Maternal and Paternal Grandmothers and Child Nutrition: Exploring Potential Differences in the Effect of Intergenerational Co-Residence in Sub-Saharan Africa

Several studies have explored the role played by grandmothers for child welfare in low-income countries. Most studies find that co-residence with maternal grandmothers improves child wellbeing, while the association with co-residence with paternal grandmothers appears to be either null or negative. However, recent studies document a Sub-Saharan African (SSA) exceptionality, with paternal rather than maternal grandmothers being beneficial for child welfare measured as child schooling outcomes. The male dominance characterizing this region has been hypothesized to drive this result, via grandmothers’ preferential investment in their sons and, by extension, their sons’ children. This paper contributes to the literature by employing Demographic and Health Surveys and Ethnographic Atlas (EA) data to test this hypothesis, by looking at the role of different-lineage grandmothers for child nutrition. Using EA data to retrieve information on individuals’ ancestral kinship system, I explore whether the association between co-residence with different-lineage grandmothers and children’s stunting risk depends on whether children live in ancestrally matrilineal versus patrilineal areas. While no support for the male dominance hypothesis is found, evidence of a negative selection mechanism as the main driver of the positive effects of co-residence with paternal grandmothers on children’s nutritional status is provided. Indeed, children appear to co-reside with maternal grandmothers for the most part in event of fatherlessness – i.e., when the father does not reside in the same household as the child because of single motherhood, divorce, or death. Thus, the nutritional advantage of children living with paternal grandmothers is actually a nutritional disadvantage of children living with maternal grandmothers, experimenting the negative effects of fatherlessness.