Construction of the Trans-Atlantic Slave Trade Database: Sources and Methods

David Eltis (Emory University), 2010


1 David Eltis and David Richardson, eds., Extending the Frontiers: Essays on the New Transatlantic Slave trade Database (New Haven: Yale University Press, 2008).

2 Joseph Inikori has given “a preferred global figure of 15.4 million for the European slave trade.” Adjusting for those carried to the offshore islands and Europe, this implies 14.9 million headed for the Americas. See Cahiers d’Etudes africaines, 32 (1993):686.

3 H.C.V. Leibbrandt, Precis of the Archives of the Cape of Good Hope: Journal 1662–1670 (Cape Town: Richards, Government Printers, 1901), 127-128.

4 British slave ships trading from Africa to Lisbon include the Kent (1731), the Mary (1737), and the Betsey and Hennie (1755). For sources see the data set. For the removal of slaves from Ambriz to St. Helena and Sierra Leone, see Kelly Muspratt to Aberdeen, 31 July 1843, British National Archives (henceforth BNA), FO84/501.

5 A separate discussion of tonnage is to be found below.

6 One frequently cited shipping list reports that there were no children on board several British slave voyages in the 1790s (House of Lords Record Office, House of Lords, Main Papers, 28 July 1800). This document, however, omitted to report the children embarked (cf. BNA, T70/1574; House of Lords Record Office, House of Lords, Main Papers, 14, 25 June 1799). In Luanda and Benguela Portuguese customs reports of departures for Brazil report very low numbers of children embarked. But in this instance, “children” refers to infants only, and was above a tax category that indicated exemption from customs duties.

7 Mediterranean passes were issued by most European nations as a result of treaties with the Barbary powers. In theory, these documents allowed the vessels of the signing nation to pass freely through the “Mediterranean” waters frequented by Barbary corsairs. The passes record vessels’ and captain’s names, tonnage, the date the pass was issued, and intended trading location, such as “Africa” or “Africa and the Americas” or “Barbary” or “Madeira.” See David Richardson, The Mediterranean Passes in the Public Records Office (East Ardsley, UK: EP Microform Ltd., 1981). On different numbers reported in the Americas, in the date given for the arrival of slave vessels might be the date the vessel cleared customs, but in fact this could easily occur 2-4 weeks after the actual arrival.

8 Robert Louis Stein, The French Slave Trade in the Eighteenth Century: An Old Regime Business (Madison: University of Wisconsin Press, 1979), xv. Some armateurs also may have owned the vessel. French dictionaries define armateurs firstly as those merchants who fit out the ship or expedition and secondly as (ship)owners. See Emile Littré, Dictionnaire de la Langue Française, vol. I (Paris: Hachette, 1881), 194.

9 One major aid in identifying produce vessels is the Seaman’s Sixpence ledgers (BNA, ADM 68 series). One of our team, Jelmer Vos, went through this large series with great care.

10 One of the most widely used contemporary surveys of African regional preferences was Lt. Edward Bold, The Merchants and Mariners’ African Guide (J. W. Norie and Co.: London, 1822). For a very detailed private record, see the manuscript in the Sidney Jones Library, University of Liverpool, “Memorandum of African Trade, 1830–1840,” for W.A. Maxwell and Co.

11 See Pierre Verger, Trade Relations Between Bahia de Todos os Santos and the Bight of Benin, 17th to the 19th Century (Ibadan: Ibadan University Press, 1976), 358-361; and David Eltis, “The Export of Slaves from Africa, 1820–43,” Journal of Economic History 37 (1977): 417-420, for a fuller discussion.

12 In the downloadable version they would use the data variables “Percentage male embarked*” (MALRAT1), “Percentage male disembarked*” (MALRAT3), “Percentage children embarked*” (CHILRAT1) and “Percentage children disembarked*” (CHILRAT3) variables instead of the variables, “Percentage male*” (MALRAT7) and “Percentage children*” CHILRAT7.

13 For discussion of the general problem see Frederick C. Lane, “Tonnages, Medieval and Modern,” Economic History Review, 17 (1964–5): 213–33.

14 The 1773 legislation is 13 Geo III, c. 74. See W. Salisbury, “Early Tonnage Measurements in England: I, H.M. Customs and Statutory Rules,” Mariner’s Mirror, 52 (1966): 329–40. To convert registered tons into measured tons, we used the formulae in Christopher J. French, “Eighteenth Century Shipping Tonnage Measurements,” Journal of Economic History 33 (1973): 434–43. The 1786 act is 26 Geo III, c. 60, and its 1835 counterpart is 5 and 6 Will IV, c. 56, which introduced different rules for empty ships (s. 2) and those with cargo (s. 6). As the latter appears to have been used on slave ships, it is the one adopted here, and a further regression equation allows us to convert post-1836 tonnages into the measured ton of 1773–1835. It is:
Y = 52.86 + (1.22 x X) N = 63, R² = 0.77
where Y = measured tons, 1773–1835, and X = measured tons after 1835.

15 Palmerston to Kennedy, May 4, 1840 (circular dispatch), BNA, FO84/312.

16 H. Chamberlain to Canning, 18 Sept. 1824 (enc.), FO84/31; W. Cole and H. W. Macaulay to Palmerston, 1 Jan. 1835 (enc.), BNA, FO84/169; W. W. Lewis and R. Docherty to Palmerston, 9 Sept. 1837 (enc.), BNA, FO84/214; J. Barrow to Aberdeen, 16 May 1842 (enc.), BNA, FO84/439; G. Jackson and F. Grigg to Aberdeen, 2 Jan. 1841 (enc.), BNA, FO84/350.

17 For Spanish into British tonnage, data are limited. The equation is:
Y = 71 + (0.86 x X) N = 32, R² = 0.66.
Where Y = British measured tons, 1773–1835, and X = Spanish tons.
For US and British, see “An Act for Registering and Clearing Vessels, Regulating the Coasting Trade, and for other purposes,” Statutes at Large of the United States of America, 1789-1873, 1 (1789): 55. For a discussion, see W. Salisbury, “Early Tonnage Measurements in England: IV, Rules Used by Shipwrights and Merchants,” Mariner’s Mirror 53 (1967): 260–64.

18 See David Eltis and David Richardson, “Productivity in the Transatlantic Slave Trade,” Explorations in Economic History 32 (1995): 481, for the formula and a discussion.

19 See Lane, “Tonnages, Medieval and Modern,” 217-233 for a discussion.

20 Stein, French Slave Trade, 40–1; Patrick Villiers, “The Slave and Colonial Trade in France just before the Revolution,” in Slavery and the Rise of the Atlantic System, ed. Barbara L. Solow (Cambridge, UK: Cambridge University Press, 1991), 228.

21 See Herbert S. Klein, The Middle Passage: Comparative Studies in the Atlantic Slave Trade (Princeton: Princeton University Press, 1978), 29-31.

22 See, for example, H. Chamberlain to Canning, 7 July 1824 (enc.), BNA, FO84/31.

23 David Eltis and David Richardson, “Slave Prices of Newly Arrived Africans in the Americas, 1673-1807: A Quinquennial Series,” in Historical Statistics of the United States, Susan Carter et al., eds. (Cambridge, UK: Cambridge University Press, 2006), 5: 690-691; idem, “Markets for Newly Arrived Slaves in the Americas, 1673-1864,” in Slavery in the Development of the Americas, David Eltis, Frank Lewis, and Kenneth Sokoloff, eds. (Cambridge, UK: Cambridge University Press, 2004), 183-221.

24 More specifically, the price presented here is the average of the first ten males sold off the vessel.

Copyright 2008, 2009 Emory University. Software licensed under GNU General Public License 3.0 or later version. Some content licensed under Creative Commons Attribution-Non-Commercial 3.0.