The gender gap in mathematics: Evidence from Chile

Motivated by the existence of a gender gap in math in many countries, the dynamics and underlying causes behind the gender gap using detailed microdata in Chile was investigated. The roles played by parents, classroom environments including teacher's gender and class composition, and individual characteristics of the students were explored. Similarly, controlling for school, and even class, fixed effects does very little to the gender gap, suggesting that sorting across schools or classes is not what causes this gap. A unique feature of the data from Chile is the ability to measure self-assessed ability in math in boys and girls in fourth grade. What is perhaps surprising is that even conditional on math scores, across the wide range of questions, girls are much more likely to be pessimistic about their math abilities. Boys are 10% more likely to say that they are good at math and 9% more likely to say that they get good grades in math without studying much. Class size has a small negative effect on girls and a small positive effect on boys in fourth grade and a small positive effect on both sexes in eighth grade, even controlling for school fixed effects.