Abstract

Methodology is often a point of contention in gender-based salary studies. Although this debate seems at first to be merely about technical issues, it also has an important conceptual dimension. We argue that there are two competing implicit conceptions of discrimination, one institutional and the other individual, that underlie many such debates. We first contrast the preferred methodologies advanced by each side, the policy capturing approach and the flagging approach, and explore the theoretical meaning of their statistical models. We then describe a practical application of both methodological approaches in one specific salary inequity study. In conclusion, we reflect on the implications of such practical statistical choices, discuss how such models can be combined, and make suggestions for sociologists who act as statistical experts or work with them in gender-based salary inequity studies on their own campuses.