¹ “Economic Concentration and Monopoly in Japan,” Journal ofPolitical Economy, LXXII (June 1964), pp. 262–77.

² See Economic and Scientific Section, GHQ, SCAP, Mission and Accomplishments of the Occupation in the Economic and Scientific Fields (Tokyo, Sept 1949), pp. 20–21; Eleanor Martha Hadley, “Concentrated Business Power in Japan” (unpublished Ph.D. dissertation, Radcliffe, 1949), pp. 7, 13–14; and Bisson, T. A., Zaibatsu Dissolution in Japan (Berkeley: Univ. of California Press, 1954). pp. 12–13Google Scholar, and chap, iii, esp. pp. 47–57.

³ Although in her 1949 study (n. 2 above) Hadley had asserted that in the prewar period “most Japanese markets” were “oligopolistic,” in a more recent study (Antitrust in Japan, Princeton: Princeton Univ. Press, 1970, p. 331 and n. 21), she argues that such markets were not generally “concentrated,” since they frequently rested “on oligopoly positions [presumably the positions of individual firms] in the 10–20% range” of industry output. However (setting aside the point that reference to such individual market occupancies, especially without specification of the ranks of the firms involved, does not suffice to indicate whether the markets were concentrated or not), the term “oligopoly” itself typically suggests at least moderately high concentration.

⁴ The growth in concentration began to appear at the top 10 firm level somewhat earlier. A Fair Trade Commission [hereafter FTC] study of 141 “major” (and it may be assumed, predominantly manufacturing) industries for the period 1960–1966 indicates that concentration at the top 10 firm level rose slowly throughout this period: Report of the Activities of the Fair Trade Commission, submitted to the Restrictive Business Practices Committee of the Organization for Economic Cooperation and Development (Tokyo, 1969), p. 35. This is particularly relevant here, for to a substantial degree the FTC sample (on which the results in Table II above are also based) consists of oligopolistic industries, Firms ranking below the top 10 (in several cases 10 or fewer firms occupy the entire market) would often be small relative to the leading firms. They would thus find the obstacles to expansion especially formidable as the leading firms consolidated their positions. The size of the lower-ranking firms as compared with the industry leaders also helps explain the virtual absence of any drop in the top-10 concentration ratios during the entire period covered in Table II.

⁵ My analysis here is very similar to the analysis of Kozo Yamamura; see his Economic Policy in Post-war Japan (Berkeley: Univ. of California Press, 1967), pp. 97–107. Yamamura stresses the disposition of the government to permit or encourage merger. This is a relevant consideration, although—generally—if business itself were not strongly inclined to merge, it may be presumed that the massive growth of merger would not have occurred.

⁶ Material on annual merger rates given me by the FTC shows that, among mergers taking place between 1965 and 1970, 75% of the resulting firms had a total capital of less than 100 million yen (about $300,000).

⁷ Firms in the “larger-size class” would be those with 100 million yen of paid-in-capital or more (which conforms to the definition of the “large” firm in Japan). In addition to the small firms acquired by these firms (28% of their acquisitions), 27% of their acquisitions were firms with a paid-in-capital of 100–500 million yen (about $300,000-$1,500,000). The largest-size class for firms (either acquired or acquiring) listed in the study from which these figures were compiled is $300,000,000 and over. The study used was a table of mergers, prepared by the FTC; it appeared in Nihon no Kigyō-Shūchū: Daikigyō ni Yoru Shihon-Shūchō Kabushiki-Shoyō Gappei nojittai (Concentration Among Japanese Firms: The Facts Con-cerningi Concentration, Stockholding and Merger), Tokyo, 1971, pp. 200–12.

⁸ FTC, Nenjihōkoku (Annual Report) 1972, p. 113. “Horizontal” mergers being those between firms producing competing products, “vertical” mergers involving firms operating at different stages of the process required to produce a single product, and “conglomerate” mergers encompassing the remainder, it is thus only the “horizontal” merger that immediately and directly reduces competition in the sale of a product. Other mergers may do so, but only indirectly, in certain cases.

⁹ The reference here is to the Big Three descendants of the prewar zaibatsu—Mitsui, Mitsubishi, and Sumitomo, which are considered below—and to several other smaller and considerably weaker business groupings (e.g., Dai-Ichi Kangin, Sanwa, Nihon Kōgyo Ginkō) commonly classified in the same category.

¹⁰ Rotwein (n. 1 above), pp. 266, 272. Though I called attention to basic differences between the prewar and postwar Big Three, in that study I termed the postwar groups “zaibatsu” for purposes of indicating their special prewar ancestry and lineage. I was under the impression that many Japanese used this term in the same way. This, however, was an error. It has led one observer to assert that I shared the common view that, in the postwar period, essentially the same older zaibatsu organizations were being revived. See Hadley (n. 3 above), p. 341.

¹¹ Rotwein (n. 1 above), pp. 269–72.

¹² See Hadley (n. 3 above), ch. vii.

¹³ For example, Hadley (n. 3 above), pp. 250–53; and Yamamura (n. 5 above), pp. 126–28.

¹⁴ The 1957 list is taken from a Marxist study entitled Genda't Nippon no Dokusen Sbibon (Monopoly Capital in Present-Day Japan) (Tokyo: Nippon Hyoron Shinsha, 1958), pp. 132–43. The 1970 list is drawn from Keiretsu no Kenkyū (Study of Business Groupings) [hereafter KK] by the Economic Research Association (Tokyo: Keizai Chōsa Kyōkai, 1971), pp. 9–14.

¹⁵ The figures on total non-financial corporate assets were secured from the Japan Statistical Year-book published by the Bureau of Statistics of the Office of the Prime Minister—p. 273 of the 1958 ed. for the 1957 data, and p. 316 of the 1971 ed. for the 1970 data. For 1957, the total assets of the 120 affiliates of the Big Three were secured from the 1959 edition of Kaisha Nenkan (Company Yearbook), published by Nihon Keizai Shinbunsha; for 1970, they are from KK 1971 ed., pp. 9–14.

¹⁶ Mitsui increased its proportion from 3.9% to 4.0%, and Sumitomo (the largest gainer) from 3.0% to 3.8%. Mitsubishi's proportion fell from 4.4% to 4.0%.

¹⁷ The proportion represented by Mitsuis Class I affiliates increased from 1.5% to 2.1%, while that of their Class II affiliates dropped from 2.4% to 1.9%. The equivalent percentage changes for Sumitomo were 1.1 to 1.5 and 1.9 to 2.3; for Mitsubishi, 3.9 to 3.5 and 52 to 49.

¹⁸ The 1958 data are drawn from market concentration tables given in the FTC study Shuyō Sangyōni okeru Seisanshōchōdo (The Extent of Concentration of Production in Major Industries), 1960. The 1968 data are taken from market concentration tables in a special study in Tōyō Keizai Tokei Geppō (Oriental Economist Monthly Statistics) [hereafter OE], June 1970, pp. 6–45. The former study covers 124 markets, the latter 188. It would have been desirable to use FTC data for both periods, but after 1958 the FTC, for reasons of confidentiality, ceased identifying the firms covered in its concentration studies. However, I am informed that the OE study is reliable; although for the same year there appear to be moderate differences between its concentration ratios and those of FTC, the comparison reveals no systematic bias. The FTC study ranks all firms among the top 10 in the industry (except, of course, when fewer firms occupy the whole market); the OE does likewise only in the majority of cases. Wherever necessary, therefore, I have reduced the FTC list of leading firms in individual markets to make it comparable with the coverage by the OE.

¹⁹ Although it may be presumed that coordinated activity among the presend-day Big Three is far less extensive than it was among their prewar zaibatsu ancestors, the studies of Hadley and Yamamura do not show that the types of cooperative activities mentioned above no longer exist. While intercorporate stockholding within each of these groups varies widely and is often very small, Hadley's figures (n. 3 above, pp. 213, 217, 219) indicate that in 1966 the average cross-ownership of stock among the “core companies” (comparable to what we have designated as Class 1 affiliates) of Mitsui, Mitsubishi, and Sumitomo was 15%, 20%, and 33–34% respectively. These percentages are considerably higher than the proportion of total corporate assets represented by these firms (see n. 14 above), indicating that here some preference for the acquisition of each others' stock still persists. With respect to the availability of credit, other figures given by Hadley (ibid., pp. 223, 225) indicate that there is some measure of attachment between the “core companies” in each of these groups and the grou p bank. These figures show that on the average such core affiliates among Mitsui, Mitsubishi, and Sumitomo secure respectively 31%, 39%, and 60% of their loans from their group bank. But the proportion of total bank business accounted for by any of the largest private banks does not exceed 7%. See FTC, Nihon no Sangyō Shūchū (Concentration of Industries in Japan) (Tokyo, 1969), p. 291. Also, if one considers the borrowings of the largest of the “core companies,” Hadley's figures would show a noticeable difference between their loans from the group financial institutions (bank and non-bank) and those from other and especially non-governmental financial institutions; see Hadley (n. 3 above), pp. 226–32. As to the markets affiliates of a given group afford for each others' products, my earlier study indicated that the affiliates displayed a distinct preference for the products of others in the same group (although of course they also made purchases from outside the group as well); see Rotwein (n. 1 above), p. 267. (The unavailability of data precluded a similar study for a more recent period.) Speaking more generally, the force of custom and tradition appears to remain of more than token importance between important affiliates of these groups, and the weekly meetings (which still continue) of presidents of the principal companies within each group may well afford opportunities to consider methods of advancing their common interests.

²⁰ As indicated in Table III, while the average market occupancy of Mitsubishi Class I affiliates grew by 11%, that of the Class II affiliates grew by 30%. However, only four of the Mitsubishi Classn affiliates appear among the leaders in these markets, an extremely small sample. Moreover, the increase among these Class 11 affiliates was due wholly to the very substantial growth of one firm. The market occupancies of the remaining three actually declined.

²¹ Both are examined in detail by Hadley (n. 3 above, pp. 269–90, 301–15).

²² The list of kinyū keiretsu affiliates appears on pp. 79–84 of KK, 1970 ed.

²³ See Hadley (n. 3 above), pp. 291–301, 342, 346. It should be emphasized that the dominant company in these complexes is not a pure holding company, since the latter is prohibited by the anti-monopoly law.

²⁴ See Report of the Activities of the Fair Trade Commission, submitted to OECD Restrictive Business Practices Committee (Tokyo, 1971), p. 49.

²⁵ Of the 28 holding company complexes covered in this analysis, 25 are the largest of such organizations listed in a special study of the shihon keiretsu, “Kyodai Kigyō Yonjusha nimiru Keiretsu Shihai no Jittai” (The State of Group Control Among 40 Large Firms) in Shūkan Tōyō Keizai, Nov 1970, pp. 117–33. The remaining three (Mitsui & Co., Sumitomo Shoji, and Mitsubishi Corp.) were drawn from a list appearing in KK, 1970 ed.; they were added becaus e they are larger than some on the Shūkan Tōyō Keizai list. This edition of KK also contains the list of the largest 650 corporations (pp. 62–70) and the subsidiaries of the 28 firms here considered (pp. 377–715).

²⁶ OE, June 1970; see n. 18 above.

²⁷ Hadley (n. 3 above, pp. 353–56), cites an FTC study showing that, when account is taken of subsidiaries of major firms, concentration ratios rise significantly in several markets in the steel industry. My own study shows that it is in the steel industry that the horizontal relations within the holding company complexes are most pronounced. But this case appears to be distinctly atypical—at least for subsidiaries that are leading firms (and where the effect on market concentration would be greatest).

²⁸ Of particular concern at present are the rapidly expanding acquisitions of the trading companies (the sōgōshosha), the three largest of which—affiliated with Mitsui, Mitsubishi, and Sumitomo—are included in our sample study of 28 shihon keiretsu.

²⁹ See Kōsei Torihiki (Fair Trade), Aug 1972, pp. 9, 11. As is shown here, in this year, 734 cartels in small-scale industry represented 64 industry sub-divisions. It should be noted that none of the figures given in the text includes government-financed car-tels in agriculture. It is estimated that about. 5% of the national income is spent to provide price supports in agriculture—the predominant portion of which is devoted to supporting the price of rice.

³⁰ See Gendai Infure to Shotoku Seisaku (Modern Inflation and Incomes Policy), ed. Keizai-ikakuchō (Toyko, 1972), p. 330.

³¹ Kōsei Torihiki (n. 29 above), p. 9.

³² Of the total of 45 industries in which these cartels existed in Apr 1973, 17 were “self-controlled” and 28 were subject to control by MITI, which has the power to impose fines to compel outsiders to abide by the cartel agreement. Officials of the Smaller Enterprise Division of MITI inform me that such fines have rarely, if ever, been imposed, though in some cases (which were not specified) MITI has made attempts to enforce the provisions of the cartel.

³³ Consider the interesting case of the match industry (small-scale and competitive in Japan), which avidly sought to have its products taxed so that it could use the required sales reports to the government as a means of monitoring its cartel restrictions. The tax was imposed; but despite this (as confidential data I have seen indicate) the cartel, for quite a while, experienced difficulty in bringing output at all under control.

³⁴ The data of this study, made by MITI, were given to me in multigraphed form.

³⁵ It perhaps deserves emphasis that our tests judge cartel “effectiveness” by success in meeting output adjustments regarded as necessary to control the general industry price level. A finding of “ineffectiveness” therefore does not imply that the cartel may not check relatively minor and short-run price reductions.

³⁶ These findings essentially confirm portions of the findings of my earlier study (n. 1 above), pp. 275–76.

³⁷ The relevant U.S. findings are given in Nutter, G. W. & Einhorn, H. A., Enterprise Monopoly in the United States: 1899–1958 (New York: Columbia Univ. Press, 1969), p. 83Google Scholar, Table 22, “Estimate II,” note c; it indicates that in 1958, 15.8% of the U.S. national income was produced under “monopo-listic” conditions. The criteria on which this estimate is based (there is a broader-based and higher estimate of 20.2%) are defined on pp. 48–49. Einhorn and Nutter classify Transportation as wholly competitive. If this were done for Japan, our estimate of the proportion of the Japanese national income produced under “high concentration” conditions would decline by about 2 percentage points to 15% in 1956 and 17% in 1968. It appears that the estimates for the two countries are suitable for rough comparison, but they should not be used for purposes of finer comparison. For, as between the two studies, there may be differences in the definition of the industry; and the relative importanc e of regional market concentration in the two economies may vary. as may the adjustments in concentration ratios required to allow for the effects of imports. Since it eliminates the first of the three possible difficulties mentioned, a narrower study I made of 61 matching (and largely oligopolistic) manufacturing industries is of relevance here. It shows that the average concentration ratio (for the top 4 firms) was 57% for Japan (in 1966) and 53% for the U.S. (in 1967). The Japanese data are taken from the 1969 FTC study Nihon no Sangyō Shūchū (Concentration of Industries in Japan), the U.S. data from the 1971 U.S. Department of Commerce study Concentration Ratios in Manufacturing.

³⁸ A recent confidential FTC study for the period 1960–72 shows that the number of alleged illegal cartels investigated, and the number found in fact to exist, have generally been noticeably higher since 1963. In interpreting these findings, the growing vigor of the Japanese anti-monopoly program is of manifest relevance.

³⁹ While they both contain much of value, the studies of both Hadley (n. 3 above) and Yamamura (n. 5 above) reflect views of this general character, They do not, however, share all these views in common; and the emphasis on various developments differs. The concern with large-scale or oligopolistic industry is most pronounced in Hadley's treatment; she argues (pp. 319–20) that, in judging market concentration, attention should be given only to Japanese industries that are “capital intensive.” The explicit contention that Japanese industry has been growing more concentrated over the entire postwar period (which is based ultimately on growth move- ments in manufacturing firms of differing absolute size) is Yamamura's (ch. vi, esp. pp. 96, 105). Both stress the effects of the Japanese merger movement and regard a growth in concentration as the expected “trend” (Hadley, pp. 340, 342, 344, 348; Yam-amura, pp. 96–99, 103). Yamamura is noticeably more circumspect in his treatment ofJapanese cartels (pp. 65–67) than Hadley (pp. 340, 342, 346; and ch. xv, esp. pp. 373–89); it is in Hadley's study (pp. 291–301, 340, 342, 346) that one finds the view that a specific type of business grouping—the shihon keiretsu —constitutes a serious threat to competition, See also the concluding chapters of both studies.

⁴⁰ See note 38 above.

⁴¹ The proposal would reduce from 10% to 5% the maximum proportion of a company's stock that banks could acquire; in the remaining cases, corporations would be prohibited from using more than their paid-in capital or half their net assets to purchase stock in other firms. The principal purpose of this latter provision is to check the growth of the shihon keiretsu, in particular the acquisitions by the large trading companies which (nurtured by government action in the earlier postwar period with a view to promoting exports) have, as indicated previously, become especially aggressive in amassing holdings in other corporations.