In this talk I motivate and defend the use of causal methods for the conceptual and empirical study of racial discrimination. Causal methods are crucial both for clarifying what it means for a person to have been discriminated against and for empirically testing claims of discrimination. Nevertheless, evaluating racial discrimination raises challenges not present in more standard applications of causal inference. For instance, whereas one tests a drug’s effectiveness by varying whether an individual takes the drug or a placebo (while holding all else fixed), it is far from obvious what it would mean to vary just an individual’s race. Addressing such questions requires one not to focus narrowly on experimental methodology, but to further connect it to philosophical, sociological, historical, and legal analyses of race and discrimination. I will offer graphical causal methods as a unifying framework for bringing together these domains, reject various skeptical arguments against the application of these methods to discrimination, and highlight the open questions that provide a fruitful basis for future research.
Refreshments will be served preceding the talk, beginning at 1:45.