Team Gender Composition and Firm Performance: Experimental Evidence from Solar Entrepreneurs in Rwanda

How does team gender composition affect the performance of small technology-enabled firms in the Global South? We study this question using a field experiment with 270 village-level solar enterprises in rural Rwanda, where teams of four entrepreneurs were randomly assigned to be all-male, all-female, or mixed-gender. Over 18 months, mixed-gender teams underperform all-male teams by approximately 40%, while all-female teams also underperform all-male teams by 24%, with point estimates placing all-female teams between the all-male and mixed arms. To probe mechanisms, we pair the 18-month administrative output data with minute-level GSM timestamp data over the first six months. Mixed teams generate fewer recharges during evening peak-demand hours, while all-female teams do not exhibit the peak-hour coordination deficit. Despite lower business revenues, all-female teams generate distinctive household-level spillovers, with children studying approximately 74 additional minutes per week. Our findings suggest that performance in small entrepreneurial firms depends critically on whether teams can organize around conflicting and often binding temporal constraints.