He wishes to thank Donald McCloskey and anonymous referees for their numerous helpful comments on a previous draft. Summer research support from the Department of Economics at UNC-Chapel Hill is acknowledged gratefully.Google Scholar

¹ Thomas, R. P. and Bean, Richard N., “The Fishers of Men: The Profits of the Slave Trade,” this JOURNAL, 34 (12 1974), pp. 885–914, especially 885–909.Google Scholar Thomas and Bean neglect to deal with the implications of spillover effects onto competitiveness at other stages of slave production created by the existence of an externality at the “fishing” stage. It is significant that they present their argument by moving step by stp from all the stages of production that they believe fully satisfy the criteria for perfect competition to the first stage—the initial capture of the slave—where they uncover the externality. Their argument might have looked quite different if their presentation had moved in reverse order. Then the consequences of the theory of the second best might have been apparent. See Lipsey, R. G. and Lancaster, Kelvin, “General Theory of the Second Best”, Review of Economic Studies, 24, no. 63 (1956–1957), pp. 11–32.CrossRefGoogle Scholar

² Inikori, J. E., “Market Structure and the Profits of the British African Trade in the Late Eighteenth Century,” this JOURNAL, 41 (12 1981), pp. 745–76.Google Scholar

³ See Anderson, B. L. and Richardson, David, “Market Structure and Profits of the British African Trade in the Late Eighteenth Century: A Comment,” this JOURNAL, 43 (09 1983), pp. 713–21;Google Scholar and Inikori, J. E., “Market Structure and the Profits of the British African Trade in the Late Eighteenth Century: A Rejoinder,” this JOURNAL, 43 (09 1983), pp. 723–28.Google Scholar

⁴ Thomas and Bean, “The Fishers of Men,” p. 886, n. 7.Google Scholar

⁵ p. 898, n. 37.Google Scholar

⁶ A further difficulty is that all participants have bound up the debate with the neoclassical market structure conception of competition. As a consequence they must belabor themselves with establishing the number and size of firms, concentration ratios, market shares, and such, in an attempt to determine “how much” competition existed in the slave trade. There are, however, alternative conceptions of competition that would render this entire line of argument moot. For example, the classical conception of competition can be interpreted to mean that competition exists if there is a tendency for all rates of profit to equalize. In this context, the notions of perfect versus imperfect competition are not relevant. Competition is an economy-wide characteristic rather than a characteristic of a firm or an industry. The emergence of monopoly or other such trends might facilitate the mobility of capital toward the activity promising the greatest rewards. For a comprehensive exegesis see Clifton, James A., “Competition and the Evolution of the Capitalist Mode of Production”, Cambridge Journal of Economics, 1 (1977), pp. 137–51.Google Scholar

⁷ Inikori, “Market Structure,” p. 747.Google Scholar

⁸ pp. 772Google Scholar

⁹ pp. 772–77.Google Scholar

¹⁰ Anderson and Richardson, “Market Structure: A Comment”, p. 715.Google Scholar

¹¹ Thomas and Bean, “The Fishers of Men”, p. 897.Google Scholar

¹² Inikori, “Market Structure”, p. 747.Google Scholar

¹³ Anstey, Roger, “The Volume and Profitability of the British Slave Trade, 1761–1807”, in Engerman, S. L. and Genovese, E. D., eds., Race and Slavery in the Western Hemisphere: Quantitative Studies (Princeton, 1975), pp. 3–31;Google Scholar and Anstey, Roger, The Atlantic Slave Trade and British Abolition 1760–1810 (Atlantic Highlands, N.J., 1975).CrossRefGoogle Scholar

¹⁴ pp. 720.Google Scholar

¹⁵ Williams, Eric, Capitalism and Slavery (1st ed., Chapel Hill, 1944; New York, 1961), p. 36.Google Scholar

¹⁶ “Account of the Liverpool Slave Trade, 1795” in Donnan, Elizabeth, ed., Documents Illustrative of the History of the Slave Trade to America: The Eighteenth Century (Washington, D.C., 1931), vol. 2, pp. 625–30.Google Scholar

¹⁷ One of three referees indicated that Anstey's estimates are weakened because the n journeys occurred over ten-year periods, but Anstey's formula treats the earlier and later voyages in each decade as identical in weight. The referee suggested that Anstey might have escaped his difficulty by calculating internal rates of return. The internal rate of return calculation, however, is prospective whereas the Anstey r is retrospective. Regardless, the relevant discount rate from the perspective of the merchants would be the rate of interest foregone on funds committed to the trade during the course of the journey, which is precisely the discount rate Anstey seeks to incorporate in his calculations.Google Scholar

¹⁸ Inikori, “Market Structure”, pp. 761–62, n. 52.Google Scholar

¹⁹ The “numbers game” for economic historians began in earnest with Philip Curtin's well- known effort to arrive at estimates of imports of slaves to the New World over the course of the entire slave trade. Curtin utilized secondary sources and a technique involving demographic inference based upon available data on growth in the slave population in the Americas. (Curtin, Philip, The Atlantic Slave Trade: A Census [Madison, 1969]).Google Scholar Curtin estimated that, during the interval 1761–1807, 1.29 or 1.3 million Africans had been taken for export by way of the Atlantic slave trade.In contrast, Anstey focused directly on an important primary source—shipping data for the period—to construct his estimates. The details of his procedure appear in Anstey, “The Volume and Profitability,” especially pp. 3–12. Anstey's procedure for calculating slaves imported to the New World by British vessels pushed the estimate toward 1.42 million for the same interval. Curtin implicitly accepted Anstey's upward revision of the figure for the British slave trade for 1761–1807 because of the greater precision of the shipping data, with the proviso that the revision had negligible implications for Curtin's “grand total” of close to 10 million Africans imported to the New World over the course of the entire Atlantic slave trade. Curtin, Philip, “Measuring the Atlantic Slave Trade,” in Engerman, S. L. and Genovese, E. D., eds., Race and Slavery in the Western Hemisphere, p. 107. The fly in the ointment of peaceful accord between Curtin and Anstey was Inikori. Inikori's newer minimum estimates pushed the figure still higher for slaves landed by the British slave trade between 1761 and 1807—toward 1.89 or 1.9 million. The latter figure was about 25 percent above Anstey's estimate and 46 percent above Curtin's estimate. Inikori identified two major reasons for his higher estimates: “our total number of tonnage of slave ships is much higher than that of Anstey; our main cargo size or slaves-to-ton ratio is also much higher than one employed by Anstey.”Google ScholarInikori, J. E., “Measuring the Atlantic Slave Trade: An Assessment of Curtin and Anstey,” Journal of African History, 17, no. 2 (1976), p. 215.Google Scholar Inikori justified these assumptions on the basis of an important archival discovery not available to Anstey. He unearthed an additional customs list, Customs 17, at the Public Record Office that enabled him to identify the actual number of ships clearing from England for the period 1771–1807. The figures in Customs 17 are uniformly higher than Anstey's numbers, providing a justification for an upward revision of all slave loadings. Inikori retained Anstey's assumptions on Middle Passage mortality and Anstey's assumption that a 5 percent deduction should be made for nonslave-carrying vessels. (Inikori, “Measuring,” p. 217, n. 99, n. 101, n. 103, n. 105.) He thereby produced higher estimates for slaves landed. Anstey acknowledged that Inikori's discovery of the Customs 17 series justified a correction of numbers and tonnages of British ships clearing to Africa between 1777 and 1807. Still he was not persuaded that his previous estimates of slaves carried per ton or landed per ton merited alteration.Google Scholar He finally settled on a figure for 1761–1802 in the range of 1.57 million slaves landed (Anstey, Roger, “The British Slave Trade, 1751–1807: A Comment,” Journal of African History, 117, no. 4 (1976), p. 606.CrossRefGoogle Scholar) I derived the revised estimates below from the Inikori-Anstey exchange (Anstey, “British For a valuable survey of the controversy see Lovejoy, Paul F., “The Volume of the Atlantic Slave Trade: A Synthesis,” Journal of African Histoty, 23, no. 4 (1982), pp. 473–501.CrossRefGoogle Scholar Lovejoy concludes that Curtin's global figure for the total number of Africans landed in the Americas was reasonably accurate, despite upward revisions in estimates for various portions of the trade.

²⁰ Inikon, “Market Structure”, p. 761–62, n. 52. Inikon notes that the £36 average slave price figure for 1781–1790 led Anstey into the anomaly of reporting “a higher rate of profit for the decade 1781–1790 than the rate for the preceding decade”. (See my Table 1,) The “glory years” for profits from the slave trade, according to Inikori (p. 761), was the interval between 1779 and 1788. Inikori adds, “There are other difficulties with the Anstey estimates” (p. 762). He is not specific about these.Google Scholar

²¹ As Lovejoy observes (“The Volume of the Atlantic Slave Trade”, pp. 486–87), the core of the dispute over numbers landed between Inikori and Anstey involves conflict over slaves carried per ship. Lovejoy prefers Anstey's more conservative estimate because a similar ratio was arrived at independently by Drescher, Seymour in Econocide: British Slavery in the Era of Abolition (Pittsburgh, 1977), p. 28.Google Scholar

²² Inikori accepts Anstey's estimates of the rate of Middle Passage mortality. However, obviously a case can be made that the rate would have increased with greater crowding of slaves on board the vessels. This would lower the percentage of slaves loaded that actually arrived in the Americas, and thereby reduce profitability estimates based upon Inikori's statistics. How sensitive mortality was to the extent of crowding, as opposed to other adverse psychological and physical aspects of shipboard captivity, remains a deeply unsettled question. David Eltis (“Mortality in the Nineteenth Century Transatlantic Slave Trade”, [unpublished manuscript, Algonquin College, 1983]), after a careful survey of the literature contends that communicable disease was the major cause of slave mortality during the Middle Passage rather than “tight packing” or “psychic shock.”.

²³ Inikori, “Market Structure: A Rejoinder”, p. 728.Google Scholar

²⁴ Anstey, The Atlantic Slave Trade, pp. 22–23.Google Scholar

²⁵ Solow, Barbara L., “Caribbean Slavery and British Growth: The Eric Williams Hypothesis”, Journal of Development Economics, 17 (1985), pp. 99–115.CrossRefGoogle Scholar

²⁶ Darity, William Jr, “A General Equilibrium Model of the Eighteenth Century Atlantic Slave Trade: A Least-Likely Test for the Caribbean School”, Research in Economic History, 7 (1982), pp. 290–92, 321 n. 8,Google Scholar

²⁷ See Darity, William Jr, “Mercantilism, Slavery, and the Industrial Revolution”, Research in Political Economy, 5 (1982), pp. 1–21.Google Scholar

²⁸ For example, see Hobsbawm, Eric, Industry and Empire (Harmondsworth, 1968), especially chap. 1.Google Scholar

²⁹ Williams, Capitalism and Slavery, p. 36.Google Scholar