This article quantifies the role played by education in the reduction of global poverty. I propose tools for identifying the contribution of schooling to economic growth by income group, integrating imperfect substitution between skill groups into macroeconomic growth decomposition. I bring this “distributional growth accounting” framework to the data by exploiting a new microdatabase representative of nearly all of the world’s population, new estimates of the private returns to schooling, and historical income distribution statistics. Education can account for about 45% of global economic growth and 60% of pretax income growth among the world’s poorest 20% from 1980 to 2019. A significant fraction of these gains was made possible by skill-biased technical change amplifying the returns to education. Because they ignore the distributional effects of schooling, standard growth accounting methods substantially underestimate economic benefits of education for the global poor.