ABSTRACT
Affirmative Action remains a highly controversial issue in the United States. Recent court decisions (such asAdarand v. Pena, USSupremeCourt) or state-level referenda (such as Proposition 209 in California) are likely to influence the use of Affirmative Action in employment and/or education, and other actions by courts and legislatures are pending.1 Yet the effects of Affirmative Action on employment outcomes are not particularly well understood to date. The seminal work by Jonathan Leonard (summarized in his paper in 1990) provides evidence that Affirmative Action affects the employment and occupational status of women and minorities, which has been confirmed elsewhere (e.g. Leonard, 1985a; Rodgers and Spriggs, 1996; Holzer and Neumark, 1999).2 But the exact mechanisms by which these effects occur, and their consequences for a wide range of performance measures of firms or employees, have received little attention. The reason for this is that much of the empirical analysis of Affirmative Action
to date treats the program as a “black box,” and generates only reduced-form estimates on employment (or occupational status) by race and gender. For instance, Leonard (as well as Rodgers and Spriggs) have analyzed the effects of being a Federal contractor on employment outcomes across establishments, while Holzer and Neumark analyze the relationship between self-reported use of Affirmative Action and employment outcomes in a cross-section of establishments.3
