Downloads
In this repo you can find the main dataset for the replication in Juhász, Réka and Lane, Nathaniel and Oehlsen, Emily and Pérez, Verónica C., Measuring Industrial Policy: A Text-Based Approach (2025).
We gratefully acknowledge the use of the Global Trade Alert (GTA) dataset, which was kindly shared with us by the GTA team. We are particularly thankful to Johannes Fritz for helpful discussions about the GTA project.
The main file used in our analysis contains information at the "Measure" or "Intervention" level, we transform this dataset in multiple ways to obtain the desired figures and tables in our main text and appendix. If you have more questions regarding specific tables or figures please reachout to authors.
Note: In our dataset (MeasureID), it does not necessarily match the current online version of the GTA dataset, as this is updated often and IDs change.
In this page we provide a summary of our findings as well as excel files that replicate the contents of the main figures in our Working Paper.
We have made available our BERT model so other researchers are able to use it. You should be able to use this to predict industrial policy on a diversity of datasets that include policy text.
In our paper we developed an Industrial Policy codebook that instructs annotators to identify policy goals either through direct statements (e.g., "in order to boost domestic industry by making Egyptian cars more competitive") or implicit indicators (e.g., "China's 'Major Technical Equipment' policy grants tax-free imports to firms in certain sectors involved in the production of said equipment."). We have made these codebook available for other researchers to use.