Bayesian Multilevel Modelling of Women’s Decision-Making Capacity on Intimate Partner Violence in Tanzania

Introduction: Despite the focus on gender equality and women's empowerment in the Sustainable Development Goals (SDGs), Intimate Partner Violence (IPV) remains a significant issue in Tanzania, with 38% of women experiencing IPV. This poses a challenge to achieving SDG targets by 2030. The power imbalance in family decision-making is a major risk factor for IPV, yet the relationship between women's empowerment and IPV remains unclear. This study centres on addressing the question of how women's decision-making capacity influences IPV. Methods: Bayesian Multilevel mixed effects generalized linear models were used to analyse the influence of women’s decision-making capacity on IPV using the Tanzania Demographic and Health Survey (TDHS) 2015/16. Results: Compared to women not involved in decision-making, those who participated in household purchases exhibited a 13% (95% CI: 0.767, 0.978) lower likelihood of experiencing physical IPV. Similarly, women who decided on social visits were 7 % (95% CI: 0.873, 0.985) less likely to experience physical IPV. Women who decided on their health had a 15% (95% CI: 0.734, 0.979) lower risk of experiencing sexual violence. Women decision-making capacity on health was associated with a 10% (95% CI: 0.843, 0.951) reduction in emotional IPV. Women who decided on household purchase had a 26% (95% CI: 0.669, 0.820) lower risk of experiencing emotion IPV and 44% (95% CI: 0.473, 0.672) decrease in sexual IPV. Regional variations in IPV prevalence were also observed, with significant differences in physical (4.5%, BPOR - 0.045), sexual (9.7%, BPOR - 0.097), and emotional IPV (13.3%, BPOR - 0.133) odds at the regional level. Conclusion: The findings suggest that improving awareness of women’s decision-making capacity is crucial. This is especially important because effective interventions in this field appear to face inherent challenges.