Bruno Merlevede and Angelos Theodorakopoulos, "Productivity Effects of Internationalisation through the Domestic Supply Chain", Journal of Applied Econometrics, Vol. 36, No. 6, 2021, pp. 808-832. Data ---- In this paper we rely on two datasets: 1) The Orbis-Amadeus firm-level dataset used in the paper is purchased from the commercial provider Bureau van Dijk (see https://www.bvdinfo.com) This is a Moody's Analytics company, and thus we are not allowed to make the data public. Any interested researcher can purchase the data from the provider. Thus, access to data implies that the interested researcher either purchases the data directly or has access via her/his affiliated institution. For the analysis we rely on this source to extend the dataset constructed by Merlevede et al. (2015) with newer data. Merlevede et al. (2015) describes the procedure, which is along the lines of Kalemli-Ozcan et al. (2015). Also, in the code files discussed below, we provide all information and guidance needed to handle the data and replicate all results, tables, and figures presented in both the main manuscript and online appendix. References: Merlevede, B., M. de Zwaan, K. Lenaerts, and V. Purice, 2015, Multinational networks, domestic and foreign firms in Europe. Working Paper D/2015/7012/02, Working Papers of Faculty of Economics and Business Administration, Ghent University, Belgium. Sebnem Kalemli-Ozcan, Bent Sorensen, Carolina Villegas-Sanchez, Vadym Volosovych, and Sevcan Yesiltas, 2015, How to Construct Nationally Representative Firm Level Data from the Orbis Global Database: New Facts and Aggregate Implications, NBER Working Paper Paper No. 21558 (Revised December 2019) 2) The WIOD input-output-table data used in the paper are directly downloaded from the WIOD website (http://www.wiod.org/database/wiots16). Specifically, we use the "WIOT tables in STATA format" download. Similarly, in the code files discussed below, we provide all information and guidance needed to handle the data and replicate all results, tables and figures presented in both the main manuscript and online appendix. Software -------- Stata/MP v. 16.0 Hardware ------- CPU: 3,20 GHz Cores: 16 Memory: 320 GB Folders ------- There are two main folders within the project folder: 1) "code" contains Stata .do files used to produce all results, figures and tables. 2) "data" contains all data used. Note that the "orbis_amadeus" and "wiot_stata_Nov16" sub-folders are empty. As discussed above, the proprietary firm-level data cannot be shared, but the code provided below describes how to process and clean it. Also, the publicly available WIOD data can be directly downloaded from WIOD as explained above and detailed in the .do files below. Description of code (Stata .do files) ------------------------------------- 1) In the "code" folder you can find all Stata .do files with all code used to handle the data and replicate all results, tables and figures presented in both the main manuscript and online appendix. 2) The .do files are enumerated based on the order in which they were executed. 3) The master .do file "[_] master.do" explains in detail the contents and purpose of each .do file thereafter. 4) Each subsequent .do file contains a detailed explanation both at the beginning, but also, throughout the code, of all relevant information about the data sources, variables used and choices made. 5) If you wish to run files separately, please make sure that all previous .do files have been executed up to that point. 6) Please make sure to update the path directory name of your local project folder in line 16 of "[0.0] directories.do". Contact ------- Please address any questions to: bruno.merlevede [AT] ugent.be angelos.theodorakopoulos [AT] oxfordmartin.ox.ac.uk Orm for the most up-to-date contact info, please visit the following web pages: Bruno Merlevede: https://sites.google.com/site/merlevedebruno/ Angelos Theodorakopoulos: https://sites.google.com/site/theodorakopoulosangelos/