Effects of polygyny on child health and determinants of birth registration : the case of Nigerian children

  • Uchechi Shirley ANADUAKA

Student thesis: PhD Thesis (Lingnan)


This thesis contains three empirical essays focused on the impact of parental socioeconomic characteristics on early childhood outcomes, namely health and birth registration in Nigeria. The focus is motivated by the growing body of literature showing the crucial role the former play in shaping early childhood outcomes. Nigeria is of utmost concern because of the significant differences in parental socioeconomic characteristics (e.g. polygyny is widespread) and high rates of poor child outcomes (poor health and low birth registration). Identifying the impact of parental socioeconomic characteristics can allow more effective policy interventions for achieving sustainable development goals: health (SDG target 2.2) and birth registration (SDG target 16.9), respectively. The first essay examines the effects of polygyny on a child’s health measured by growth and survival outcomes. I use a sample of 43108 children (growth) and 54859 children (survival) under-5 years of age drawn from three rounds of the Demographic and Health Survey (DHS) dataset. The results of the Ordinary Least Squares (OLS) and district fixed-effects regression techniques show a strong negative relationship between polygyny and child health. However, I identify evidence of self-selection into polygyny by young and poorly educated women, which is addressed using the Lewbel Heteroskedasticity-based Instrumental Variable approach. The results of the instrumental variable analysis suggest that only the effects on child mortality can be defined as causal. Investigation into mechanisms suggest poor parental investments as a possible explanation. For instance, polygynous children are less likely to be delivered in health facilities, with skilled birth attendants or be vaccinated. They are less likely to use treated bed nets (anti-malaria) or proper treatments for diarrhea or fever/cough. Their households are also more likely to be crowded and their mothers are also more likely to experience high-levels of control and domestic abuse. These factors can impact child health negatively despite the protective behaviors of longer birth-intervals. The second essay examines the determinants of birth registration using a sample of 38652 children under-five years of age, drawn from the 2008 and 2013 Nigeria DHS data. Using OLS and district fixed-effects regression models, I find that children with more educated and richer parents are less likely to be non-registered nor non-certified. Their level of health service utilization (prenatal visits, skilled attendants at birth and vaccination), and higher birth rank also improve their registration chances. Finally, the third essay examines the effect of place of birth on birth registration with a two-part research objective. First, I examine equity in the distribution of registration centers. Second, I examine whether the distribution for birth non-registration and non-certification status are spatially correlated. Descriptive analysis techniques suggest inequity in the distribution of registration, and exploratory spatial regression methods (Moran’s I) suggest moderate spatial clustering of the distribution. In summary, the essays show that the differences in parental socioeconomic characteristics translate to inequalities in early childhood. To achieve the health and birth registration targets, these differences need to be incorporated into already existing or future health and birth registration policies, to ensure that no children are left behind.
Date of Award15 Jul 2019
Original languageEnglish
Awarding Institution
  • Lingnan University
SupervisorPing LIN (Supervisor) & Ho Lun Alex WONG (Co-supervisor)

Cite this