Is there any relation between education and democracy? Once we correct for weak instruments and identify education as "weakly exogenous" we find new evidence that education systematically predicts democracy. Our results are robust across model specification, instrumentation strategies, and samples.