A1990
Title: An improved inference for IV regressions
Authors: Liyu Dou - Singapore Management University (Singapore)
Pengjin Min - Singapore Management University (Singapore)
Wenjie Wang - Nanyang Technological University (Singapore) [presenting]
Yichong Zhang - Singapore Management University (Singapore)
Abstract: Empirical instrumental variables (IV) studies often report separate results based on low-dimensional instruments and many base instruments. A combination test is proposed that integrates these commonly reported statistics. The test linearly combines a cluster-robust Wald statistic based on low-dimensional IVs with leave-one-cluster-out Lagrangian multiplier (LM) and Anderson-Rubin (AR) statistics constructed from many IVs. Joint asymptotic normality and asymptotic optimality of the proposed test are established. The procedure yields costless efficiency improvements, automatically adapts to weak identification of many instruments, and is accompanied by a practical rule of thumb for assessing efficiency gains.
