Události

Let us invite you to a Brown Bag Seminar by Alena Skolkova (CERGE-EI PhD student)
on November 2, 2022, at 13:00 in room 402

You can join also online: 
Lifesize link: https://call.lifesizecloud.com/16176658, Passcode: 5594

Presenter: Alena Skolkova

Title: "Elastic-Net for Instrumental Variables Regression"

Abstract: Instrumental variables (IV) are commonly applied for identification of treatment effects and policy evaluation. The use of many informative instruments improves the estimate accuracy. However, dealing with high-dimensional sets of instrumental variables of unknown strength may be complicated and requires instrument selection or regularization of the first-stage regression. Currently, lasso is established as one of the most popular regularization techniques relying on the assumption of approximate sparsity. I investigate the relative performance of the lasso and elastic-net estimators for fitting the first stage as part of IV estimation. As elastic-net involves ridge-type regularization, it generally improves upon lasso in finite samples when correlations among the instrumental variables are significant. In addition, by attaining a balance between lasso and ridge penalties, elastic-net accommodates deviations of the first-stage equation from a sparse structure, thus being a robust alternative to lasso that heavily relies on the sparsity assumption. I show the asymptotic equivalence of the IV estimators that employ the lasso and elastic-net first-stage estimates under sparsity. Via a Monte Carlo study I demonstrate the robustness of the IV estimator based on the elastic-net first-stage estimates to correlation among the instruments, and deviations from sparsity. Finally, I provide an empirical example that employs the elastic-net IV estimator for estimation of return to schooling.

This project is co-financed by the European Union.