Abstract
In developing countries, women exhibit substantially lower cognitive abilities than males. Using data from the national baseline of the Longitudinal Aging Study in India, we examine two hypotheses of the female disadvantage: lack of education and discriminatory experience in earlier life. We estimate regression models with and without education to assess what fraction of the gender gap in cognitive ability is explained by education. We perform a Blinder-Oaxaca decomposition to distinguish the gap attributable to sex differences in observable characteristics and the part attributable to differences in coefficients. The latter represents the gap unexplained by differences in “endowments” which may reflect the influence of other factors such as discrimination. To test the discrimination hypothesis, we introduce interaction terms between gender and caste and between gender and region of residence. This allows us to assess whether women in castes and areas where discrimination is greater, exhibit larger cognitive disparities.