Assessing the Causal Effect of Gendered Market Structures on Wage Attainments and in Evaluating the Gender Wage Gap: An Intention-to-Treat Analysis

Scott Eliason, University of Minnesota
Melissa Bonstead-Bruns, University of Wisconsin at Eau Claire
Jennifer C. Lee, University of Minnesota

Previous research has documented a persistent gender wage gap. On average, women earn less than men, and empirical evidence indicates that gendered market structures play an important role in creating this gap. Recent research, however, has cast doubt on this conclusion, suggesting that the effects of gender structures on wages dissolve with properly specified models. In this paper we seek to inform this debate by using Hirano, Imbens, Rubin, and Zhou’s (2000) formulation of the intention-to-treat design to assess causal effects. By considering gender-of-job as treatment and gender-of-person as the mechanism encouraging respondents to obtain gendered jobs, we obtain the compliers average causal effect of gender-of-job on wages, and isolate that portion of the gender wage gap due to gender-of-job. Using data from 1988 and 2000 Current Population Surveys, we show that a substantial amount of the observed gender wage gap can be causally attributed to gendered market structures.