One Instrument to Rule Them All: The Bias and Coverage of Just-ID IV

We argue that in microeconometric applications, just-identified instrumental variables (IV) estimators are virtually unbiased and the usual inference strategies are likely adequate. Confidence interval undercoverage exceeds 5% only for endogeneity beyond that seen even when IV and OLS estimates differ by an order of magnitude. Three widely-cited applications are used to explain why endogeneity is likely low enough for IV estimates to be reliable. IV identification typically implies a first-stage sign restriction; most analysts probably screen their estimates accordingly. We show that screening on the estimated first stage sign halves median bias of conventional IV without reducing coverage.