Sources of Fertility Decline in Modern Economic Growth: Is Aggregate Evidence on the Demographic Transition Credible?

The large declines in total fertility rates that have occurred in many low income countries since 1960 are thought to have been caused by subsidized birth control and economic development, in roughly equal proportions. Studies of these issues are not generally guided by economic approaches to the determinants of fertility. Consequently, income is not disaggregated according to its sources, which according to some economic theories should exert different effects on fertility and child mortality. The assumed exogeneity of child mortality as a determinant of fertility has also not been subjected to statistical tests, and national family planning programs are assumed to occur independently of consumer demands for these services. Because of these statistical and economic shortcomings of the existing evidence on these issues, and the social significance of child mortality, fertility and population growth on a global scale, this paper analyzes the aggregate evidence and confirms many microeconomic findings about the determinants of child mortality and fertility. Based on data for 62 low income countries in 1972, 1982, and 1988, the educational attainment of adult women is the powerful engine of demographic change, associated with lower levels of child mortality and fertility, and on balance slower population growth. Factors that permit higher levels of calorie consumption per capita, given a country's education and income, are associated with lower levels of child mortality, and Wu-Hausman tests confirm that child mortality is endogenous in exerting its substantial positive effect on fertility. Male education is associated with higher levels of fertility. Male education is associated with higher levels of fertility, while income levels from nonhuman capital sources, compared on a purchasing power parity basis per adult, are associated with lower child mortality but higher fertility. Decreases in the share of the labor force employed in agriculture are associated with decreases in fertility. If family planning programs are assumed to be independent of parent demands for birth control and other unobserved variables affecting fertility, family planning exerts a significant, but quite small, effect on fertility. If family planning programs are more realistically treated as endogenously affected by transfers to a country by a major international donor agency (IPPF), the role of domestic family planning programs is an insignificant determinant of fertility.