This paper investigates household preferences over who should work and whether these preferences are malleable. We document that men and women prefer that husbands work over wives. To understand why, we randomly assign a six-week job to either the husband or wife and document asymmetry: women’s work improves their own wellbeing but not their husbands’, while men’s work improves both partners’ wellbeing. One year later, we surprise households with a work opportunity. Both women and men in households where women were previously employed are more likely to prefer the woman take the job and express fewer concerns about women’s employment in general.