Theory of the Growth of the Firm

by Edith Penrose

  It follows that if concentration increases as an economy grows, the absolute expansion of the given number of larger firms (however selected) must have been greater than the expansion of all other firms. Therefore, if an economy (or industry) is growing at a constant rate, the larger firms must be growing at a faster rate. If concentration remains unchanged or declines, on the other hand, the large firms must be growing at the same or a slower rate relative to the rate of growth of the economy as a whole.

  If we maintain the assumption that the older and larger firms have a competitive advantage over newer and smaller firms, it is clear that concentration will continue to increase so long as the limits on the rate of expansion of the larger firms do not prevent them from taking advantage of the opportunities for expansion in the economy at a greater rate than these opportunities develop. All circumstances that lift the maximum rate of growth of the larger firms relative to the economy (or industry) will increase the rate at which concentration can develop, raise the ‘level’ of concentration, and encourage its persistence. Thus merger, economies of large-scale production, extensive diversification as opportunities for expansion in existing fields become reduced, restriction on entry, and fluctuations in the rate of growth of the economy, are all conducive to the rapid emergence or prolonged persistence of concentration.

  These are, of course, the considerations usually put forward to explain the emergence of concentration, but if our analysis of the factors determining the rate of growth of individual firms is valid, we should expect a tendency for concentration to decline in a steadily growing economy once the larger firms have reached a point where their rate of growth slackens. This decline would show up in an increase in the number of large firms, a reduction in their overall importance and a reshuffling in the relative positions of the firms at the very top, but not necessarily, nor even probably, in any decline in the size of the largest firms, exclusive of their purely financial size.248

  On the other hand, no economy can be expected to grow indefinitely at a steady compound rate. Not only will there be secular fluctuations in the rate of growth, but some tendency for the rate of growth to decline is apparently already evident in many older industrial countries.249 In these circumstances, the course of concentration will depend, first upon the length of the periods of rapid expansion relative to the periods of decline, and second, on the size of the largest firms. It seems highly probable that the rate of growth of many of the large firms in the United States economy today has already started to decline, although the statistical investigations are too few to permit generalization. The depressed period of the 1930s has still such an important effect on statistical measurements of average rates of growth both of the economy and of individual firms that interpretation is hazardous. Consequently it is difficult to say whether or not we should expect concentration to be declining as yet. But it is interesting to note that in spite of all of the developments which have raised the rate of growth of large firms, and in spite of the powerful position of these firms in the economy, there is so little evidence of any substantial increase in concentration that economists can debate with vigour and conviction the question whether it is in fact increasing or declining.

  If there has been a levelling off or an actual decrease in concentration, however, it is not because there are no new fields to conquer; on the contrary, as the economy grows, the total opportunities for the expansion of firms also grow. Concentration slackens because it becomes more difficult for large firms as a group to take advantage of the same percentage of total opportunities as they did when they could sustain a more rapid rate of growth. The unreliability of the statistics and the difficulties of interpreting them are very great, but the evidence is worth looking at.

  Some Shaky Evidence

  Such studies as are available of the course of concentration over time, dangerous as they are to interpret because of the nature of the data, lend support to the hypothesis that concentration tends to develop rapidly to a high point, increase fairly slowly thereafter, and eventually, when existing large firms become very large, to decline. Each of the various studies of the question measures concentration in a different way, some being concerned with inequality, rather than concentration in the sense used here, and some relating concentration to assets, some to output, and some to employment. Consequently, none of them can directly and unequivocally be used to support the present analysis. All we can say is that in general, their results are not inconsistent with the results we should expect from the line of reasoning presented here.

  In the United States the weight of the evidence seems to point to a slow decline in concentration in the last forty to fifty years, with the exception of the period surrounding the 1930s, although a recent Federal Trade Commission report shows a slight increase in concentration between 1935 and 1950 measured in terms of the total value of product accounted for by the 200 largest companies.250 Thus the interpretation of the trend of concentration in the United States is still a subject of great controversy, and this fact alone is sufficient to support the presumption that the fears commonly expressed twenty or thirty years ago that concentration would continue to increase at a rapid rate were unfounded.251

  With respect to Great Britain, an older industrialized country with some very large firms, many of which, however, produce for international markets even more than for the national market, I have found no studies of changes over time of the type of concentration discussed here. The only extensive analysis I know of that is at all relevant to the problem, attempted to measure changes in ‘relative concentration’ (or inequality) from 1885 to 1950 using the log-normal distribution and dealing with firms quoted on the stock exchanges.252 The authors concluded that, in general, inequality among surviving firms has tended to increase in the last fifty years, but that the rise of new industries and new firms, together with a substantial decline in inequality in the period 1939-1950, offset this tendency to such an extent that ‘changes in business concentration in the economy as a whole over the past half century may not be very great.253 The decline in the period 1939-1950 is attributed to the ‘exceptionally profitable record’ of smaller firms, which means, of course, that in this period small firms were able to improve their relative position, a state of affairs which we should expect on the basis of our general analysis. Unfortunately, however, this study tells us little about changes in absolute concentration.

  Comparisons between countries are especially hazardous because of differences in industrial structure, rates of growth, the importance of international trade in total activity, and similar matters, but a recent careful study of concentration in Canada is worth noting. In some ways Canada can be looked upon as an economy similar to the United States, though in an earlier stage of industrial development, where the rate of growth of the economy in relation to the size of firms is still such that concentration may be expected to be increasing. On the average, Canadian firms are not much smaller than are United States firms, but Canadian industries are smaller.254 Thus concentration was found to be higher in Canada than in the United States, that is, in general a smaller number of firms in Canada controlled the same or a greater percentage of the output of industry; on the other hand, inequality was higher in the United States, that is, in general, a higher percentage of the industry was accounted for by an equal or lower percentage of the firms.255 In other words, there are more big firms in the United States than in Canada, but there are also more smaller firms and the big firms control a smaller percentage of total output. As the Canadian economy grows, if it follows the pattern of the United States, more firms will grow into the large category, more new firms will appear, and concentration will cease increasing and probably decrease.256

  Concentration Within Industries

  The above remarks refer primarily to concentration in industry as a whole, specifically in manufacturing. This type of concentration is often held to be a useful measure of concentration of control over industrial assets, but from the point of view of
the market behaviour of firms, it is of less significance than concentration of output in individual industries. There is, of course, a relation between industry concentration and concentration in the economy as a whole, although I think estimates of the latter from a weighted average of the former are of little value, for the significance of concentration in particular industries is fundamentally different from that of concentration in the economy as a whole.257 If ‘concentration is itself highly concentrated’,258 for example, if a few large but concentrated industries account for the bulk of it, surely the significance of a high concentration index is very different from the significance of an index that is high because a large number of small industries are heavily concentrated while no firm or industry is very large.

  Nevertheless, if industrial concentration is very high in industries which are themselves quantitatively very large in the economy, not only will the large firms in these industries be large, but concentration in the economy as a whole will be high; for the most part this is the characteristic of high over-all concentration, since many industries (though by no means all) that tend to be large in any industrialized economy tend also to have characteristics which are conducive to concentration. Unless diversification of the largest firms in the economy is so extensive that few of them bulk large in any single industry, it is to the processes of concentration in individual industries that we must look in order to explain concentration in the economy as a whole.

  The interstices in the economy which provide the opportunities for the growth of smaller firms appear as opportunities to expand the production of specific products or to enter specific industries. Since under our assumptions these are, by definition, the opportunities the large firms ignore, their significance for smaller firms depends upon the type of difficulty which must be overcome in taking advantage of them—upon barriers to entry. The competitive advantages of the older and larger firms in an industry may or may not be of a kind which make it difficult for newcomers to enter the interstices, but the most common type of advantage seems, as we have seen, to be associated either with the amount of capital required or with the possibilities of attaching consumers’ loyalty through branding and advertising. Although there may not be significant technological advantages associated with large multiplant operations, the mere fact that individual plants are fairly large militates against the easy entry of new firms and facilitates the rapid expansion of existing large firms. And branding and advertising give competitive advantages very largely because they impede easy entry of new comers. Thus we should expect concentration to develop most rapidly in industries where this type of advantage is greatest or where the size and growth of the market are such that large firms can expand sufficiently rapidly to leave no interstices.

  The rate at which the large firms can expand in an industry is itself closely related to the rate of growth of demand in the industry in relation to the growth of total capacity. 259 In industries where the advantage of the larger firms is not overwhelming (for example certain branches of the textile and apparel industry) and where entry is fairly easy, the constant competition of newcomers may prevent the larger established firms from expanding rapidly, even though they have a sufficient advantage to maintain themselves and perhaps grow slowly in conditions where the rate of failure among new and small firms is very high. Of all the factors that we have considered which keep down the rate of growth of firms, those related to the necessity of meeting competition are of the greatest significance with respect to their effect both on the managerial services required for existing operations and on those required per dollar of expansion. Thus if newcomers seriously compete with existing firms, even though they are eventually eliminated, the necessity of meeting constant new competition will keep down the rate of growth of the established firms and result in low or only slowly growing concentration. A decline in the profitability of the industry reducing the rate of entry of newcomers may lay the groundwork for the development of a semblance of concentration, perhaps through the consolidation by merger of some of the surviving firms.

  On the other hand, even where there are substantial advantages of large-scale production to be obtained, the rate of expansion of the larger firms may not be sufficient to prevent the rapid entry of smaller high-cost producers if the level and growth of demand is very high. But in this case the expansion of the larger firms need not be inhibited by the smaller producers since they present no serious competition and will be quickly eliminated as the large firms expand, especially when the initially high rate of growth of the industry tends to decline.260

  The growth of an economy is marked by substantial shifts in the relative importance of different industries, which implies that industries grow at different rates. Since rapid growth has often promoted concentration in a new industry, while a decline in the rate of growth promotes concentration in old industries, industrial shifts will offset to some extent the tendency of concentration to decline with continued growth of an economy, and may result in industry concentration moving in a direction opposite to that of concentration in the economy as a whole, where the latter is defined as the percentage of total non-financial assets held by a given number of firms.261 The growth of the leading firms in the newer industries, even though sufficient to produce great concentration in these industries, need not affect the position of the dominant firms in the older industries. Thus, if the older industries, though losing in relative importance, are still the largest industries in the economy, the leading firms in these industries may remain the largest firms in the economy, but they will be controlling a smaller percentage of the economy’s total assets. On the other hand, a growing concentration in the newer industries and no reduction in concentration in the older industries implies an increase in industry concentration.

  ‘Since firms are not confined to particular industries, however, industry concentration cannot effectively be analysed with reference to the factors controlling the rate of growth of the firm as a whole. Rather one must look to competitive relationships and to the conditions of supply and demand for particular products, an examination of which would lead us far from the central problem of this study. However, some of the questions we have discussed in relation to the diversification of firms are relevant for the process of industry concentration.

  Diversification and Industry Concentration

  We have already analysed in some detail the factors determining the nature of diversification, the limitations on the rate of diversification, and the kinds of new fields that the large firms will in general find the more profitable. Large firms can move into industries where entry is difficult for smaller firms and will tend to avoid products where margins over direct cost are too low to support the long-range investment, research and development programmes, advertising, and other types of ‘overhead’ that characterize their operations. For these and similar reasons, many of the profitable opportunities for diversification will bring the large firms into competition with each other; to the extent that oil firms move into chemicals, automobile firms into locomotives and airplane engines, steel firms into shipbuilding, soap producers into cosmetics, etc., industry or product concentration may be reduced without any reduction in the concentration of the economy as a whole.262

  The diversification of large firms into new industries may in the early stages keep down industry concentration, but the course of competition between the large firms will tend to involve a ‘shaking down’ process analogous to that which occurs in any new industry where large numbers of producers enter in the beginning and are weeded out as competition becomes more intense. Where the bulk of the new ‘entrants’ are large firms diversifying into a new and growing industry, the weeding out process may occur more rapidly than it would if the entrants were small, because it is not necessary that prices and profits sink low enough to cause ‘failures’ of whole firms; as we have seen, larger firms finding themselves at a competitive disadvantage in any field can often sell their new ‘business’ without loss, someti
mes at a profit, long before they actually start losing money on it.263

  Concentration and Dominance

  One of the outstanding characteristics of an industrialized economy is the dominant position in the economy as a whole occupied by a relatively small number of large firms. We have enquired how this comes about, relating our enquiry to the processes of growth of individual firms in a growing economy. So far as I am aware no theory of the emergence of dominance or the development of concentration has so far been attempted from this point of view, with the possible exception of what might be called a ‘theory of the headstart’, which states simply that the earlier established and larger firms have a headstart over younger and smaller firms that gives them certain advantages enabling them to maintain their position and to grow at the expense of the younger firms and of firms not yet established.264 Clearly this has much in common with the analysis we have developed, but it has never been examined in any detail nor have the logical consequences been at all worked out. We have gone further and, in the light of the possibilities for growth open to individual firms, have enquired how far the advantages derived from a favourable start will lead to an increase in concentration. We have concluded that in a steadily growing economy, or in an economy where expansion is more prevalent than stagnation, the process of concentration will come to an end and eventually reverse itself.

  In general the empirical evidence available, weak and uncertain as it is, at least does not contradict either the main propositions of our analysis or our conclusions. The analysis of the growth of firms that we have developed is fully applicable only to corporations in an industrial economy, and consequently the historical period to which it can be applied reaches no further back than the last quarter of the 19th century. Although corporations were widely prevalent in certain fields, notably railroads and public utilities, they did not become the dominant form of business enterprise until the 1870s and ‘80s in the United States or the UK; for many other countries, for example, Canada and Australia, it has been only since just before the Second World War that industrial activity, let alone industrial corporations, began to play a dominant role in the economy as a whole. Consequently, we can look at a period not longer than 80 years, and in this period have occurred two major wars, numerous minor ones, and one of the most severe depressions in history, bringing in their train after-effects which persisted longer than the disturbances themselves. Furthermore, there are extremely meagre data available before 1900 and no really reliable data until at least the end of the first quarter of the century. An adequate ‘testing’ of the theory advanced is, in the present state of our knowledge, virtually out of the question, but the approach is, I think, a useful one for the theoretical analysis of the subject and one which has the advantage of relating the process of concentration (or ‘deconcentration’) to the economic characteristics of a growing economy.