This paper demonstrates how a natural experiment in education can be used to estimate causal effects. The Swedish compulsory school reform extended basic education gradually across cohorts and municipalities, allowing for a difference-in-differences analysis. The paper summarizes the literature...