Theory of the Growth of the Firm

by Edith Penrose

  At all times the productive opportunity of a firm is being shaped by the circumstances mentioned at the beginning of this chapter—internal and external inducements, and internal and external obstacles. In a certain sense each one is decisive, but nothing can be determined by looking at one of them in isolation. The significance of existing resources may not be noticed by particular firms when their management is preoccupied with satisfying the demand for existing products, but it becomes very much in evidence when demand falls off or when entrepreneurs no longer want to put all of the resources available for expansion into existing product lines. But internal inducements to growth are not by themselves profitable opportunities for expansion, nor are external inducements by themselves, as we shall see when we analyze the economics of diversification and examine the interplay of internal and external inducements to expand as well as the effect of both internal and external obstacles. Before turning to this subject, however, we ought to make a slight detour to analyze somewhat more closely the relation between the ‘internal economies of growth’ that are implied in this discussion and the economies of ‘size’.

  VI

  The Economies of Size and the Economies of Growth

  The economies of size. Technological economies. Managerial economies. Economies in operation and expansion. The economies of growth. Disappearing versus enduring economies.

  THE proposition that enterprising firms have a continuous incentive to expand and that there is no limit to their absolute size (other than that imposed by our conception of the nature of an industrial firm) stands in sharp contrast to the notion of an ‘optimum’ size of firm. We have argued that the expansion of firms is largely based on opportunities to use their existing productive resources more efficiently than they are being used. In so far as a firm’s opportunities are not based on the possession of gross monopoly power to exploit suppliers or consumers, expansion that uses resources more efficiently may be an efficient process from the point of view of society as a whole as well as from the point of view of the firm. As a result of the process, firms become larger and larger, and the question arises whether the large firms, because they are larger, are more efficient than smaller firms would be. The biggest firms in the economy today will continue to grow; does this justify the assumption that they will become more efficient as they grow, that each of their product-lines will tend to become cheaper, of better quality, or more adapted to the wants of consumers because they are produced within the administrative framework of a larger organization? Are there economies still to be obtained which relate directly to the increased size of the firms? Growth is a process; size is a state. Our task now is to examine whether there may be economies from the point of view of the efficient utilization of the resources of society which relate to the process but which do not pertain to the ‘state’, to the size that is the byproduct of the process.

  I shall first describe very briefly the nature of the economies of size as they are usually presented. I shall not go into detail, or attempt to evaluate the relative significance of the different kinds of economies or the conditions under which they arise. The economies of size have been extensively and competently described by others, and my purpose in the following summary is merely to provide a background for some important distinctions and relationships between the economies of size and those of growth.91

  The Economies of Size

  Economies of size are present when a larger firm, because of its size alone, can not only produce and sell goods and services more efficiently than smaller firms but also can introduce larger quantities or new products more efficiently. In discussions of the economies of size, so-called ‘technological economies’, derived from producing large amounts of given products in large plants, are commonly distinguished from ‘managerial’ and ‘financial’ economies, derived from improved managerial division of labour and from reductions in unit costs made possible when purchases, sales, and financial transactions can be made on a large scale. Moreover, one can distinguish the economies of size applying to plants from those applying to firms, but the distinction between plant and firm is not coterminous with the distinction between technological and managerial economies.92 The size of plant is not independent of managerial and financial economies; nor is the size of firm independent of technological economies, although technological economies relate most directly to the organization of plants.

  Technological Economies

  A plant, unlike a firm, is necessarily confined to a given geographical location. From the point of view of the present discussion a ‘plant’ or ‘factory’ is characterized chiefly by the fact that the activities contained within it and the products produced by it, are technically related to each other in the process of production.93 Technological economies arise, when, under given conditions, for given products, changes in the amounts and kinds of resources used in production permit a larger output to be produced at lower average cost. Thus, technological economies arise when costs can be reduced through an increase in the specialization of labour; the introduction of automatic machinery, assembly-line techniques, or mechanized internal transport systems; the installation of large units of equipment capable of producing larger quantities at lower unit cost if used to capacity; and other similar technical alterations in the organization of production.

  The effect of any of these technological changes on costs depends not only on the physical productivity of the combination of ‘inputs’, but also on the prices of the factors of production required. Hence the ‘technically optimum’ size of plant is as much a function of prices as of technology, and the concept of technological economies of scale can only mean that with given prices of productive resources a larger scale of output permits changes in the productive techniques or resources used which reduce the average cost of output.94 If the change in the type of input required to produce a larger output of given product in a larger plant calls for an increased use of more expensive resources, costs may not fall as the scale of production is enlarged; while if the same resources were cheaper, the larger size of plant could produce at lower cost. Thus, where capital equipment, or the skilled labour necessary to operate it, is relatively expensive, the introduction of more capital-using mass-production techniques may not reduce costs, and the most efficient size of plant will be smaller than it would have been if capital had been cheap and skilled labour less scarce and expensive. Furthermore, if transportation costs of either the raw materials or product are high, a given plant may be faced with increasing costs on this account as output expands, and the most profitable size will be appropriately limited.

  Moreover, management varies in ability, and the size of plant that can be operated most efficiently by one type of management may not be an efficient size for another type. Small plants run by unspecialized and relatively untrained men may compete successfully with larger plants run by highly skilled, specialized, but more expensive managers, if the lower cost of management offsets the technical disadvantages of the smaller plants.95

  Nevertheless, it often happens that when the scale of output is increased technological considerations are of such overwhelming importance that changes in managerial or transport costs may be of negligible significance. In this case we may neglect them, and plants taking advantage of technological economies will always be able to produce at lower costs than plants that do not, and will therefore tend to dominate the industry. It may be that this is a fairly common situation in some industries, particularly in the largescale mass-production industries, such as automobile assembly, and in industries in which the most efficient units of productive machinery are very large. Thus, where technological economies of size are very great and can be achieved without the aid of exceptionally scarce managerial or other productive factors, and when they more than offset any increases in transport costs as output increases, the size of plant that can survive in an industry will have to be large enough to take advantage of the production methods which make possible the bulk of the economies. Techno
logical economies will affect the minimum size of plant, and therefore of firm in such industries. Furthermore, where a large plant is necessary to achieve low-cost production, the minimum amount of expansion planned by firms will have to be fairly large.

  The size of plant can be measured in different units, with different results. A measure of plant size in terms of employment will understate the effects of increasing mechanization, while a measure in terms of capital equipment will distort comparison of plant sizes between regions where the relative prices of labour and capital are different, or between periods of time in which the ‘utilization’ of plant has changed.96 Within the same industry, the volume of output is the simplest measure, but between industries ‘outputs’ are difficult to compare and some other measure must be adopted. In considering the effect of the size of plant on the size and growth of firms we must, of course, measure plant size in the same units as we do the size of firms and, as we shall see in Chapter IX, if we measure both by the capital equipment employed, there is reason to believe that the amount of expansion a given firm can plan is greater when the most profitable plant requires a large amount of capital.

  Managerial Economies

  Large firms are for the most part multi-plant firms, and economies of multi-plant operation must in general be sought in other sources than technology—in what are known as ‘managerial economies’ which, in the broadest sense, include marketing, financial, and research economies. Managerial economies are held to result when a larger firm can take advantage of an increased division of managerial labour and of the closely allied mechanization of certain administrative processes; make more intensive use of existing managerial resources by the ‘spreading’ of overheads; obtain economies from buying and selling on a larger scale; use reserves more economically; acquire capital on cheaper terms; and support large-scale research.97

  When the scale of production is sufficient to justify a specialized production manager, a sales manager, a financial expert, or a specialist in raw-material buying, for example, each function is performed more efficiently than it would be if all of them were performed by one person. An excellent plant manager may make a poor financial manager indeed. For any given degree of specialization, further economies may often be obtained by the spreading of managerial overhead cost, thus reducing average cost as output increases. This may be possible because existing personnel have been used below capacity or because a given function or service required for one scale of operation need not be increased proportionately for larger outputs. For example, if a firm employs a specialized market-analyst, a research staff, specialized salesmen, etc., it may be able to plan an output appreciably larger than the minimum that would justify their original employment, thereby reducing their cost per unit of product. The same market forecast may be as applicable to an output of 500,000 units as for an output of 50,000 units.

  When a larger output can be produced more cheaply than a smaller output without any change in the basic techniques of production, simply because at the larger scale of operations it is possible to employ specialized managerial talent and so to improve the efficiency of operations that savings are made in materials cost, labour cost, fuel or any other cost, these savings may properly be classed as a managerial economy.

  If some of the reduced cost can be attributed to basic changes in production techniques (e.g., to increased mechanization), managerial economies may in some sense be part of the total economies of the changed scale of operation, but they could never be identified. This is, of course, the more usual case in reality. Managerial and technological reorganization proceed pari passu with increasing size of firm, and an increased division of managerial labour is often necessary to keep costs from rising at higher levels of output; it is an adaptive procedure undertaken to permit economies to be achieved elsewhere. Average cost of output may fall, but average managerial cost may remain constant or even rise as additional managerial personnel are acquired to fulfil more and more specialized functions. Even when it is possible to prevent the emergence of managerial diseconomies by appropriate distribution of the various managerial functions among a larger number of people, it does not necessarily follow that any part of the lower cost of the greater output can be traced to managerial economies.

  Economies in selling come from the increased efficiency resulting when specialized sales personnel can be employed and from producing on a scale sufficiently extensive to use their full selling capacity. There may also be economies in large transactions and in handling bulk quantities; in maintaining an advertising programme sufficiently extensive to ensure not only that all potential consumers are aware of the kind and quality of product available, but also to persuade them that they need the particular brand of product produced; and in maintaining a sales organization that reaches well forward towards the final consumer.

  A great deal of the superior selling strength of large firms as compared with small firms is undoubtedly of a ‘monopolistic’ variety in the economist’s sense. This is especially true of advertising and certain types of sales technique. Yet, given the natural imperfections of the market—the difficulties consumers have in knowing what is available at what prices and where, the physical arrangements necessary to bring seller and buyer together, the education and instruction of the consumer in the use and care of products, and similar problems—selling efforts on a sufficient scale to meet the real needs of consumers are surely an economy for the consumer as well as an economy of scale for the firm. There is little doubt that the smallest firms often connot support sales organizations and programmes that can serve consumers as effectively in such matters as do those of larger firms.

  A firm needs a variety of ‘reserves’ for its operations, whether they be financial reserves, inventory reserves, or labour reserves. With large-scale operations, economies may be obtained in the use of reserves because the proportion of total reserves to total operations can be reduced. This is perhaps most clearly seen in the case of inventories. A small repair shop, for example, may have to carry reserves of parts of all kinds but cannot easily adjust the size of the stock of each part to requirements because the small scale of operations makes it difficult for the firm to predict with any accuracy the demand for any particular part. The larger shop, on the other hand, can, because of the ‘law of large numbers’, predict demand for each kind of part more successfully and can more accurately trim its inventories accordingly.98

  As to financial economies and the ability to support research, the advantages of the large firm are self-evident. The greater security offered to investors, the easier access to capital markets, the greater public knowledge of the firm’s existence and operations, the fact that it is often cheaper within limits to borrow large than small amounts of funds, all combine and interact to place the larger firm in a better financial position than the smaller firm. When research personnel and laboratories are expensive, the large firm can support more of them than can the small firm; when research operations require a large organization, the larger firm can administer them with greater ease than can a smaller firm. We shall have more to say in the next section about the question of treating the ability to support research as an ‘economy of size’.

  Economies in Operations and Economies in Expansion

  Clearly the several economies of size refer to different types of operation: some economies apply to the large-scale production of given products in large plants; some relate to the improved utilization of an administrative organization and have no relevance to any particular products; some are economies in expanding into new fields. But when we speak of ‘economies’ that apply to different sets of existing products, or to new products yet to be created or produced, we have departed a considerable distance from the traditional economies of large-scale production, commonly called ‘increasing returns’.

  The theoretical analysis of the economies of large-scale production has been developed most rigorously within the context of the ‘theory of the firm’ or ‘theory of the indus
try’ and refers, not to the growth of an administrative organization producing many products, but rather to the production of a given product on an increasing scale. In this context, the costs of production as the ‘firm’ grows in size, as well as the costs of production of different ‘firms’ in the same ‘industry’, are always comparable because ‘firm’ and ‘industry’ are defined as producers of a given product. Within a firm the existence of economies of scale can be shown by a ‘reversible cost curve’ that declines as output increases in volume (i.e., the ‘firm’ increases in size) and rises as one traces the curve backwards to smaller outputs.

  In the more descriptive analyses, on the other hand, many of the economies of size apply to firms defined differently and presumably influence the cost of any number of products; further, some of them, notably economies in the ability to support research and certain economies of selling, apply primarily to the development or introduction of new products. The word ‘economies’ implies that in some sense output is ‘cheaper’, and this in turn implies a comparison with some other output of the same or very similar products. Here we have two possibilities: first, the average cost of a larger output produced by a firm may be compared with the cost of a smaller output produced by the same firm. If the cost of the larger output is lower, economies of large-scale production are present and we can unequivocably say that the same firm is more efficient when it is larger than when it is smaller. Second, the average cost of additional output may be compared with the cost of the same output in some other firm. Here the ‘additional output’ may consist of products very different from those the firm has been producing and the appropriate comparison is with the cost of producing the additional output in some other firm. If this cost is cheaper in a larger firm than in a smaller one, where size is the only variable, then economies of size are present, though not necessarily economies of large-scale production, for the lower cost of the additional output may be due only to the size of the firm that undertook it and not to the scale on which the new output itself is produced. In this case we can say that a large firm is more efficient than a smaller firm, but we cannot say that the same firm is more efficient when it is larger than it was when it was smaller.