Son preference and child under nutrition in the Arab countries: is there a gender bias against girls?

Although son preference has been demonstrated in the MENA region with different manifestations and at several phases of human development, the literature remains sparse with studies that examined the early childhood phase. The current study aims to explore the presence of a gender bias in child nutrition status and its association with maternal son preference in three Arab countries; namely, Egypt, Jordan, and Yemen on which limited research has been conducted. Child nutritional status is measured using the Height-for-Age z-score (HAZ). To examine the presence of gender bias across the entire nutritional distribution, we utilized a quantile regression framework which characterize the heterogeneous association of each determinant across the different percentiles of the nutrition distribution. We use data from the most recent rounds of the Demographic and Health Survey on a nationally representative sample of children aged 0-4 years, for which we observe their health measures. The multivariate analyses include a set of HAZ determinants that are widely used in the literature. Descriptive statistics show that 21.5% of the mothers have son preference in Yemen compared to 19.10% in Jordan and 13.26% in Egypt. Results of the baseline OLS model demonstrate a robust pro-girl nutrition bias in the three countries. However, results of the quantile regression model show that this pro-girl nutrition bias is only prevalent at the lower segment of the conditional HAZ distribution for Jordan and Yemen and is prevalent across the whole conditional HAZ distribution for Egypt. We also find no statistically significant association between maternal son preference and gender bias in child nutrition in the three countries. Although son preference is manifested in several phases of human development in the MENA region, the current study finds no nutritional bias against girls in the examined countries at early childhood.