The Language of Performance Evaluations

Abstract

Performance evaluations of male and female junior attorneys in a Wall Street law firm were analyzed. Male supervisors judged male attorneys more favorably than female attorneys on numerical ratings that mattered for promotion but offered narrative comments that showed either no sex effects or greater favorability toward women. Judgments of male attorneys were more consistent overall than they were for female attorneys, and predictors of numerical ratings differed by sex: Narrative ratings of technical competence mattered more for men than women, and narrative ratings of interpersonal warmth mattered more for women than men. Open-ended use of positive performance words—the only outcome that favored women—did not translate into positive numerical ratings for women. The data suggest subtle patterns of gender bias, in which women were harmed by not meeting gendered expectations of interpersonal warmth but were less benefited than men by meeting masculine standards of high technical competence.