So far in this section the data have been examined entirely in terms of the relative contributions of mergers to the assets of the acquiring firms (or of all the firms in a size class) when the acquiring firms were classified into six different size classes. A smiliar analysis has been made of the contribution of mergers to the assets of acquiring firms in the Department of Commerce's list of the 1,000 largest manufacturing concerns ranked according to their total assets as of the end of 1946. This analysis shows that:

(1) The average percentage growth in the assets of the acquiring company per acquisition was regularly and substantially smaller for larger firms than for smaller firms (chart 4 and table 9).—On the average, each acquisition made during these 8 years by any 1 of the 10 largest manufacturing concerns in the country added only 0.2 percent to the assets of the acquiring company at the start of the period. The average contribution per merger increases regularly up to a level of approximately 15 percent for the acquisitions among the firms ranking 751 to 1,000. These general relationships also hold consistently when the 1,000 largest companies are broken down into separate industry groups.

(2) The average percentage growth in the assets of the acquiring company for all its acquisitions taken together was regularly and substantially less for larger than for smaller acquirers (chart 5 and table 9).—All the mergers of the acquiring companies among the 10 largest concerns added only 1.5 percent to their 1939 CHART 4.-AVERAGE PERCENTAGE INCREASE IN ASSETS OF 1,000 LARGEST MANUFACTURING COMPANIES DUE TO MERGERS, PER ACQUISITION, BY RANK GROUPS

assets; the figure for the acquiring companies in the 250-500 rank group was over 40 percent. The increasing relative importance of mergers was quite regular up to this group but showed no further increase for smaller concerns. CHART 5.-AVERAGE PERCENTAGE INCREASE IN ASSETS OF 1,000 LARGEST MANUFACTURING COMPANIES DUE TO MERGERS, ALL ACQUISITIONS, BY RANK GROUPS As % of acquirers' 1939 assets

%

40

This relationship, however, was not found consistently in the separate industry groups.

TABLE 9.-Relative growth of 1,000 largest manufacturing companies from acquisitions during 1940–47, by rank groups

1

Total assets acquired through

merger as a percentage of

growth of

acquiring

growth,

1939-46 2

companies,

(3) Through the top 500 firms, the total assets acquired by firms in each rank group, when expressed as a percentage of the aggregate assets of all the firms in each rank group, were regularly and substantially smaller for larger firms than for smaller firms, but this relationship is reversed for the second 500 firms (table 9).-It should be noted that as of the end of 1946 the five hundredth firm held assets of about 18.4 million dollars.

The findings for the first 500 firms hold whether the aggregate assets of all the firms are measured in 1939, before the wartime expansion took place, or in 1946, after this growth had largely occurred. The aggregate assets acquired by firms ranking 251-500 were about 17 times as large in relation to the groups' 1946 assets as were the assets acquired by the largest 10 concerns. When 1939 assets were used as a base, the assets acquired by the firms ranking 251-500 were nearly 27 times as large as were the acquisitions of the top 10 companies. On this point and the immediately following point, however, the distributions by separate industry groups tend to be erratic.25

(4) The total assets acquired by companies in each rank group form a rapidly increasing fraction of the total growth (a) of the acquiring firms (chart 5) and (b) of all firms in each rank group, through the top 500 manufacturing concerns; their relative importance then falls off substantially.—The assets acquired through merger and acquisition accounted for only 4.2 percent of the total growth between 1939 and 1946 of the 25 largest concerns in the latter year. The corresponding fraction for the next 25 largest firms was 12.1 percent; virtually 25 percent of the growth of the firms ranking two hundred and fifty-first through five hundredth in size in 1946. Even though the percentage drops off after the five hundredth firm, it is still 3 times as large for the companies ranking 7511,000 as for the top 25 companies.

EFFECTS OF MERGERS ON CONCENTRATION OF CORPORATE ASSETS

Changes in three different aspects or types of concentration of corporate assets due to mergers are measured in this study. Data and conclusions have already been presented on the first type-that which reflects, in the most general way, the concentration of previously existing assets in the smaller remaining total number of firms resulting from mergers.26

25 On points 3 and 4 the data for the foods, chemicals, and petroleum and coal groups show a definite, though not entirely regular, tendency to follow the above pattern at least for the 50 largest firms in each industry; textiles reverse the pattern; the ratios for the other industry groups show no marked trend.

26 See pp. 37 ff.

The impact of mergers on the second, and generally more significant, type of concentration has also been measured. This type may be called absolute concentration because it measures the proportion of assets accounted for by some particular absolute number of the largest firms in an industry or industry group. In measuring changes in this type of concentration directly, we can use the data underlying the percentage rates of growth in the assets of the largest firms, which have already been given, to show the increases in the percentage of all assets held by such firms.27 These data show that all the assets acquired in mergers during 1940-47 by the 130 firms with more than 100 million dollars of assets would raise the percentage of all manufacturing and mining assets held by these firms in 1943 by only 0.8 percent, or one-fiftieth of the share (42.1 percent) which they held in 1943. The corresponding increase for the 245 firms with assets of more than 50 million dollars in 1943 would be from 49.9 to 51.2 percentless than one thirty-fifth of its base value. Except for foods and beverages, the relative increases in each of the major industry groups were comparable in magnitude.

28

This absolute measure of concentration leaves out of account the effect of all mergers and acquisitions which are not made by the firms in the selected group of largest firms. If attention is confined to the largest 200 manufacturing concerns, two-thirds of all mergers noted in the financial journals as taking place within the 8 years 1940-47 would be left entirely out of account and, by implication, presumed to have no effect whatsoever on concentration and (by frequent implication) on competitive behavior. Similarily, if attention is confined to concerns with over 100 million dollars in assets, more than threefourths of all the mergers which occurred would be left out of account.

The third type of concentration here measured may be called the relative concentration of corporate assets. In contrast to measures of absolute concentration, relative concentration measures the different percentages of all assets held by various proportions of all companies. It is a more general measure because it summarizes changes in concentration occuring throughout all asset size classes rather than solely within the top size classes.29

We have computed the usual "Gini concentration ratio" 30 for each industrial group studied, for all manufacturing and mining as a whole, and for the largest 500 and 1,000 manufacturing companies. The results for the first two computations (table 10) correspond broadly to those found in our absolute measures of concentration; some increases in relative concentration due to mergers occurred between 1940 and 1947, but, with only one or two exceptions in individual industries, the increases were relatively small. For manufacturing and mining as a whole, the index of concentration on a 1943 base increased from 0.857 to

31

27 The data on concentration reported in this entire section have been adjusted only to allow for changes due to mergers. No allowance whatsoever has been made for other factors affecting corporate growth and industrial concentration such as the large accumulations of retained earnings since 1940, the impact of the accelerated amortization of emergency wartime facilities, and the sales of publicly financed war plants and equipment to private corporations. All these other factors involved substantially large dollar amounts, especially for very large companies. They have, therefore, probably had a greater effect than mergers on over-all changes in the concentration of corporate assets holdings since 1940. The effect of all these factors, however, has been rigorously excluded from our calculations in order to isolate and appraise the effect which merger activity has had on the level of concentration.

28 The exception, of course, reflects the particularly heavy merger activity in the distilling industry.

29 In more detail this measure allows fully for the influence of the decreased number of firms and for the loss of assets in each size class resulting from mergers as well as merely for the assets acquired by firms in each size class; only the latter are covered by measures of absolute concentration.

30 Graphically, this ratio measures the fraction of the total area lying below the 45° line (representing equal distribution) which lies above the locus of cumulated percentages of firms and of total assets on a Lorenz chart. The formula used is given in Dwight B. Yntema, Measures of Inequality * * (Journal of the American Statistical Associa

tion, December 1933, p. 428). The ratios were computed from data grouped in the six size classes shown in the charts, and are consequently net of intraclass inequalities. The significance of this feature was checked with respect to the largest size classes, where its effect is most likely to be significant, by computing the change in relatiove concentration within the 1,000 largest manufacturing firms due to mergers and acquisitions.

31 Changes in concentration were computed by adjusting the Statistics of Income data of the base year, 1943, algebraically for the reduction in number of firms and the net increase or decrease in total assets of firms in each size class due to acquisition by and sales of firms initially in the class. Data for 1943 were used as the base distribution because it seemed appropriate to have much of the wartime growth in assets included in the baseespecially in view of the further fact that about two-thirds of all the mergers studied occurred after 1943. A check calculation from a 1939 distribution suggests that the relative magnitude of changes in concentration was about the same on either base.

0.866, or by 1 percent of its base level. On a 1939 base the index changed from 0.809 to 0.816 or by less than 1 percent.

TABLE 10.—Changes in relative concentration due to mergers, 1940–47

Since particular importance attaches to the distribution of assets among the largest companies, we also computed the index of concentration for the 500 and the 1,000 largest manufacturing companies as of the end of 1946. In both instances relative concentration was reduced as a result of all the acquisitions of these companies over the 8-year period. The index for the 1,000 largest was reduced from 0.639 to 0.632.32 For the 500 largest the decrease was substantially greater, from 0.550 to 0.534, or by about 3 percent as compared to the increases of about 1 percent for all manufacturing and mining as a whole.

The reduced relative concentration of assets within the 500 or 1,000 largest companies directly reflects the relatively greater percentage increase in assets attributable to mergers for smaller firms within the 500 largest companies, as indicated in earlier sections. The reasons for the increase in the relative concentration of assets for all companies, however, are more complicated, but the principal factors can be condensed essentially to the following points:

33

(a) Companies with assets of over 50 million dollars acquired 664 firms of less than this size which held total assets of 1,243 million dollars. No firm with assets of less than 50 million dollars acquired a company with assets of more than 50 million dollars.

(b) In sharp contrast, both the number and assets of firms in each smaller size class which were sold exceeded the number of firms and assets acquired by other firms in the same size class.34 Even though the decreases were not large

32 Here and in the next sentence, the second figure gives the Gini index computed for the assets data as given in the Department of Commerce list. The first figure represents the ratio as recomputed after subtracting all assets acquired by firms in each rank group through merger. The subtraction was made because 1946 was the next to the last year of the 1940-47 period. If growth subsequent to acquisition was at all comparable for the "acquirees" of each rank group, the reductions in concentration will be understated by this procedure.

33 As an analytical point, comparisons of relative rates of growth indicate directly whether relative concentration is increasing as a result of expansion either from retained earnings or outside financing or from acquisitions by the 500 or 1,000 largest companies because in both cases the population of firms remains substantially constant. For two reasons, however, relative rates of growth due to mergers are not rigorously correlated with changes in relative concentration over the entire asset distribution. First, full allowance must be made for differences in the relative number of firms disappearing from different size classes as a result of mergers. Secondly, the total assets of all the firms in each size class which lost their independence through mergers must be offset against the assets of all the firms acquired by firms in each size class.

34 For instance firms with assets of 10 to 50 million dollars acquired 584 other firms with total assets of 1,051 million dollars but companies in other size classes acquired 57 firms in the 10 to 50 million dollar size class with aggregate assets of 1,334 million dollars. Thus, instead of gaining 1.1 billion dollars in assets as a result of merger activity, this size class actually lost 83 million dollars in assets.

The net increase in assets due to mergers of manufacturing and mining acquirers with assets over 50 million dollars was 2.5 percent of their 1943 assets as given in Statistics of Income; the corresponding net decrement in the 10 to 50 million dollar bracket was 0.4 percent in the 5 to 10 million dollar bracket, 2 percent; and under 5 million dollars, 4.2 percent. Except for foods, textiles, and electrical machinery (where the increment for the over 50 million dollar asset size class was 5.6 percent, 55.3 percent, and 17.6 percent, respectively) separate industry groups had about the same experience.

in relation to the total number of firms or to the total assets in these size classes, they were sufficient in conjunction with the net increases noted in the top brackets to increase over-all levels of concentration.

66

In conclusion, the Federal Trade Commission correctly appraised the recent merger movement in describing its outstanding characteristic insofar as its impact on concentration is concerned as * * * the absorption of smaller, independent enterprises by larger concerns." " But it drew an erroneous inference as to the significance of this characteristic. Rather than promoting concentration this very characteristic explains why the recent merger movement has had such slight effect on over-all levels of concentration.

This same characteristic also distinguishes the recent merger movement from its major predecessors which were characterized by combinations in which very large corporations acquired other very large corporations. In fact, the combination of very large companies was the dominant characteristic of the great waves of mergers of from 50 to 75 years ago which gave "American industry its characteristic twentieth century concentration of control." 36 Such combinations also played an important role in the extraordinarily active merger move. ment of the 1920's.". In striking contrast, in the recent merger movement combinations between very large companies have been negligible in number.

ILLINOIS MANUFACTURERS' ASSOCIATION,
Chicago, Ill., 3, October 7, 1949.

Celler-Kefauver proposal, H. R. 2734.

Hon. HERBERT R. O'CONOR,

Chairman, Senate Judiciary Subcommittee,

Senate Office Building, Washington, D. C.

DEAR SENATOR O'CONOR: Pursuant to the understanding with the members of the subcommittee during the course of my appearance before the committee on Friday, September 30, in relation to the above legislation, I am sending you herewith the original copies of letters from 49 Illinois manufacturers in which they set out their views regarding this legislation. I am also sending you a copy of my letter to these Illinois manufacturers. I shall, of course, be deeply grateful if you will arrange to have these enclosures incorporated in the record of the hearings.

I deeply appreciated your courteous consideration during my appearance before the committee.

Cordially yours,

JAMES L. DONNELLY,
Executive Vice President.

ILLINOIS MANUFACTURERS' ASSOCIATION,
Chicago 3, September 23, 1949.

Celler-Kefauver (antimerger) proposal, H. R. 2734.

GENTLEMEN: I am sending you herewith a digest of a statement regarding the above legislation which I plan to make in Washington before a subcommittee of the Senate Judiciary Committee next week.

In event you share the views regarding this measure which I have expressed in the enclosure, I shall be grateful if you will write a letter to the IMA in

35 F. T. C., op. cit., p. 28.

36 Paul T. Homan, article on Trusts, Encyclopedia of the Social Sciences, vol. XV, p. 114. 37 The 72 industrial corporations which appeared among the 200 largest nonfinancial companies in both 1919 and 1927 acquired through merger or purchase 15 other industrial concerns which appeared in the list of the 200 largest nonfinancial companies at some time during the period, 1922-29. The assets of these 15 concerns amounted to 1,928 million dollars at the time of acquisition or merger; this represented 28.6 percent of the entire growth of the 72 corporations and 13.4 percent of their total assets in 1919. Cf. Gardner C Means, The Large Corporation in American Economic Life, The American Economic Review, March 1931, pp. 10-42. Other acquisitions by these companies would doubtless raise the percentage still higher. In contrast, all the acquisitions of the 100 largest manufacturing companies during the recent movement added only 3.2 percent to their 1939 assets and accounted for only 8.3 percent of their growth through 1946.

It should also be noted that mergers reported in the financial manuals were much more numerous in the 1920's than during recent years. The peak rates of 1946 or 1947 were lower than those in each of the 13 years ending with 1931, except for 1922, 1923, and 1924. Less than one-half as many mergers were reported in either 1946 or 1947 as in 1920, 1926, or 1927, and only one-third as many as in 1929.