LAUSR

Empirical findings on the relationship between women’s employment and intimate partner violence (IPV) in low- and middle-income countries (LMICs) are mixed. These varied findings may arise because research thus far has given insufficient attention to how individual attributes and community context shape the pathways between women’s employment and IPV. Using publicly available Demographic and Health Survey (DHS) data from 20 LMIC settings (n = 168,995), we investigate (1) how women’s employment is associated with past-year IPV and (2) if associations differ by household- or community-level structural drivers of IPV: women’s attitudes toward IPV, women’s participation in household decision-making, and relative wealth. We fit mixed-effects logistic regression models exploring the total, individual, community, and contextual effects of women’s employment on past-year IPV; effect measure modification by structural drivers; and cross-level interactions between community-level structural drivers and individual employment. Our analyses reveal positive associations between total (odds ratio [OR] = 1.31; 95% CI [1.27, 1.35]), individual (OR = 1.23; 95% CI [1.19, 1.27]), community (OR = 1.06; 95% CI [1.06, 1.07]), and contextual effects (OR = 1.04; 95% CI [1.03, 1.05]) of women’s employment for IPV. Only individual wealth demonstrated statistically significant effect measure modification for the relationship between individual employment and past-year IPV (ratio of OR = 0.95; 95% CI [0.92, 0.99]). These findings suggest interventions that focus only on increasing women’s employment may be associated with harmful increases in the occurrence of IPV, even when these interventions enable a large proportion of women in a community to be employed. Structural interventions that change norms of women’s autonomy or attitudes toward IPV at the household or community levels may be insufficient to ameliorate these negative effects, whereas interventions that increase household wealth partly may buffer these effects.