Pathways Versus Pipelines into the STEM Workforce: Race and Gender Effects

This paper fills a critical gap our knowledge about the differential determinants by gender, race and ethnicity of entering and persisting in science, technology, engineering, and mathematics (STEM) fields.  We use data on middle-school to college to workforce transitions from the National Educational Longitudinal Survey (NELS88), which tracks students periodically from 1988 - 2000, to isolate the differential determinants across gender, race and ethnicity of entering and persisting in STEM fields. This data set provides longitudinal information on coursework, majors, test performance, graduation rates, college attendance, college major and post-baccalaureate employment in STEM vs. other fields.

The theoretical foundation for our empirical analysis of these transitions distinguishes between pathway models and pipeline models. There may be multiple pathways to STEM fields, which often differ across race/ethnicity and gender groups. Examples include enrollment in two-year and community colleges, which provide an alternative path to STEM careers than do four year colleges and universities. Pipelines to STEM careers historically have been routed in science course selection in high school where there are wide variations in course availability across racial and ethnic communities. Thus, the econometric modeling of the determinants of entering and persisting in STEM fields must account for these alternative structures of pipelines vs. pathways that influence choices and decisions in isolating race, gender and ethnicity effects.  The paper catalogues the relative determinants at different points along the pipeline/pathway to STEM employment and provides an assessment of what factors matter for each race/ethnic/gender group.