Environmental Provisions and the Innovation–Productivity Paradox: Evidence from Trade Agreements Across the Global North and South
Description
Data include ERPs data, Domestic and Foreign Patenting data, and all control variables elaborated in the attached data file.
Files
Steps to reproduce
Details to run the regression using the attached file are posted. This study uses multiple public and restricted sources to construct a bilateral country-pair panel (2001–2018). The submitted dataset contains the cleaned and merged panel data used in the empirical analysis. • Environmental provisions (ERPs): from the Trade and Environment Database (TREND, Morin et al., 2018, updated 2022). • Patent applications: from the INPACT-S database (LaBelle et al., 2024). Access to INPACT-S requires request from the authors; therefore, the raw patent data are not included in this submission. Instead, we provide the cleaned variables derived from INPACT-S. • Productivity: total factor productivity from Penn World Table v9.1. • Controls: CEPII Gravity (GDP, distance, contiguity, language), WTO membership, FTA indicators (TREND/DESTA), GDP growth and R&D expenditures (World Bank WDI). • Classification: IMF GNI categories for Global North vs. Global South. The final unbalanced panel covers 72 origins and 81 destination countries (87,373 observations). All cleaning and merging were performed in Stata 17. Researchers wishing to replicate our analysis may: 1. Request INPACT-S from its authors, 2. retrieve the remaining public datasets, and 3. Apply the workflow described above 4. Estimate treatment effects with Propensity Score Matching and staggered Difference-in-Differences (Callaway & Sant’Anna, 2021).
Institutions
- Sungkyunkwan University