Estimation and Inference for Heckman Selection Models with Cluster-Robust Variance

Tools for the estimation of Heckman selection models with robust variance-covariance matrices. It includes functions for computing the bread and meat matrices, as well as clustered standard errors for generalized Heckman models, see Fernando de Souza Bastos and Wagner Barreto-Souza and Marc G. Genton (2022, ISSN: < https://www.jstor.org/stable/27164235>). The package also offers cluster-robust inference with sandwich estimators, and tools for handling issues related to eigenvalues in covariance matrices.

heckmanGE

heckmanGE

The heckmanGE package has functions for modeling data with selection bias problems. It includes the generalized Heckman model, introduced by Bastos, Barreto-Souza, and Genton (2022), which allows the inclusion of covariates to the dispersion and correlation parameters, allowing the sample selection bias and the dispersion parameters to depend on covariates. More than that, our package allows the inclusion of sample weights and the grouping of data into clusters, in such a way that between clusters the errors are independent, but correlated within each cluster. The package also allows the adjustment of the classical Heckman model (Heckman (1976), Heckman (1979)) and the estimation of the parameters of this model via the maximum likelihood method and the two-step method.

Installation

You can install the released version of heckmanGE from CRAN with:

install.packages("heckmanGE")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("fsbmat-ufv/heckmanGE")

Code of Conduct

{heckmanGE} is released with a Contributor Code of Conduct.

References