Causal Impact of Being an Unwanted Child on Survival in Matlab, Bangladesh

David Bishai, Johns Hopkins University
Abdur Razzaque, ICDDR,B

We develop a theory that highlights dueling statistical biases from biological removal of least fit households and socioeconomic removal of most fit households from contributing to the sample of births of unwanted children. We test the effect of infant wantedness on survival and schooling using data on 7,946 women from Matlab, Bangladesh who were asked if they wanted more children while not pregnant in 1990. Using an unadjusted bivariate model, we find that unwantedness increases child mortality (OR 1.21 p<0.1). The results of unwantedness on survival are insignificant in multivariate adjusted models. We show evidence of biological culling of unwanted pregnancies through miscarriage and stillbirth as well as socioeconomic culling with more unwanted pregnancies being aborted by women with higher schooling. Even with a large high-quality data set it is not possible to demonstrate that unwanted children are more likely to die as a result of their unwantedness.