Theory of the Growth of the Firm

by Edith Penrose

  183 The T.N.E.C. Monograph (No. 27, op. cit.) in 1941 did not mention this type of acquisition at all and it appeared in the Federal Trade Commission statistics for the first time in 1955. In the report of that year the Commission stated: ‘Such acquisitions during 1940-47 appear to have been rare. However, owing perhaps to recent changes in tax laws, this type of action has become increasingly important, accounting for about 9 per cent of the total number of mergers and acquisitions reported during the 1948-54 period.’ The Commission now includes these in the statistics of acquisition and merger that it reports. Report on Corporate Mergers and Acquisitions (May 1955), op. cit., p. 18.

  184 The impossibility of measuring the ‘extent’ of diversification has been discussed in the preceding chapter, and there are no statistics which classify mergers in a manner relevant to the above analysis. Hence the argument cannot be statistically supported.

  185 The decentralized system of General Motors, for example, ‘contemplates that the responsibility of each individual unit in the system is complete. The objective is to make each separate unit self-contained, complete with all the functions essential to its full development. Its chief executive is charged with the responsibility for its success or failure’. Annual Report, 1942. And of Proctor and Gamble Company, the American Institute of Management noted with respect to its three major products divisions: ‘Each of these three divisions is a separate business with its own operating departments; and each in effect has its own general management with its separate recruitment and training programs’. American Institute of Management, Management Audit, Special Audit No. 148, Vol. VI, No. 14 (Feb. 1957), p. 5.

  186 This is an increase that statistics of concentration not separating the activities of a diversified firm into the several appropriate industries will be too crude to pick up.

  187 Many of the earlier studies of entrepreneurial activity gave sensational emphasis to empire-builders, and were often concerned largely with either attacking or glorifying the leading figures. There may have been some justification for this in earlier times, but certainly today the role of the empire-builder is of marginal importance for the understanding of business growth. There is now a respectable body of serious literature on the nature of entrepreneurs. One of the leading studies is by R. A. Gordon, Business Leadership in the Large Corporation (Washington: Brookings, 1945). See also, Mabel Newcomer, The Big Business Executive (New York: Columbia University Press, 1955) and the works cited by her.

  188 There is an extensive literature dealing with these and similar problems, to which economists as well as sociologists, political scientists, historians, and even philosophers have contributed, and which is far too diverse to cite. One of the leading research groups is the Harvard Research Centre in Entrepreneurial History, the studies of which should be consulted by the student. The subject has received renewed attention since the resurgence of interest in the problems of underdeveloped countries, and in almost every applied analysis in this area some discussion of the availability, attitudes, and abilities of entrepreneurs is required.

  189 And in my view economists should frankly recognize this. Attempts to reduce all behaviour to mathematical utility functions are surely as unenlightening as they are complicated and add little to the solution of our problems. I have no objection to economists putting on other hats to handle such non-economic questions provided they look in the mirror occasionally to see which hat they have on.

  190 William Lever, the founder of Unilever, for example, stated frankly:’... I don’t work at business only for the sake of money.... I work at business because business is life. It enables me to do things.’ Charles Wilson, The History of Unilever (London: Cassell, 1954), p. 187. But, in general, this merely meant that Lever was more ruthless and imaginative in taking chances than a man more preoccupied with pecuniary calculations might have been; it does not mean that he took action that he believed, at the time when he took it, would turn out to be unprofitable.

  191 Such a crossbreed is perhaps found in Unilever, the story of which provides an excellent example of ‘empire-building’, which was none the less largely based on industrial operations.

  192 This point is discussed more fully in the last chapter which deals with concentration and the rate of growth of firms.

  193 The vice-president of a relatively new firm, but one which had recently been growing extremely fast through acquisition, was complaining to me about the difficulty of maintaining a satisfactory administrative organization of such a rapidly growing firm, He asked how the conflict between speed of expansion and adequate organization could be resolved. I told him I thought it could not be resolved except by reducing speed to managerial capacities. ‘But,’ I asked him, ‘what is the hurry?’ He replied that Mr. X, the head of the firm, ‘is trying to build one of the biggest in the field before he dies. He has to be in a hurry, and he intends to acquire more and more firms, expecting me to solve the problems.’ This is, I think, a not unusual attitude to be found among able but impatient businessmen.

  194 None of the old established industrial firms in the United States, large as they are, seems as yet to have approached this ‘final stage’ in its life as an industrial firm. The organization of DuPont, for example, is largely governed by what has been called the ‘DuPont investment formula’: ‘This formula, as is well known, is based upon the idea that management, through its financial officer, acts as a kind of holding company for its several operating divisions. It furnishes the most capital to those divisions which promise the greatest rate of return’. (Alfred N. Watson, ‘Operations Research and Financial Planning’, pp. 5-6 in ‘Techniques and Data for Planning Financial Policy’, Financial Management Series, No. 102, American Management Association, 1953). Nevertheless, the degree of financial control and its pervasive influence on operating policy and practice seems to be such that the DuPont organization as a whole can properly be classed as an industrial firm in our sense. There seemed to have been no such control exercised by DuPont over General Motors in which it had a controlling interest until the Supreme Court required it to dispose of its General Motors stock. The absence of pervasive administrative control is sufficient reason to refuse to treat the two firms as one firm in spite of the fact that DuPont apparently could, through its influence on General Motors, ensure for itself a preferential position as a supplier.

  195 Again the history of Unilever provides an excellent example. Asa Briggs in reviewing Charles Wilson’s book summarized this aspect of the firm’s history as follows: ‘Bold enterprise and wide-ranging imagination made Lever successful in the long run: the continued success of Lever Brothers and ultimately of Unilever demanded more specialized and more orthodox qualities of management and control. ... The qualities which made Lever great could not necessarily keep Unilever strong.’ Lever’s successor, d’Arcy Cooper, took up the task of creating an integrated administrative organization and ‘... this was the beginning of a new phase, in which first Lever Brothers and then Unilever passed into the hands of new men and became increasingly “institutionalized”.’ Asa Briggs, ‘Essays in Bibliography and Criticism: XXXV. Business History’, The Economic History Review, Second Series, Vol. IX, No. 3 (April 1957), pp. 490–1.

  196 Consider, for example, the following description of the early years of General Motors, which was the result of a consolidation promoted in 1908 by William Durant in association with Buick. ‘Buick, Oldsmobile, and Oakland—three of the largest plants and three of the “best names” in the industry—had been brought together in less than three months’. Subsequently ‘in what must be reckoned a terrific outburst of corporate energy’ the company proceeded to acquire other companies: ‘By the end of 1909 General Motors had acquired or substantially controlled more than twenty automobile and accessory companies. Two other prospective purchases narrowly missed fire—Thomas and Ford’. ‘Acute indigestion followed.’ Control was poor, accounting not organized, and the exact amount of liabilities was uncertain. Finally reorganization and consoli
dation were required; the bankers took over and a ‘general plan’ was worked out in 1913 by ‘ultra-conservative men’, which ‘still stands ‘though still subject to revision as circumstances dictate’. Arthur Pound, The Turning Wheel: The Story of General Motors (New York: Doubleday, 1934), pp. 119 ff.

  197 A. S. Dewing in his study of corporate reorganizations found that ‘every crisis studied was the result of financial embarrassment’ but pointed out that this in turn was the result of two other ‘sets of causes’: one, ‘the difficulties attending the administrative management of a large business… The other. the difficulties attending the creation of a business organization sufficiently powerful to dominate an industry in the presence of actual or potential competition’. A. S. Dewing, Corporate Promotions and Reorganizations (Cambridge: Harvard University Press, 1914), p. 558.

  198 There is no need to review here the methods or results of the several empirical investigations into the success of mergers. In addition to the study by Dewing cited above is the same author’s ‘A Statistical Test of the Success of Consolidations’, Quarterly Journal of Economics. XXXVI (Nov., 1921), pp. 84–101, and Shaw Livermore, ‘The Success of Industrial Mergers’ also Quarterly Journal of Economics, Vol. L (Nov. 1935), pp. 68–93. J. Fred Weston has given a useful review and critique of these and other studies in Chapter V of his Role of Mergers in the Growth of Large Firms, op. cit.

  199 See the discussion of some aspects of this question in my ‘Foreign Investment and the Growth of the Firm’, Economic Journal, Vol. LXVI (June 1956). However, E. R. Barlow in his study, Management of Foreign Manufacturing Subsidiaries (Cambridge: Harvard, 1953), which was concerned with United States subsidiaries in Mexico, found that considerable co-ordination and control was exercised by the parent corporation. My impression of United States subsidiaries in Australia was that there was rather less control, but I did not make an extensive study of this particular problem. Distance and the supply of local entrepreneurial and managerial resources would, of course, account for some of the differences.

  200 For some time after the United States Steel Corporation was created, for example, it ran into extensive administrative problems, faced many difficulties, and its growth was slow. In consequence the company was frequently held up as an example of the diseconomies of size. Although it is apparently true that the fear of public opinion exercised a retarding influence on its expansion, there can be little question, it seems to me, that the basic difficulties could be traced to an excessive early rate of growth, that is to say, to a rate of growth in a short period exceeding the rate at which the organizational structure could be efficiently adapted; ‘diseconomies of rapid growth’ would be a more accurate diagnosis than diseconomies of size. The absolute size of the firm would not, under present-day standards, be considered remarkable. To be sure, there have been fundamental innovations in administrative ‘technology’ since 1901, but even to-day an equivalent administrative task could probably not be carried through without set-backs.

  201 The distinction is valid in spite of the problems of identifying the boundaries’ of a firm and in spite of the existence of a grey’ area which obscures on the periphery any clear-cut line between administrative’ and market’ decisions.

  202 Of those economists to whom this difficulty indicates that the analysis is meaningless, I would only ask that instead of throwing the book down at this point they would read on to see if their skill in quantifying the unquantifiable, which has already been so ingeniously demonstrated in other fields, has equal scope here.

  203 See the useful discussion of the accounting problems in the interpretation of the financial statements of firms in Edgar O. Edwards, Funds Statements for Short- and Long-Run Analyses’, The Journal of Business of the University of Chicago, Vol. XXV, No. 3 (July 1952), pp. 156–74. Also the discussion of various measures of the size of firms in Gertrude Schroeder, The Growth of Major Steel Companies, 1900–1950 (Baltimore: Johns Hopkins Press, 1952), pp. 24–35.

  204 With the exception noted, a rate of increase of total services slower than the rate of growth of the firm implies that a constantly increasing proportion of services would be absorbed in operations and eventually there would be no managerial capacity left for expansion.

  205 In a study of administrative overhead in United States manufacturing Seymour Melman found that worker productivity grew faster than administrative overhead (the ratio between administrative and production workers—A/P ratio) between 1899 and 1947. He also found that productivity increases were to a large extent a function of mechanization. It would seem from this that the A/P ratio should fall with increasing productivity, but he found no correlation between changes in the A/P ratio and changes in productivity when he examined 21 American industries for the period 1909-1937, nor when he examined this ratio for manufacturing industries as a whole in three different countries at the same time. He therefore concluded that there was no causal relation between changes in productivity and changes in the A/P ratio and ascribed the changes in the ratio to changes in the nature of the managerial function. That is to say, the effect of increasing mechanization was more than offset by an enlargement of the managerial tasks and by increases in managerial functions. Nevertheless, he did find that the larger firms tended to have lower A/P ratios than smaller firms. This could be explained partly by the considerations discussed above. Seymour Melman, ‘The Rise of Administrative Overhead in the Manufacturing Industries of the United States, 1899-1947’, Oxford Economic Papers (New Series), Vol. III, No. 1 (Feb. 1951). Also his Dynamic Factors in Industrial Productivity (New York: Wiley, 1956), pp. 69 ff.

  206 In one empirical study dealing with the ratio of administrative expense to size of firm, it was found that this ratio was lower in the larger firms than in the smaller firms, but that as increasingly larger firms were examined the rate of fall of the ratio slowed up. It was suggested that there may be ‘an exponential relation between administration expense as a percentage of net sales and the size of firms. Viewed in this manner, the ratio at first declines steeply with increasing size at the lower end of the asset size scale. As one moves to the larger size groups the rate of decrease in the ratio lessens, tapering off asymptotically for the largest size firms. In the absence of detailed studies of management practices by firms along the size range it is not possible to offer a more comprehensive statement of the administration expense-size relationship’. Seymour Melman, Production and Administration Cost in Relation to Size of Firm’, Applied Statistics, Vol. III, No. 1 (March 1954), p. 8.

  207 I am, of course, ignoring the sharp rises in the difficulty of the managerial task that may occur after an expansion has taken place and before the new activities have become thoroughly familiar, the new staff (or the old staff in new jobs) working smoothly, and the various ‘bugs’ ironed out in whatever administrative reorganization has been required. It is perhaps more appropriate to consider such increases in managerial requirements as requirements for expansion rather than as requirements for normal operations.

  It should be pointed out, however, that managerial services in the sense used in this study cannot be accurately measured either by the number or the cost of managerial personnel.

  208 I do not mean to imply, however, that all, or even any firms, do in fact adopt the optimum managerial techniques available at any time. The rate at which the best managerial practices are adopted may well fall far short of the optimum rate, especially when firms are to some extent protected against competitive pressures. This is just another example of the well-known fact that entrepreneurs, through ignorance or inertia, do not always adopt the lowest cost methods of production, even when it would be profitable to do so, unless forced to by competition. See, for example, the dicussion in The Wall Street Journal, Feb. 26, 1957, of the response of firms to the pressures created by the high levels of economic activity in the United States at that time to cut costs, particularly managerial and labour costs.

  209 I want here to make a distinction between those
changes in environmental conditions which can reasonably be related to the fact that a firm has grown bigger and those changes which are unrelated to the increasing size of the firm. Only the first of these changes is dealt with here. The second will be dealt with in our discussion of the relation between the growth of the firm and changes in the economy. In neither case, however, are we concerned with the kind of external change which might be called ‘accidental’ in that it affects a particular firm only; rather we are interested in those changes in external conditions which affect whole groups of firms and are significant because of their effect on the relationships between firms of different sizes.

  210 See the discussion in the next chapter of the relative position of small and large firms.

  211 Referring to the rate of growth of firms, E. A. G. Robinson writes: ‘As to what is the maximum efficient rate, individuals in the same industry appear to hold widely divergent views. A firm, it must be remembered, is an organism in itself. Its parts depend upon each other. Smooth working arises, we have seen, from the most perfect co-ordination of many individuals. A too rapid expansion will introduce so many disharmonious elements that efficiency will be destroyed.... The most efficient rate of growth would appear to be lower in those industries where a high degree of planning and co-ordination is necessary, higher in those where the technical processes of production are fairly straightforward, and the different departments not closely interdependent.’ E. A. G. Robinson, The Structure of Competitive Industry (New York: Harcourt Brace, 1932), p. 56.