We measure the impact of employment by assessing its effect on the individual and their marital partnership in the Rohingya refugee camps. To do so, we randomly assign either the husband or wife to a six-week job. We find that partnerships in which men are employed create greater positive spillovers for their partners compared to those in which women are employed, leading to improved psychosocial well-being for both partners and better IPV outcomes for women. However, households with formerly employed women later exhibit a stronger preference for female work. These findings highlight that employment is a social experience, with implications for labor supply decisions and employment targeting policies.