Income growth and gender bias in childhood mortality in developing countries

Mika Ueyama
ifpri discussion paper

With poverty studies having shifted their focus from household poverty to individual poverty, a number of studies have started to examine intrahousehold resource allocation, especially gender bias within the household as potential causes of poverty. The literature has highlighted the existence of gender inequalities in South Asia, attributed to strong preferences for male offspring stemming from cultural and traditional customs. Only a few studies focused on the regional difference in the extent of gender bias and its response to income growth. To fill a void in previous studies, this study analyzes regional differences in gender discrimination, taking into account time-series variations. Furthermore, we test whether economic factors are responsible for gender bias in child mortality. There are two main objectives in this study. First, through a comprehensive literature review and a careful treatment of data compilation, regional features and recent trends in gender bias in children’s health outcomes are updated. We find strong evidence of severe disparity in child health against girls in South Asia; in contrast, no such anti-female gender bias exists in Sub-Saharan Africa. Second, this paper empirically tests the relationship between gender biases in child mortality and income growth using carefully-compiled new country-level panel data, paying attention to the possibility that such relationship differs between regions and changes over time. To investigate the relationship, two types of data sets are used: (1) new cross-country panel data of childhood mortality rates by sex, collected from various sources of macro statistics, such as DHS stat and WHO statistics; and (2) our own estimates for age-specific child mortality rates of children, constructed from the retrospective information on birth and death histories included in micro data of each round of the Demographic and Health Surveys (DHS). The empirical result suggests that income growth is correlated with the reduction of the anti-female bias in childhood mortality in most regions of the developing world—including South Asia. This result is reasonable, since income growth leads to an increase in nutrition intake (food consumption) and in health related inputs. In sharp contrast, the regression result does not show any significant correlation between gender biases in child health outcomes and income growth in Sub-Saharan Africa. While previous studies focused on the severe gender bias in South Asia, this study examined the correlation between income growth and gender bias and found a new dimension of regional contrast between Sub-Saharan Africa and other regions." from -- Authors' Abstract