A0172
Title: A conditional linear combination test with many weak instruments
Authors: Yichong Zhang - Singapore Management University (Singapore) [presenting]
Abstract: A linear combination of jackknife Anderson-Rubin, jackknife Lagrangian multiplier (LM), and orthogonalized jackknife LM tests for inference in IV regressions are considered with many weak instruments and heteroskedasticity. The weights in the linear combination are chosen based on a decision-theoretic rule that is adaptive to the identification strength. Under both weak and strong identifications, the proposed test controls the asymptotic size and is admissible among specific classes of tests. Under strong identification, the linear combination test has optimal power against local alternatives. Simulations and an empirical application to Angrist and Krueger in the 1991 dataset confirm the good power properties of the test.