This dissertation examines social determinants of fertility attitudes and behavior in the United States. It consists of three separate studies, each relying on a different large-scale panel study that follows the same set of individuals over time. Two of the studies examine predictors of fertility attitudes and behavior, while the other examines predictors of change in fertility attitudes.

In the first chapter, I examine the relationship between school enrollment and unintended pregnancy during early adulthood. I find that the risk of unintended pregnancy is significantly lower among women enrolled in school than among those not enrolled. My results suggest that expectations for employment and pregnancy partially mediate this relationship, but the primary reason why women enrolled in school have a lower risk of unintended pregnancy is because they use contraception more effectively than non-enrollees.

The second chapter investigates the influence of women's working conditions on change in fertility expectations. Compared to working full-time, I find that working part-time is associated with greater declines in expected births. Self-employment also is associated with greater declines in expected fertility compared to employment in organizations, but this association washes out at higher parity levels. I find a marginally significant association between lack of paid vacation leave and a decline in fertility expectations after controlling for part-time hours and self-employment.

In the third chapter, I examine how involvement in sibling care during adolescence affects early childbearing and childbearing attitudes in young adulthood. The results indicate that greater involvement in sibling care increases the likelihood of having a child before age 20. Findings on the relationship between sibling care and childbearing attitudes are somewhat mixed. Among those who have not had a child, involvement in sibling care is positively associated with perceived parenting ability and the number of children expected. Only involvement in instrumental sibling care is associated with the higher perceived likelihood of having children, but only involvement in emotional sibling care is associated with a lower expected age at first birth.