Sequential Sales As a Test of Adverse Selection in the Market for Slaves

We propose an alternative test for adverse selection using notarial records for slaves sold in New Orleans in 1830. The experiment is simple and mimics the used car example originally proposed by Akerlof (1970). When first sold in New Orleans, buyers of imported slaves were uninformed of the slaves' unobservable characteristics. In time, the new owners learned more about their slaves and the "lemons" were sold and the "peaches" retained. Because buyers anticipate that the slaves offered for sale were of lower quality on average, they reduce their bids for these slaves. Consequently, we should observe a lower price for the slaves who were resold in the market. We test this proposition by linking the sequential sales records of 833 slaves sold in New Orleans. Through a comparison of initial and resale prices, we find that prices increased which suggests that adverse selection had a relatively small effect on the prices of slaves.