Instrumental Variables Lab — When OLS Lies

Three real published cases — schooling, the Vietnam draft, colonial institutions — walked end to end. Estimate naively with OLS, diagnose the bias, apply the canonical instrument, then defend the exclusion restriction against published critiques. A credibility-revolution lab built on Angrist-Krueger 1991, Angrist 1990, and Acemoglu-Johnson-Robinson 2001.