The Final Victims: Foreign Slave Trade to North America, 1783-1810.

The Final Victims: Foreign Slave Trade to North America, 1783-1810. By James A. McMillin (Columbia, South Carolina: University of South Carolina Press, 2004. 207 pp. +.1 CD-ROM, $39.95).

Less is known of the details of US slave voyages than about the slave trade of any other national group except the Portuguese. James McMillin has addressed this issue for the period 1783-1810, as well as taken on one of the major un-resolved demographic questions on the relationship between the transatlantic flow of African peoples and the black population of the Americas. Several years' research have allowed him to accumulate a database of what he claims as 1,764 separate slave voyages between 1783 and 1810--supplied to readers via a CD-ROM in PDF format slipped inside the back cover of his book. On the basis of these new data he offers three re-interpretations of US involvement in the slave trade. First, "far more foreign slaves were imported than previously thought" into the USA; second, "the Revolution did little to stop ... the slave trade"; third, "conditions captives encountered worsened rather than improved" between 1783 and 1810.

On the size of the late US slave trade the author quite properly points to the wide discrepancy that currently exists between the high estimates of African arrivals supported by the demographic approach to measuring the slave trade (291,000), and the low estimates supported by voyage-based data. (92,000 in one case and 113,000 in another). He makes some new assumptions about demographic patterns and calculates a new estimate of 187,843 slaves imported. Turning to his new voyage data set he presents breakdowns of slaves carried into the US (and elsewhere by US ships) from which he derives an estimate of "North American slave-carrying capacity" of 146,000 slaves for the period. He then examines the two separate estimates by region within the US and by decade and develops a preferred series (amalgamating the two approaches) which yields 170,300 slaves disembarked in all the US between 1783 and 1810.

There are two problems with these new estimates. On the demographic side McMillin offers no new data, but rather offers some fresh assumptions. Most of these seem perfectly reasonable, but of course as with any demographic model, dozens of other assumptions that would give quite different results would be equally reasonable. Readers would be unwise to assign too much weight to McMillin's outcomes.

For the voyage-based count, for which he does have new data, there is a different problem. The database he has assembled certainly includes many voyages that had simply escaped the dragnet cast over the years by Elizabeth Donnan, Coughtry and the authors of the 1999 transatlantic slave trade CD-ROM, among others. I estimate that 147 entries in his data set comprise previously unknown transatlantic slaving voyages, and there are at least as many again new "intra-American" slaving voyages--that is vessels coming into US ports from other parts of the Americas rather than from across the Atlantic. In addition McMillin provides new information on many voyages that scholars already knew about. Given the holes in what we know of the US slave trade, this is a major contribution. Unfortunately many of the voyages are entered more than once and what appear to be separate voyages are often really fragments of other voyages already included. I have not been able to check all 1,764 entries in the McMillin database but I can verify that one voyage shows up as no less than five times, twenty-two others are entered three times, and 224 voyages are included twice. A further thirty-one are either not slave ships (or lack any identifying marks), but have dates that suggest probable double-counting. Moreover, some other entries have references that I cannot track down.

At this point I am inclined to set aside a total of 364 entries--one fifth--of the new database. I say "at least" because so far I have examined only what I think are the transatlantic slave voyages in the McMillin database. The intra-American arrivals--which McMillin does not attempt to distinguish from the transatlantic voyages--may have similar problems. Finally, as McMillin himself points out, not all the voyages were carrying slaves to the US. Many of his slave voyages were sailing to Surinam, the Caribbean, Rio de la Plata, and a few are not transatlantic slave traders at all, but were rather in the Indian Ocean trade. Thus, neither his new assumptions using the demographic approach, nor his new voyage data help narrow the aforementioned wide range in the existing estimates of African arrivals into the US between 1783 and 1810. The new empirical data will eventually be used to resolve the issue.

His other major conclusions might also be questioned. The Revolution may have done little to halt the slave trade immediately, as, indeed, is already well known, but the connection between the Revolution--or at least the Constitutional Convention--and the abolition legislation of 1807 is direct and irrefutable. As for worsening conditions in the 1783-1810 period, McMillin curiously ignores the standard measurements of crowding, voyage length and shipboard mortality, despite his extensive work pulling together new information on voyages. McMillin rests his case instead on anecdotes, comments from contemporary observers, and arguments about the impact on the already awful experience of a slaving voyage of first, the war at sea, second, the rush to buy slaves as abolition approached, and third the lengthened time on board ensured by the rise of the New Orleans market. The actual data themselves point to a different interpretation. It has been possible to combine his new data with the 1999 CD-ROM and recalculate some elementary statistics. The middle passage length of an average US slave voyage turns out to have been shorter after 1783 than it had been before; voyage mortality remained about the same (though in other national trades it declined), and slaves carried per measured ton actually declined by about one third--mainly, one assumes, because American slave traders resorted to using larger vessels after 1804 when South Carolina reopened the traffic. McMillin is right that the US trade differed in some ways from other branches of the traffic, but it always had. Proper comparisons with earlier US patterns would indicate much less that was new in 1783-1810 (or 1804-1808) than McMillin suggests.

In conclusion, there are new data here, but readers should be prepared to do a lot of additional work before buying into the reinterpretations that McMillin derives from them.

David Eltis 

Emory University