In this dissertation, I argue first that for policy and research purposes, identifying the agent who controls the child's labor reveals more about child work than simply identifying the child's attachment to the labor market. Children are not perfectly substitutable between different types of work either; the child's sex determines the kind of work assigned to the child. Thus in addition to the importance of agent controlling the child's labor, the dissertation asserts that at the margin, the trade-off between school and work partly depends on the task assigned to the child.

The important findings in this dissertation are as follows: By using non-parametric analysis, I uncover that as household wealth improves, children are withdrawn from third party employment, but continue to work under parental control. Often, but not always, children working for third parties work harder compared to children working under parental control; and in some cases, children from better-off households work far more hours than the children of poor households. I then build a model in which a child's time is allocated to school, market and domestic work. The model predicts that, at the margin, the trade-off between school and work depends on the task, which may not fall under market boundaries. I then test the two theses of the dissertation using a dataset from Turkey. The results reveal that in Turkey, girls lose the priority in schooling in the presence of brothers, while boys gain by having sisters. Having assets complementary to child's work shifts boys' time away from third party employment to work under parental control, and parents adjust only their daughters' time when the household infrastructure is less developed.

The dissertation contributes to our understanding of work and schooling outcomes of children living in the so-called Third World in three ways: first, by highlighting the importance of the agent controlling the child's work; second, by emphasizing that the trade-off between domestic work and schooling may be as crucial as the trade-off between market work and schooling, and last, by highlighting the need for different policy tools in order to improve Turkish children's schooling.