¹ Trattner, Walter I., From Poor Law to Welfare State: A History of Social Welfare in America (New York, 1974), p. 44.Google Scholar

² The skilled-unskilled wage ratios presented by Williamson and Lindert suggest that the distribution of income changed little from the time of the Revolution to 1816, but that the period from 1816 to 1856 brought a rise in inequality. Real wage series developed by Williamson and Lindert (for urban common laborers), by Adams (for manufacturing workers), and by David and Solar (for common laborers) all show slower growth rates for the final three or four decades of the antebellum period than for either the proceding or the following periods. The patterns observed in the height-by-age data summarized by Fogel, Engerman, and Trussell also are consistent with an increase in inequality from 1820 to 1860, with the heights and wages of common laborers falling relative to those of other groups. See Williamson, Jeffrey G. and Lindert, Peter H., American Inequality: A Macroeconomic History (New York, 1980), chaps. 4–5;Google ScholarAdams, Donald R. Jr, “The Standard of Living During American Industrialization: Evidence from the Brandywine Region, 1800–1860”, this JOURNAL, 42 (12 1982), 903–17;Google ScholarDavid, Paul A. and Solar, Peter, “A Bicentenary Contribution to the History of the Cost of Living in America”, Research in Economic History, Uselding, Paul, ed., Vol. 2 (Greenwich, 1977), Tables 1, 5, and B1;Google Scholar and Fogel, Robert W., et al. , “Exploring the Uses of Data on Height: The Analysis of Long-Term Trends in Nutrition, Labor Welfare, and Labor Productivity”, Social Science History, 6 (Fall 1982), 401–21.Google Scholar

³ For similar interpretations, see Heale, M. J., “Patterns of Benevolence: Charity and Morality in Rural and Urban New York, 1783–1830,” Societas, 3 (Autumn 1973), 337–50;Google ScholarSchneider, David M., The History of Public Welfare in New York State, 1609–1866 (Chicago, 1938), pp. 233–35;Google ScholarKlebaner, Benjamin J., “Poverty and Its Relief in American Thought, 1815–1861,” Social Service Review, 38 (12 1964), 382–99;CrossRefGoogle ScholarRimlinger, Gaston, “Welfare Policy and Economic Development: A Comparative Historical Perspective”, this JOURNAL, 26 (12 1977), 563;Google ScholarMencher, Samuel, Poor Law to Poverty Program: Economic Security in Britain and the United States (Pittsburgh, 1967), pp. 131–47;Google Scholar and Mohl, Raymond A.. Poverty in New York, 1783–1825 (New York, 1971), p. 261.Google Scholar

⁴ Changes in the public relief system might increase relieved pauperism either because qualification standards were relaxed or because the quality or quantity of relief granted increased. Income qualifications were never codified so it is impossible to determine how they might have changed overtime. The state poor law of 1788 did include settlement requirements, and these were simplified when a new poor law was adopted in 1824. This was the only antebellum poor law reform that might have made it easier to qualify for relief. However, the 1824 poor law also made the relief system less generous by prohibiting the provision of outdoor relief and requiring that the poor be relieved in county almshouses. An 1827 revision allowed the provision of temporary outdoor relief, not to exceed ten dollars, and no county stopped granting outdoor relief entirely. However, by 1835, almost every county operated an almshouse and almost 40 percent of all recipients were sent to an almshouse. Surely, the rise of the almshouse represented a qualitative decline in the generosity of relief. Moreover, the benefit-earnings ratio (relief expenditures per full-year recipient divided by the annual earnings of common labor) declined steadily across the antebellum period. See Joan Underhill Hannon, “Poor Relief Policy in Antebellum New York State: The Rise and Decline of the Poorhouse”, Explorations in Economic History (forthcoming);Google Scholar and Hannon, Joan Underhill, “The Generosity of Antebellum Poor Relief,” this JOURNAL, 44 (09 1984), 810–21.Google Scholar

⁵ 1823, 1835, 1849, and 1859 are all years in which the level of business activity was either on or very close to the long-run trend. See Vedder, Richard K., The American Economy in Historical Perspective (Belmont, California, 1976), pp. 316–17.Google Scholar

⁶ See U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 99 (July 1975), pp. 17–18;Google Scholar and Patterson, James, America's Struggle Against Poverty, 1900–1980 (Cambridge, Mass., 1981), p. 171. There are a wide range of public welfare programs encompassed by the term Public Assistance, but it does not include social insurance programs such as unemployment insurance.Google Scholar

⁷ Hannon, “Poor Relief Policy”, Table 5.Google Scholar

⁸ The exceptionally high figure for 1844 and low figure for 1849 given in Figure 2A are somewhat suspicious. Since 1842–1844 were depression years, one would expect some increase in the relieved pauperism rate. But the increase in total pauperism reported for New York City in those years (Figure 2A) was accompanied by a decline in almshouse recipients as a fraction of the population (Figure 2B); and in the rest of the state, the relieved pauperism rate leveled off by 1844. It should be noted, however, that business downturns typically had a greater effect on relieved pauperism in New York City than in the rest of the state and that New York City typically responded by increasing outdoor, rather than almshouse, relief. The 1849 figure, however, is not easily explained and may reflect an error in the published data. The relieved pauperism rate reported for New York City dropped sharply from 13.03 percent of the population in 1847 to only 6.58 percent in 1849, while real per capita relief expenditures reportedly fell by only 6.7 percent. By contrast, the number of relief recipients reported for the rest of the state rose from 2.26 percent of the population in 1847 to 2.95 percent in 1849.Google Scholar

⁹ In another paper I estimated that about 28 percent of New York City's relief recipients were sent to the almshouse in 1724–1729, and about one-third were relieved in the almshouse in 1736–1746. By comparison, 37 percent were relieved in the almshouse in 1823. See Joan Underhill Hannon, “Industrialization, Urbanization, and the Origins of Jacksonian Poor Relief Policy: New York State's 1824 Poor Law Reform”, paper presented to the Economic History Seminar, Indiana University (Nov. 10, 1983), 58–60.Google Scholar

¹⁰ Residents in the almshouse on January 1, 1823, represented 17.97 percent of total recipients in 1823. Almshouse residents on December 1, 1846 were only 7.76 percent of total recipients in 1847, a 10.21 percentage point decline. If the fraction in the almshouse had fallen by the same amount between 1800 and 1823, then the number in the almshouse on January 1, 1800 (874) would have been 28.18 percent of total recipients. The total number of recipients in 1800 would then be 3,101 or 5.12 percent of the city's 1800 population, compared with 6.47 percent in 1823. These estimates imply relatively slow growth in pauperism in New York City over the first quarter of the nineteenth century compared with either the succeeding or preceding period. See Hannon, “Jacksonian Poor Relief”, pp. 10–12;Google ScholarHannon, Joan Underhill, “Contours of Public Pauperism and Poor Relief in Antebellum New York State,” paper presented to the Economic History Workshop, University of Chicago (11 11, 1983), Table Cl;Google Scholar and Nash, Gary B., “Urban Wealth and Poverty in Pre-Revolutionary American,” Journal of Interdisciplinary History, 6 (Spring 1976), 557–60.Google Scholar

¹¹ The pauperism rate outside New York City grew at an average annual rate of 3.8 percent from 1823 to 1835. If growth had occurred at that rate from 1800 to 1823, the pauperism rate outside New York City would have been 0.24 percent in 1800. Applying 1800 population weights implies a statewide pauperism rate of 0.74 percent in 1800, compared with 1.13 percent in 1823.Google Scholar

¹² For a discussion of the origins of this view and its persistence in the historical literature, see Katz, Michael B., Poverty and Policy in American History (New York, 1983), chaps. 2–3.Google Scholar

¹³ The statewide estimates are weighted averages of the figures for New York City and those for the other 249 towns that provided information on the numbers of males, females, adults, and children, where the weights are the fractions of recipients in New York City and in the rest of the state.Google Scholar

¹⁴ Yates provides no source for these estimates. It is possible that the information was included in the individual town reports, but it is not included in the published tables. Yates, John V., New York Secretary of State, “Report of the Secretary of State in 1824 on the Relief and Settlement of the Poor”, reprinted in The Almshouse Experience: Collected Reports (New York, 1971), pp. 1058–59. Hereafter cited as Yates Report.Google Scholar

¹⁵ See also the comments of various local poor relief officials in Yates Report, pp. 980, 987, 1010, 1018–1019, 1022, 1036, 1041, and 1046.Google Scholar

¹⁶ See Schneider, The History of Public Welfare, pp. 317–79; Katz, Poverty and Policy, pp. 72–89; and Rothman, The Discovery of the Asylum, pp. 287–90, for a discussion of attempts to remove children and the physically and mentally handicapped from the almshouses into specialized institutions. All of these authors, however, conclude that almshouses remained the central institutions for the care of the elderly.Google Scholar

¹⁷ It is possible that widows, who may be included among the indigent and destitute, accounted for some of this growth, though in one year (1858) widows and their children were listed separately and accounted for less than I percent of all recipients.Google Scholar

¹⁸ By comparison, Mary McKinnon estimates that adult males (both able- and non-able-bodied) were only 20 percent of all relief recipients in England in 1860s. In the United States in 1980, males made up slightly under 40 percent of all persons below the official poverty line, and males between the ages of 18 and 64 were only 18 percent. McKinnon, Mary, “Moral Hazard and the Poor Law” (unpublished ms., Oxford University, 1983), p. 2;Google Scholar and U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 133 (Washington, D.C., 1982), p. 46.Google Scholar

¹⁹ The statewide estimates, presented in Hannon, “Generosity,” Table 3, are based on the assumption that the 366 reporting towns (excluding New York City) are representative of the state (excluding New York City).Google Scholar

²⁰ Three months may be a high estimate of the average length of occasional relief, but the assumption that permanent recipients remained on the rolls for 52 weeks may introduce an offsetting downward bias since they may have stayed on the rolls for several years. Moreover, other data from the Yates Report suggest that the average length may have been even higher. The average weekly cost of supporting a pauper in the 18 almshouses in the state in 1823 was $0.68. Annual expenditure on relief divided by the number of recipients was $21.91. Dividing annual expenditure per recipient ($21.91) by weekly cost per recipient ($0.68) yields an estimated average length of relief of 32 weeks. See Hannon, “Generosity,” Tables 2–3.Google Scholar

²¹ See New York Secretary of State, “Annual Reports on Statistics of the Poor,” in New York State Legislature, Assembly Documents, 1834, Vol. 3, No. 173, pp. 20–21; 1835, Vol. 3, No. 185, p. 4; and 1837, Vol. 3, No. 270, p. 9.Google Scholar

²² Hannon, “Contours,” p. 42.Google Scholar

²³ See Hannon, “Generosity,” Tables 4–5.Google Scholar

²⁴ These estimates understate the average length of relief throughout the period because recipients who remained for more than one year would be counted for only 52 weeks. Evidence collected from the records of the Erie County poorhouse by Katz indicates that over the period from 1829 to 1844, an average of one-fifth of all poorhouse residents remained for at least one year. Katz's data imply an average length of poorhouse residence in Erie County (again counting those who stayed at least a year as 52 weeks) of about 14 weeks over the period 1829–1844. By comparison, my procedure yields an average length of 13.5 weeks for Erie County, 1840–1844. See Katz, Poverty and Policy, Table A 16.Google Scholar

²⁵ This assumption may bias the result, but it is difficult to say which way the bias would go. Contemporary proponents of the almshouse argued that weekly expenditures would be lower for almshouse than for outdoor relief. While there are a priori reasons for expecting this to be true, there are equally compelling reasons for expecting outdoor relief to have been less expensive. See Hannon, “Contours,” 29–30.Google Scholar

²⁶ Full-year-equivalent recipients for 1823 are estimated by counting each part-year recipient as one-quarter of a full-year equivalent. For 1855–1859, the total number of recipients was multiplied by the average length of relief (expressed as a fraction of a year). The same procedure yields a statewide growth rate for full-year equivalents (as a fraction of the population) of only 0.28 percent per year from 1823 to 1840–1844, compared to a 5.42 percent annual increase in the relieved pauperism rate. In New York City, estimated full-year equivalents declined as a fraction of the population from 1823 to 1840–1844, while in the rest of the state they rose at an annual rate of 1.85 percent.Google Scholar

²⁷ See Heale, “Patterns of Benevolence;” Schneider, History of Public Welfare, pp. 233–35; Kelbaner, “Poverty and Its Relief;” Mencher, Poor Law to Poverty Programs, pp. 131–47; and Mohl, Poverty in New York, pp. 14–34.Google Scholar

²⁸ Katz, Poverty and Policy, pp. 10–14, 200–201.Google Scholar

²⁹ The analysis is limited to 250 of the 367 towns that reported to Yates because the data for the other 117 towns are incomplete. The 250 towns represent 49 percent of the state's population. Their size distribution and labor force composition are almost exactly the same as the state's as a whole. See Hannon, “Jacksonian Poor Relief Policy,” p. 60–61. All equations are estimated with ordinary least squares. A logistic probability model was also estimated where the dependent variable is relieved paupers as a fraction of a town's population, and the explanatory variables are those in Table 3. The standard errors of the estimated coefficients in the logistic model are generally smaller than those obtained for the linear model, but the substantive conclusions remain unchanged.Google Scholar

³⁰ See Hannon, “Jacksonian Poor Relief Policy,” for a discussion of the various relief systems used in New York State in 1823.Google Scholar

³¹ FRACEMP does have a significant negative coefficient in the logistic model for all samples, as do FRACCHILD and FRACOLD. These results are consistent with the evidence presented above that children and elderly adults represented a declining fraction of all recipients as the relieved pauperism rate rose. They are also consistent with some recent work on family and demographic history which suggests that the family did not break down as a result of immigration and urbanization; in fact the family cushioned its members in their adjustment to new environments and unfamiliar living conditions. Hareven, Tamara K., ed., Family and Kin in Urban Communities, 1700–1930 (New York, 1977), p. 8.Google Scholar

³² Immigration appeared on the lists of “ten major causes of pauperism” published by the New York Society for the Prevention of Pauperism in 1818–1821; and it was number one on the list in 1819. See New York Society for the Prevention of Pauperism, Annual Reports, 1818–1821.Google Scholar

³³ The difference resulting from the elimination of one observation is quite striking, but New York City's population was more than 20 times that of the next largest town in the sample and almost 60 times the mean population of the other 249 towns. Thus, it is possible that POP simply is serving as a proxy for New York City which, with a relieved pauperism rate more than 10 times that in the rest of the sample, clearly was qualitatively different. In the logistic model for the 250-town sample, POP remains barely significant at the 10 percent level even when New York City is excluded. FRACNAG has a significant positive coefficient in the logistic model for all samples except the 40-town sample excluding New York City, but its quantitative impact remains small. The largest estimated coefficient on FRACNAG in the logistic model (250-town sample, reduced form) implies that a 10 percentage point increase in the non-agricultural labor force share would increase the relieved pauperism rate by only 0.12 percentage points.Google Scholar

³⁴ An appendix containing data on regional characteristics is available from the author on request.Google Scholar

³⁵ In most regions the most rapid growth in relieved pauperism preceded the period of most rapid urbanization, immigration, and industrialization. The most rapid period of urbanization in New York came in the 1840s to 1860s, the growth in the nonagricultural share of the labor force accelerated to a peak in the 1840s, and immigration peaked in 1846–1855. See Williamson, Jeffrey G., “Urbanization in the American Northeast, 1820–1870,” in Fogel, Robert and Engerman, Stanley, eds., The Reinterpretation of American Economic History (New York, 1971), p. 429;Google ScholarLebergott, Stanley, Manpower in Economic Growth (New York, 1964), Table Al;Google Scholar and Easterlin, Richard A., Population, Labor Force, and Long Swings in Economic Growth (New York, 1968), Table B2.Google Scholar

³⁶ Since PCTNAG is not available for 1860, industrialization in the later period is measured by IND (regional share of the state's manufacturing labor force divided by regional share of the state's population). The difference between the results for 1840–1844 and those for 1855–1859 does not seem to be attributable to this shift in variables. Using IND for 1840–1844 does not change the results for that period.Google Scholar

³⁷ Regression analysis in which the same explanatory variables used in Table 3 are related to permanent recipients as a percent of the population, occasional recipients as a percent of the population, adult recipients as a percent of the adult population, child recipients as a percent of the child population, male recipients as a percent of the male population, and female recipients as a percent of the female population in 1823 confirm this conclusion. When New York City is included, POP has a larger positive impact on occasional pauperism than on permanent pauperism, and its impact on adult, child, male, and female pauperism rates are about equal. But when New York City is excluded, POP does not appear as a significant variable in any of the equations. FRACNAG is significant and positive in the male, but not in the female pauperism equation; and its positive impact on adult pauperism is much larger than its impact on child pauperism. FRACNAG has a significant positive impact on permanent, but not on occasional pauperism. See Hannon, “Jacksonian Poor Relief Policy,” Tables 5–6 and pp. 19–21.Google Scholar

³⁸ Michael Reich estimates that wage labor increased from about 20 percent of the U.S. labor force in 1780 to 62 percent by 1880, but neither the timing of nor cross-sectional patterns in growth have been documented. See Edwards, Richard, Reich, Michael, and Weisskopf, Thomas, The Capitalist System, (2nd ed; Englewood Cliffs, N.J., 1978), p. 180.Google Scholar

³⁹ It was also negatively correlated with industrialization. However, the rank correlation coefficient between per capita household production in 1821 and PCTNAG in 1820 in New York State (-0.69) was barely significant at the 5 percent level, and excluding New York City, the correlation coefficient (-0.54) was not significant. Similarly the rank correlation coefficient between IND in 1860 and per capita household production in 1855 was significant only with New York City included. The only period in which PCTNAG and household production were strongly and significantly correlated even without New York City was the 1840s, when the correlation coefficient with New York City was -0.90 and excluding New York City, -0.86.Google Scholar

⁴⁰ The following draws heavily on McNall, Neil A., An Agricultural History of the Genesee Valley, 1790–1860 (Philadelphia, 1952).CrossRefGoogle Scholar

⁴¹ Rochester Daily Democrat, May 26, 1857, cited in McNall, An Agricultural History, p. 229.Google Scholar

⁴² See Mohi, Poverty in New York, pp. 28–29; Yates Report, p. 1010; Katz, Poverty and Policy, pp. 11, 45–46, 83;Google ScholarMontgomery, David, “The Working Classes of the Pre-Industrial American City, 1780–1830,” Labor History, 9 (Winter 1968), 3–22;Google ScholarWare, Norman, The Industrial Worker, 1840–1860 (New York, 1964), pp. 26–70.Google Scholar