3+parity in Egypt: a multilevel decomposition of wealth-based inequality

BackgroundWealth disparities in levels of fertility are well documented in Egypt. Data from the Egypt Demographic and Health Survey (2014) showed that 61% of births to mothers from the poorest wealth quintile were of the third order or higher compared to 33% among mothers from the richest wealth quintile. The current paper investigates the main individual and socio-contextual level determinants of having more than two living children among women aged 30 and older, and decomposes its wealth-based inequality.MethodsThe secondary analysis was based on the 2014 Egypt Demographic and Health Survey. Multilevel linear regression was used to model the number of additional living children a woman has after her first two living children. A decomposition analysis of the wealth-based concentration index was applied using a multilevel model.ResultsIndividual level and area level attributes explained 83 and 17% of the variation in 3+ parity, respectively. Only areas not well served by the health system were significantly related to 3+ parity.Decomposition of the wealth-based concentration index revealed that 55.7 and 44.3% of the 3+ parity inequality were attributed to individual level attributes and area level attributes, respectively. At the individual level, early marriage accounted for 26% of the inequality in 3+ parity inequality. At the area level, living in areas not well served by the health system accounted for 22.3% of the 3+ parity wealth- based inequality, while highly gendered areas contributed 5.8% to this inequality. Areas' random effects contributed 7.1% to the 3+ parity inequality, assist in identifying specific areas that require targeted policies.ConclusionMultilevel decomposition allowed the contributions of both the individual and area level attributes to be quantified. The decomposition highlighted the need for more tailored policies that target different social groups and different areas. Decomposition analysis also pinpointed specific areas that require additional targeted policies. This targeted approach can be used to support the efficient use of limited resources within any society.