This paper evaluates a multifaceted female empowerment program in urban Liberia that combined intensive psychosocial therapy with vocational skills training over 12 months. The evaluation uses a randomized controlled trial with follow-up surveys at 1 year and 3.5 years post-program (2 and 4.5 years after enrollment). At the first endline, treatment women experienced large reductions in emotional and physical/sexual intimate partner violence (IPV) and in perceived community norms supporting IPV; effects on psychological well-being were null. By 3.5 years, the treatment group’s IPV levels remained close to their first-endline values, but the treatment–control difference had closed because the control group also experienced substantial reductions. Only the gap in perceived community norms persisted. Qualitative evidence points to the psychosocial component as the primary active ingredient, while the economic arm was operationally compromised by the non-delivery of promised capital grants. We interpret the longer-term convergence as durable individual- and interpersonal-level change for treatment women, set against contemporaneous structural interventions in the study region that plausibly affected the control group in particular.

https://doi.org/10.1016/j.jdeveco.2026.103823