Zusammenfassung:
Bellemare et al. (2023b) examine discrimination against individuals with physical disabilities in the labor market in Quebec, Canada. Their findings indicate that callbacks from potential employers decrease by 25 percentage points if physical disability is (randomly) revealed. Callbacks increase by 10 percentage points if there is a video resume (randomly) sent to potential employers. In this document, we first conduct a computational reproduction using the replication package. Then, we test the robustness of the findings to the inclusion of different covariates, selecting them with a Double Selection Lasso approach. We complement the analysis estimating heterogeneous treatment effects using Causal Forests, which allow us to uncover data-driven subgroups and test their responses to the treatment. We find that Bellemare et al. (2023b)'s estimates are stable across these robustness checks.