ABSTRACT

We use data from Indonesia and a demographic modeling strategy to estimate the total effect of increasing women’s schooling for the schooling of the next generation. This approach differs from standard approaches in that we include an estimate of how changes in women’s schooling affect children’s schooling not only directly but also through women’s choice of mate, marriage timing, fertility timing, fertility levels and the mortality of women and children. Each of these demographic factors can affect the relative number of children who will achieve different levels of schooling in the subsequent generation as a result of increasing women’s schooling in the previous generation. Some mechanisms, such as assortative mating, have very strong positive effects while others, such as marriage timing and fertility levels have offsetting effects. Differential mortality has a positive effect. Our results also demonstrate that the effects of expansions in women’s schooling depend on both the starting distribution of women’s schooling and where in the distribution women’s schooling increases.