Determinants of early marriage among women in South Africa : a multilevel analysis

Abstract

Background: Early marriage is a social and cultural issue that affects women in many sub-Saharan African countries. In the context of South Africa, early marriages are still valued in terms of cultural aspects and continue with varying degree of practice across the provinces. Additionally, recent information on the determinants of early marriage in rarely available. The study therefore aimed to examine individual-and community-level determinants of early marriage among women in South Africa. It is important to investigate the multilevel determinants of early marriage because of the social and legal issues early marriages have on society. In the context of South Africa, early marriages are valued in terms of cultural aspects and are still in practice in provinces such as KwaZulu-Natal, Eastern Cape, and Limpopo. The main objective of this study was to determine the multilevel determinants of early marriage among women in South Africa. The information available on this topic in the country is a bit dated; as a result, there is a need for more recent studies of this nature in South Africa. Methods: The study used cross-sectional data involving 7087 women aged 20-49 years extracted from the most recent 2016 South Africa Demographic and Health Survey dataset. In this study, early marriage is defined as women who first got married or had union before 18 years. Data were analysed using descriptive statistics including percentages. Chi-square test was used to examine the association between each of the independent variables and the outcome variable. Multilevel analysis was employed to examine effects of individual- and community-level characteristics on early marriage. Results were presented as odds ratio and their 95% confidence intervals (95% CI). This study used secondary data from the 2016 South Africa Demographic and Health Survey (SADHS). The research design used in this study is a descriptive research design. The analysis was based on women aged 20-49 years at the time of the survey. Three types of analyses were selected for the study. These included univariate, bivariate, and multilevel logistic regression analyses. On the multilevel logistic regression analysis, a two-level model, individual/household and community level model, was fitted. Results: The bivariate findings showed that population group, level of education, spousal/partner educational differences, age at first sex, parity, household wealth, community poverty, place of residence and province were statistically associated with early marriage. The findings further showed that 6.5% of women reported they were married before age eighteen. Among these women, the majority (58.7%) were aged 16-17 years, followed by those aged 14-15 (24.5%) and the least percentage was among those less than 12 years. The multilevel analysis showed that population group, level of, education, spousal/partner educational difference, parity (number of children ever born), HIV status, household wealth, and province were important determinants of early marriage. Conclusion: Although the prevalence of early marriage is low in the country, it is still of concern considering that the country has had various strategies of dealing with issues surrounding early marriages. The findings showed that women with primary education, early sexual debut, from poor households and women from Limpopo, Mpumalanga, Gauteng and Free State had higher odds of early marriage. Several strategies have been recommended in order to reduce early marriage (marriage before eighteen) among women in South Africa. These include providing awareness in protection of children towards early marriage especially in the tradition tribal societies. Additionally, it is important to strengthen, monitor and revisit laws and policies that protect children from harmful, traditional practises such as child marriage regularly and whenever necessary.