OBJECTIVETo understand patterns of gender discrimination by exploring whether the risk of dying during infancy and childhood is correlated with not only the sex and birth order of the child but also the sex composition of previous siblings.
METHODSEvent history analysis (Cox proportional hazards model) is applied to pooled data from the 2006-2007, 2012-2013, and 2017-2018 rounds of the Pakistan Demographic and Health Survey to highlight patterns of mortality risk during infancy (0-11 months) and childhood (1-4 years) by the sex composition of previous siblings, birth order, and the sex of the index child.
RESULTSFemales lose the survival advantage early. Evidence of generalized discrimination against all girls is reflected in higher-than-expected mortality risk during infancy and childhood. Evidence of selective preferential treatment of first sons is reflected in a survival advantage, even as early as the first year of life in large families and in all family sizes during childhood.
CONTRIBUTIONThis paper examines sex differentials in infant and child mortality in a context with strong son preference but with a lack of evidence on sex-selective abortions. It gives theoretical and empirical attention to the birth order and sibling composition contexts within which births and discrimination occur. The specific pattern of discrimination provides clues to the rationale for, and mechanisms causing, differential mortality risk among siblings and across genders.