Prenatal Smoking Cessation and Infant Health: Evidence from Sibling Births

This article uses a large data set of sibling births to examine when mothers must quit smoking in pregnancy to deliver healthy babies. It applies sibling fixed effects models to provide robust evidence that smoking cessation in the first trimester has a negligible effect on infant health, but cessation as late as second trimester or smoking throughout pregnancy is associated with substantially lower birth weights and higher risks of delivering low birth weight babies. In particular, about two thirds of the total detrimental smoking impact on birth outcomes occurs in the second trimester. Therefore, reallocating resources on prenatal smoking cessation towards the first trimester can lead to a significant efficiency gain. This study also shows when the timing information of prenatal smoking cessation is improperly used, it will introduce a new nontrivial downward bias in estimating the causality between the conventionally used group measure "prenatal smoker" and infant health.