Das Dokument ist öffentlich zugänglich im Rahmen des deutschen Urheberrechts.

We take advantage of recent advances in behavioral genetics to revisit a classic question in economics: how do parents respond to children's endowments and to differences in endowments among siblings? Parental investment decisions depend both on parental preferences regarding inequality in the distribution of their children's quality and on how costly it is for parents to add to their children's quality by investing in their human capital (or the price effect). Our empirical strategy allows us to isolate the effect of parental preferences regarding equality from the price effect, a distinction that cannot be made when relying on sibling or twin fixed-effects models. Importantly, we use genetic variants that predict educational attainment as a measure of children's educational endowments. Individuals' genetic makeup is fixed at conception, so these indicators cannot be affected by parental investment decisions. We find evidence that parents of non-twin siblings display inequality aversion and, given the absolute endowment level of one child, they invest more in him/her if his/her sibling is better-endowed. In contrast, parents of twins do not significantly react to endowment differences among their children, likely because it is difficult to provide differential parental investments across children of exactly the same age.