Education systems in developing countries are often centrally managed in a top-down structure. In environments where schools have different needs and where localized information plays an important role, empowerment of the local community may be attractive, but low levels of human capital at the local level may offset gains from local information. This paper reports the results of a four-year, large-scale experiment that provided a grant and comprehensive school management training to principals, teachers, and community representatives in a set of schools. To separate the effect of the training from the grant, a second set of schools received the grant only with no training. A third set of schools served as a control group and received neither intervention. Each of 273 Gambian primary schools were randomized to one of the three groups. The program was implemented through the government education system. Three to four years into the program, the full intervention led to a 21 percent reduction in student absenteeism and a 23 percent reduction in teacher absenteeism, but produced no impact on student test scores. The effect of the full program on learning outcomes is strongly mediated by baseline local capacity, as measured by adult literacy. This result suggests that, in villages with high literacy, the program may yield gains on students' learning outcomes. Receiving the grant alone had no impact on either test scores or student participation.