An alternative test for conditional unconfoundedness using auxiliary variables

This paper proposes an alternative test procedure for testing the conditional unconfoundedness assumption which is an important identification condition commonly imposed in the literature of program analysis and policy evaluation. We transform the conditional unconfoundedness test to a nonparametric conditional moment test using an auxiliary variable which is independent of the treatment assignment variable conditional on potential outcomes and observable covariates. The proposed test statistic is shown to have a limiting normal distribution under the null hypothesis of conditional independence. Monte Carlo simulations are conducted to examine the finite sample performances of the proposed test statistics. Finally, the proposed test method is applied to test the conditional unconfoundedness in the real example of the return to college education. •Testing the conditional unconfoundedness assumption.•Using an auxiliary variable.•Transforming the conditional unconfoundedness test to a nonparametric conditional moment test.•Test statistic is shown to have a limiting normal distribution under the null hypothesis.