Since 1991, Chile has provided large, renewable cash grants to indigenous children in lower-income households, conditional on school enrollment. We estimate intent-to-treat effects of grant exposure on indigenous adults and their children, leveraging variation in expected grant exposure across birth cohorts and never-treated adults, and using fixed effects to absorb unobserved variables shared by adults born in the same year and community. Cohorts with the greatest exposure had 0.6 more years of schooling, 10% more hours worked, and 22% higher labor earnings, reducing pre-treatment ethnic differences. Mothers' exposure increased their children's early-grade test scores and reduced second-generation grant receipt.