Robust misspecification tests for the Heckman’s two-step estimator

We construct and evaluate LM and Neyman’s C(α) tests based on bivariate Edgeworth expansions for the consistency of the Heckman’s two-step estimator in selection models, that is, for the marginal normality and linearity of the conditional expectation of the error terms. The proposed tests are robust to local misspecification in nuisance distributional parameters. Monte Carlo results show that instead of testing bivariate normality, testing marginal normality and linearity of the conditional expectations separately have a better size performance. Moreover, the robust variants of the tests have better size and similar power to non-robust tests, which determines that these tests can be successfully applied to detect specific departures from the null model of bivariate normality. We apply the tests procedures to women’s labor supply data.