Martin Spindler, "Lasso for Instrumental Variable Selection -- A Replication Study", Journal of Applied Econometrics, Vol. 31, No. 2, 2016, pp. 450-454. Note: It is intended to publish the functions with some additional functionality and updated versions as an R package in the next time. Further information will be provided on my website: https://sites.google.com/site/spindlermartin/programming The file ms-codes.zip contains six R codes. These are ASCII files in DOS format, so Unix/Linux users should use "unzip -a". fuller.R: implementation of the fuller estimator (help function) tsls.R: implementation of the 2SLS estimator (help function) hetero_se.R: function for calculating heteroskedastic robust standard errors (help function) LassoShooting2.R: function for conducting Lasso estimation with variable dependent penalty terms (help function) Replication_cutoff_design.R and Replication_exponential_design.R are the main files for replicating the results (except some minor transformations) If you have any questions, please do not hesitate to contact me at: spindler [AT] mea.mpisoc.mpg.de