This study aims to investigate the relationship between consanguineous marriages and children ever born (CEB) in Pakistan and the moderating effect of working women in the relationship between consanguineous marriages and CEB. Furthermore, decomposition analysis was used to find out the factors that influence the likelihood of the child ever born.
Data from the Pakistan Demographic and Health Survey were utilized, which was conducted between 2017 and 2018. The sample includes 15 671 households, with 63.8% reporting consanguineous marriages. Zero-inflated negative binomial regression was employed to check the association between consanguineous marriages and children ever born, followed by the moderating role of working women in the relationship between them, and multivariate decomposition analysis was used to find out factors that influence the likelihood of CEB.
Our results show that consanguineous marriages significantly increase fertility (AIRR = 1.055, 95% CI: 1.034–1.076). While working women initially exhibit higher fertility in the bivariate model, this effect diminishes in the multivariate model (AIRR = 0.986, 95% CI: 0.960–1.013). Second-cousin marriages are associated with higher fertility (AIRR = 1.025, 95% CI: 1.009–1.042), and husband's education reduces fertility (AIRR = 0.767, 95% CI: 0.746–0.787). Rural residence and regions like Balochistan and FATA show higher fertility rates. Decomposition analysis reveals that working women slightly increase the CEB likelihood, while husbands' higher education and rural residence reduce it. Female children and having the last child alive lower CEB. Age, region, and education significantly influence fertility, with notable regional disparities across Pakistan.
Our finding suggests that consanguineous marriages and working women are positively associated with CEB. The findings suggest several policy implications and recommendations for government and policymakers, including family planning initiatives, educational campaigns, and informed family planning decisions. However, the study's cross-sectional design limits its ability to infer causality. Future research using longitudinal data is recommended for more accurate predictions.