What Marginal Outcome Tests Can Tell Us About Racially Biased Decision-Making
Marginal outcome tests compare the expected effects of a decision on individuals who are of different races but at the same indifference point of the decision-maker. I present a simple formalization of how such tests can detect racial bias, defined as a deviation from accurate statistical discrimination. Namely, the tests can reject that the decision-maker ranks individuals according to some accurate prediction of a mandated outcome, given some unspecified race-inclusive information set. The frontier of marginal effects can furthermore rule out canonical taste-based discrimination. I relate this analysis to other interpretations of marginal outcome tests, other notions of racial discrimination, and recent identification strategies.
I thank David Arnold, Ivan Canay, Will Dobbie, Nicolás Grau, Jim Heckman, Conrad Miller, Magne Mogstad, Jack Mountjoy, Damián Vergara, and Crystal Yang for many helpful comments and conversations. Jerray Chang provided excellent research assistance. The views expressed herein are those of the author and do not necessarily reflect the views of the National Bureau of Economic Research.