Different predictors of intimate partner and natal family violence against women

Violence against women is often studied in the context of violence from intimate partners. However, women receive violence from a wider range of individuals-such as their natal kin-including their siblings, parents, uncles and cousins. Applying insights from evolutionary theory, we examine whether cousin marriage, which has been hypothesized to both reduce the risk of partner violence but increase the risk of natal family violence, associates differently with each type of violence. Second, we test whether common risk factors for partner violence, such as wealth, associate similarly with natal violence.

We analyse over 16 000 Jordanian women from three cohorts of the Jordan Demographic Health Surveys. Predictor variables include type of cousin marriage (patrilateral or matrilateral), education, wealth, number of children, urban living and polygyny. Outcome variables include whether a woman's husband or her natal family has ever been physically violent towards her.

Being married to a patrilateral cousin but not a matrilateral cousin is associated with a reduced risk of reporting intimate partner violence (IPV). By contrast being married to a matrilateral cousin but not a patrilateral one is associated with a reduced risk of reporting natal family violence. As expected, wealth is negatively associated with reporting partner violence, but we find no association with reports of natal family violence. Finally, individuals with more children are more likely to report IPV.

Findings indicate the importance of distinguishing between types of cousin marriage and highlight substantial differences in risk factors for intimate partner compared to natal family violence.

Sociodemographic risk factors, such as wealth, may associate differently with intimate partner and natal family violence. Results suggest that whether cousin marriage is protective of violence may depend on the type of cousin and secondly, that violence can have fitness relevant outcomes.