The study set out to estimate the effect of breastfeeding on child mortality and stunting. We used nationally representative Uganda Demographic and Health Survey (UDHS) of 2011 merged with the community section of the Uganda National Household Survey (UNHS) of 2012/13 to include community variables, such as distance to the health facility, that are potential instrumental variables. We used various techniques: ordinary least squares (OLS), Instrumental Variables (IV) approach and control function during the analysis. We found that breastfeeding reduced child mortality but was weakly associated with child stunting. Just as previous literature documents, the OLS estimate of the coefficient on breastfeeding was biased downwards compared to the IV and control function estimates. Health knowledge was important in influencing child mortality. Government efforts towards sensitizing citizens about the importance of breastfeeding should be strengthened. Additionally, government needs to prioritize dissemination of health knowledge to women of reproductive age who are no longer in school. Furthermore, school curricula should be improved to include communication of health knowledge to students during early education to mitigate poor child health outcomes for future generations.
