The Relevance of Statistics to Prove Discrimination: A Typology

Substantial confusion exists among lawyers, litigants and the courts about how to identify and prove discrimination. The difficulty is primarily a result of the absence of a definition of discrimination in the federal civil rights statutes, and uncertainty as to what social policies the anti-discrimination statutes seek to accommodate. This Article provides an approach for identifying and proving discrimination through the use of statistical evidence. The Article begins by classifying discrimination cases into five basic types, based on the nature of the harm involved in each type of case.

The Article then describes the potential uses of statistical evidence by each type of plaintiff and analyzes the probative value of statistical evidence in proving the various types of discrimination. The Article then concludes with the argument that statistics are of great practical value as evidence of discrimination.