This paper draws upon joint work with Robert W. Fogel financed under National Science Foundation Grants GS-27262 and GS-3262. Needless to say, many of the points discussed here reflect this work and the discussions we have had over the course of several years. In writing this paper I have also benefited from frequent discussions with Sherwin Rosen, as well as from suggestions made by Walter Oi, Claudia Goldin, Eugene Genovese, and Nathan Rosenberg.

¹ For a presentation of the philosophical problem of reconciling slavery with recognition of the humanity of the slave, ranging across continents and centuries, see Davis, David Brion, The Problem of Slavery in Western Culture (Ithaca: Cornell University Press, 1966).Google Scholar For a discussion of this problem in the antebellum South, see Sio, Arnold A., “Interpretations of Slavery: The Slave Status in the Americas,” Comparative Studies in Society and History, VII (1965), pp. 289–308.CrossRefGoogle Scholar

² Between 1856 and 1860, six states passed legislation permitting free Negroes to become slaves voluntarily. See Gray, Lewis Cecil, History of Agriculture in the Southern United States to 1860, Vol. I (Gloucester, Massachusetts: Peter Smith, 1958), p. 527.Google Scholar See also the reference in footnote 42. While slavery is widely accepted as being an involuntarily achieved status (although there were cases of voluntary entry and sales of children in ancient and medieval Europe), other forms of what are sometimes called “forced labor” are the result of voluntary agreement. Recently economic historians have reopened the discussion of whether European serfdom represented a voluntary exchange—protection for labor services—or whether it was a form of forced labor imposed from above. See, in particular, North, Douglass C. and Thomas, Robert Paul, “The Rise and Fall of the Manorial System: A Theoretical Model,” The Journal of Economic History, XXXI (1971), pp. 777–803CrossRefGoogle Scholar; SirHicks, John, A Theory of Economic History (New York: Oxford University Press, 1969), Ch. VIIGoogle Scholar; and Gerschenkron, Alexander, “Mercator Gloriosus,” The Economic History Review, XXIV (1971), pp. 653–666.CrossRefGoogle Scholar There seems agreement that serfdom in Russia and eastern Europe came to more closely resemble slavery than it did in western Europe; thus, the equation of slavery and serfdom in Domar, Evsey D., “The Causes of Slavery or Serfdom: A Hypothesis,” The Journal of Economic History, XXX (1970), pp. 18–32.CrossRefGoogle Scholar

³ For recent presentations of the arguments of a most outspoken southerner on this issue, see the discussions of George Fitzhugh in Genovese, Eugene D., The World the Slaveholders Made (New York: Pantheon Books, 1969)Google Scholar, Two, Part, and Woodward, C. Vann, American Counterpoint: Slavery and Racism in the North-South Dialogue (Boston: Little, Brown and Company, 1971), Ch. 4.Google Scholar

⁴ For a discussion of other forms of involuntary labor than slavery, see the worldwide survey by Kloosterboer, W., Involuntary Labour Since the Abolition of Slavery (Leiden, Netherlands: E. J. Brill, 1960).Google Scholar In the preface (p. 1) the author remarks: “It is indeed not clear why precisely slavery should exist if there is not sufficient voluntary labour. As labour systems, serfage, debt-slavery, and even contract labour under penal sanction fulfil exactly the same function. The form compulsory labour takes will in the first place depend on the spirit of the times.” Kloosterboer's survey was, in part, intended to modify the conclusions of H. J. Nieboer on the occurrence or non-occurrence of slavery in different societies. Nieboer's analysis has recently been utilized by Domar, and will be discussed below.

⁵ See the discussions in Gray, History of Agriculture, Ch. XX, and Phillips, Ulrich B., American Negro Slavery (Baton Rouge: Louisiana State University Press, 1966)Google Scholar, Ch. XVIII, which summarize the writings of economists and others on the relative efficiency of slave labor.

⁶ Rather than defend these latter propositions here I refer the reader to Fogel, Robert W. and Engerman, Stanley L., “The Economics of Slavery,” in Fogel, Robert W. and Engerman, Stanley L., editors, The Reinterpretation of American Economic History (New York: Harper and Row, 1971) pp. 311–341Google Scholar; and the same authors', “The Relative Efficiency of Slavery: A Comparison of Northern and Southern Agriculture in 1860,” Explorations in Economic History, VIII (1971), pp. 353–367.Google Scholar For a collection of articles dealing with arguments on profitability, efficiency, and growth, see Aitken, Hugh G. J., editor, Did Slavery Pay?: Readings in the Economics of Black Slavery in the United States (Boston: Houghton Mifflin, 1971).Google Scholar

⁷ See the path-breaking article by Conrad and Meyer, reprinted in Conrad, Alfred H. and Meyer, John R., The Economics of Slavery and Other Studies in Econometric History (Chicago: Aldine Publishing Company, 1964)Google Scholar, Ch. 3, and the evaluation of the subsequent debate in Fogel and Engerman, “The Economics of Slavery,” pp. 311–341.

⁸ For a discussion of these points, and the presentation of the typical age-price and age-net income profiles for slaves, see Robert W. Fogel and Stanley L. Engerman, “The Market Evaluation of Human Capital: The Case of Slavery,” (unpublished, presented at Twelfth Annual Cliometrics Conference, Madison, Wisconsin, April 1972).

⁹ Note the argument by D. R. Goodloe that “slavery merely serves to appropriate the wages of labor—it distributes wealth, but cannot create it.” Quoted in Phillips, American Negro Slavery, p. 349.

¹⁰ On the question of “productive wage increases,” see Becker, Gary S., Human Capital: A Theoretical and Empirical Analysis, With Special Reference to Education (New York: Columbia University Press, 1964), pp. 33–36.Google Scholar

¹¹ For a discussion of human vs. non-human capital, see Friedman, Milton, Price Theory: A Provisional Text (Chicago: Aldine Publishing Company, 1962)Google Scholar, Ch. 11. In the non-human capital case, given that the asset is returned in equivalent condition, its owner will not be concerned with the location, etc., in which the renter uses the asset.

¹² On indentured servants and convict labor in America, see Smith, Abbot Emerson, Colonists in Bondage: White Servitude and Convict Labor in America, 1607–1776 (New York: W. W. Norton & Company, 1971).Google Scholar For illuminating discussions of the move to slave labor see Gray, History of Agriculture, Ch. XVI; Jordan, Winthrop D., White Over Black: American Attitudes Toward the Negro, 1550–1812 (Baltimore: Penguin Books, 1969)Google Scholar, Ch. II, and Morgan, Edmund S., “Slavery and Freedom: The American Paradox,” The Journal of American History, LIX (1972), pp. 5–29.CrossRefGoogle Scholar The debate over whether racism led to enslavement of blacks or enslavement led to racism continues; Jordan stresses their interaction and mutual reinforcement. That there was some initial difference in attitude toward blacks and toward whites is seen both in the fact that only the former were enslaved, and also the particular allocation of work among the sexes. While there has been much written on the imbalance of the sex ratio of slave imports throughout the Americas, apparently that for indentures was even more unbalanced. Black females did field work; white females apparently did not. On this, see, in particular, Morgan, “Slavery and Freedom,” pp. 26–27.

¹³ Engerman, Stanley L., “Some Economic Factors in Southern Backwardness in the Nineteenth Century,” in Kain, John F. and Meyer, John R., editors, Essays in Regional Economics (Cambridge: Harvard University Press, 1971), pp. 279–306Google Scholar

¹⁴ A quote from The Political Economy of Slavery, taken from Phillips, American Negro Slavery, p. 352.

¹⁵ An additional economy permitted by the consolidation of households was the greater likelihood of the slave pairing within the plantation, reducing costs of visiting as well as enhancing morale. This point has been suggested by George Stigler and Richard Steckel. In our forthcoming book Fogel and I present evidence showing a greater stability of the slave family and a somewhat more restrained attitude toward sexual activity than has heretofore been commonly held. On these points, see also the important forthcoming work of Herbert Gutman on the black family during slavery and afterwards.

¹⁶ J. Frederick Dewhurst & Associates, America's Needs and Resources (New York: The Twentieth Century Fund, 1947), p. 695.Google Scholar

¹⁷ This argument from The Political Economy of Slavery is quoted from McKitrick, Eric L., editor, Slavery Defended: The Views of the Old South (Englewood Cliffs: Prentice-Hall, 1963), pp. 71–73.Google Scholar It is presumed that the marginal product from additional hours worked was positive and in excess of the costs to the slaveowner.

¹⁸ Olmsted had raised the question of why there were no large plantations based upon free labor, since he seemed to accept the economic advantages of “associated labor.” Among his suggestions to improve southern agriculture was “especially they [the whites] should become, or should desire to become, richer, more comfortable, than they are.” Olmsted, Frederick Law, The Cotton Kingdom, Vol. II (New York: Mason Brothers, 1861), pp. 234, 264–271.Google Scholar Some historians see this lack of desire for more income as reflecting a demoralization of white labor by the presence of unskilled black workers and a wealthy plantation class.

¹⁹ Black rates of internal migration fell relative to those of whites in the postbellum period, while measures of health and life expectancy show deterioration in the first quarter-century after emancipation. Taste differences include differences in the rates of time preference affecting the desire to invest in human capital, while the slaveowners found it easier to borrow on human capital than did free workers, reducing the costs of such investments. Although medical care under slavery might have been less than desired by the slaves since the financial cost was zero, it might have exceeded the amounts they would have spent from a comparable free labor income. Interregional migration often took the form of movement or plantations, and of family units even if entire plantations did not move, and thus did not cause the breaking up of slave families. The case for systematic slave-breeding for sale to the west is most dubious. See Fogel and Engerman, “The Market Evaluation.”

²⁰ The pattern of intergenerational transfers will differ, affecting not only levels of investment but also the persistence of status differentials over time. In a slave society, with the costs of childcare being provided out of a pool rather than parents’ income, the level and distribution of such expenditures will differ from that in the free labor society.

²¹ For a discussion of the Soviet use of forced labor, see Swianiewicz, S., Forced Labour and Economic Development (New York: Oxford University Press, 1965).Google Scholar

²² This distinction no doubt underlies much of the discussion of exploitation in the debate over the standard of living during the Industrial Revolution. For a somewhat similar issue made about late nineteenth-century imperialism in Africa, see Landes, David, “Some Thoughts on the Nature of Economic Imperialism,” The Journal of Economic History, XXI (1961), pp. 496–512.CrossRefGoogle Scholar The legislation (restricted land settlement, etc.) amounted to a tax which forced people to work for others, or else moved them down a backward-bending labor supply curve, again increasing the availability of labor.

²³ See the survey in Kloosterboer, Involuntary Labour.

²⁴ For a critical discussion of Adam Smith's view on the relationship between slavery and innovation, see Rosenberg, Nathan, “Adam Smith on the Division of Labour: Two Views or One?,” Economica, XXXII (1965), pp. 127–139.CrossRefGoogle Scholar

²⁵ Domar, “Causes of Slavery.” See also. North, Douglass C. and Thomas, Robert Paul, “An Economic Theory of the Growth of the Western World,” The Economic History Review, XXIII (1970), pp. 1–17CrossRefGoogle Scholar, and Hicks, A Theory.

²⁶ Nieboer, H. J., Slavery as an Industrial System (New York: Burt Franklin, 1971).Google Scholar

²⁷ On the importance of restricting the alternatives open to the labor force, see Evans, Robert Jr, “Some Notes on Coerced Labor,” The Journal of Economic History, XXX (1970), pp. 861–866.CrossRefGoogle Scholar While slavery does not eliminate the movement of labor, the effect of any increased demand for labor remains within the owning class and does not go to the worker. Slavery, which permits labor mobility, can thus be more efficient than the stereotyped version of serfdom in which the worker is tied to the land.

²⁸ Georgescu-Roegen has a model of feudalism, based upon Eastern European experience, in which feudalism solves the distribution problem in an overpopulated economy. The lords are sacrificing income to maintain the population, perhaps for military service, and the induced onset of a free labor system leads to starvation among the peasantry. See Georgescu-Roegen, N., “Economic Theory and Agrarian Economics,” Oxford Economic Tapers, XII (1960), pp. 1–40.Google Scholar Whatever its relevance to the cases considered, it seems doubtful that this model can be applied to many other cases of agrarian labor control.

²⁹ See the sources cited in footnote 5. See also the essay by Thompson, Edgar T., “The Natural History of Agricultural Labor in the South,” in Jackson, David D., editor, American Studies in Honor of William Kenneth Boyd (Durham: Duke University Press, 1940).Google Scholar

³⁰ Tucker, George, Progress of the United States in Population and Wealth in Fifty Years as Exhibited by the Decennial Census from 1790 to 1840 (New York: Press of Hunt's Merchant's Magazine, 1855), Ch. XIII.Google Scholar For a discussion of many related points of southern population analysis, see Spengler, Joseph, “Population Theory in the Ante- Bellum South,” Journal of Southern History, II (1936), pp. 360–389.CrossRefGoogle Scholar

³¹ “The scarcity of hands occasions a competition among masters, who bid against one another, in order to get workmen, and thus voluntarily break through the natural combination of masters not to raise wages.” Smith, Adam, The Wealth of Nations (New York: Random House, 1937), p. 68.Google Scholar

³² For a summary of recent literature on the decline of serfdom in England, see Hilton, R. H., The Decline of Serfdom in Medieval England (New York: St. Martin's Press, 1969).CrossRefGoogle Scholar

³³ While central government provides centralized law enforcement, the rise of a central power has ambiguous effects upon labor control. Blum points out that in Western Europe centralization occurred with increased labor freedom, with the opposite being the case in Eastern Europe and Russia. The distinction is that in the latter case the central power was more dependent upon landowners, and thus provided ‘labor grants” to maintain control. Blum, Jerome, “The Rise of Serfdom in Eastern Europe,” The American Historical Review, LXII (1957), pp. 807–836.CrossRefGoogle Scholar There is some analogy in the discussion of slave treatment in the Americas, where it has been argued that the greater local control in the English colonies made for harsher treatment than did the stronger control exercised by the Iberian governments.

³⁴ For a description of the English Vagrancy Act of 1547 which permitted enslavement for limited periods of time, and its rather quick failure, see Davies, C. S. L., “Slavery and Protector Somerset; the Vagrancy Act of 1547,” The Economic History Review, XIX (1966), pp. 533–549.CrossRefGoogle Scholar On the different attitudes toward black and white, see Jordan, White Over Black.

³⁵ The presumed drying-up of “sources of supply” as an explanation for the disappearance of slavery is not fully satisfactory. It could explain a shift in the proportion or slaves, if reproduction rates for slaves were below those of free laborers, out unless the reproduction rates were zero and the slaves die out, the incentive to end the system is unclear.

³⁶ This argument has been used to explain why slavery ended in the northern United States at the start of the nineteenth century. The persistence of positive slave prices, and the infrequency of voluntary manumissions before legislation, suggest that more than costs to slaveowners were involved. See Zilversmit, Arthur, The First Emancipation: The Abolition of Slavery in the North (Chicago: University of Chicago Press, 1967).Google Scholar On the role of supervision costs in the choice between slavery and serfdom, see North and Thomas, “Manorial System,” p. 779.

³⁷ See Bergstrom, Theodore, “On the Existence and Optimality of Competitive Equilibrium for a Slave Economy,” The Review of Economic Studies, XXXVIII (1971), pp. 23–36.CrossRefGoogle Scholar

³⁸ See, for example, Richard Sutch, “The Profitability of Ante-Bellum Slavery-Revisited,” reprinted in Aitken, Did Slavery Pay?, pp. 221–241.

³⁹ Kloosterboer, Involuntary Labour, p. 1. On the decline of Roman slavery, see Westermann, William L., The Slave Systems of Greek and Roman Antiquity (Philadelphia: The American Philosophical Society, 1955).Google Scholar I must admit to some difficulty in interpreting this case. Bloch's description of the end of slavery in the early Middle Ages seems to be that a decline in entrepreneurial and administrative talent led to the break-up of large farms and a shift to small farms, with the labor force left to live in their own huts and owing services to the lord, whether or not legally freed. The decline in administrative ability, not a “drying-up” of supply sources, seems to be the crucial determinant in this case. See Bloch, Marc, “The Rise of Dependent Cultivation and Seigniorial Institutions,” in Clapham, J. H. and Power, Eileen, editors, The Cambridge Economic History of Europe, Vol. I. The Agrarian Life of the Middle Ages (Cambridge: University Press, 1941), Ch. VI.Google Scholar

⁴⁰ See the comment by John E. Moes, reprinted in Conrad and Meyer, The Economics of Slavery, pp. 99–107, and the response by Conrad and Meyer, pp. 107–109.

⁴¹ Aimes, Hubert H. S., “Coartación: A Spanish Institution for the Advancement of Slaves into Freedmen,” Yale Review, XVII (1909), pp. 412–431.Google Scholar He estimates that even in the 1850's, at its peak, annually only six-tenths of one percent of the slaves could have achieved freedom via this mechanism. For the U.S. in 1860, the ratio of all manumissions to the slave population was about one-eighth that amount, and selfmanumission trivial. On Brazilian manumissions, see Klein, Herbert S., “The Colored Freedmen in Brazilian Slave Society,” The Journal of Social History III (1969), pp. 30–52.CrossRefGoogle Scholar

⁴² For a discussion of the reverse movement, voluntary reenslavement, at the end of the antebellum period in one state, see Franklin, John Hope, The Free Negro in North Carolina, 1790–1860 (New York: W. W. Norton & Company, 1971), pp. 218–220.Google Scholar

⁴³ Does the initial allocation of property rights in man effect the economic outcome? Harold Demsetz has argued that this should not occur if all costs were internalized by property rights. See Demsetz, Harold, “Toward a Theory of Property Rights,” The American Economic Review, LVII (1967), pp. 347–359.Google Scholar Two points suggesting the opposite can be made. First, the initial allocation might affect the attitudes of the population, as the debate concerning the relationship between enslavement and prejudice suggests. Second, the Demsetz proposition depends upon demand being independent of wealth distributions. While this may have appeal applied to some commodities, my demand for my freedom will probably differ from the reader's demand for my freedom, so the wealth redistribution affects the pattern of demand. The slave will probably not be able to borrow capital to finance his purchase of his freedom without, in effect, repledging himself to his creditors.

⁴⁴ For a description of the problem created by this overthrow, and the attempts by the new rulers to reintroduce forced labor, see Kloosterboer, Involuntary Labour, Ch. 11. In the Danish West Indies a general strike by blacks brought a quicker end to the system, the insurrection occurring at a time when the government was discussing gradual abolition with compensation. See Genovese, Slaveholders, pp. 38–39.

⁴⁵ See Blum, Jerome, Lord and Peasant in Russia From the Ninth to the Nineteenth Century (New York: Atheneum, 1964), Ch. 26 and 27Google Scholar, and A. Gerschenkron, “Agrarian Policies and Industrialization: Russia 1861–1917,” in Habakkuk, H. J. and Postan, M., editors, The Cambridge Economic History of Europe, Vol. VI. The Industrial Revolution and After: Incomes, Population and Technological Change (II) (Cambridge: University Press, 1965), pp. 706–800.Google Scholar

⁴⁶ Thus northern capitalists may have taken actions leading to the Civil War to overthrow southern control over legislation (as a Beardian would argue) or else to preserve the western lands for free labor settlers. It seems doubtful, however, that there was a direct concern for the black or that the North was merely trying to have the South phase out a loss-making liability. In the British case Eric WilHams has argued that the ending of slavery was due to the economic interests of “mature industrial capitalism,” and not to the “saints.” For a critique of Williams on this point, see Anstey, Roger T., “Capitalism and Slavery: A Critique,” The Economic History Review, XXI (1968), pp. 307–320.CrossRefGoogle Scholar

⁴⁷ See the paper by Claudia Goldin, as well as an unpublished paper by West, Robert Craig, “Social, Political and Economic Factors Concerning the Feasibility of Compensated Emancipation During the 1860's,” (Northwestern University, 1972).Google Scholar Racism and economics cast a cruel dilemma for the southerner. Racism pointed to colonization of the freedmen; economic considerations then posed the question of where the labor force to make land productive would come from. This latter worry was raised in all antebellum discussions of compensated emancipation.

⁴⁸ Kloosterboer, Involuntary Labour, p. 1.