From ‘20. and odd’ to 10 million: the growth of the slave population in the United States

ABSTRACT

This research note describes the growth of the slave population in the United States and develops several new measures of its size and growth, including an estimate of the total number of slaves who ever lived in the United States. Estimates of the number of births and slave imports are provided in ten-year increments between 1619 and 1860 and in one-year increments between 1861 and 1865. The results highlight the importance of natural increase to the rapid growth of the U.S. slave population and indicate that approximately 10 million slaves lived in the United States, where they contributed 410 billion hours of labor. A concluding discussion highlights a few descriptive statistics historians might find useful, including the cumulative number of slaves who lived in the United States by decade and the proportion of slaves who were living at various moments in U.S. history, including shortly after the ratification of the Constitution in 1788 and at the start of the American Civil War in 1861.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Acknowledgments

The author would like to thank Adam Rothman and USA Today graphic editor James Sergent for suggesting this study, Clifton Berry at the Unpaid Labor Project for encouragement, Amy Murrell Taylor, Alex Borucki, David Eltis and Gregory O’Malley for assistance, and Saje Mathieu for comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes on contributor

J. David Hacker is in the Department of History and Minnesota Population Center, University of Minnesota, Twin Cities, MN, USA. Email: [email protected]

Notes

1 The origins of African slavery In British North America may have been earlier than the arrival of the White Lion. A census of Virginia in March 1619 counted 32 individuals of African descent. See William Thorndale, ‘The Virginia Census of 1619’, Magazine of Virginia Genealogy 33 (1995): 155–70; and John Thornton, ‘The African Experience of the “20. and Odd Negroes” Arriving in Virginia in 1619’, The William and Mary Quarterly 55, no. 3 (1998): 421–34. The origins of African (and Indian) slavery are even earlier in parts of Spanish territory that eventually became part of the United States. A small number of African slaves lived in St. Augustine, Florida from the beginning of its settlement in 1565. Paul E. Hoffman, Florida’s Frontiers (Bloomington, University of Indiana Press, 2002).

2 David Eltis, ‘The U.S. Transatlantic Slave Trade, 1644–1867: An Assessment’, Civil War History 54, no. 4 (2008): 347–78, and Stanley L. Engerman et al., ‘Slavery’, in Historical Statistics of the United States, Earliest Times to the Present, Volume 2: Part B, Work and Welfare, eds. Susan B. Carter et al. (Cambridge: Cambridge University Press, 2006), 369–74.

3 Good summaries of slave demography in British North America and the United States include Lorena S. Walsh, ‘The African American Population of the Colonial United States’, in A Population History of North America, eds. Michael R. Haines and Richard H. Steckel (Cambridge: Cambridge University Press, 2000), 191–239; Richard H. Steckel, ‘The African American Population of the United States, 1790–1920’, in A Population History of North America, eds. Michael R. Haines and Richard H. Steckel (Cambridge: Cambridge University Press, 2000), 433–81; John J. McCusker and Russell R. Menard, The Economy of British North America (Chapel Hill: University of North Carolina Press, 1991); and Michael R. Haines, ‘The Population of the United States, 1790–1920’, in The Cambridge Economic History of the United States, Volume II: The Long Nineteenth Century, eds. Stanley L. Engerman and Robert E. Gallman (Cambridge: Cambridge University Press, 2000), 143–205. On the recent debate on the impact of slavery on capitalism and economic growth in the United States, see Sven Beckert, Empire of Cotton: A Global History (New York: Vintage Books, 2014); Edward E. Baptist, The Half Has Never Been Told: Slavery and the Making of American Capitalism (New York, 2014); Alan L. Olmstead and Paul W. Rhode, ‘Cotton, slavery, and the new history of capitalism’, Explorations in Economic History 67 (2018): 1–17; and Gavin Wright, ‘Slavery and Anglo-American capitalism revisited’, Economic History Review (2020): 1–31. For a recent estimate of the present value of slave labor with regards to possible reparations, see Thomas Craemer, ‘Estimating Slavery Reparations: Present Value Comparisons of Historical Multigenerational Reparations Policies’, Social Science Quarterly 96, no. 2 (2015): 639–55.

4 The graphic can be found at https://www.usatoday.com/pages/interactives/1619-african-slavery-history-maps-routes-interactive-graphic/?utm_campaign=s2seriesrecirc&utm_content=news&utm_medium=onsite&utm_source=oembed.

5 The estimated number of slaves born in the United States, for example, is robust to a wide range of possible birth rates in the seventeenth century. Under most reasonable assumptions, less than 0.3 percent of all slave births occurred prior to 1700.

6 John J. McCusker, ‘Population’, in Historical Statistics of the United States, Earliest Times to the Present, Volume 5: Part E, Governance and International Relations, eds. Susan B. Carter et al. (Cambridge: Cambridge University Press, 2006), Series Eg41, 653. The size of the free black population was first measured for the nation in 1790, when it was 7.87 percent of the total black population. An early census count for Maryland recorded 4 percent of the colony's blacks as free in 1755, but the result is surely unrepresentative of other colonies with large numbers of slaves. I assumed that 2.0 percent of blacks between 1620 and 1770 were free, increasing to 4.93 percent in 1780 (halfway between the assumed 2.0 percent for 1770 and the 7.87 percent measured by the 1790 census). See Ira Berlin, Slaves Without Masters: The Free Negro in the Antebellum South (New York: Vintage Books, 1974), 3–4 for the Maryland estimates and comments on the low numbers of free blacks in the colonial period.

7 Among other places, census counts can be found in Susan B. Carter, ‘Slave Population,’ in Historical Statistics of the United States, Earliest Times to the Present, Volume 2: Part B, Work and Welfare, eds. Susan B. Carter et al. (Cambridge: Cambridge University Press): Table Bb1-98, 375–9. For adjustments to the overall size of the black population for territories not covered by the census, see Peter D. McClelland and Richard J. Zeckhauser, Demographic Dimensions of the New Republic: American Interregional Migration, Vital Statistics, and Manumissions, 1800–1860 (Cambridge: Cambridge University Press, 1982): 120.

8 I assumed the first 375,000 of these emancipated slaves were evenly distributed between June 1, 1861 and May 30, 1864, with the final increment of 125,000 emancipations occurring between June 1, 1864 and April 9, 1865, consistent with research describing emancipation as a slow process encompassing the entire course of the war and estimates of the number of former slaves living in ‘contraband’ camps or taking part in some form of federally-sponsored free labor in Union-occupied areas of the South in the spring of 1865. Ira Berlin, Barbara J. Fields et al., Slaves No More: Three Essays on Emancipation and the Civil War (Cambridge: University of Cambridge, 1992), 182–3.

9 If P_t is equal to the size of the slave population on at the beginning of the decade and P_t+10 equals the size of the slave population at the end of the decade, the number of person-years, assuming constant population growth = (P_t+10−P_t)/ln((P_t+10 /P_t)/10).

10 On average, 81.7 percent of the slave population in 1850 and 1860 was age 15–64. Over the censuses when the age structure can be observed (1820–1860), it appears to be relatively constant. For the typical slave workweek of 54 hours, see John F. Olson, ‘Clock-Time vs. Real-Time: A Comparison of the Lengths of the Northern and the Southern Agricultural Work Years’, in Without Consent or Contract: The Rise and Fall of American Slavery, Technical Papers, Vol. 1, eds. Robert W. Fogel and Stanley L. Engerman (New York: Norton, 1992), 216–40. The estimated millions of hours worked in each interval = person-years lived * 0.817 *54 * 52 / 1,000,000.

11 Melvin Zelnik, ‘Fertility of the American Negro in 1830 and 1850’, Population Studies 20, no. 1 (1966): 77–83; Reynolds Farley, Growth of the Black Population: A Study of Demographic Trends (Chicago: Markham Publishing Company, 1970); Jack Ericson Eblen, ‘New Estimates of the Vital Rates of the United States Black Population during the Nineteenth Century’, Demography 11, no. 2 (1974): 301–19; Peter D. McClelland and Richard J. Zeckhauser, Demographic Dimensions of the New Republic: American Interregional Migration, Vital Statistics, and Manumissions, 1800–1860 (Cambridge: Cambridge University Press, 1982), Series D23, 184; Michael R. Haines, ‘The Population of the United States, 1790–1920’, in The Cambridge Economic History of the United States, Volume II: The Long Nineteenth Century, eds. Stanley L. Engerman and Robert E. Gallman (Cambridge: Cambridge University Press, 2000), 158. All crude birth rate estimates cited above are for the black population, not the slave population. Given that the free black population was, at its largest, just over 10 percent of the black population, fertility differences would have to be very large before biasing the overall results. Stable population estimates are based in part on an assumed mortality schedule or life table. Until recently, the only available life table was Paul H. Jacobson's 1850 life table for Massachusetts and Maryland, which was based on relatively few slave decedents. If Richard Steckel is correct that age pattern of slave mortality did not conform to the age pattern of white mortality—compared to the white population, slave infant and child mortality rates were elevated relative to adult slave mortality rates—birth rates derived using stable population methods and Jacobson's life table may be too low. Given that the estimated nineteenth-century birth rates are already near 60 births per thousand, however, it is difficult to see how much higher slave birth rates could have been. Richard H. Steckel, ‘Birth Weights and Infant Mortality among American Slaves’, Explorations in Economic History 23, no. 2 (1986): 173–98.

12 Russell R. Menard, ‘The Maryland Slave Population, 1658 to 1730: A Demographic Profile of Blacks in Four Counties’, The William and Mary Quarterly 32, no. 1 (1975): 29–54.

13 Allan Kulikoff, ‘A “Prolifick” People: Black Population Growth in the Chesapeake Colonies, 1700–1790’, Southern Studies 16, no. 4 (1977): 391–428.

14 Philip D. Morgan, Slave Counterpoint: Black Culture in the Eighteeenth-Century Cheseapeake and Lowcountry (Chapel Hill: Omohundro Institute of Early American History and Culture, Williamsburg, Virginia, by the University of North Carolina Press, 1998), 79–95.

15 On slave owner pressures and incentives for childbearing, see Marie Jenkins Schwartz, Born in Bondage: Growing Up Enslaved in the Antebellum South (Cambridge, MA: Harvard University Press, 2000), 191.

16 I used estimated regional slave populations estimated by McCusker and Menard to weight the regional child-woman ratios synthesized by Morgan. McCusker and Menard, The Economy of British North America. Morgan, Slave Counterpoint, 83. The 1850 census tabulated the number of slave children in sufficiently detailed age categories to create a similar ratio of slave children age 0–14 to slave mothers age 15–44 used by the colonial historians. Although trends in child-woman ratios are typically dominated by changes in fertility rather than changes in child mortality, Menard concluded that ‘both were probably at work’ in the seventeenth and early eighteenth century. I assumed that two-thirds of the trend in child-woman ratios observed by Menard prior to 1730 was the result of trends in slave fertility and one-third was the result of trends in child mortality and that trends in child-woman ratios in the eighteenth century accurately reflected trends in the birth rate.

17 See also, Herbert S. Klein and Stanley L. Engerman, ‘Fertility Differentials between Slaves in the United States and the British West Indies: A Note on Lactation Practices and Their Possible Implications’, The William and Mary Quarterly 35, no. 2 (1978): 357–74 and Philip D. Morgan and Michael L. Nicholls, ‘Slaves in Piedmont Virginia, 1720–1790’, The William and Mary Quarterly 46, no. 2 (1989): 211–51.

18 A description of the Transatlantic Slave Trade Database can be found in David Eltis and David Richardson, ‘A New Assessment of the Transatlantic Slave Trade’, in Extending the Frontiers: Essays of the New Transatlantic Slave Trade Database, eds. David Eltis and David Richardson (New Haven: Yale University Press), 1–60. Details on the 2010 estimates made by Eltis and Lachance can be found in David Eltis and Paul F. Lachance, ‘Estimates of the Size and Direction of the Transatlantic Slave Trade,’ (2009), available at https://slavevoyages.org/documents/download/2010estimates-method.pdf. For the Intra-American slave trade and the published estimate of the number of slaves in each decade, see Gregory E. O'Malley, Final Passages: The Intercolonial Slave Trade of British America, 1619–1807 (Chapel Hill: Omohundro Institute of Early American History and Culture, Williamsburg, Virginia, by the University of North Carolina Press, 2014), 351–81.

19 Steven Ruggles, Sarah Flood, Ronald Goeken, Josiah Grover, Erin Meyer, Jose Pacas and Matthew Sobek. IPUMS USA: Version 9.0 [dataset]. Minneapolis, MN: IPUMS, 2019. https://doi.org/10.18128/D010.V9.0.

20 In 1860, the schooner Clotilda arrived in Mobile Bay with approximately 110 smuggled African slaves on board, the last known smuggled slaves to the U.S. Sylviane A. Diouf, Dreams of Africa in Alabama: The Slave Ship Clotilda and the Story of the Last Africans Brought to America (New York: Oxford University Press, 2007).

21 For the age distribution of slaves on captured vessels and estimates of slave smuggling, see McClelland and Zeckhauser, Demographic Dimensions, 48, 123. I assumed 50 percent of smuggled slaves died within 10 years of their arrival in the U.S. and that slaves who survived the first 10 years subsequently experienced mortality conditions equivalent to a Model West life table with life expectancy at age 35 of 24.6 years, a level suggested by application of the two-census mortality estimation method of the black population enumerated in the 1850 and 1860 censuses. See McClelland and Zeckhauser, Demographic Dimensions, 123 for the wide range of slave smuggling estimates.

22 Beyond a few thousand former slaves who found their way to Canada after the Revolution and in the 1840s and 1850s, there were likely few out-migrants from the U.S. slave population that escaped measurement in Table 1. Fugitives who successfully escaped slavery and slaves who were manumitted by their owners or who purchased their own freedom and lived in the United States, however, are ‘out-migrants’ from the slave population in a figurative sense and were likely significant enough in number to bias the natural increase rate. For the period 1800–1860, I relied on McClelland and Zeckhauser's estimates of ‘manumissions’ to adjust the estimated natural increase rate. McCelland and Zeckhauser, Demographic Dimensions, 19. For the period 1770–1800, I made similar estimates, rounded to the nearest 5,000, and assumed the level of manumissions and fugitives was negligible in all previous decades.

23 For each decade before 1750, I averaged the rate of natural increase for the current decade, the two prior decades, and the two following decades. For decades between 1750 and 1800, I averaged the prior, current, and following decades.

24 J. David Hacker, ‘New Estimates of Census Coverage in the United States, 1850–1930’, Social Science History 37, no. 1 (2013): 71–101.

Additional information

Funding

The author gratefully acknowledges support from the Minnesota Population Center (P2CHD041023) funded through a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). This project was also supported by a research award from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01-HD082120-01).