1 Pessen, Edward, “The Egalitarian Myth and the American Social Reality: Wealth, Mobility and Equality in the ‘Era of the Common Man’”, American Historical Review, 76 (October 1971)CrossRefGoogle Scholar; Riches, Class and Power before the Civil War, (Lexington, Mass., 1973); “Who Rules America? Power and Politics in the Democratic Era, 1825-1975,” Prologue, 9 (Spring 1977).

2 Gallman, R.E., “Trends in the Size Distribution of Wealth in the 19th Century: Some Speculations,” in Studies in Income and Wealth (New York, 1969), vol. 34: Six Papers on the Size Distribution of Wealth and Income, Soltow, Lee, ed.Google Scholar

3 See Davis, L. E., Easterlin, R. A., Parker, W. N., et al., American Economic Growth (New York, 1972), 50–51Google Scholar. The variable is not wealth, but a close correlate, income, and the measures are less comprehensive than one could wish. The United States’ data, which are very rough, refer to 1860 and 1867. The European evidence relates to the 1870s and 1880s, a period somewhat later than the period under discussion. But according to Lee Soltow, the concentration coefficient of British income was roughly the same in 1801-3, 1812, 1867, 1880 and 1913. (“Long-Run Changes in British Income Inequality,” The Economic History Review, 21, 2nd series (April 1968), 17-29.)

4 See Atkinson, A. B., “The Distribution of Wealth and the Individual Life-Cycle,” Oxford Economic Papers, 23 (July 1971), 239–54CrossRefGoogle Scholar. Income and wealth vary directly with age through most of life, but vary indirectly with age toward the end of life, as the subsequent discussion makes clear.

5 This assumes that each individual would have an identical lifetime pattern of family responsibilities or that differentially heavy family costs would be shared out among all members of the population. In the same way, it assumes that the debts of those dying in debt would also be shared out among the entire population.

6 See, for example, Mincer, Jacob, Schooling, Experience and Earnings (New York, 1974)Google Scholar. I thank my colleague, Solomon Polachek, for suggesting this source to me.

7 The American and British age distribution statistics were taken from: U.S. Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington, D.C., 1960)Google Scholar, series A-71-85 and Mitchell, B. R., Abstract of British Historical Statistics (Cambridge, 1962), 12Google Scholar. Data referring to males were used, since initially the calculations were viewed as referring to heads of households, who were chiefly males. But this view raises complications over the meaning of the model and it would be more reasonable to shift to the age distribution of males and females, combined. These recalculations have not been carried out, but they are not at all likely to change the conclusions in any material way, in any case, because the age distribution for males is not markedly different from the distribution for males and females, combined. The French age distribution data (males and females) are based on data on page 134 of Part I, Volume VI of the Cambridge Economic History of Europe (Cambridge, 1965), interpolated to obtain narrower age divisions, on the basis of stable population “West” model life tables, with mortality level 11 (Ansley J. Coale and Paul Demeny, Regional Model Life Tables and Stable Populations, (Princeton, N.J., 1966). I am indebted to my colleague, Boone Turchi, for these calculations. The underlying French data source (which is not presently available to me) may contain the necessary details. But, in any case, the “West” model almost certainly produces results of sufficient accuracy for present purposes. The assumption was made that each population contained 1000 people and, thus, the wealth, income, and savings of each population (as well as the distribution of each) could be computed, given the average experience of members of each age class. To simplify calculations, the median age of the population in the age group 15-19 was assumed to be just 17 years (i.e., half of the group were 15 and 16; the other half, 17, 18, 19); the median age of the group 20-29 was just 24, etc. For purposes of interpolating within age groups, it was assumed that 15 year olds were as numerous as 16 year olds; 17 year olds, as numerous as 18 year olds; 18 year olds as numerous as 19 year olds, etc.; that is, it was assumed that half of the population in each age class—except age class 70 and over—was divided equally among the years below the median age of the class; the other half, equally among the years above the median age of the class. In the case of age class 70 and over, estimates were interpolated on the tail of the “West” model (see above). More elegant (and accurate) procedures would, without much doubt, produce results that would modestly strengthen the position adopted in the text, but the improvement obtained would not be worth the effort.

8 But the model does a much poorer job accounting for the way wealth was divided among the rich. For example, according to the model, the richest 1 percent of people fifteen and over held 5.5 percent of wealth, the richest 10 percent, roughly 50 percent, while according to available direct estimates, the actual figures (for families) were 24 percent and 72 percent (See Gallman, op. cit.). But note that the universes are also different (see text, below).

9 Lampman, R. J., The Share of Top Wealth-Holders in National Wealth, 1922-1956 (Princeton, 1962), 213Google Scholar. When the 1950 data are restricted to population twenty years old and older, the model produces the result that the richest 31.7 percent held 73.5 percent of total wealth.

10 In order to see the type of economic structure implied by the model, predictions of the capital-output ratio and the savings rate were also made, with the following results: The model predicts U.S. capital-output ratios of 1.6 for 1830, 1.8 for 1860, and 2.9 for 1950, while independent direct estimates place the actual ratio at 1.6, 1.8 and 2.6, in 1840, 1850 and 1950. The model also predicts net savings rates of 9.1 percent, 11.1 percent and 9.5 percent for the United States in 1830, 1860 and 1950, respectively, while independent direct estimates yield figures of 9.5 percent, 12.1 percent and 9.0 percent for the U.S., 1834-1843, 1849-1858, and 1946-1955, respectively. (See Tables 1 and 3 in Davis, L. E. and Gallman, R. E., “Savings and Investment in the United States During the 19th Century,” in Cambridge Economic History, Volume VII (in press)Google Scholar; Goldsmith, Raymond, “The Growth of Reproducible Wealth in the United States of America from 1805-1950,” in Income and Wealth, Series II (Cambridge, 1952), 298Google Scholar. The Goldsmith measurement is apparently in real magnitudes. The Davis and Gallman measurements for 1840 and 1850 are expressed in prices of 1860, but current price values would be almost identical.) The model predicts a net savings rate of 9.2 percent for Great Britain in 1861, whereas an independent direct estimate yields a value of 8.9 percent for the United Kingdom, 1855-1864. On the other hand, the predicted value of the capital-output ratio—2.1—is much lower than the probable actual value. (See Kuznets, Simon, “Quantitative Aspects of the Economic Growth of Nations: VI. Long Term Trends in Capital Proportions” Economic Development and Cultural Change, 9 (July 1961), 58, 62.CrossRefGoogle Scholar) This discrepancy may reflect the importance of inherited wealth in Britain, a variable left out of the model. American growth was so rapid that the share of wealth attributable to inheritance from previous generations was probably much smaller. (See below.)

11 See Gallman, Robert E., “The Social Distribution of Wealth in the United States of America,” Proceedings of the Third International Conference of Economic History (Paris, 1968), 330Google Scholar. The figures are in current prices and include net claims on foreigners, fixed reproducible capital, houses, land, and inventories. In constant prices, the rate of increase was about 6.1 percent per year. Domestic wealth (as distinct from national wealth) increased at a rate of about 6.4 percent per year (current prices).

12 Indeed, they were typically well under 6 percent per year, according to Homer, Sidney, A History of Interest Rates, (New Brunswick, N.J., 1963), 287–88.Google Scholar

13 Inferred from Pessen, Riches…, Tables 3-1 and 3-2 and from U.S. Bureau of the Census, Historical Statistics of the United States, (Washington, D.C., 1949)Google Scholar, series A-100 (Value of land and buildings of Manhattan Island). Incidentally Pessen’s Table 7-6 suggests that the very rich in New York typically did even less well than this. Of the 440 richest New York families in 1856-1857, only 47 represented the very rich of 1828-i.e., those who held at least $100,000 in 1828. In the twenty-eight or twenty-nine year interval between 1828 and 1856-1857, 27 of these families apparently increased their fortunes by less than VA fold, which is at a rate very much smaller than that at which the United States’ national wealth stock was increasing. The other 20 may or may not have held their own. One cannot be sure, from the Table. Table 7-5 suggests the same finding. Of 55 families who held at least $250,000 in 1845, 23 held less than this in 1856-1857, while another 16 held between $250,000 and $500,000, which may or may not have represented an absolute improvement during the period, but almost certainly represented a relative deterioration of position in the United States. But, of course, Pessen only takes into account wealth assessed in New York and these people may have had substantial holdings elsewhere.

14 Davis, et al., American Economic Growth, 38.

15 Simon Kuznets makes a similar, but distinct, point in his presidential address to the American Economic Association, American Economic Review, 45 (March 1955), 10. Kuznets argues that industry growth patterns tend to follow the logistic. Rapid aggregative growth is accompanied by pronounced changes in the industrial composition of the economy. This tends to produce constant changes in the personnel of the class of very rich and economically powerful. A successful entrepreneur rises to wealth and power by a fortunate association with a new industry. If his heirs retain their investments in this industry, they eventually lose in relative standing, compared with those who invest in the next group of new industries. Only if the heirs are prepared to shift their investments and only if they are skillful and lucky enough to choose the correct new industries will they retain place.

16 See appendix.

17 Mitchell, B. R., Abstract of British Historical Statistics (Cambridge, 1962), 455, 456, 460.Google Scholar

18 Of course they also took up the opportunity to invest in the securities of the Bank of the United States and the Commonwealth of Pennsylvania, decisions that produced a certain “flux” in English fortunes. But see Jenks, L. H., The Migration of British Capital to 1875 (New York, 1927)Google Scholar, 415, for a very favorable view of the returns on foreign investments at a slightly later date.

19 The calculated return (7.3 percent) is the ratio of average “rents, profits, interest and mixed incomes,” 1860 through 1869, to the value of the United Kingdom wealth stock in 1865 (see appendix). The value of “rents, etc.” in 1860 through 1869 was computed by multiplying current price national income at factor prices (Deane, Phyllis and Cole, W. A., British Economic Growth, 1688-1959: Trends and Structure, 2nd ed. (Cambridge, 1967), 329Google Scholar) by the share of “rents, etc.” in national income at factor prices (Deane and Cole, 247). The results are necessarily chancey, since Deane and Cole took “profits, etc.” as a residual; that is, they estimated “wages,” “salaries,” and “rents” directly, and subtracted these figures from national income to obtain “profits, etc.” Also, “profits, etc.” include some components (“mixed incomes”) which are not strictly returns on property, but this is unlikely to be a major source of error.