Sisters, Not Brothers: Gender and the Household Division of Educational Resources

Abstract

Female education has become one of development policy’s most familiar devices. Educate girls, the argument goes, and early childbearing falls. Beneath that proposition there are two questions that receive noticeably less attention: who inside the household gets the education in the first place, and how much causal work can schooling itself really do? This paper examines both using the Cape Area Panel Study and sibling sex composition as a source of variation in the educational attainment of young adults aged 18 to 26 in metropolitan Cape Town, South Africa. The first-stage results reveal a striking gender asymmetry. Young women raised with any sisters complete 0.78 fewer years of schooling on average, controlling for household size and family background. Male schooling, on the other hand, barely moves with sibling composition. The sign differs across countries and cohorts, and the gender asymmetry appears more stubborn. Collectively with the wider literature, that instability points toward a gendered household allocation rule: girls’ educational investment appears more contingent on family composition, more available for adjustment when circumstances inside the household change. Sisters expose an inequality they did not create. The second stage results complicate another familiar story. Once education is instrumented, the large negative relationship between schooling and teenage pregnancy shrinks sharply and loses statistical significance. The Cape Town evidence therefore places gender inequality at two distinct points in the development story: in the household distribution of education itself, and in the distance between educational resources and reproductive agency. Gender responsive policy has reason to look further inside the household. Increasing what families have available is only part of the question. Who gets what still remains to be a question of its own.