Replication package for: Open-source participation as an emerging mode of digital innovation: Firm-level evidence from GitHub

This replication package accompanies the manuscript "Open-source participation as an emerging mode of digital innovation: Firm-level evidence from GitHub." It provides Python scripts for collecting public GitHub records, Stata analysis code, the 159-firm GitHub cohort list, a GitHub field dictionary, and non-restricted aggregate outputs. The underlying firm-year panel combines public GitHub records with Crunchbase and China Stock Market and Accounting Research (CSMAR) data. Because Crunchbase and CSMAR are commercial databases available under licence, their extracts and any row-level derived panel containing their fields are not redistributed. Raw GitHub API activity files are likewise excluded because some responses contain third-party account details and email addresses. Researchers with lawful access to the licensed data can use the included scripts and documentation to reproduce the analysis.