Lasso Methods for Gaussian Instrumental Variables Models

• DOI: 10.2139/ssrn.1908409
• Publication Date: 2012-01-05T17:38:13Z
• Authors: Alexandre Belloni; Victor Chernozhukov; Christian Hansen
• Keywords: Instrumental variable; Lasso

In this note, we propose the use of sparse methods (e.g. LASSO, Post-LASSO, p and Post-p LASSO) to form first-stage predictions estimate optimal instruments in linear instrumental variables (IV) models with many canonical Gaussian case. The apply even when number is much larger than sample size. We derive asymptotic distributions for resulting IV estimators provide conditions under which these sparsity-based are asymptotically oracle-efficient. simulation experiments, a estimator data-driven penalty performs well compared recently advocated many-instrument-robust procedures. illustrate procedure an empirical example using Angrist Krueger (1991) schooling data.