Replication package for: Mandatory sustainability reporting, carbon performance and greenwashing in Southeast Asia.
Description
This dataset is the replication package for the article "Mandatory sustainability reporting, carbon performance and greenwashing: Evidence from staggered mandates in Southeast Asia" (Business Strategy and the Environment, [year], https://doi.org/[to be added]). Research question. Do mandatory sustainability-reporting rules change what firms do, or only what they report? The study uses the staggered introduction of reporting mandates in six Southeast Asian markets (Indonesia, Malaysia, the Philippines, Singapore, Thailand and Vietnam) between FY2016 and FY2021, with listed Korean firms as a never-treated comparison group, and estimates the effects with the Callaway–Sant'Anna difference-in-differences estimator. Greenwashing is measured with a firm-level decoupling index (GW): a firm's standardised Environmental pillar score minus its standardised carbon performance, within industry and year. Main findings. In the main sample (441 firms, FY2014–2020), the mandates raised the share of firms publishing a sustainability report by 11.6 percentage points, mostly in comply-or-explain markets. ESG and Environmental pillar scores rose by 1.8 and 2.7 points (significant only at the 10% level), carbon intensity did not change, and the decoupling index did not rise on average (0.03 standard deviations). The decoupling effect was 0.43 standard deviations smaller for firms with externally assured reports. In Indonesia (43 firms, FY2014–2024), firms that had to start reporting did so without a relative rise in the decoupling index. Contents. - data/: author-coded files: mandate cohorts with their legal sources; the Indonesian bank cohort; the main estimation sample (LSEG instrument codes with cohort coding); and a hand-coded classification of 43 Indonesian sustainability reporters for FY2017–2018. - code/: Python scripts that retrieve the data in LSEG Workspace (CodeBook), build the firm-year panel and estimate all models (Callaway–Sant'Anna, two-way fixed effects with wild cluster bootstrap, Sun–Abraham, stacked difference-in-differences, Goodman-Bacon decomposition, Rambachan–Roth sensitivity bounds, and the Indonesian treatment-intensity design with entropy balancing); an equivalent R implementation; validation tests on synthetic data; and the figure scripts. - results/: the output files of the final estimation runs (aggregate estimates and test statistics only) with SHA-256 checksums, and a table of the hypotheses against the evidence. README.md maps every table and figure in the article to its script and output file. Data restrictions. ESG scores, emissions and financial data come from LSEG Workspace under licence and are not included; licensed users can rebuild them with the scripts provided. The package contains no firm-level LSEG values.
Files
Institutions
- Padjadjaran UniversityWest Java, Bandung