may be standardized while services rendered in connection with it permit some price differential, or it may be standardized by segregation of 'seconds.' "2. The number and size, distribution of producers: There may be many, all small, or few, all large, or a few large and numerous small, in which case the large producers may or may not be the 'dominant' factor in the market. "3. The general method of price making: The producer may create a supply and then sell it at whatever the market will bring ('supply-governed price'); or he may quote a price and sell what he can at that price until he sees fit to change it ('quoted price'). In the latter case, he may produce to stock or to order. Ör he may sell on sealed bids. "4. The general method of selling: Sale may be by brokers in a produce market, by exclusive agents, through general dealers or direct via traveling salesmen or by mail. "5. The character and means of market information. "6. The geographical distribution of production and consumption: This is important where transportation cost is a material factor. "7. The degree of current control of output: Production may be seasonal, or nonseasonal with a significant lag between the initiation of a change in the rate of production and the resulting change in supply available for sale, or inventory changes may be sufficient to cushion the effect of this lag and so minimize its importance. "8. Variation of cost with varying size of plant or enterprise. "9. Variation of cost with short-run fluctuations of output. "10. Flexibility of productive capacity: Increase may be easy and rapid, or difficult and slow; while rapid decrease may be virtually impossible, owing to fixed capital investment." 23 Notice that the size of the business units which enter the market affects one (i. e., number 2), but only one of these factors which determine the character of competition. The problem of preserving "workable" competition in the real world is far broader than the problem of optimum business size. Thus, business size and competition are both questions of very great importance to the public interest, but they are by no means merely different aspects of the same question. They overlap to some extent, but each goes far beyond the other. VII. BUSINESS SIZE AND ADMINISTERED PRICES In recent years the argument of those who favor a limitation of business size has come to be based on an alleged difference in price behavior between "concentrated" and "unconcentrated" industries. The former are supposed to operate under "administered" prices. The meaning of this phrase is explained as follows by Dr. John D. Clark, a member of the President's Council of Economic Advisers: 16* * * We talk about administered-price industries, because we wish to emphasize the important economic difference between those industries in which a few large firms are dominant and those lines of business in which many firms are true competitors. This is the difference. Where the business is divided up among many small competitors, each business manager realizes that the market price will not be materially affected by changes in his own volume of production. If he enjoys as much as 5 percent of the market, a change of 20 percent in his output will change the market supply only 1 percent. His normal ambition to grow when conditions are favorable leads him to endeavor to market additional production by aggressive salesmanship, without worrying lest market prices will be affected. On the other hand, if he faces a declining market, he knows that no reduction in output which he makes will help to sustain market prices, and instead of reducing production, which means reducing employment, he fights for the market, cuts prices as deep as he can, and maintains output as long as possible. "Where three or four large firms control 70 or 80 percent of the market, each manager knows that market price will be materially affected by his decision about the volume of his production, and he knows that each of the others has the same understanding. Each restrains his impulse to grow when business is booming and keeps his expansion within limits which will protect the market price. When prices weaken, each reduces his production and employment rather than his price, confident that each of the others will do likewise. That may be prudent and it may be good business, as those now participating in the practice warmly assert, but it is not the practice of a competitive business. "Although I have described the policies of managers of administered-price industries in terms of noncollusion, I am not so naive as to believe that in these "Ibid, pp. 243-244. days of business camaraderie and of trade institutes each manager's confidence that each of the others will follow what is for them the orthodox course is not fortified by adequate communication of ideas about the market situation and what it calls for. But the bold days of the ‘Gary dinners' of 40 years ago are long past, and the interchange of ideas is now carried on with finesse which frustrates the prosecutor except in the rare case when some carelessness places a damning memorandum in a file which is investigated. Irrespective of the possibility that ways might be found to discover collusion of this covert character, however, the point is that there need be no collusion. Inherent in the administered-price situation is the failure of the forces of competition to work effectively, and the remedy must be found by attacking the structure of the industry. * * * 24 The remedy Dr. Clark proposes is a drastic one. He wants not to prohibit specific trade practices, but to "attack the structure of industry"; that is, evidently, to limit the size of business enterprises. Nevertheless, it must be admitted that this course is justified if the condition he alleges does actually exist. Especially serious is his charge that, in industries which are characterized by a few large producers rather than many small producers, a preference exists for maintaining prices even at the expense of production and employment. It is easy to see that such a policy is not the road to economic progress and full employment. The pattern of behavior which Dr. Clark describes, if it exists, might conceivably be the cause, or at least an aggravating factor, in business depressions such as that which occurred in the 1930's. The thesis Dr. Clark advances is a very clear and explicit one, which readily lends itself to a check by statistical tests. If his theory is correct, it should be found that in a period of declining prices and national income the industries in which production is concentrated in the hands of a few firms should show a smaller decline in price and a greater decline in physical volume than the unconcentrated industries. The recorded facts show no such tendency. A very careful and scholarly study has been made by an official Government agency in order to determine whether any difference in price and quantity behavior existed as between concentrated and unconcentrated industries in the business cycle between 1929 and 1937. It yielded negative results. The study in question formed a part of the TNEC papers.25 It was prepared by Walter F. Crowder, at that time Chief of the Special Research and Analysis Section, Bureau of Foreign and Domestic Commerce, Department of Commerce. Mr. Crowder required 60 pages to describe his methods and results, so that they can be summarized only very briefly here. His basic material consisted of the record of prices, physical output, and degree of concentration, for 407 separate manufactured products. This record was derived chiefly from the census of manufactures. Two separate periods were studied: the business decline between 1929 and 1933, and the recovery between 1933 and 1937. The degree of concentration in each product was measured by the percentage of the total output of that product in 1937 which was accounted for by the four largest concerns. These "concentration ratios" varied all the way from virtually zero to actually 100 percent. Mr. Crowder compared the concentration ratio, as so measured, with: 1. The percentage decline in the price of that product between 1929 and 1933. 2. The percentage increase in the price of that product between 1933 and 1937. 3. The percentage decline in the physical output of that product between 1929 and 1933. 4. The percentage increase in the physical output of that product between 1933 and 1937. ccording to the administered-price theory, as promulgated by Dr. John D. Clark, a definite and systematic relation should exist in each of these four comparisons. Mr. Crowder's study revealed no such relation in any of the four cases. It is true that, in different products, different patterns of price and quantity behavior were observed in these data between 1929 and 1937. But these differences were unrelated to differences in the "degree of concentration" in the industry. Rather they were found to be determined by certain other economic characteristics of the product. Goods sold to producers, for example, showed a greater decline in volume between 1929 and 1933 than goods sold to consumers, but no significant difference in price behavior was found between these two groups. Durable goods showed a greater quantity decline than nondurable goods but again the price 24 Statement of John D. Clark, before the Subcommittee on the Growth of Monopoly, House of Representatives Committee on the Judiciary, July 13, 1949. 25 TNEC Monograph No. 27, The Structure of Industry; Pt. V, The Concentration of Production in Manufacturing, ch. V. behavior was about the same. An especially significant difference in behavior was found as between goods fabricated from agricultural materials and goods fabricated from minerals. The products manufactured from agricultural materials showed the greater price drop and the smaller quantity drop during the depression. This is apparently a reflection of the different behavior of the raw materials themselves. The following paragraph summarizes, in Dr. Crowder's own words, his conclusions and their economic significance: "Over the past few years there has been considerable interest in the effect of relatively high- or low-price flexibility on production during the recession and recovery phases of the business cycle. One group of writers has maintained that there are two essentially different types of markets in operation—the traditional market in which supply and demand are equated by a flexible price and the administered market in which production and demand are equated at an inflexible administered price. In the traditional free market no individual buyer or seller alone has any significant power over either price or total volume of production for the industry, while in the administered market the number of competing concerns is so small that the individual concern has a significant power to choose within limits between changing its prices and changing its volume of production or sales. This does not necessarily imply monopoly but rather fewness of sellers (oligopoly). For these products with inflexible, administered prices, the full force of the adjustments in the depression period has fallen upon production. At the other end of the scale there are products produced by many sellers where the prices are flexible. For these products with flexible prices the depression adjustments take the form of lower prices and relatively well-maintained output and, of course, employment. "The second group of writers, while not minimizing the importance of price flexibility in cyclical movements, contends that price behavior and pricing policies are in general to be associated with the economic characteristics of the products rather than arbitrary business policies. Since writers in both schools emphasize the difficulties of adjustment to cyclical movements when some products have flexible prices and others have inflexible prices, the heart of the controversy centers in the diverse reasons advanced in explaining the causes of price flexibility. "In the first group, one of the primary causes of inflexibility or rigidity of prices is seen to be the concentration of control over the supply of products. By this control it is held that a few sellers by agreement, 'understanding,' or through some form of leadership' are able to maintain prices in a period of recession by curtailing output. In the second group, the price system is seen as a composite of many different kinds of prices, the relative flexibility or inflexibility of which is conditioned by a multiplicity of factors associated with the economic characteristics of the products themselves as well as the cost conditions under which they are produced and distributed. Obviously, a different public policy is indicated when the causes of rigidity are those advanced by the first group from that which is indicated when the causes are those advanced by the second. "From the material presented in this chapter and over the complete list of products, there appeared to be no strongly marked relation between the conditions of concentration under which products were produced and their quantity and price behavior; high and low concentration and large and small changes in price and quantity appeared together almost as if by chance. And further, different quantity behavior tended to be associated more directly with some particular economic characteristic of the product than with the amplitude of price changes." 28 [Emphasis added.] A more recent study, which reaches essentially the same negative conclusion as that described above, was prepared by Alfred C. Neal.27 Mr. Neal uses much the same material as Dr. Crowder, and much the same methods with one important refinement. Mr. Neal points out that, obviously, to a large extent the different price behavior of different industries results from the different behavior of their costs. He therefore seeks a relationship between concentration and that part of price which is within the administrative control of the producer; that is, the differential between price and unit direct costs. But even with this improvement in statistical methodology, the relationship to be expected under Dr. Clark's thesis is not detectable. The following is Mr. Neal's conclusion: "We find, in summary, that concentration cannot, as alleged, be considered to exercise a significant influence during depression on either price behavior, production 26 Ibid. p. 395. 27 Industrial Concentration and Price Inflexibility by Alfred C. Neal, published by the American Council on Public Affairs in cooperation with Brown University. behavior, or price-production behavior for individual industries taken separately. While it does affect the flexibility of price relative to direct cost, it does so only to a minor extent, and even then, as far as we can judge, this slight influence of concentration does not have direct repercussions on the production behavior of individual industries. To a more important extent, concentration tends to enable industries to maintain unit overhead-plus-profit margins in depressions. But this maintenance of the unit overhead-plus-profit margin in turn seems to have had no direct repercussion on production so far as individual industries are concerned." 11 28 [Emphasis added.] Once again Dr. Clark stands refuted. He has advanced a very neat theory of the effects of business concentration on price and on volume of production. It is not, however, in accord with the facts. TABLE 1.-Size-distribution of producing units in the American economy, 1937 NOTE.-Producing units include Government units, farms, professional enterprises, single proprietorships, partnerships, corporations, nonprofit organizations, etc. Subsidiaries of a corporation are treated as separate units. Source: U. S. Department of Commerce. (Totals for the years, 1900, 1910, and 1920, are rough estimates prepared by the Department of Commerce on the basis of data published by Dun & Bradstreet, Inc. See Survey of Current Business, May 1948, p. 15). 28 Ibid. p. 140. TABLE 3.-Percentage break-down of the business population, 1929 compared with 1947 and 1948 TABLE 4.-Business population, compared with human population and with industrial production Sources: Number of firms: U. S. Department of Commerce. Nonfarm population: U. S. Bureau of Census. (Figures for 1900 estimated by NAM.) Industrial production: Federal Reserve Board (FRB series extended backward by use of Pearsons Index). |