typical” rather than innovative and suggests further research into earlier account¬ing records. Having read and compared both the Johnson and Stone studies, it was the belief of this author that the 50 year gap between the earliest known English and American industrial costing systems could well be narrowed. Research by William Holmes shows that in at least one other area of accounting, that related to governmental entities, early American accounting practice more than held its own compared with contemporary European practice.5 By examining the original financial records of several of America’s first large-scale corporations, it was hoped that evidence establishing the use of American industrial cost accounting, on par with costing systems at the Charlton Mills, could be found well before 1850.

For several reasons the accounting records of the Boston Manu-facturing Company, (BMC), were chosen as a starting point.6 Founded in 1813, the Boston Manufacturing Company was the first of the large scale “second stage” textile firms in America.7 As such, it furnished the laboratory wherein many early American solutions to several peculiarly industrial problems were pioneered. Its many technical and administrative innovations and its resounding success made the Boston Manufacturing Company the precedent after which the later Lowell mills, and even Johnson’s Lyman mills, were modelled.

The BMC was the first mill either here or in England to integrate loom spinning with weaving, bleaching and dyeing under one roof. The Waltham factory changed the unit of textile production from “small yarn mill to all purpose factory.”8 This integration of the various processes which yielded finished cloth from raw cotton under one roof logically called for an accounting system capable of accurately assigning and accumulating costs throughout the many departments and stages of production.

The history of the BMC and its merchant founders is itself one of the best examples of the transition from mercantile to industrial capitalism in America.

Unlike the technicians and mechanics behind the earlier smaller mills in New England, the men behind the creation of the BMC were financiers and merchants. It was these men who grafted onto the fledgling American textile industry the financial wherewithal and the equally important managerial competence it previously lacked. The initial’ capitalization of the BMC in 1813 was a staggering $300,000. But it is clear that these merchants shared an apprecia¬tion for the value of accounting control commensurate with the scale of their venture. Nathan Appleton, one of the founders and a trained mercantile accountant, voiced this appreciation when he ascribed the most common cause of business failure to “a want of knowledge of the proper principles of bookkeeping.

The host of technological, organizational, financial, and market¬ing innovations introduced by the firm’s Waltham factory and later adopted by other large scale mills is collectively referred to as the “Waltham System.” Though there is much written on the influence of the “Waltham System” in the abovementioned areas, little study of the BMC’s cost accounting records or the degree to which the BMC’s costing systems influenced other mills has been made. The latter consideration is particularly significant given that one fifth of all cotton cloth produced in this country in 1850 was from firms be¬longing to the “Waltham System.”

Previous references to the accounting systems of the BMC credit the firm with bringing to manufacturing industry the conscientious record keeping of the merchant countinghouse. S. Paul Garner described the BMC as “definitely cost conscious” but characterized the firm’s accounting methodology as “exceptionally crude in com¬parison with techniques already illustrated by text writers such as Payen and Cronhelm.”11 This characterization is in keeping with Garners’ view that “during the decades 1820-1880 little can be found which is of interest in the history of cost accounting.”12 Upon closer examination, however, the records of the BMC and affiliated mills are seen to contain an integrated costing system which, aside from its failure to include depreciation expense as we now define it, is quite arguably superior to the costing methodologies of Payen and Cronhelm.

Treasurer with “keeping his accounts in a perspicacious manner.14 This he did. As early as 1817, General and Administrative overhead was allocated to the cost of manufacturing and distinctions were made between current and capitalized expense. Unexpired ex-penses in the form of general supplies, cotton, cloth, and even labor charges to cotton and cloth work in process, were inventoried at cost.15 Costs were accumulated in a systematic fashion and transcost.15¬ferred at actual cost or close approximation from one cost center to the next through the various stages of production.

A brief look at the basic production accounts in 1817 shows us the following:

Cotton costs including interest, freight, insurance, etc. were accumulated as debits in the cotton account. To prepare raw cotton for spinning, the cotton had to be picked, drawn, and carded. Labor costs of this preparation were accumulated by type in the spinning account. Beginning in 1817 these labor charges associated with work in process were inventoried at a standard per pound charge equal to the average per pound cost of preparing cotton during the period. The weaving account accumulated charges and inven¬toried work in process labor in the same manner. The later Lyman mills failed to inventory such work-in-process labor charges.

Most notable of the accounts in 1817 which were closed out to the cloth account is the General factory expense account. Debits to this account included management salaries and miscellaneous supplies. The inclusion of general expense charges in the cloth account is the first indication of the BMC’s efforts to include over¬head as well as direct costs in calculating the cost of cloth manu¬facture.

Thus, by 1817 we see the rudiments of a costing system; a con-scientious effort by the BMC to accumulate costs throughout the production process. Period costs such as watchmen’s salaries were properly expensed in the period incurred and product costs such as spinning labor were carefully inventoried. Although we begin to see use of cost centers, allocation of overhead, and strict dis-tinction between period and product costs as early as 1817, there is no evidence that unit costs by type of finished cloth were calcu¬lated this early.
Between 1816 and 1822 the BMC grew markedly in sales volume, organizational structure, and in the level of requisite accounting sophistication. Beginning in 1818, an Accounts Current ledger was kept. Unlike the earlier journals and general (ledgers, the Accounts Current ledger allowed for a detailed semi-annual summary of each of the current accounts. Cotton on hand at the beginning and end of the period and cotton purchased during the period was listed by bale, pound, and accumulated cost. Spinning and weaving accounts calculated labor applied to WIP using standard costs based on actual cost to the thousandth of a cent per pound. Cloth inventory by type, (bleached and unbleached) was valued at between 141/40 and 260 a yard indicating a fairly precise form of costing as early as 1822.

In 1820 the company built its own bleachery which performed bleaching services both for the company itself and for outside customers. From the start, it was established as its own cost and profit center. Intracompany bleaching was performed at the bleachery’s calculated average per pound cost. The intracompany transfer price, initially rounded to the nearest cent, was calculated to within one ten thousandth of a cent by the 1830’s. Outside bleach¬ing was performed at a profit.

Like the bleachery, the Barn was accounted for as a distinct cost center. The Barn was the cost center responsible for intracompany transportation and teaming. All hauling on behalf of other cost centers, such as the cotton, cloth, bleaching or machine shop ac-counts, was charged to that cost center at a standard rate. It appears that the company attempted to fully allocate barn expense. Every six months the debit and credit entries were totalled and the excess of costs over allocation calculated. This ending debit balance was brought forward to the next period. But ending debit balances remained relatively small during the period reviewed indicating that intercompany allocation of barn costs, at least in aggregate, closely approxi mated cost.

An additional cost component which was included in cost of cloth and cloth inventory by 1822 was that of repairs. Through deprecia¬tion charges, as Payen described them in 1817, were not calculated by the BMC, or any other American company at this time as far as we know, it is arguable that the BMC’s inclusion of Repair charges in the costing of cloth was made with the same intent as deprecia¬tion charges are made today; namely to recognize on a periodic basis the reduced value of fixed assets due to use and aging.

One aspect of modern depreciation not adequately reflected in the cost accounts of the BMC and other early manufacturing com-panies was that of obsolescence. Though rapid technological ad-vances in the textile industry in the first half of the nineteenth century caused many firms to go bankrupt due to obsolete ma-chinery, the industry was too young and the future of technological advancement too uncertain to allow reasonable estimates of obso-lescence. Instead of systematically providing for anticipated obso-lescence as a cost of manufacturing, the BMC and other mill com-panies made periodic appraisals of fixed assets, writing them up or down as they saw fit.21 These appraisals were usually coincident with the issuance of additional stock. Write ups or downs were charged directly through capital stock and therefore did not affect cost accounts.

The next quantum improvement in the quality of surviving costing records of the BMC occurred in 1836. This was the earliest year for which “Semi-Annual Accounts” of the firm are available. The Semi-Annual Accounts are significant because they provide the first explicit evidence that the BMC was calculating unit costs of pro-duction in aggregate and because they give strong implicit evidence to the supposition that unit costs of production by individual type of cloth were being accurately calculated rather than simply guessed at.22
The schedule within the Semi-Annual Accounts entitled “Cost of Materials and Expense” gives totals of each major cost com¬ponent (cotton, labor, general expense, and repairs). These com-ponent costs are then divided by total production in pieces, in yards, and in pounds to arrive at average costs per class of expense and in total. The dollar figures of expenses and the production quantities used in this calculation are the same figures used else¬where in the semi-annual accounts and in the firm’s general ledger. Cost calculations were thus derived from and applied in the ac¬counting records of the BMC in a fully integrated manner

The semi-annual accounts also gave management and ownership detailed schedules as to the source, use, and valuation of all cotton owned during the period by lot. An equally detailed “Memo of Cloth Made’’ gave strict account of the production, disposition, and in-ventory of cloth by type. Ending inventory was valued at cost by type. These semi-annual accounts are very similar in format to those kept by the Lyman mills 20 years later.

The accuracy of the valuation of inventory and production sug¬gests that detailed calculations of the cost of each cloth were made. For the first semi-annual period of 1836 I multiplied the yard¬age of each type of cloth produced by the standard costs, as used to price ending inventory, and found that the standard costs differed from actual by only .6%. (For two other years 1846 and 1854 this same test was performed similarly suggesting that valuation prices by cloth type were based on actual production costs calculated elsewhere.) Beginning in 1836 other indications of the use of account¬ing information for management control appear. The percentage of waste in cotton during production was included as a regular feature of the semi-annual accounts.

Although the first surviving semi-annual accounts for the BMC begin in 1836 there is strong evidence that semi-annual accounts calculating unit costs were prepared earlier than this by the BMC . The first surviving semi-annual report is number #36 indicating that semi-annual reports might have been prepared as early as 1818, the year in which the accounts current were first kept. The accounts current contained all the information necessary for the preparation of the semi-annual accounts. The assignment of inven¬tory valuation costs to each grade of cloth as early as 1822 indicates that detailed unit costs of production were likely calculated out¬side of the semi-annual accounts and other formal accounting records.

Miscellaneous papers in the Baker Library Archives indicate that the BMC actively endeavored to share its technology and manage¬ment systems with affiliated mills by signing formal agreements entitling these affiliates to patent rights, knowledge of machinery processes and “any knowledge, skill, art or information of any sort in any way relating to the business of manufacturing.”23 Account¬ing systems were clearly included in the transfer of technology from the BMC to affiliated mills.

Because existing records of the BMC appear incomplete, more detailed costing calculations were sought (and found) in the con¬current accounting records of two mills directly descended from the BMC. The costing systems and financial statement presentation of the two affiliated firms studied differed only slightly from those of the BMC . The first large mill to copy the Waltham System was the Merrimack Mfg. Co. , a firm started in 1821 by the founders of the BMC and 80% owned by either the BMC or its principle share-holders. The surviving accounting records of the Merrimack Mfg. Co. prior to 1840 are very spotty and contain little evidence of cost¬ing calculations beyond what was derived from the BMC records.
In 1831 another large mill in the BMC tradition, the Lawrence Mfg. Co. , was founded with an initial capitalization of $1,350,000. Surviving records for this firm are very complete and include the earliest known American examples of unit production costing calculations by type of cloth. The structure of the accounting sys¬tem and the format of presentation of this firm’s semi-annual accounts differ little from the BMC .

The one notable departure from the BMC semi-annual account format is the inclusion of a new schedule, the “Memo of Aver-ages.”24 The “Memo of Averages” provides proof that detailed costing control by grade of cloth and by production location was kept in the 1830’s and that the costs calculated were used in valuing inventory and calculating profit. Each of the four mills in operation in 1834 was treated as a separate cost center. Cotton and payroll to each mill were charged at actual. Repairs and General Expense were allocated to the individual mill cost centers based on a pre¬determined overhead percentage rate. The per yard cost of the various grades of cloth produced in each mill was determined to a fraction of a cent. See exhibit 1a and 1b.

From the commencement of full scale production in 1834, the Lawrence Mfg. Co. maintained detailed semi-annual accounts and from the beginning these semi-annual accounts included the “Memo of Averages.” These surviving unit cost calculations are 52 years earlier than the earliest unit cost calculations noted by Johnson in the Lyman Mill records and two years earlier than the first surviving BMC semi-annual accounts.25 There is a strong likelihood that the BMC, the well from which the Lawrence Mills drew their account¬ing systems, was making calculations similar to those noted in the Lawrence “Memo of Averages” but that these calculations were simply not included and therefore not preserved in the semi-annual accounts.

The Lawrence Mills soon outstripped other Waltham related factories in the quality of costing information prepared. Beginning in 1848, the semi-annual accounts of the Lawrence Mills included an “Analysis of Profits” which precisely calculated the profit as well as the cost on each of nine varieties of cloth and for each of five mill cost centers.26 This schedule proves dramatically the trans¬formation of cost accounting into a “managerial instrument of pre¬cision” prior to 1850. See Exhibit 2a and 2b.

In conclusion, the “Waltham System” did more than simply bring to textile accounting the conscientious bookkeeping of the mer-cantile countinghouse. The accounting records of the Boston and Lawrence manufacturing companies illustrate in rich detail the swift transition from mercantile to industrial accounting in the twenty years from 1815 to 1835. The advanced costing systems in use at these mills during this period are the earliest known in America, 30 years earlier than those noted at Johnston’s Lyman Mills. Knowledge of the existence of sophisticated costing within the “Waltham System” narrows by half the lag between the earliest known English and American costing applications. Though little cost accounting literature was produced between 1820 and 1885, it is clear that men af practice were rapidly advancing the applica¬tions of costing in America well before the much heralded “Costing Renaissance” in the final quarter of the 19th century. Virtually from the beginning of large scale industry in America the precursors of the costing systems now considered essential tools of manage¬ment were in use.

FOOTNOTES

1Littleton, p. 359.

2R. H. Parker in his Management Accounting: An Historical Perspective, pp. 138-146, shows the traditional reliance on published works to date the development of cost accounting as well as the dearth of published material on this subject from 1820 to 1880. The first mention of cost accounting in this chronology is Wardlaugh Thompson’s 1777 article in the “Accountants Guide” (York, England) describing process costing. The second mention is dated 1817, the year in which Anselm Payen published his “Essai sur la Tenue des Livres d’un Manufacturier”: the first significant French work on cost accounting. The next significant published work on costing cited by Parker was not until the “Costing Renaissance” nearly 60 years later. This work by Henry Metcalf, an American, entitled “The Cost of Manufac¬turers”, was published in 1885.

3Johnson, p. 474.

4Stone, p. 74

5William Holmes, “Governmental Accounting in Colonial Massachusetts”, unpub-lished paper, 1978

6A major consideration in the selection of the records of the BMC for the pur¬poses of this study was the availability and completeness of those records now housed at the Baker Library at the Harvard Business School. The BMC records be¬gin in 1813 whereas equally complete records of another “second stage” mill do not start until 1832.

All original records of the Boston Manufacturing Company and the Lawrence Manufacturing Company to which reference is made in this study are housed in the Baker Library at the Harvard Graduate School of Business Administration. In the following notes these records are referred to as the “Boston Manufacturing Collection” and the “Lawrence Manufacturing Collection” respectively. The Baker Library manuscript reference number is also given.

7For an excellent history of the origins the American textile industry and the significant role the Boston Manufacturing Company played in that history, see Caroline F. Ware’s The Early New England Cotton Manufacture.
8Ware, p. 63.

9William Bagnall, page 2316 of notes prepared for the second volume of his Textile Industries of the United States (never published). Copies of these notes

The Accounting Historians Journal, Spring, 1980 are held by the Manuscripts Division of the Baker Library of the Harvard Graduate School of Business Administration.

10Taylor, p. 231.

11Garner, The Evolution of Cost Accounting to 1925, 1954, p. 83.

12Garner, “Highlights in the Development of Cost Accounting Thought”, p. 216.

13For a brief synopsis of the cost accounting systems of Payen and Cronhelm see Littleton, pp. 320-324

14Boston Manufacturing Collection, v. 2, Directors’ Records.

15Boston Manufacturing Collection, v. 23, General Ledger.
16Johnson, p. 470.

17Boston Manufacturing Collection, v. 34, Accounts Current

18Boston Manufacturing Collection, v. 34, Accounts Current.

19Boston Manufacturing Collection, v. 34, Accounts Current.

20Johnson noted a similar absence of any systematic depreciation of fixed as¬sets in the Lyman Mill accounts forty years later. Johnson, p. 470.

21Boston Manufacturing Collection, v. 42, Semi-Annual Accounts.

22Though it does not appear that weight was the overriding factor in calculating the cost of each variety of cloth, it does appear that there was a positive correla¬tion between weight and the value added to bleached inventory. All varieties of cloth under three yards to the pound received a standard differential of 10 for bleaching and all over three yards to the pound received a 1/40 differential.

23Boston Manufacturing Collection, v. 187, Misc. Papers.

24Lawrence Manufacturing Collection, v. 41, Semi-Annual Accounts.

25Although Johnson found the “raw data needed to estimate product costs . . . as early as 1875”, actual unit cost calculations at the Lyman Mills were not found until 1886. Johnson, p. 472.

26Lawrence Manufacturing Collection, v. 48, Semi-Annual Accounts.

Bagnall, William, Textile Industries of the United States Volume II unpublished notes.

Boston Manufacturing Company records.

Chandler, Alfred D. Jr., The Visible Hand; The Managerial Revolution in American Business, Harvard University Press, Cambridge, Mass., 1977.

Eldridge, H. J., The Evolution of the Science of Bookkeeping. The Institute of Bookkeepers, London, 1931.

Garner, S. Paul, The Evolution of Cost Accounting to 1975, University of Alabama, 1954.

Garner, S. Paul, “Highlights in the Development of Cost Accounting Thought” in Contemporary Studies in the Evolution of Accounting Thought, Michael Chatfield, Ed., Belmont, California, 1968.

Holmes, William, “Governmental Accounting in Colonial Massachusetts,” unpub-lished paper, 1978.

Johnson, H. Thomas, “Early Cost Accounting for Internal Management Control:” Lyman Mills in the 1830’s, Business History Review, Vol. XLVI, No. 4, p. 474.
Lawrence Manufacturing Company records

Littleton, A. C., Evolution of Cost Accounting to 1900, New York, 1933.

Littleton, A. C. and Yaney, B. S. ed., Studies in the History of Accounting, Irwin, Homewood, IlI., 1956

Mailloux, Kenneth, “The Boston Manufacturing Company” in the Textile History Re-view, Vol. 5 (1960).

Parker, R. H., Management Accounting: An Historical Perspective, London, 1968. Solomons, David ed., Studies in Cost Analysis, Sweet & Maxwell, London, 1968. Stone, Williard E., “An Early English Cotton Mill Cost Accounting System:

Mills, 1810-1899,” Accounting and Business Research, Winter 1973. Szepesi, Eugene, Cost Control for Textile Mills, Bragdon, Lord and Nagle, New York, 1922.

Taylor, George Rogers, The Transportation Revolution, 1815-1860, New York, 1951. Ware, Caroline F., The Early New England Cotton Manufacture, Riverside Press,
1972.