Studies of gender inequality in science have not systematically considered the effects of occupational sex segregation, although it is known from decades of economics and sociological research that a major source of inequality is because jobs predominately held by women are consistently paid less than those held by men. In this paper the focus is the workforce that controls funding resources for many U.S. scientists: employees of 8 major federal science policy agencies. In this unique dataset of 2.8 million federal employee records from 1994-2008, not only are the occupational and percent female in occupation variables available (along with controls for race, age, disability, education attainment, location), but also the agency specific occupational fixed effects in the models. Findings from these data show that , although these are government contexts where gender inequality is expected to be less, women’s work is consistently devalued and occupational sex segregation is a major source of income inequality among science policymakers. Importantly, there are different gendered earnings effects by agency, raising the question of how STEM policy agencies may vary in their influence on gender inequalities in the national science fields they support.  Findings from focus-group discussions and interview with policymakers show the organizational inhibitors to gender parity salaries and policy influence.