Women and minorities are generally underrepresented in science research careers. Significant underrepresentation is found in different science, technology, engineering, and mathematics (STEM) disciplines. A wide array of explanations has been offered for the underrepresentation of women and minorities in the STEM workforce and/or among those majoring in STEM fields.  These explanations have helped to inform interventions that have received millions of dollars of support over the years.  A core underlying premise of virtually every major intervention designed to increase the representation of women and racial and ethnic minority group members in STEM careers is that there exists a dominant pipeline ­toward those careers.  The premise is that there is a conventional sequence of educational and training procedures for a specific career profile, and that such a sequence is effective in producing the desired results of increased representation of underrepresented groups. A good case study of the predictive validity of the pipeline model is chemistry profession. This profession includes many persons at different points along the educational pipeline and has witnessed dramatic shifts in the racial and gender make-up of the field.    

This  paper tests the hypothesis that the pipeline metaphor is the correct representation of the production of the increased diversity within the chemistry profession. Using data from the IPUMS-CPS March Supplement, estimate post-baccalaureate (supply-side) effects and wage impacts (demand side effects) on the relative presentation of women and minorities among those employed as chemists. We find large differences across racial, ethnic and gender groups. We find very limited evidence to support the supply-side argument. The  responsiveness to demand side factors tends to be larger for minority group members than for others, suggesting that the pipeline model is inadequate for explaining underrepresentation in all professions.  Finally, we show that women and minorities are underrepresented at different critical transition points from high school to college to graduate school to the workforce.  The most salient transition point is often viewed as post-baccalaureate training.  In the pipeline model, increases in the supply of persons with graduate degrees in STEM fields ought to increase the number of persons employed in the field.  Our findings do not support this view.  An alternative model – the pathway model – posits that there are multiple routes towards the required training for science careers and that the underlying problem is not the under-supply of graduates in science but barriers that undervalue these alternative routes taken by women and minorities.