Marriage Market, Parents' Bargaining Powers, and Children's Education

Whether the family makes decisions as a unit or through a collective decision making process has been tested elsewhere by examining whether the income pooling hypothesis holds or by examining whether premarital assets have significant effects on household consumptions. The results, however, may be confounded by endogeneity of the income and asset variables. This paper suggests that the marriage market conditions, summarized by the sex ratio, can be used to test the household models and that the test may be combined with the income pooling test to produce more reliable inferences. Using data from Indonesia, this paper conducts the combined test of the household models by estimating the effects of the provincial sex ratio and the parents' nonlabor incomes/premarital assets on Indonesian household's investment in children's education. I find that in urban areas the sex ratio has a strong positive effect on education expenditures, but not in rural areas. Being consistent to the sex ratio effect estimates, the income pooling hypothesis is mostly rejected in urban areas, but not in rural areas. In addition, premarital assets are found to have significant effects on investment in children's education in urban areas, but not in rural areas. I find that the estimation results are robust against alternative definitions of the sex ratio and additional controls for women's fertility choices and community-level income/wealth differences