Twin peaks: the emergence of bimodal fertility profiles in Latin America

Everton E. C. Lima, Universidade Estadual de Campinas (UNICAMP)
Kryštof Zeman, Wittgenstein Centre (IIASA, VID/ÖAW, WU) and Vienna Institute of Demography
Ruben Castro, Universidad Diego Portales
Mathias Nathan, Universidad de la República, Uruguay
Tomas Sobotka, Vienna Institute of Demography

A decline in period fertility to sub-replacement levels in Latin America did not lead to a sharp decline in fertility rates at young reproductive ages. This evidence suggests that Latin America may differ from most other low-fertility regions, where a shift to low fertility was accompanied by a fall in early childbearing. Reconstructing period fertility rates by age, birth order and level of education, we investigate changes in the age pattern of childbearing in four Latin American countries that experienced a decline in period total fertility rates below the replacement level in the early 2000s—Brazil, Chile, Costa Rica and Uruguay. Our analysis shows that all the four countries display a combination of continuing high rates of childbearing at younger ages with a parallel increase in first birth rates at later reproductive ages. This trend results in a rise in the standard deviation of the age distribution of first birth rates, suggesting a pattern of reproductive polarization by social status where women with different social groups increasingly differ in their timing of birth. This pattern is also manifested by the emergence of bimodal fertility schedules by age, especially in Chile and Uruguay. We show that this reproductive polarization is more pronounced than the bimodal profiles identified earlier for selected countries of Europe and the United States. We suggest that Latin American low fertility pattern is linked to a high level of income inequality and wide social status differences in the region that go hand in hand with a high rate of unplanned early pregnancies and births, especially among women with lower education. We support our argument by showing massive differences in the age profiles of fertility by level of education, which we analyze on the examples of Brazil, Chile, and Uruguay.