Social scientists have long been aware that as education levels rise, marital fertility declines. This relationship is statistically robust, is of sufficient magnitude to be empirically important, and holds crosssectionally, over time, and across countries. Further, it holds for parent education and family size and for child schooling and the number of siblings that children have. Given the pervasiveness and apparent strength of these relationships, it is natural to consider schooling as a potential policy instrument to reduce fertility in areas of rapid population growth. However, before large-scale efforts to increase either parent or child schooling are undertaken, a number of issues must be settled. Among these issues are the following:

• Separating causation from correlation. It hardly need be said that a strong negative simple correlation between schooling and family size does not necessarily imply that public policy-induced increases in parent or child schooling will cause parents to want and to have fewer children.