We reply to the comments of Roodman (2022) regarding the data and econometric specifications of Duflo (2001). We replicate Roodman’s results and agree with his comments about the errors in the coding of variables. We run additional specifications that seek to correct for the age-differentiated wage dilation issue he describes in the original 1995 data, finding mixed results. We further add data from the 2011-12 SUSENAS survey and find that the treatment effects on long-run wages become substantially larger, in line with the results of Hsiao (2021). We also discuss additional weak instrument corrections.