Lasso Methods for Gaussian Instrumental Variables Models
In this note, we propose the use of sparse methods (e.g., LASSO, Post-LASSO, p LASSO, and Post-p LASSO) to form first-stage predictions and estimate optimal instruments in linear instrumental variables (IV) models with many instruments in the canonical Gaussian case. The methods apply even when the...
| Main Authors: | , , |
|---|---|
| Format: | Working Paper |
| Language: | en_US |
| Published: |
Cambridge, MA: Department of Economics, Massachusetts Institute of Technology
2011
|
| Online Access: | http://hdl.handle.net/1721.1/65142 |
| _version_ | 1826188073189244928 |
|---|---|
| author | Belloni, Alexandre Chernozhukov, Victor Hansen, Christian |
| author_facet | Belloni, Alexandre Chernozhukov, Victor Hansen, Christian |
| author_sort | Belloni, Alexandre |
| collection | MIT |
| description | In this note, we propose the use of sparse methods (e.g., LASSO, Post-LASSO, p LASSO, and Post-p LASSO) to form first-stage predictions and estimate optimal instruments in linear instrumental variables (IV) models with many instruments in the canonical Gaussian case. The methods apply even when the number of instruments is much larger than the sample size. We derive asymptotic distributions for the resulting IV estimators and provide conditions under which these sparsity-based IV estimators are asymptotically oracle-efficient. In simulation experiments, a sparsity-based IV estimator with a data-driven penalty performs well compared to recently advocated many-instrument-robust procedures. We illustrate the procedure in an empirical example using the Angrist and Krueger (1991) schooling data. |
| first_indexed | 2024-09-23T07:54:10Z |
| format | Working Paper |
| id | mit-1721.1/65142 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T07:54:10Z |
| publishDate | 2011 |
| publisher | Cambridge, MA: Department of Economics, Massachusetts Institute of Technology |
| record_format | dspace |
| spelling | mit-1721.1/651422019-04-09T15:24:33Z Lasso Methods for Gaussian Instrumental Variables Models Belloni, Alexandre Chernozhukov, Victor Hansen, Christian In this note, we propose the use of sparse methods (e.g., LASSO, Post-LASSO, p LASSO, and Post-p LASSO) to form first-stage predictions and estimate optimal instruments in linear instrumental variables (IV) models with many instruments in the canonical Gaussian case. The methods apply even when the number of instruments is much larger than the sample size. We derive asymptotic distributions for the resulting IV estimators and provide conditions under which these sparsity-based IV estimators are asymptotically oracle-efficient. In simulation experiments, a sparsity-based IV estimator with a data-driven penalty performs well compared to recently advocated many-instrument-robust procedures. We illustrate the procedure in an empirical example using the Angrist and Krueger (1991) schooling data. 2011-08-15T17:09:35Z 2011-08-15T17:09:35Z 2011-02-25 Working Paper http://hdl.handle.net/1721.1/65142 en_US Working paper (Massachusetts Institute of Technology, Department of Economics);11-14 An error occurred on the license name. An error occurred getting the license - uri. application/pdf Cambridge, MA: Department of Economics, Massachusetts Institute of Technology |
| spellingShingle | Belloni, Alexandre Chernozhukov, Victor Hansen, Christian Lasso Methods for Gaussian Instrumental Variables Models |
| title | Lasso Methods for Gaussian Instrumental Variables Models |
| title_full | Lasso Methods for Gaussian Instrumental Variables Models |
| title_fullStr | Lasso Methods for Gaussian Instrumental Variables Models |
| title_full_unstemmed | Lasso Methods for Gaussian Instrumental Variables Models |
| title_short | Lasso Methods for Gaussian Instrumental Variables Models |
| title_sort | lasso methods for gaussian instrumental variables models |
| url | http://hdl.handle.net/1721.1/65142 |
| work_keys_str_mv | AT bellonialexandre lassomethodsforgaussianinstrumentalvariablesmodels AT chernozhukovvictor lassomethodsforgaussianinstrumentalvariablesmodels AT hansenchristian lassomethodsforgaussianinstrumentalvariablesmodels |