The institutional character of the buyers' market is clearly indicated. Eighty-eight percent of the first public sales of these publicly issued securities were made to institutions. Furthermore, the major part of the sales to security dealers undoubtedly found their way to institutions within a short time. The difference between distribution effected through public offering and that effected by private placement is blurred when, as in the case of one issue, the distributing group sold 74 percent directly to insurance companies and sold another 19 percent to banks.

Further studies by the Securities and Exchange Commission indicate that the banks are only temporary stopping-places for these bonds.34 The banks resold from one-half to four-fifths of their purchases from the distributing group within 3 months. As might be expected, life insurance companies were the principal purchasers, taking from two-fifths to three-fourths of the total amount resold.35 In addition, the life insurance companies bought blocks of these issues on the open market and from security dealers; they continued to buy during the period studied.

It is, therefore, not surprising to discover that the life insurance companies engaged in a large number of transactions to acquire their holdings. In the case of 2 issues of twenty-five and thirty million dollars, the companies on the average required 67 and 66 separate transactions, respectively; and in the case of 2 larger issues of one hundred and forty and one hundred and thirty million dollars they required 93 and 101 transactions on the average, respectively. Most of these transactions were small. In all, the insurance companies made 4,294 separate transactions in connection with the 5 issues studied. There were 32 purchases in $1,000 lots and 142 in blocks of $2,000. More than one-third of the transactions were in blocks of $5,000 or less. More than three-quarters were in amounts of less than $30,000, though they accounted for only 21 percent of the total purchased.

In the cases studied, insurance companies were the largest single group of purchasers, but they were obviously put to much time, effort, and expense to acquire the amounts they did. If one may judge by other evidence, they probably did not succeed in purchasing as much of these issues as they wanted to. If the issues had been privately placed, the probability is that the investment bankers' commission would have been divided between the issuing company and the life insurance buyers, with the former getting more for the bonds and the latter paying less for them.

These facts by no means describe all the elements in the controversy between private placement and public offering. They do indicate, however, that private placement has a solid institutional base derived from the concentration of savings and the coming-of-age of

34 Ibid., pp. 13021-13035.

35 In the case of one issue, the United States Steel Corporation 32's of 1948, the amount resold to insurance companies was only 7 percent of the total. This issue was different from the others, however, because insurance companies had bought only a small part from the distributing group. The explanation in both cases was that the issue was regarded by the "trade" as a "banking issue" by reason of the short maturity and the industry involved.

both the insurance companies and the issuing corporations. Private placement is merely another change, this time in the field of investment banking, resulting from concentration in industry and finance. The impact of these forces is such as to take away from investment bankers some part of the power of deciding what industries shall be able to borrow and vest it in the executives of insurance companies and of large industrial corporations. But the banks still determine how a good deal of individual savings shall be invested. Moreover, the investment bankers still exert a large measure of control over our largest industries and even over insurance companies.

WHAT TYPES OF GOODS DO INVESTMENT DECISIONS PRODUCE?

The end product of the investment process is physical capital, the implements of production, items such as homes, offices, bridges, and highways that give off services for a period of years.

The demand for capital is in the main a derived demand. No one wants a steel mill or cement works or glue factory for its own sake, but for the consumable goods it produces. As Fortune points out:

The tools and extensions of industrialization do not exist for their own sake. They exist for the individual, known in this connection as the consumer. The entire producers' goods industry whose purpose is the making of tools is quite secondary to the real purpose of industrialization. That real purpose of indus trialization may be defined as an increase in the power to consume,

The central economic problem is not a revival in the producers' industry, although that would help. Nor can it be "investment" in the old sense of the word. The central economic problem is simply the conversion of a high potential power to consume into an actual power to consume.

36

New aircraft plants or rayon plants or bridges or homes are built because there is an active consumer demand for the product or service. The larger the volume of consumption, the faster goods move and the more capital formation takes place. (See table 55.) In the words of the Brookings Institution:

We found from a study of our industrial history that the growth of capital is closely adjusted to and dependent upon an expanding demand for consumption goods. * ** Fluctuations in the construction of capital goods have usually followed rather than preceded fluctuations in the output of consumption goods. The controlling importance of consumption was, however, more conclusively revealed by the discovery that the rate of growth of new plant adjusted to the rate of increase of consumptive demand rather than to the volume of savings available for investment purposes.

37

36 Fortune, "United States Industrialization," February 1940, pp. 50, 160.

Harold G. Moulton, Income and Economic Progress, Brookings Institution, Washington, 1935, p. 43. [Italics in original.]

TABLE 55.-Income-producing expenditures that offset saving and gross national income, 1921-39

1 Estimated by George Terborgh.

2 Estimates by D. L. Wickens and R. R. Foster for the National Bureau of Economic Research, the Department of Agriculture, and the Department of Commerce.

3 Principally from Simon Kuznets, Commodity Flow and Capital Formation (1938).

4 From the Department of Commerce.

Computed by the Board of Governors of the Federal Reserve System.

6 Estimates for 1923-37 by Rolf Nugent; other years, by the Board of Governors of the Federal Reserve System.

7 Equal to 60 percent of the current year, plus 40 percent of the preceding year.

• Estimated.

Preliminary.

Source: Hearings before the Temporary National Economic Committee, Part 9; Savings and Investment, pp. 4010-4018, 4122, as revised, where sources and methods are described.

From the last column of table 55 it is probable that less capital was required on the average in the 1930's to produce a given volume of national income than in the late 1920's.

Some of the more important factors leading toward increased productivity in recent years have been increasing utilization of largecapacity equipment, accompanied by decreases in equipment expenditures per unit of capacity; use of industrial measuring, recording, and controlling devices; improvements in the composition of metals, varnishes, and lacquers, and in concentration processes; biological improvements; and managerial improvements resulting in better factory lay-out, and more effective flow of production.

The effects of increases in productivity upon investment are strikingly clear in many industries.38 For example, investment in fixed capital in the automobile industry in 1938 was 38 percent less than in 1926, while output (of vastly improved quality) increased by 22 percent.39

In 1926 the fixed capital invested in the iron and steel industry was valued at 3.8 billion dollars; in 1937, at 3 billion dollars; yet capacity

38 See Work Projects Administration, National Research Project, Production, Employment, and Productivity in 59 Manufacturing Industries, 1939-40. 80 Spurgeon Bell, Productivity, Wages, and National Income, Brookings Institution, shington, 1940, pp. 288-290, 299.

was 57.8 million tons in the former year, compared with 69.8 million tons in the latter.40 The data presented by the United States Steel Corporation to the Temporary National Economic Committee illustrate these trends for one company. Between 1926 and 1937 the book value of fixed assets decreased from 1.7 billion dollars to 1.4 billion dollars, but ingot capacity increased from 22 to 25 million tons.

The estimated value of fixed capital in the privately owned segment of the electric light and power industry was 7 billion dollars in 1926 and 10.9 billion dollars in 1938; output (measured in index numbers) rose from 106 in 1926 to 239 in 1938.41 Thus, from 1926 to 1938 investment increased by 56 percent, but output increased by 125 percent.

In the railroad industry, with a 17 billion dollars investment in plant and equipment, creosoting has more than doubled the life of ties and heavier rails and steel rolling stock have reduced replacement costs. Locomotive tractive power has increased, more efficient locomotive designs and the widespread use of water treatment have reduced repairs and increased both the capacity and the life of steam engines. The decrease in passenger traffic, the decline in less-than-carload-lot shipments, and the increase in the average length of haul have reduced the wear and tear. Utilization of existing plant has increased. Train speeds have increased sharply, and terminal facilities operate with greater rapidity.

Finally, in the machine tool industry a recent survey indicated that of a total of 11,610 machines purchased in 1936-37, the 4,666 acquired for the specific purpose of replacing old ones were substituted for 7,377 machines. "It may well be assumed that the total capacity of the machines used for replacement was at least equal to that of the machines which were scrapped." 42

Plant and Equipment.

As table 55 indicates, investment in plant and equipment by all business enterprises totaled 10.2 billion dollars in 1929 (4.6 billion dollars for plant, 5.6 billion dollars for equipment), 7.6 billion dollars in 1937 (2.3 and 5.3 billion dollars respectively), and 6.2 billion dollars in 1939 (1.9 and 4.3 billion dollars respectively). Equipment outlays of 5.3 billion dollars for all enterprises in 1937 were higher than those of any year in the 1920's except 1929. In view of decreases in prices and increases in productivity during the period, the equipment outlays in 1937 undoubtedly reflected more real investment and substantially more productive capacity than in 1929. But plant outlays followed a different course. The 2.3 billion dollars of plant outlays in 1937 were less than those in every year in the 1920's, and only half the outlay in 1929.

Many contend that increases in productivity, particularly those designated as "managerial," have reduced the demand for plant in relation to equipment. Following an exhaustive study of changes in productivity, it was reported that

In the automobile industry particularly, but in other manufacturing industries as well, improvements in plant lay-out appear to have been greatly

40 Data compiled by the American Iron and Steel Institute, as of January 1. op. cit., pp. 288-289.

Cf. Bell,

41 Bell, op. cit., pp. 275-277. 42 David Weintraub, Effect of Current and Prospective Technological Developments Upon Capital Formation, Report G-4, National Research Project, Work Projects Administration, Pp. 12-13. (Reprinted in American Economic Review Supplement, vol. 29, 1939, pp. 15-32. From estimates of George Terborgh, "Estimated Expenditures for New Durable Goods, 1919-38," Federal Reserve Bulletin, September 1939 and February 1940. For evaluation and greater detail see Temporary National Economic Committee Monograph No. 37, especially appendixes III, IV, and VI.

stimulated by the depression, with resulting better continuity of the flow of work and savings in direct and supervisory labor, equipment, floor space, and inventories."

For example, in 1934 through changing its lay-out, Packard cut floor space per unit of output nearly in half, and was left with a vacant building. When Western Electric substituted straight-line for functional manufacture it reduced its required floor space by 17 percent.45 At the levels of industrial production which have prevailed in the last decade it has been possible to modernize machinery and equipment without adding substantially to plant floor space.

Annual expenditures for equipment did not increase greatly from 1923 to 1928. Currie testified that during this period "despite rapidly increasing production, despite rapidly increasing consumption, and despite the smallness of the increase in equipment expenditures, there was no evidence of any growing strain on our productive facilities." 48 Plant and equipment outlays in mining and manufacturing, and in agriculture, recovered almost completely between 1929 and 1937, but investments therein by railroads and transit companies in 1937 were 63 percent and 75 percent, respectively, of the 1929 totals, while those in the electric power and telephone industries were only 52 percent and 57 percent, respectively.

Agricultural outlays for equipment were 14 percent higher in 1937 than in 1929, despite the fact that outlays for plant were more than onethird lower. The growing strides of mechanization are indicated by the continued high level of post-depression equipment expenditures: 1936-39 was 6 percent higher than 1926-29, despite the increases in productivity and the decreases in prices that had occurred in the meantime.47

Railroad outlays for plant are far short of the level of the late 1920's; the abandonment of trackage is continuing, while the construction of new unified railroad terminals in the 1920's did not need to be repeated or extended in the 1930's. Equipment expenditures in 1937 were as high as in 1929, though less than half of what they had been in 1923. The volume of carloadings has decreased by one-third from 1926-29, the average number of serviceable freight cars and locomotives needed has decreased because of striking increases in the speed and efficiency of railroad transportation.

Residential Construction. ́ ́

Since a separate chapter is devoted to housing, 48 the discussion here is restricted to a few pertinent facts. Housing does not follow the cycle of general business, but traces a pattern all its own of roughly 15-year swings. The volume of construction rose from 2 billion dollars in 1919 to a peak of 5.1 billion dollars in 1925, declined steadily during the next 8 years to a low of 375 million dollars in 1933, and rose steadily to 2.1 billion dollars in 1939.49 The trend is still upward.

In 1939, 76 percent of all new dwelling units were one-family houses, as opposed to but 59 percent in the years 1924-26. The rapid shift

44 Weintraub, op. cit., pp. 12-13.

45 Ibid., pp. 12-13, footnote 29.

46 Hearings before the Temporary National Economic Committee, Part 9, p. 3524.

47 George Terborgh, "Estimated Expenditures for New Durable Goods, 1919-38," Fed. Res. Bull., Sept., 1939.

48 Ch. XIII.

49 Terborgh's estimates (Temporary National Economic Committee Monograph 37, appendix III). Kuznets' estimates (ibid., appendix I) show a similar pattern.