Chapter 3 described a set of econometric procedures that were used to estimate the prices of slaves in different demographic categories from the Royal African Company invoices in the face of an unexpected problem. The problem, which had not previously been discussed in the historical literature on the slave trade, was that the prices of slaves within given demographic categories were consistently found to decline throughout the course of sales of individual cargoes of slaves. The procedure followed in Chapter 3 was to assume that this decline implied that slaves were ordered within sales by quality, with the best slaves sold early in a sale, and the quality of slaves declining thereafter as the sales progressed. The order of transactions within a sale could then be used as an independent variable that functioned as a proxy for the unobserved quality variable.

This procedure yielded plausible results; yet the validity of those results depends on the validity of the assumption that slave sales were ordered by quality. Further investigation of this phenomenon is therefore called for, in order to determine whether this assumption is accurate. If it is correct, the result is an intriguing one, for both historical and theoretical reasons. Historically, little is known about the conduct of slave sales in this early period of the history of the slave trade. An investigation of this conduct, and why it might have resulted in a systematic ordering of the sale of slaves by quality, can provide previously neglected information about the nature of relations between buyers and sellers in early American slave markets.