What causes a depression !

The Statistical Setting

The setting of our story is of necessity statistical. And statistics have the merit of being succinct. I am aware, however, that many readers cannot face statistics, a fact that leads seasoned editors to advise writers to dispense with them or to hide them in the back of the book. Apparently childhood encounters with arithmetic under inferior school conditions have developed in many people (even the cultivated) a distaste for numbers, and when they see them they merely skip. But it will repay readers to study and ponder carefully the following figures.

While good studies have been made for some decades, three recent high-level inquiries have developed the picture in sharper and more exact detail than ever before. They represent a long series of analyses of the extent and concentration of American wealth that was begun by G. K. Holmes in 1893. These analyses, showing greater and greater precision with the passing years, are listed in the chapter notes. 5

The three recent studies were made, independently, by Professor Robert J. Lampman of the University of Wisconsin for the National Bureau of Economic Research, by the Survey Research Center of the University of Michigan as a continuing project in 1947, 1952, 1956, 1960 and 1963; and by the Harvard historian Gabriel Kolko as presented in his Wealth and Power in America (1962). I will touch upon these, as well as a resounding official clincher, in this order.

Running to 286 pages, containing 138 formidable tables and 37 charts (including 13 Lorenz curves) and employing the most sophisticated applicable mathematics, the Lampman study was published by Princeton University Press in 1962. 6

What Professor Lampman did was to obtain basic data from federal estate tax returns for the years 1922, 1929, 1933, 1939, 1945, 1949, 1953 and in some cases for 1954 and 1956; but he concentrated attention on 1953. Such tax returns are required by law of all decedents with estates exceeding the level of exemption, which was $50,000 for 1922-26, $100,000 for 1926-32, $50,000 for 1932-35, $40,000 for 1935-42 and $60,000 after 1942.

With the data in hand, Professor Lampman then employed the established estate-multiplier method. This requires that one multiply the number and property of decedents in each age-sex group by the inverse of the general mortality rate for each such group. One thereby arrives at an estimate of living persons and the amount of estate in each age-sex group and in each estate size.

Professor Lampman illustrates the method as follows: "Suppose that out of a population of 1,000 men aged 40 to 50, two men died in one year with estates of $100,000 or more. Suppose further that it is known that 5 per cent of all the 1,000 men aged 40 to 50 died in that year. Then it may be assumed that the two men who died with $100,000 were 5 per cent of all the living men in the group with $100,000. Hence, to estimate the number of living men with $100,000, we should multiply two by twenty (the inverse of 5 per cent) to get the answer of forty living men with $100,000 or more."7



The Lampman Findings

What Lampman found was as follows:

1. More than 30 per cent of the assets and equities of the personal sector of the economy (about 20 per cent of all wealth in the country being government-owned) in 1953 was held by 1.6 per cent of the adult population of 103 million.8

2. This group of 1.6 per cent owned 32 per cent of all privately owned wealth, consisting of 82.2 per cent of all stock, 100 per cent of state and local (tax-exempt) bonds, 38.2 per cent of federal bonds, 88.5 per cent of other bonds, 29.1 per cent of the cash, 36.2 per cent of mortgages and notes, 13.3 per cent of life insurance reserves, 5.9 per cent of pension and retirement funds, 18.2 per cent of miscellaneous property, 16.1 per cent of real estate and 22.1 per cent of all debts and mortgages.9

3. The following table shows the percentage of national wealth-holdings for the top 1/2 of 1 per cent and 1 per cent for the indicated years.10

1/2 of 1 Per Cent 1 Per Cent
of Adult Population of Adult Population
(per cent) (per cent)

1922 29.8 31.6
1929 32.4 36.3
1933 25.2 28.3
1939 28.0 30.6
1945 20.9 23.3
1949 19.3 20.8
1953 22.7 24.2
1954 22.5 ....
1956 25.0 26.0

4. The estimated gross estate size for the total adult population in 1953, obtained by extension of the same methods, was as follows:11

Gross Estate Number of Average Total Gross
Size (dollars) Persons Aged Estate Size Estate
20 and Over (dollars) (billion
(millions) Percentage dollars) Percentage

0 to 3,500 51.70 50.0 1,800 93.1 8.3
3,500-10,000 19.00 18.4 6,000 114.0 10.2
10,000-20,000 21.89 21.2 15,000 328.4 29.3
20,000-30,000 6.00 5.8 25,000 150.0 13.4
30,000-40,000 2.00 1.9 35,000 70.0 6.3
40,000-50,000 0.80 0.8 45,000 36.0 3.2
50,000-60,000 0.35 0.3 55,000 19.3 1.7

All under 101.74 98.4 7,900 810.8 72.4
60,000

60,000-70,000 0.18 0.1 61,000 10.5 0.9

over 60,000 1.66 1.6 186,265 309.2 27.6

All estate 103.40 100.0 10,800 1,120.0 100.0
sizes

Median estate size 3,500

In this table is found one verification of my initial paragraph. It shows that 50 per cent of the people, owning 8.3 per cent of the wealth, had an average estate of $1,800--enough to cover furniture, clothes, a television set and perhaps a run-down car. Most of these had less; many had nothing at all. Another group of 18.4 per cent, adding up to 68.4 per cent of the population, was worth $6,000 on the average, which would probably largely represent participation in life insurance or emergency money in the bank. Perhaps this percentage included some of the select company of "people's capitalists" who owned two or three shares of AT&T.

Another 21.89 per cent of adults, bringing into view 92.59 per cent of the population, had $15,000 average gross estates--just enough to cover a serious personal illness. This same 92-plus per cent of the population all together owned only 47.8 per cent of all assets.



Top Wealth-Holders

The number of persons in the top 1 per cent of wealth-holders through the decades was as follows:12

Years Number of Persons Percentage Share
(thousands) of Gross Estates

1922 651 32
1929 744 38
1939 855 33
1945 929 26
1949 980 22
1953 1,030 25



But the top 11 per cent of persons in the magic 1 per cent (or 0.11 per cent) held about 45 per cent of the wealth of this particular group while the lower half (or 0.50 per cent) held only 23 per cent.13

Says Lampman: "The personally owned wealth of the total population in 1953 amounted to about $1 trillion. This means that the average gross estate for all 103 million adults was slightly less than $10,000, The median would, of course, be considerably lower. In contrast the top wealth-holder group had an average gross estate of $182,000. The majority of this top group was clustered in estate sizes below that average. Of the 1.6 million top wealth-holders, over half had less than $125,000 of gross estate and less than 2 per cent (27,000 persons) had more than $1 million."14

There were, then, in excess of 27,000 millionaires in the country in 1953--not only the greatest such aggregation at one time in the history of the world but a number greater than the aggregation throughout all of history before 1875 (as of 1966, millionaires numbered about 90,000). If consumer prices had remained stable from 1944 to 1953 there would have been fewer. "In 1944 there were 13,297 millionaires," says Lampman. "In 1953 there were 27,502 millionaires in 1953 prices, but only 17,611 in 1944 prices."15

What of the 1965-67 year-span? As the prices of stocks advanced tremendously in the preceding dozen years, one can only conclude that the proportion of wealth of the top wealth-holders also advanced impressively. For this small group, as we have seen, owns more than 80 per cent of stocks. The Dow-Jones average of 65 industrial stocks stood at 216.31 at the end of 1950; at 442.72 in 1955; at 618.04 in 1960; and at 812.18 in March, 1964. As of May, 1965, it was well above 900. The less volatile Securities and Exchange Commission index of 300 stocks shows the same quadrupling in value, standing at 41.4 in 1950; 81.8 in 1955; 113.9 in 1960; and 160.9 in March, 1964. How many employees have experienced a fourfold increase in salaries in the same period?

The rise in value of stocks, however, surely invalidates one of Lampman's speculations, to this effect: "Our finding that the share of wealth held by the top 2 per cent of families fell from about 33 to 29 per cent from 1922 to 1953, or about one-eighth, would seem compatible with . . . the general belief that there has been some lessening of economic inequality in the United States in recent decades."16 The more recent rise in stock prices and in corporation earnings shatters even that slight concession.

Professor A. A. Berle, Jr., has rushed forward to hail the Lampman showing that the upper 1 per cent saw its participation reduced from 32 per cent of all wealth in 1922 to 25 per cent in 1953; but his celebration was premature and he did not fully report Lampman, who indicated that the participation had been reduced from 1922 to 1949 but thereafter was again increasing.17

The Lampman findings were extended to 1958 in an extremely sophisticated statistical critique presented in 1965 to the American Statistical Association by James D. Smith and Staunton K. Calvert of the Statistics Division of the Internal Revenue Service.18

After reviewing Lampman, revising him in a minor particular, Smith and Calvert conclude that "top wealth-holders owned 27.4 percent of gross and 28.3 percent of net prime wealth in 1953, but increased their share to 30.2 and 32.0 percent respectively by 1958. These data support Lampman's conclusion that the share of top wealth-holders has been increasing since 1949." Prime wealth, as they explain, is total wealth less the value of assets in trust funds and pension reserves.

This is where the question rests on the basis of the most recent data supplied by leading authorities in the field: Concentration of wealth in a few hands is intensifying.

Actually, in view of market valuations, the share of top wealth-holders at this writing is easily the greatest in history. It is my hypothesis that the share of the top 1/2 of 1 per cent now exceeds the 32.4 per cent of this group for 1929. Later studies should show that the proportions for all groups of top wealth-holders studied by Lampman have been significantly exceeded. So much for Lampman although there is much else in his razor'-sharp book that merits attention.19



The University of Michigan Study

Although showing some minor variations, the continuing University of Michigan survey dovetails with the Lampman study and fully supports it.

First, it was found that 3 per cent of spending units in 1953 had $60,000 or more total assets; this compares with 2.3 per cent of individuals in the Lampman study. A "spending unit" consisted of any one or more persons established as a household.

According to this University of Michigan "Survey of Consumer Finances," the upper 11 per cent of the nation's 54 million spending units held 56 per cent of the total assets and 60 per cent of the net worth of all private holdings in the country. "While this group held only 30 per cent of consumer capital," Lampman comments (p. 195), they held 80 per cent of business and investment assets." 20

According to the 1960 University of Michigan "Survey of Consumer Finances," 86 per cent of all spending units in the country owned no stock whatever. Of incomes under $3,000, 95 per cent owned no stock; of incomes of $3,000-$5,000, 93 per cent owned no stock; and of incomes of $5,000-$7,500, 87 per cent owned no stock. The class of $7,500-$10,000 incomes was 78 per cent without stock ownership, while even in the $10,000-$15,000 income class 61 per cent owned no stock. In 1963 a total of 83 per cent owned no stock. Stock ownership, it is clear, was being somewhat more widely diffused as long-term holders gradually unloaded at rising prices. Whereas in 1953 only 44 per cent of the income class above $15,000 owned no stock, in 1960 this same broad class included only 26 per cent without stock ownership.21

For some years the New York Stock Exchange and the Advertising Council, as part of a campaign to show that a "people's capitalism" exists with a widely diffused ownership in American industry, have been busily pyramiding figures. These computations show that in 1956 there were 8,630,000 American shareholders, and in 1962 there were 17,010,000.22 The figure more recently being cited is 20 million .23

Even though the method of their compilation is challenged by statisticians, these computations could all be true and still not after the implications of the Lampman analysis and University of Michigan surveys. For if 17 per cent of spending units owned stock in 1963, as the University of Michigan survey indicates, that would be well over 17 million persons. And anyone would qualify as a stockholder if he owned only one share worth 10 cents.

That most stockholders own trivial amounts of stock is shown by the University of Michigan figures for 1963. The 17 per cent of spending units holding stock broke down in this way: 3 per cent held less than $500 worth; 2 per cent held $500 to $999 worth; 4 per cent held $1,000 to $4,999 worth; and 2 per cent held $5,000 to $9,999 worth. As far as stock ownership goes, these are all insignificant figures. Yet they make up 75 per cent of the households holding stock. Only 4 per cent of all spending units owned more than $10,000 of stock.24 But most of this group, exceeding four million people, also owned little stock; for we are already aware that a group consisting of 1.6 per cent of the population owns more than 80 per cent of all stock, 100 per cent of state and local government bonds and 88.5 per cent of corporate bonds. Less than 20 per cent of all stock in 1963, then, was owned by some 15.4 million people.

Throughout this study, therefore, it is going to be taken as fully established that 1.6 per cent of the adult population own at least 32 per cent of all assets, and nearly all the investment assets, and that 11 per cent of households (following the University of Michigan study) own at least 56 per cent of the assets and 60 per cent of the net worth. It is even possible, as we have seen, that 1/2 of 1 per cent own more than one-third of all productive assets as of 1965-67. It is evident that this leaves very little to be apportioned among 90 per cent of the population. It will be recalled that Lampman showed 50 per cent owning virtually nothing, with an average estate size of only $1,800 as of 1953. This same study, according to my tabulation numbered 4, showed that 89.6 per cent of the adult population had available to it only 47.8 per cent of the assets, while 50 per cent had only 8.3 per cent. The University of Michigan figures and the Lampman figures, in short, coincide rather closely although developed by different methods.



Supporting Studies

Every other serious study supports these findings. The Senate Temporary National Economic Committee (TNEC) just before World War II inquired into the distribution of stock among 8.5 million shareholders in 1,710 major companies as of 1937-39 and found that 4 per cent of all common stockholders held 74.9 per cent of the stock, and 4.5 per cent of the preferred stockholders held 54.8 per cent. 25 Looking into the same situation as of 1951, the Brookings Institution of Washington, D. C., found that in 2,991 major corporations only 2.1 per cent of the holders owned 58 per cent of the common stock and 1.1 per cent of the holders owned 46 per cent of the preferred stock. Two-thirds of all common stockholders owned only 10 per cent of the shares.26 Harvard's J. Keith Butters estimated that in 1949 the spending units (households) that owned $100,000 or more in marketable stock, comprising 1 /5 of 1 per cent of all spending units and 2 per cent of stockholders owned between 65 and 71 per cent of all marketable stock held by individuals.27 None of these studies took into account the beneficial interest of individuals in stock held by institutions for the account of individuals, which swells the percentages proportionately.

The Lampman estate studies do not necessarily reveal the sizes of fortunes. 'This is because many of the fortunes are systematically distributed during the lifetime of the owner, mainly for the benefit of heirs. At the time of death the fortune is reduced.

Again, in extrapolating from the estates to the rest of the population, at least two distortions are discernible. First, only adults are considered by Lampman, whereas a considerable number of children are millionaires owing to having had trust funds settled upon them. Second, the economic position of age groups is not strictly comparable between the affluent and the poor because of an average earlier death rate for the latter.

But, on the whole, the Lampman study came closer than anyone had yet come to showing the asset position of all adult age-sex groups.



Definitive Data from the Federal Reserve

Strongly persuasive though all these studies are, it is possible to be definitive about the distribution of wealth in the United States, on the basis of findings put forth recently under the highest official auspices.

In a complex and comprehensive study prepared for the Board of Governors of the Federal Reserve System on the basis of Census Bureau data under the title Survey of Financial Characteristics of Consumers, the cold figures are officially presented on asset holdings as of December 31, 1962, removing the entire subject from the realm of pettifogging debate.

On that date the number of households in the country worth $500,000 or more was carefully computed at about 200, 000.28 The number of millionaires at the year-end was more than 80,000, compared with Lampman's 27,000 as of 1953. Only 39 per cent of these 200,000 had no inherited assets. 29 These 200,000 at the time held 22 per cent of all wealth, while 57 per cent of the wealth was held by 3.9 million individual consumer units worth $50,000 or more.

The panorama of wealth-holding throughout the populace was as follows (in millions of units):30

Percentage of
Millions Households
All consumer units
(households) 57.9 100.0

Size of wealth:
Negative 1.0 1.8
Zero 4.7 8.0
$1-$999 9.0 16.0
$1,000-$4,999 10.8 18.0
$5,000-$9,999 9.1 16.0
$10,000-$24,999 13.3 23.0
$25,000-$49,999 6.2 11.0
$50,000-$99,999 2.5 5.0
$100,000-$199,999 .7 1.25
$200,000-$499,999 .5 Less than 1.0
$500,000 and up .2 Less than 0.4

In stating that 200,000 households held 22 per cent of the wealth there is some danger of suggesting that the power of these 200,000 is less than it actually is. The nature of the wealth held is of determining importance here. in general, the lower wealth-holders mostly own inert assets such as automobiles, small amounts of cash and some residential equity, while the upper wealth-holders mostly own corporate equities in an aggregate amount sufficient to show that they are in full control of the productive side of the economic system.

Households in the number of 200,000 worth $500,000 and more held 32 per cent of all investment assets and 75 per cent of miscellaneous assets, largely trust funds, while 500,000 worth $200,000 to $499,999 held 22 per cent of investment assets. The 700,000 households worth $100,000 to $199,999 held 11 per cent of investment assets.31