WHAT CAN(NOT) EXPLAIN THE GAP? Evidence and Decomposition of Gendered Stream Choice in India

Abstract

Gendered pattern in stream choices is well established in the education literature. Males are over represented in the mathematically oriented courses while females are more likely to opt for life sciences and non-science courses. We use three cohorts of student results data from the Central Board of Secondary Education, the single largest education board with an all-India presence, to first quantify and subsequently decompose the gender gap in the very first stream choices made by students at the school level in India. We use our rich dataset to explore a large set of explanatory factors proposed in the literature to explain the gender gap, namely, student ability, attributes of cohort peers, "chilly" climate and socioeconomic characteristics of students. We employ a novel way to use a student’s immediate seniors in schools to elicit the "chilly" climate aspect of stream choice. Our measure of the expected "chilliness" of the climate in a prospective course accounts for the largest portion of the observed gender gap in male dominated subjects. It explains 13.48% of the gap in the take-up of Mathematics. Ability related attributes explain none of the gender gap in Mathematics, but 12% of that in Biology. The contribution of peer related variables, on the other hand, is negligible.