Rejecting Wife-Beating Norms Cuts Intimate Partner Violence Risk Across 21 African Countries

Intimate partner violence remains one of the most pervasive violations of women’s rights worldwide, and a sweeping new analysis of survey data from 21 sub-Saharan African countries suggests that not all forms of women’s empowerment protect equally against it. The study, published in Frontiers in Global Women’s Health, pooled Demographic and Health Survey data collected between 2010 and 2023 and analyzed responses from 90,807 married or cohabiting women aged 15 to 49. Its central finding is striking: women who firmly rejected the justifications commonly used to excuse wife-beating had roughly 41 percent lower odds of reporting recent intimate partner violence, while women’s participation in household decision-making showed no independent protective effect at all.

The researchers measured empowerment using the SWPER Global Index, a validated survey-based instrument that captures three distinct domains: attitudes towards violence, social independence, and decision-making. Rather than collapsing empowerment into a single proxy such as education or employment, the index distinguishes whether a woman rejects violence-supportive norms, where she stands in broader social terms including schooling, access to information, and timing of marriage and childbirth, and how much say she has over household choices. Each domain was classified into low, medium, and high levels using established cutoffs derived from 14 questions in the DHS women’s questionnaire.

The outcome was intimate partner violence experienced during the 12 months preceding each survey, constructed as a composite of physical, sexual, or emotional abuse reported by the woman herself. The pooled prevalence came to 12.4 percent, but the range across countries was dramatic, from just 2.0 percent in Mauritania to 30.8 percent in Sierra Leone, with Liberia close behind at 26.3 percent. South Africa, Malawi, and Burundi all sat below 8 percent. The authors note that such variation likely reflects differences in gender norms, economic conditions, legal protections, and disclosure patterns, although their analysis did not directly measure these contextual factors.

When the researchers ran survey-weighted multivariable logistic regression, adjusting for age, marital status, religion, education, occupation, parity, wealth, residence, and partner characteristics, the attitudes-towards-violence domain emerged as the strongest empowerment-related signal. Women with high empowerment in this domain had an adjusted odds ratio of 0.59 for past-year IPV compared with women with low empowerment, a 95 percent confidence interval of 0.55 to 0.63. High social independence was also protective, with an adjusted odds ratio of 0.84. Decision-making empowerment, by contrast, was essentially at the null, with high versus low empowerment yielding an adjusted odds ratio of 1.05.

The bivariate patterns reinforced the same hierarchy. Among women with high empowerment in attitudes towards violence, 22.7 percent reported past-year physical violence compared with 38.3 percent of those with low empowerment, and 8.2 percent reported sexual violence compared with 15.2 percent. Emotional violence followed the same gradient, affecting 25.5 percent of highly empowered women versus 36.0 percent of those with low empowerment. For decision-making, the pattern was inconsistent and non-monotonic, with physical violence actually slightly more common among women with high decision-making empowerment than among those with low empowerment.

The study’s covariate findings add nuance to the empowerment story. Education showed a mixed pattern: primary education was associated with slightly higher odds of IPV, an adjusted odds ratio of 1.08, while higher education was associated with lower odds, an adjusted odds ratio of 0.75. The authors suggest several possible explanations for the primary-education signal, including greater willingness among women with some schooling to recognize and report abusive behavior, and a transitional position in which modest gains in autonomy outpace changes in partners’ attitudes or household power dynamics. Older age, cohabitation, parity of two or more children, several occupational groups, and partner’s secondary education were all associated with higher odds of IPV, while Islamic affiliation and wealth in the richer or richest quintiles were associated with lower odds.

The authors are careful to frame these as associations rather than causal effects, and they flag the possibility of reverse causation. Women experiencing violence may withdraw from household decisions, reduce social participation, or shift their views about the acceptability of abuse as a coping mechanism, meaning that lower measured empowerment could partly be a consequence of violence rather than a precursor. Similarly, the relationship between education and IPV is difficult to disentangle, since schooling can be interrupted by early marriage, pregnancy, or controlling partners. A country-adjusted sensitivity analysis broadly confirmed the main SWPER findings, suggesting the results were not driven solely by which countries were included.

Methodologically, the analysis was rigorous for a pooled cross-sectional design. The Hosmer-Lemeshow goodness-of-fit test indicated adequate model calibration, the area under the receiver operating characteristic curve was 0.624, and variance inflation factors showed no problematic multicollinearity among the SWPER domains or covariates. A preliminary multilevel model found that clustering by country accounted for a negligible share of variance in IPV, supporting the use of standard survey-weighted logistic regression with country as a fixed covariate in sensitivity analyses. Roughly 98 percent of the descriptive sample had complete data for the primary regression model.

The policy implications are pointed. The authors argue that education and rights-based information should be integrated with, rather than substituted for, interventions addressing unequal power within relationships, community norms that tolerate violence, and barriers to confidential, survivor-centered support. Because decision-making empowerment showed no independent association with IPV, participation in selected household decisions alone may not adequately represent women’s actual power or safety. Sustainable reductions in intimate partner violence, they conclude, will require coordinated action that supports women’s agency while simultaneously challenging harmful social norms, enforcing laws against gender-based violence, and improving access to health, legal, and psychosocial services for survivors.

The study also carries important caveats. IPV may be underreported due to fear, stigma, or social desirability bias, and cultural differences in how violence is understood and disclosed may complicate cross-country comparisons. The SWPER index, while validated, may not capture context-specific forms of empowerment such as legal literacy, property ownership, or collective mobilization. The authors did not test whether the associations varied across individual countries, so the pooled estimates should be read as average associations rather than uniform effects. Still, the core message stands out clearly: empowerment is multidimensional, and the domains that matter most for reducing violence risk are those that reshape internalized norms about whether abuse is ever acceptable, and that expand women’s broader social position beyond the household.

Subject of Research: Association between multidimensional women's empowerment and intimate partner violence in sub-Saharan Africa

Article Title: Women's empowerment and its association with intimate partner violence in sub-Saharan Africa: a study using recent demographic and health surveys

Article References: Mugisha, J., Dushimirimana, T., & Koray, M. H. (2026). Women's empowerment and its association with intimate partner violence in sub-Saharan Africa: a study using recent demographic and health surveys. Frontiers in Global Women's Health, 7, Article 1929589. https://doi.org/10.3389/fgwh.2026.1929589

Image Credits: AI Generated

DOI: 10.3389/fgwh.2026.1929589

Keywords: intimate partner violence, women's empowerment, sub-Saharan Africa, SWPER Global Index, Demographic and Health Surveys, gender norms, attitudes towards violence, social independence, decision-making, public health, gender-based violence, Women's