Weak identification robust tests in an instrumental quantile model

摘要

We develop a testing procedure that is robust to identification quality in an instrumental quantile model. In order reduce the computational burden, a multi-step approach is taken, and a two-step Anderson-Rubin (AR) statistic considered. We then propose an orthogonal decomposition of the AR statistic, where the null distribution of ea component does not depend on the assumption of a full rank of the Jacobian. Power experiments are conducted, a inferences on returns to schooling using the Angrist andKrueger data are considered as an empirical example.