Replication Materials for Does Tight Money Slow the Green Transition?
Description
This dataset provides the replication materials supporting the article “Does Tight Money Slow the Green Transition? Carbon Exposure, Monetary Policy, and the Speed of Corporate Decarbonization.” It contains a de-identified processed analytical panel, analysis code, variable dictionary, annual monetary-policy shock series, regression outputs, robustness-test results, documentation, and publication-quality figures. The processed panel covers 154,918 firm-year observations from 14,187 firms. The primary two-year estimation sample contains 6,457 observations from 1,121 firms across 11 monetary areas over 2011–2022. Company names and original securities identifiers have been removed and replaced with stable pseudonymous firm identifiers. The principal result indicates that a 25-basis-point unexpected monetary tightening interacted with one standard deviation greater predetermined carbon exposure predicts 1.82 log percentage points less cumulative carbon-intensity reduction after two years. The results indicate a temporary transition delay rather than permanent derailment. Monetary-policy surprises are based on Bolhuis, Das, and Yao (2024), “A New Dataset of High-Frequency Monetary Policy Shocks,” IMF Working Paper 24/224. The unprocessed firm-level source database is not redistributed. Users wishing to reconstruct the analytical panel from the original records must obtain lawful access to the relevant source data. The included README, variable dictionary, reproducibility documentation, and SHA-256 manifest describe the files and their intended use.
Files
Steps to reproduce
1. Download and extract the repository archive. 2. Read README_FIRST.md and DATA_LICENCE_AND_ACCESS.md before using the files. 3. Decompress 02_Data/Analysis_Panel_Deidentified.csv.gz. This file contains the transformed analytical variables and stable pseudonymous firm identifiers required to preserve the panel structure. 4. Consult 02_Data/Analysis_Variable_Dictionary.csv for variable definitions, transformations and units. 5. The scripts in 04_Code document the preprocessing, fixed-effects estimation, local projections, heterogeneity analysis, inverse-probability weighting, randomization inference and machine-learning robustness procedures. Run the scripts in the sequence indicated in 05_Documentation/05_Reproducibility.md. 6. Full reconstruction from unprocessed records requires lawful access to the original firm-level source files. These unprocessed source files are not redistributed in this repository. 7. Compare reproduced estimates with 01_Results_Workbook.xlsx and the corresponding files in 03_Tables_CSV. The primary two-year coefficient is −0.01821, with a two-way clustered standard error of 0.00735. 8. Verify file integrity using the SHA-256 values reported in Package_Manifest_SHA256.csv.
Institutions
- Bangladesh Institute of Capital MarketDhaka Division, Dhaka