This study examines adverse selection directly in the cattle procurement market by analyzing the distribution of latent cattle quality inferred from a feedlot’s observed choices in an expected-utility maximization problem. Earlier studies are extended by directly testing for adverse selection, without relying on the significance of the IMR coefficient or on the specification of the widely used probit model to represent the sample selection process. We recover latent cattle quality using a framework that captures a feedlot’s optimal market choice under asymmetric information, price risk, and carcass quality uncertainty. The expected utility of profit for a feedlot is maximized with respect to the market choice probability, and the first-order condition is inverted to recover latent cattle quality from data. The Kolmogorov-Smirnov and Anderson-Darling tests are used to test differences in the empirical distribution functions of the recovered latent cattle quality for adverse selection. We find strong evidence of adverse selection in the analyzed dataset. Policies that encourage greater transparency in quality-related information for cattle marketed through the cash market could help alleviate this problem.
