Gender gaps and scientific productivity in middle-income countries : evidence from Mexico : prepared for the Institutions for Development Department

This paper provides evidence of the existence and determinants of the publication productivity gender gap in Mexico at the individual level, and its consequences for the Mexican scientific system and productivity at both the individual discipline and the aggregate levels. The paper specifies and performs a panel data econometric analysis based on a sample of Mexican researchers who are members of the National System of Researchers (SNI) of Mexico in the period 2002-13. It corrects for a selectivity bias: the existence of periods with no (or low-quality) publications, and endogeneity bias: promotion to higher academic ranks. It defines and implements counterfactual simulations to assess the magnitude of macro-impacts of existing gender gaps and illustrate the potential effects of a range of policy scenarios. The results show no significant gender gaps for an average SNI researcher. Moreover, after correcting for endogeneity and selectivity biases, the study finds that the average female researcher in public universities is around 8 percent more productive than her male peers, with most of the observed productivity being explained by gender differentials in the propensity to have periods of no (or low) quality publication. Barriers to promotion to higher academic ranks are highest among females in public research centers (PRCs). The study's macro scenarios on promotion practices, selectivity, collaboration, and age show that eliminating gender gaps would increase aggregate productivity by an average of 7 percent for university women and 9 percent for women in research centers.