Son preference and gender-biased breastfeeding in Pakistan

Abstract

We investigate gender-biased breastfeeding in Pakistan using data from the 2006–07 Pakistan Demographic and Health Survey. We replicate the analysis of Jayachandran and Kuziemko (2011) in India, where a model of breastfeeding was developed that incorporates its contraceptive properties and yields several predictions regarding when mothers wean their children. We test the validity of their theory in Pakistan, where parents’ stated preference for sons is twice as large as in India. Our replication exercise strongly supports their theory: mothers breastfeed daughters significantly less (especially if there are no elder sons) to increase the chances of becoming pregnant again with the hope to conceive a son. In Pakistan, the male advantage in breastfeeding is more than twice as large as in India (0.9 vs. 0.4 months). In addition, we find similar birth-order patterns for the mother’s subjective assessment of unwanted pregnancies as we do for breastfeeding duration.