Reproductive externalities are important for fertility behavior in Kenya. We identify from anthropology structural forms of social interaction operating across individuals belonging to different ethnic and religious groups on the number of children ever born. We use the 1998 Demographic and Health Survey, combined with primary meteorological data on Kenya, and GMM methods, to show that social interaction effects by ethnicity are important over and above an individual's characteristics such as their religion to explain variations in fertility. Our findings have implications for policy debates in Kenya and in other developing countries about ethnic, religious, and other differences in fertility behavior.

Migration is a human capital investment in which parents bear costs and children share returns. Therefore, migrants from a population with heterogeneous intergenerational discount rates will self-select on intergenerational altruism. Intergenerational altruism and fertility are arguably linked, therefore immigrants might self-select on eventual fertility. Soviet Jews who migrated to Israel despite high migration costs averaged almost one child more than members of the same birth cohorts who migrated later, at lower cost. Distinguishing selection from treatment effects using mothers' age at migration, selection accounts for most of that difference (the proportion varies with specification), even with controls for religion and religiosity. Selection on fertility may have other explanations, including cultural preservation. To probe, we conduct an alternative empirical test of immigrant selection on altruism, finding that U.S. immigrants spend more time with grandchildren than do natives. Additionally, immigrant self-selection on fertility provides an alternative explanation for Chiswick's (1978, Journal of Political Economy86(5), 897–921) earnings-overtaking result.

The Amish collective objection to high school education and refusal to comply with compulsory schooling laws can be interpreted with a religious-club-good framework. According to the religious-club interpretation, the Amish use the restriction on secular education as a religious prohibition and sacrifice to improve the welfare of sect members. I exploit the 1972 U.S. Supreme Court's decision in Wisconsin vs. Yoder, which exempts Amish children from compulsory high school education, as a policy shock to test several key predictions of the religious-club explanations. The evidence suggests that the successful restriction on high school education helped the Amish sect exclude individuals with low religious participation, lower members' shadow cost of time, and grow the sect through higher fertility.