Giovanni Millo, "Private Returns to R&D in the Presence of Spillovers, Eevisited", Journal of Applied Econometrics, Vol. 34, No. 1, 2019, pp. 155-159. All files are zipped in millo-files.zip. With one exception, they are ASCII files in DOS format. The only exception is the online appendix, which can be reproduced exactly based on the other files and is included only for convenience. The Online Appendix to the paper is provided as a self-documenting Sweave .snw file, to be compiled under R. Sweave("filename.snw") produces a file filename.tex with original tex code and numerical results; to be compiled further in, e.g., pdfLaTeX to obtain the final document as filename.pdf. Alternatively, Stangle("filename.snw") will produce an R script filename.R containing the full reproduction code for all numerical results. A copy of the R script is included here for convenience. Running the procedure requires R 3.0-1 or higher with the following add-on packages: - plm - pder - lmtest - spdep 0.5-68 or higher - xtable - reshape - MASS - car - msm and dependencies thereof. All are FOSS (free and open-source) and can be downloaded from www.r-project.org. Therefore they are not included here. One function from package 'sinkr', "dineof.R", is also needed. A copy of it has been included in the materials for convenience because the package is only available on GitHub. Please find all the citation information in said file and in the paper. Running the procedure does not require any data files, because the HPY data are included, by kind agreement of the authors of the original paper (personal correspondence with Markus Eberhardt), in the R package 'pder' as dataset "RDSpillovers". As such, they are loaded, from within R, by > library(pder) > data(RDSpillovers) and the relevant help page describing them by > ?RDSpillovers Additional .R files are required containing functionality currently not included in the relevant packages; they are contained in the "/functions" directory within these materials. Most of these functions are eventually going to become redundant as these procedures reach production quality and are included in the regular R packages the procedure loads. Compiling the Computational Appendix as a .pdf document requires the style files of the Journal of Statistical Software, to be found in www.jstatsoft.org/downloads/JSSstyle.zip, to be either in the working directory or in the search path of pdfLaTeX. A repehs.bib BibTeX file with the necessary references is also included. Giovanni Millo Trieste, September 2018