Stitching it up: accounting and financial control at J & P Coats Ltd, c1890-1960

Introduction

Britain's relative economic decline has been the subject of intensive study and debate for over three decades. Chandler, a major contributor to the debate, has alleged that part of the reason for the UK's inability to match the progress made by American business was Britain's failure to embrace "managerial capitalism','and thus it did not fully realize economies of scale or scope nor did it exploit "organizational capabilities" in its larger organizations, as had happened in the USA (Chandler, 1977,1990).

In Scale and Scope (1990, pp.43-5), building on his analysis in The Visible Hand (1977), Chandler drew attention to the "multidivisional" arrangement for the control of large manufacturing companies, which he regarded as the apex of American managerial achievement. The multidivisional organization consisted in the USA of a head office with centralized advisory staffs exercising coordination at the highest levels only, while the operating divisions, which might be organized on geographical or product lines, were virtually autonomous, with their own finance, sales and marketing, production and research and development capability (Chandler, 1977, 1990, pp.43-4). Chandler contended that this organizational structure made for greater entrepreneurial responsiveness and coordination, as well as achieving cost efficiencies, which the family-owned enterprises typical of Britain's "personal capitalism" did not realize through their failure to adopt it.

Of particular interest for the present study is the role assigned by Chandler to accounting and finance in the success and failure of companies. He stated (Chandler, 1990, p.33) that in "the modern industrial enterprise','as developed in the USA, the task of the small centralized advisory finance staff was to "co-ordinate the flow of funds through the enterprise's many units and to provide a steady flow of information to enable top management to monitor progress and to allocate resources'.' Cost accounting was necessary for ensuring the organization was realizing economies of scale and scope (Chandler, 1990, p.32), which also involved the development of uniform accounting systems.

In an important study of family-owned groups of companies in the UK, Quail (2000) corroborated Chandler's views of British entrepreneurs and their lax management, notably but not exclusively in the financial realm. He held that British firms, especially those of the holding and subsidiary variety, were "too loosely integrated and controlled to realise their potential advantages" (Quail, 2000, p.10). One key aspect of this was that the role of the accountant working in large British industrial firms in the earlier decades of the twentieth century was generally "narrow" (Quail, 2000, p.14), in that his activities were restricted to his numbers. He also tended to be divided from fellow accountants inside organizations according to specialism, namely, costing or financial accounting, which led to inefficiencies and lost opportunities for managerial control.

Another notable example of British financial backwardness was ICI, a rare instance of a British organization with decentralized divisions, whose management was "intractable" in the area of financial control (Quail, 2000, p.20). Pearcy (2001, p.23), notes from Reader (1975) that in the 1930s and 1940s, ICI's various divisions were only interested in their own financial affairs, and not those of the company as a whole. In summary, if these rather broadbrush statements are to be believed, it would appear that there was room for improvement in financial management in British big business, and that certain large, multidivisional US firms, such as Dupont, had better systems for the coordination of management in general and finance in particular. Accounting and financial control systems are, however, an under-researched aspect of large British firms in general, and the "entrepreneurial failure" debate in particular.

This article responds to the call by Dintenfass and others for more detailed case studies of alleged British "entrepreneurial failure" (Dintenfass, 1992, p.71), but its purpose is also to examine in some depth the accounting and financial controls operated by J & P Coats Ltd, the multinational, Scottish-based thread manufacturers, from 1890 to 1960. It will seek to demonstrate that, in addition to being a well-managed British-based concern, Coats' financial controls and associated accounting arrangements were an integral part of its success. Its committee-based management structures, as Kininmonth (2006) has shown, were very different to those pertaining in Chandler's "multidivisional" type of organization, and as the present study will demonstrate, so, too, were its accounting and financial arrangements, which have not been examined in detail before.

The period of Coats' history chosen for analysis straddles its commencement as a public limited liability company in 1890, up to the point of its amalgamation with William Paton, in 1960, and more or less coincides with the period over which Chandler traces the development of the "multidivisional" business form. When J & P Coats was formed in 1890, eight out of 12 of its directors, including the chairman, were members of the Coats family, and they owned one-third of all classes of issued capital, the rest being taken up by the public. Family members remained dominant on the board throughout most of the period studied (Stock Exchange Year Books for 1891, 1918,1931,1950 and 1960). Coats was a classic example of British "family capital."

As will be seen, Coats' financial systems were successful in ensuring the tight control of the firm's financial affairs through many subsidiaries operated at home and abroad from the late-Victorian period and well into the latter half of the twentieth century. The present study will show that, not only was a multidivisional organizational structure not necessary to achieve success at Coats, but neither was a devolved finance organization, coordinated at the centre, as Chandler has implied. In fairness to Chandler, he did observe that Coats was very successful, and although not organized in his favoured "multidivisional" structure, he noted that it did have a small corporate office (Chandler, 1990, p.289). Chandler had no detailed knowledge of Coats' accounting and financial systems, however, and was thus unable to comment on their contribution to its success. The present study will make that contribution clear, and go some way towards reversing the belief that major British firms were badly run from a financial point of view, perhaps encouraging further detailed historical case studies.

Of special interest in this article will be the degree of secrecy that surrounded many of Coats' accounting activities and which enhanced its financial controls. As Roberts et al (2002, p.48) note, secrecy in accounting systems is often correlated with "power distance','that is, the concentration of power at the top of an organization, and in Coats' case, as this study will suggest, this was related to family ownership and management and its desire for confidentiality and control.

The first section of the article will provide a short history of the company, as a setting for the study, looking at its growth, organization and financial success over the period. This will be followed by an examination of its arrangements for the financial consolidation of its results, after which the use of current accounts between holding and subsidiary companies will also be considered. The question of centralized cash control will be dealt with at this stage. The use of the private ledger at Coats will next be considered, after which some aspects of its management accounting arrangements will be discussed. Some final reflection will follow.

Brief History of J & P Coats Ltd

A major analysis of J & P Coats from 1830-83 by Cairncross and Hunter (1987, p. 162) highlighted the significance of family capital and management during the firm's first 50 years as a partnership. The firm's success depended on its decision to specialize in making high quality black and white six-cord thread at Paisley, reliance on the US market, and Sterling currency stability against the dollar. The absence of high tariff barriers also helped entry to this vital market (Cairncross & Hunter, 1987, p.163). The complementary functional skills of the various members of the Coats family were vital, too.

Coats grew an extensive UK business during this early period, and Andrew Coats, one of the founder's sons, was the key figure in penetrating the US market in the 1840s and 1850s (Blair, 1907, p.49). By the mid-1800s, Coats is likely to have accounted for about six per cent of total British exports to the USA (Slaven & Checkland, 1986, p.332). During the same period, the USA became responsible for around 80 per cent of Coats' production and some 90 per cent of annual profits (Cairncross & Hunter, 1987, p.159). It should be noted that, at this time, the selling of Coats thread in the USA was in the charge of agents, while its largest competitor there was none other than local rivals, Clarks of Paisley. Coats in due course resolved this situation through the purchase of the US-based Conant Thread Company in 1890 and merger with Clarks in 1896.

On his employment by Archibald Coats in the late-1870s, Otto Ernst Philippi, a German banker, set out to improve the selling function at home and abroad by removing overseas agents from the selling process, and would later go on to become one of the most influential board members in the company (Slaven & Checkland, 1986, p.329). His most significant innovation, the Central Agency (TCA) was set up in Glasgow in 1889. Its initial function was to control jointly the selling of all the goods made by Brook, Chadwick, Coats and Clarks, UK thread producers who had been rivals, but which, at Philippi's instigation, were now collaborating to fix prices, in what amounted to a cartel. These arrangements did not extend to the US business, where sales and marketing were arranged locally and whose management was entrusted to Hezekiah Conant. Conant's US firm, as already mentioned, had been purchased earlier by Coats.

Coats took up limited liability status in June 1890, with a capital of £5,750,000 (Blah, 1907, p.52). The capital consisted of £2,000,000 in debentures, £2,000,000 in six per cent preference shares, with the remaining £1,750,000 in £10 ordinary shares.1 It is clear that the decision to become a limited company, as well as being part of an increasingly popular movement, was also closely related to the need for invested capital to continue the firm's expansionary strategy (Kininmonth, 2006).

The members of TCA, Coats, Clark, Brook and Chadwick, finally merged formally on 1 July 1896. At this stage the combined firm had 17,000 shareholders, and a capital of some £11,000,000. This was not so much a merger as a complete takeover by Coats. Roy Campbell (1980, p.59) has put the commercial reality succinctly: "After the amalgamation ... a large part of the world's thread industry was controlled from Paisley'.'

Coats' strategic plans (implicit rather than explicit from the board minutes) were to increase sales, through business growth, after which it would expand sales further through growth and takeover both in the UK and overseas. A combination of acquisition and collaborative agreement ensured that selling values and shares of markets were not only maintained, but also increased. Coats expanded horizontally, in the UK, to protect against competition and preserve the firm's investments, and geographically, abroad, to utilize the firm's existing organizational capabilities in order to enter new markets and businesses (Chandler, 1990, p.37).

In addition to the change to limited company status in 1890, that same year J & P Coats took an initiative that was to prove of long-lasting significance: the introduction of committees to oversee key aspects of the business (Kim, 1994, p.5). Quail has plausibly suggested that the preference for management committees in British holding and subsidiary companies (which Coats was) was an attempt by family-owned businesses to remain involved in management through the committees so as not to lose control by delegating to senior, professional managers (Quail, 2000). Another example of a UK company run by committees involving family members was Cadburys (Chandler, 1990, p.246).

After 1890 and notwithstanding world wars and subsequent economic depressions, the huge firm continued with its expansion, including, for instance, new mills in Hungary (1903) and Switzerland (1906). Also, the company carried on buying shares as well as overseas thread manufacturers. On some occasions Coats purchased shares in countries where they were not represented, to expand sales and build on their competencies, while at other times the purchases were made to thwart competition. In addition to these objectives may be added the need to overcome the obstacle of tariff barriers, which were rising steadily in an increasingly protectionist world, but most notably in the USA. Figure 1 shows the firms owned by Coats between 1890 and 1960, which gives some indication of the number of subsidiary companies it controlled and those in which it held shares.

As Figure 1 shows, Coats became a multinational company with manufacturing interests throughout North America, South America, Europe, Russia and India. Acquisitions were often made by acquiring shares over a period of years although many were outright purchases, such as Fiducia S.A., Switzerland, for £19,000, in 1906. New mills were built on existing sites, where necessary and cost-effective. One exception was the Russian mill in St Petersburg, where a new mill was built first, after which Coats took over local competition. It was expropriated after the Revolution, one of a series of adverse political developments that would make Coats generally less successful after 1918 than it had been before.

As was indicated, Coats' strategy was to buy out existing operations in order to deal with competition, increased tariffs or high exporting costs (Kininmonth, 2006). Against this backdrop, sales, apart from the USA, were controlled by TCA in Glasgow, thus creating a formidable organization designed to maximize profits by controlling costs, eliminating as much competition as possible and controlling and scheduling the producing and pricing arrangements for thread across its international chain of subsidiaries. It was so successful that, as Farnie has observed, "it stimulated a wave of emulatory foundations in 1897-1899, especially in the finishing trades" (Farnie, 1979, p.194). Coats went a stage further, by striking up collaborative, oligopolistic arrangements with the English Sewing Cotton Company and the American Thread Company. Coats was "easily the most successful" of the three collaborating combines, as a result of its management, level of innovation and degree of vertical integration (Rose, 2000, p.173). Much of Coats' achievement up to 1917 can be attributed to Philippi, who clearly understood the market power of oligopolistic cooperation, and also the market power of the cartel, which the TCA initially was (Hunter, n.d., p.20).

The company's nominal capital was raised in 1919 from £10,000,000 to £20,250,000, again to finance expansion (Wallace, 2003). The textile industry in general was prospering at this time and the company was regularly achieving annual profits of over £2m by 1910, as Table 1 shows. By the end of the next decade of operations, the average had increased to just over £2.3m. In 1920 and 1921 Coats' profits fell substantially, after which protectionism, increased competition, two world wars and the onset of trade depressions resulted in less good but still acceptable results, compared with those achieved in the pre-1918 period.

As can be seen from Table 1, profits were made continuously across the period. Note that profit is shown after tax, and capital employed is calculated on the basis of issued share capital plus reserves and retained profits. In preparing Table 1, adjustments have been made to a number of the figures in the corresponding table in Kininmonth (2006). The errors in the earlier table were caused by conflicting information in Coats' records.

Reflecting growing international competition and a general decline in the fortunes of the British textile industry, profit to capital employed in the latter part of the period fell below the levels of the pre-First World War era, but always remained substantial. Coats survived while others declined and fell, especially in Lancashire, where diminishing international competitiveness in cotton textiles was linked with inflexible ownership and capital structures and high levels of specialization (Mass & Lazonick, 1990; Higgins & Toms, 1997; Toms, 1998; Toms & Filatotchev, 2004; Toms, 2005). This was no mean achievement, and, although manpower statistics are lacking, it is likely that Coats continuously employed well over 100,000 across its operations over the period.2

Subsequent sections will deal with the general control mechanisms and accounting arrangements at J & P Coats Ltd.

Committee control

As Figure 2 shows, control across the company was achieved by means of an array of committees, designed to ensure that, within its carefully contrived network of market arrangements intended to ensure the minimization of competition, it operated successfully on a day-to-day basis. Prior to 1931, the organization was served by four powerful committees, Finance, General Purposes, Cotton Buying and Yarn Buying, on all of which Coats' family members and senior managers were represented. Analysis (Wallace, 2003)3 shows that the Finance committee dealt with high and medium level decisions, loosely defined as appropriate for senior or middle management, respectively, with a few low level decisions being made from time to time. The Finance committee had no financial limits placed on its decisionmaking, although decisions regarding share acquisitions were subject to Board approval. The General Purposes committee covered a range of matters relating to overseas subsidiaries, selling arrangements, personnel, and even occasionally, accounting details. The financial limits applying to this committee can no longer be found, but few matters needed to be referred to the Board. While some high level decisions were made by this committee, the vast majority were medium level, such as the transfer of senior employees from one department to another. The Yarn and Cotton Buying committees controlled purchased quantities and purchase prices, in accordance with general policy set by the Board and monthly reports were sent by these committees to the Board. Given the Board's overall control, it is not surprising that most decisions made by the buying committees were of middling importance (Wallace, 2003). Committee scrutiny of the firm's international business often resulted in visits from the Glasgow head office to discuss problems.

In 1931 the company re-organized its committee structure so as to reflect the geographical grouping of its subsidiaries and also to "allocate responsibility as clearly as possible to the men who in fact made decisions, and to clear the feet of directors and senior executives of what was [sic] essentially routine matters." (Hunter, n.d., p.30). The committee system was changed radically, with the changes subsequently extending to the accounting system. In reality, the committee system became much more complex, with every type of decision requiring ratification by at least one committee.

The year 1931 saw the appearance of new committees covering merchandising, manufacturing, selling, joint-selling (with TCA), organization, executive and liaison. In spite of the declared intention of the new organization to involve Directors less, the minutes of the new committees are replete with the names of Coats and Clark family members, showing their reluctance to let go the reins of control. As might be expected, one significant feature of the post-1931 era was a reduction in meetings and responsibilities of the General Purposes committee, whose duties were dispersed elsewhere. The committee system was centred on Glasgow, and fed with relevant reports from all of the subsidiaries. The Executive and Liaison committees dealt with items related to the overseas mills, each of which had a counterpart committee feeding information for deliberation by its Glasgow parent committee. In general, the committees created after 1931 made high and medium level decisions for an organization, which by that time had become much larger than it was pre-1918.

The new committee system served the company in more or less unchanged form until its merger with William Paton in 1960, by which point it was felt that it was over-complicated (Wallace, 2003), a view consistent with the impression given by the sheer number of committees and meetings they required. Nevertheless, the committee structure was the means by which this vast company was managed over the period of study, and it is clear that the resultant control from Glasgow was very tight. An insight into the Directors' view of the company's committee system is given at the Annual General Meeting held on 29 November 1906, when the Chairman observed that the company's "advantages were not purely accidental" and that the business was so large that there were numerous "details requiring daily attention, and the necessity of a carefully devised and efficient organisation to cope with them."4

Of interest for the purposes of this article is the fact that most committees had occasional recourse to relevant financial information, but the Finance committee, which was said to be the most important committee in Coats (Wallace, 2003), quite naturally received full accounting reports and other details, as discussed later.

Control through accounting

J & P Coats Ltd's records are mainly kept at the Glasgow University Business Archives. Somewhat frustratingly, and in spite of the fact that the records are extremely voluminous, there are no complete runs of accounting books and many have been lost (for a listing of archives, see Kininmonth, 2006). Nevertheless, the main characteristics of the system are quite clear or can be reliably inferred from what remains, all of which was examined in detail.

Having examined the collection of remaining books, there were four aspects of the accounting system that stood out as significant in terms of indicating the ways in which financial information was used to manage the company: Coats' accounting for investments in its subsidiaries, the control and funding of subsidiaries, the firm's use of private ledgers and its management accounting practices Between and around these aspects of the accounting system, as was noted earlier, there was a culture of secrecy that limited access to certain accounting details and enhanced control.

Investments in subsidiaries

In terms of accounting for investments in subsidiaries, Coats never fully merged the books of Clark, Brook and Chadwick until 1950, and instead kept the books of each firm separately, transferring dividends to Coats, a practice also applied to the other home and overseas subsidiaries. In a sense, this treatment of investments was standard practice for most UK groups of companies before the 1948 Companies Act made full group consolidated accounts compulsory. Such a treatment did, however, confer a number of advantages to holding companies such as Coats. As a firm that continued to involve substantial family shareholdings and family directors throughout the period studied (Kim, 1994), there was an obvious and welcome dimension of secrecy that accrued from these arrangements (Roberts et al, 2002, p.48). Other shareholders had no idea of the overall financial assets employed by the group, and thus no conception of the potential of their indirect investments in subsidiaries, as against the information held by insider, family, Directors. As Edwards (1989, p.226) has put it, "Because subsidiaries were separate legal entities, then results were accounted for by the holding company on the cash basis ... profits were reflected in the holding company's accounts only when a dividend transfer was made ... Often there was no similarity whatsoever ... between the book value and real value of the investment'.' These arrangements came under pressure in 1903, when the querying of Coats' unchanging level of dividends was taken to parliament, after which the resultant government investigation declared the dividends reasonable (Wallace, 2003). Coats' investment accounting therefore helped instil a culture of secrecy related to the firm's finances, such that (substantially family) control from the top was made easier through the possession of exclusive information.

The J & P Coats Ltd Balance sheet for 1931, a point half way through the period studied, is given in Table 2, and will be contrasted with the broad position in 1950, after full group consolidation accounting became necessary.

This shows the investments in subsidiaries, at £16.7m, which accounts for just under half of total assets. If loans and current accounts with subsidiaries are added, these account for 70 per cent of total assets, and show the firm's dependence on its overseas operations for much of its profits. By 1950, in accordance with the 1948 Companies Act, fixed assets less depreciation had increased from £230,440, as shown, to £8.8m, reflecting the consolidation of all majority-owned subsidiaries. This is also reflected in a reduced investment figure in subsidiaries, which dropped from £16.7million, as shown, to £4.4 million. Prior to 1950, therefore, Coats' multinational group is treated, for accounting purposes, as an investment by the Glasgow-based holding company, not as the aggregated assets and liabilities of a consolidated international business empire, the position after 1948, and, indeed, the earlier accounting treatment of the group emphasized the dominance of the Glasgow head office and kept the firm's combined assets a secret until this time, consistent with the industry's known preferences for secrecy during the period, (Farnie, 1993, p.49-50), which may simply have been a facet of 'family capital'.

The control and funding of subsidiaries

The consolidated accounts show large current account balances both due to, and by, Coats' subsidiaries. As was noted earlier, all orders for subsidiaries (USA excepted) were taken and distributed by TCA, to whom they remitted commissions. TCA in turn passed dividends to Coats, its major shareholder. Dividends and cash payments from subsidiaries to Head Office were controlled at the centre, as were bank balances. As far as the USA was concerned, subsidiaries there made their own sales, collected then own revenues, ran their own overdrafts, but as with the other subsidiaries, remitted dividends and surplus cash payments back to Glasgow.5 The dividends and cash payments to be remitted by all operations to Paisley were decided by the main Board, as can be seen from the undernoted extract minute given below relative to one of the USA subsidiaries:

At the end of the year the net balance due to the Bank would be approximately $3,100,000. The Committee agreed that, nevertheless, dividends should be paid in view of the loan position.6

As far as investment is concerned, a 1944 minute of the Finance Committee notes the decision that: "no change should be made ... the capital resources of each Company should, as far as is practicable, continue to be met by credit by the Parent Company".7 Glasgow's control was very substantial, not only in the sense of all major decisions regarding subsidiaries being made by committees, but also in the sense that the liquidity and dividend position of each was dictated by the parent company.

Precise details of Coats' treasury and cash management are lacking, but some analysis of the Bought Ledger over a six-month period reveals the use of five UK banks by the firm. The banks in question were: the Bank of Scotland; the British Linen Bank; the Clydesdale and North of Scotland Bank; Parr's Bank; and, the London City and Midland Bank. Prior to 1920, Scottish banks were unable to clear deposits in foreign currency, using London banks for this purpose, and it seems likely that Parr's Bank, through whom most Bought Ledger payments were made during the period studied, was used by Coats for currency transactions as well as remittances to subsidiaries (Pollard, 1997, p.383). In many ways the Coats system prefigures modern multilateral cash netting systems, which centralize control, reduce the need for multiple cash floats, and result in less net expense (Eun & Resnick, 2004).

Accounting for subsidiaries reveals the same iron, centralizing hand. As early as 1903 entries in the Directors' Minute books indicate that the subsidiary companies were expected, if not forced, to adopt the accounting principles employed at Coats. For instance the Chairman reported in November 1903 that, "the same principles by which we have been guided in regard to depreciation at Ferguslie [Coats' original factory and head office until 1893 when the office was relocated to Glasgow] have been observed by the subsidiary companies."8 It should be made clear, though, that the final accounts of the subsidiaries were prepared at head office, such that those in charge had no knowledge of the profitability of the plants they managed. In the light of this, any accounts that required local filing must have been forwarded from Glasgow. By 1906 it was noted that"... the number of our ledger accounts far exceeds 100,000',' an indicator of the organization's (and the accounting system's) increasing complexity.9

The communications technology of the day, telephones and telegraph, were clearly indispensable in running the organization's accounting and financial arrangements in the above manner. It must nevertheless still have been heavily labour intensive, although the early adoption of accounting machinery in the late 1920s would undoubtedly have helped (Wallace, 2003). An additional problem was the question of the local taxation of subsidiaries, as the extract minute below shows:

Memorandum prepared by MacLay, Murray and Spence on the subject of U.S.A. companies, and the British Income Tax thereby involved ... it was considered that the steps required to procure taxation relief involved so complete a divorce in the management that such a course could not be recommended.10

Although precise details are lacking, this minute gives a very clear indication that control was one of the most important factors in decision making, even if it meant setting aside tax savings in order to keep full control.

Paisley's control extended as far as donations made by subsidiaries at home and abroad. A pre-First World War donation for as little as £5 to the Brough Nursing Home in Paisley had to be authorized by the Finance Committee. Conversely there was also a case where a donation had been made without authority. At the same meeting as above it was reported that the Spool Cotton Company, a US subsidiary, had given $5000 to the San Francisco Relief Fund without permission.11 This was thought to be serious enough to merit being remitted for further investigation to the next Board meeting. It is clear that authority limits, which were delegated to committees rather than individuals, were strictly adhered to, a matter further strengthened after the 1931 reorganization.

Private ledgers

Another element of the Coats accounting system contributing to the joint requirements of secrecy and maximum control was Coats' use of private ledgers in its head office accounting system. As Moss (1983) has noted, private ledgers were a longstanding and common feature of Scottish business accounting systems. The private ledger was initially reserved for transactions relating to family members of firms, covering such matters as salary, drawings, dividends or loans, details of which would be entered into the books by a confidential clerk or accountant. Typically, private ledgers would be lockable. Private ledgers came to contain more and more non-family entries of a sensitive kind, perhaps related to the expansion of business sizes and the need to assign private ledger work to a full-time employee.

Private ledgers were used at Coats over the period of this study, that is, at least up until 1960. These cut across the standard boundaries of an accounting system, as understood today, extending to include sensitive elements from the nominal ledger, purchases and sales ledgers and the cash and bank books.

Coats' Private ledger holds a number of current accounts, including those of subsidiaries, and in some cases these contain details of money that was remitted to Paisley, while in others a record of sales activity may be found. The current accounts within the private ledger occur sporadically throughout the period 1890-1930. It was difficult to determine any real pattern to the appearance of these accounts, but the larger ones tended to remain in the private ledger once they had been entered there.

The following table of remittances (Table 3) from North American subsidiaries is a prime example of key figures recorded in the private ledger, self-evidently confidential and also evidence of the scale of these businesses.

The private ledger, as well as being a phenomenon prevalent among the records of Scottish firms, was in fact a common feature of British family-owned businesses. It can be seen quite plainly from the Coats example that it could act as a mechanism for the maintenance of confidentiality, especially important as family businesses changed over to limited liability companies with shares traded on the stock exchange, as Coats had done in 1890. In these cases the need for confidentiality remained, but was related to the price sensitivity of shares as well as the privacy of shareholder proprietors.

Management accounting

Management accounting can be defined in countless ways. Some narrow definitions restrict it to costing and budgeting, but this does not mirror actual business practice. As Horngren (1978) notes:

financial accounting and management accounting would be better labelled as external accounting and internal accounting, respectively ... Management accounting emphasises the preparation of reports of an organisation for its internal users ... the same basic accounting system compiles the fundamental data for both financial accounting and management accounting. (Horngren, 1978, p.16)

Implied in Horngren's definition is the notion that financial accounting information is often used for internal purposes, thus making it part of the management accounting.

This was certainly the case at Coats. For example, the Finance Committee, which operated from 1890-1914 and then 1930-60, made frequent use of the final accounts of subsidiaries across the whole period (UGD199/1/1/22-UGD199/1/1/27). Similarly, the General Purposes Committee, which operated for the period of the study, made frequent use of subsidiaries' final accounts. Perhaps surprisingly, the Manufacturing Committee, which operated from 1930-60, also made some use of subsidiaries' final accounts, as did the Organisation Committee, which started in 1931. Unfortunately, details of the precise ways in which the information was used have not survived.

Capital expenditure controls were employed by Coats throughout the whole period of the study. These were mostly used to help the parent company, and the Counting House, keep control of such expenditures. Minutes from the various committees show that each one had an expenditure limit, or budget, up to which they could make decisions, but if the proposed expenditure was above the control level then the matter would have to be referred to a "higher" committee for consideration. For example, the Manufacturing Committee could authorize expenditure for manufacturing purposes up to £20,000, but anything costing over £20,000 had to be submitted to the Board for approval.12

As far as the control of cash is concerned, this was present from the very beginning. There are early records within minute books of the Board and the Finance Committee receiving statements of actual versus estimate, in relation to statements of cash receipts and payments.13

While there were de facto budgets for cash and capital expenditure from the earliest days, the introduction of formal budgets, which were generally becoming fashionable from the early 1930s onwards, came in at around this time in Coats.

During the 1931 reorganization it was agreed that a Budget Sub-Committee should meet once a fortnight, and would be responsible for "obtaining, co-ordinating and examining all such budgets and estimates as Finance Committee may decide, on Budget Sub-Committee's recommendation, to be necessary'.'14 Progress seems to have been slow. The Finance Committee, which met on 12 June 1933, was examining the issue of budgets and their introduction, and it was then reported that:

it would not be possible to introduce complete budgets for this company but that it would be feasible to compile budgets for certain items and that, for this purpose, it would be convenient if someone could see a complete budgetary system working in actual practice. Decided to enquire of the Federation of British Industries which companies, if any, in this country have budgetary systems.15

A later entry from the 1933 Finance Committee Minutes shows a report submitted by one of Coats' accountants, indicating that he had been visiting Unilever, Dunlop and ICI in an attempt to learn how their management accounting was operated. The resulting discussion agreed that he would not only be sent back to these companies to observe their budget setting processes, but that he might in fact be sent to the USA to learn how large companies there tackled these matters.16 Although there is no further note of the outcome of these plans there must have been an established policy by 1940, as it was reported that concern had been raised on the state of the budgeting in Mexico, with the recommendation being made that if significant improvement was not seen, the preparation of budgets would be transferred to Glasgow.17

Although no detailed budget statements survive, possibly because of then short-lived usefulness, the successful installation of a sophisticated system is implied by references to direct labour and efficiency variances in the Finance Committee minutes (Wallace, 2003). Direct material is likely to have been one of the largest elements of cost at Coats. Its cost control was handled through Yarn and Cotton Buying Committees, which carried out negotiations with suppliers, passing costs to a Cost Department, which must have been powerful since it was able to authorize negotiated prices. Other examples of the use of budgets relate to the Merchandising Committee, which was responsible for producing overall sales budgets. There is also clear evidence that expenditure budgets were drawn up for advertising and promotional work.18

There was also a Costing and On-Cost Committee at Coats, but minutes only survive for a few months. This Committee reported to the Finance Committee. Costings, in one form or another, were often discussed at committee meetings. For instance, at a 1939 meeting of the Organization Committee the decision was taken to transfer the responsibility for preparing United Thread Mills costings back to the Costing Department.19 Records are lacking, but it is clear that product costing was carried out from the earliest days.

In the mid-1930s Coats decided to introduce Standard Costing. However, committee minutes show that this was not a straightforward task. A meeting of the Manufacturing Committee in February 1936 reported that the introduction in Mexico had been delayed as the local company had found it impossible to apply the production standards in practice.20 There are also notes that indicate that Coats encountered difficulties in finding experts who could.advise on this matter.21 Eventually Coats utilized America expertise to oversee the installation process, although it is not clear whether this was an employee or an external consultant.22

In spite of the absence of records, it is clear that Coats used accounting information of many kinds throughout its committee system, where committee by committee, expenditure limits applied, and that the company was anxious to stay abreast of the latest techniques, implementing them at the earliest opportunity and by so doing, putting itself in the ranks of the most progressive British companies (Boyns, 1998a,b; Fleming et al, 2000). In this way, the company's control systems were further strengthened.

Quail (2000) has observed that the typical UK firm kept costing apart from other functions, and has implied that management accounting in these organizations followed suit, namely, it tended to be restricted to the accounting department. As is apparent, Coats, in contrast, used management accounting information throughout its committee system, thus distinguishing itself from the herd.

Conclusion

Earlier studies reveal that Coats ran a tight committee system that, if it was at times over-bureaucratic, was invariably highly disciplined (Kininmonth, 2006). This article has shown that the system depended on the provision of a range of accounting information. It is recalled by former senior employees from the latter period of the study that the Finance Committee was the most powerful of all the committees at Coats (Wallace, 2003). This article has shown that financial and other accounting information was also freely used throughout the organization's management structure, and indeed, was indispensable to it.

It should come as no surprise that such a long-lived, successful and well run company had an accounting system that helped it achieve its aims. As Horngren (cited in Otley, 1980, p.414) notes, "the design of a (management accounting) system and the design of an organisational structure are really inseparable and interdependent'.'This observation is also borne out by Pearcy (2001, p.2), who stated "that an accounting system takes its shape from the operation which it serves". The study, therefore, provides more empirical evidence for the assertion that organizational shape and philosophy express themselves in accounting systems, particularly in the respect that the strict, centralized management control preferred by Coats was expressed in tight, largely undevolved financial control.

To summarize, the four aspects of Coats' accounting system dealt with were, by their very nature, conducive to control, not only because of their compatibility with the firm's methods and structures of approval, supervision and authorization, but also because they incorporated high levels of secrecy, highly valued in the UK textile industry, probably because of the family orientation of its shareholdings. The treatment of investments in subsidiaries also lent itself to secrecy and an atmosphere of centralized control. Consolidation accounting at Coats reminded all users that the head office was at the centre of the universe and that the subsidiaries were subservient. It also kept secret, for most of the period and from the rest of the world, the true extent of the firm's assets and liabilities. Coats' accounting for the control and funding of subsidiaries self-evidently retained power firmly in the Glasgow headquarters, keeping the branch factories on what today might be termed a "cost centre" basis, where they did not even know what profits they had made, also preserving confidentiality in an atmosphere of control. The old fashioned system of private ledgers, adapted to contain more and more non-family information, almost by accident, also contributed to Glasgow's vice-like grip on the organization, and by their very design was made for secrecy. Of similar import was the firm's system of costing and budgets, crafted for detailed internal control.

In the above ways, the Coats' accounting system bears out the statements by Roberts et al. (2002) that an organizational culture of "power distance" reflects itself in the design and operation of an accounting system.

The main purpose of this article, is, however, to respond to Chandler (1990, p.32), who has observed that in the US multidivisional organization, a centrally coordinated flow of funds was achieved, that there was a steady flow of financial information provided via uniform accounting systems, including costings, enabling the monitoring of progress and the centralized allocation of resources. It will be appreciated from the account that all of these outcomes were achieved within J & P Coats. As was pointed out, head office controlled the levels of dividend to be remitted from subsidiaries, overdraft limits and the remittance of surplus cash through current accounts. This was supported by the regular provision of uniformly prepared final accounts, relating to each subsidiary, to the Finance Committee, backed up by quarterly trial balances from each subsidiary (Wallace, 2003). Later, budgets were extensively used by the Finance Committee, while costings, all prepared in Glasgow from forwarded information (Wallace, 2003), were used in the Finance and Manufacturing Committees. Capital expenditure, as was noted, was sanctioned by each committee, but mainly by the Manufacturing Committee, with referral to the Board of projects above its limit. In summary, as one ex-senior finance manager at Coats put it, 'Finance was a very powerful department' (Wallace, 2003). All this stands in clear contradiction of the generalizations of Quail (2000) regarding UK holding and subsidiary companies, where, he alleges, accounting was a divided and "backwater" discipline remote from top management.

In the multidivisional organization described by Chandler, such outcomes were achieved through the agency of devolved "divisions','which, while they had considerable operational autonomy, had their activities monitored and coordinated by staff departments, including Finance. At Coats, the mechanism was the Committee system, which allowed less discretion to the subsidiaries for family-related control reasons, but which achieved similar aims.

It will be interesting to see if future studies into British firms organized on the holding and subsidiary basis reveal similar management, accounting and financial control characteristics, particularly the successful ones which contradict the findings of writers such as Chandler (1977,1990) or Quail (2000). In the case of J & P Coats Ltd, neither Chandler's multidivisional structure nor fully devolved accounting departments (Chandler, 1990, pp.32-3) were necessary to success.