The effect of a husband's history of political imprisonment on intimate partner violence: evidence from the occupied Palestinian territory using propensity score matching

Abstract Background Little is known about the direct effect of political imprisonment on menu0027s subsequent violent behaviour within the family. Availability of these recent survey data about intimate partner violence and measures of menu0027s detention in prison and their associated physical and emotional torture provides a unique opportunity to assess the causal association between these two factors. We aimed to examine the effect of political imprisonment on intimate partner violence, using data from the 2011 Palestinian domestic violence survey, which included the West Bank and Gaza Strip. Methods The study was based on data from a 2011 cross-sectional survey of a representative sample of 4413 married Palestinian women. A total of 3988 women were included in the matching analysis, of which 640 women had husbands who had been detained. The outcome variable was the womenu0027s exposure to partner physical violence during the past year. The treatment effect was whether a husband had been detained before 2010. We used propensity score matching and propensity-based weighted regression models to assess the effect of a whether a husband had a history of detention on intimate partner violence. Score matching was undertaken with the variables of age, education, region, type of residence, refugee status, residential crowding, labour force participation, poverty, financial stress, decision making index, and husband control index. Findings Severe physical intimate partner violence during the past year was reported by 455 (11%) women and sexual violence by 439 (11%) women. 640 (16%) women reported that their husbands were detained before 2010. Analysis of model-based estimates and average treatment effect showed a positive effect of a husbands who had been detained on their physical violence (mean treatment effect 0·036, 95% CI 0·008 to 0·62), but not with sexual violence (mean treatment effect −0·014, 95% CI −0·040 to 0·11) towards their partner. Interpretation The results are important for the design of specific interventions that might mitigate the negative effect of a husbandu0027s detention on their wivesu0027s lives. A potential confounder is that Palestinian men who are more likely to be imprisoned (eg, active in demonstrations) might also be more likely to commit intimate partner violence. An important implication is the need to consider rehabilitation services or counselling for previous detainees in this context. Funding None.