A New Paradigm: A Joint Test of Structural and Correlation Parameters in Instrumental Variables Regression When Perfect Exogeneity is Violated

Currently, the commonly employed instrumental variables strategy relies on the knife-edge assumption of perfect exogeneity for valid inference. To make reliable inferences on the structural parameters under violations of exogeneity one must know the true correlation between the structural error and the instruments. The main innovation in this paper is to identify an appropriate test in this context: a joint null hypothesis of the structural parameters with the correlation between the instruments and the structural error term. We introduce a new endogeneity accounted test by combining the structural parameter inference while correcting the bias associated with non-exogeneity of the instrument. To address inference under violations of exogeneity, significant contributions have been made in the recent literature by assuming some degree of non-exogeneity. A key advantage of our approach over that of the previous literature is that we do not need to make any assumptions about the degree of violation of exogeneity either as possible values or prior distributions. In particular, our method is not a form of sensitivity analysis. Since our test statistic is continuous and monotonic in correlation, one can conduct inference for the structural parameters by a simple grid search over correlation values. We can make accurate inferences on the structural parameters because of a feature of the grid search over correlation values. One can also build joint confidence intervals for the structural parameters and the correlation parameter by inverting the test statistic. In the inversion, the null values of these parameters are used. We also propose a new way of testing exclusion restrictions, even in the just identified case.