Substitution E¤ects in Parental Investments

The paper provides causal estimates of how parents adjust bride prices to compensate sons for di¤erences in schooling investments in rural China. The benchmark estimate imply that when a son receives one yuan less in schooling investment than his brother, he will obtain 0.4 yuan more in his marital transfers as partial compensation. Accounting for measurement error in schooling expenditure,discounting of the timing of school expenditures and bride prices, heterogeneous returns to schooling, partial appropriation of schooling returns by children, or public goods’components in children’s consumption do not change the causal estimate. Although the standard income transfer derivative test of the unitary model is rejected, the assumptions necessary for the test are unlikely to be valid. The paper provides an alternative sequential income transfer test based on observing two transfers per child. This test supports the unitary model.