Minimal Dataset — Supplementary Upload
==========================================================================
Manuscript: An Integrated Framework of How Strategic Orientation Affects
ESG Engagement under China's Dual-Carbon Goals: Financial Distress Risk
Transmission and Two Types of Moderating Conditions
Authors: Shiru Sun, Malisah Latip, Ismawati Sharkawi
==========================================================================

1. CONTENTS
--------------------------------------------------------------------------
- minimal_dataset.csv        Firm-year analysis panel (12,998 obs; 1,404 firms)
- analysis_code_stata.do     Stata code reproducing variable construction,
                             winsorization, and all reported estimates
                             (bootstrap/Monte-Carlo random seeds withheld;
                             available from the corresponding author)
- README.txt                 This file

2. SAMPLE
--------------------------------------------------------------------------
Chinese A-share listed firms, 2010-2023. Main analysis sample: firm-years
with non-missing Bloomberg ESG disclosure scores (N = 12,998; 1,404 firms).
Sub-samples: mediation (N = 12,629, non-missing O-Score), event study
(N = 9,168, pilot cities with complete pre-periods).

3. VARIABLE DEFINITIONS
--------------------------------------------------------------------------
FirmID          Anonymized firm identifier (MD5 hash of the stock code).
year            Fiscal year.
Industry_code   CSRC (2012) industry classification code.
LCC_pilot       = 1 if the firm's registered city is included in any wave
                of China's low-carbon city pilot program; 0 otherwise.
                Missing values are treated as non-pilot (as in the manuscript).
LCC_first_year  First pilot year of the firm's registered city.
ESG             Bloomberg ESG disclosure score (0-100).
E_score / S_score / G_score  Bloomberg environmental / social / governance
                pillar scores.
Profit          = 1 if the firm follows a profit-oriented strategic
                orientation (Zhou-Park classification); 0 otherwise.
Growth          = 1 if the firm follows a growth-oriented strategic
                orientation; 0 otherwise (reference group in regressions).
OScore          Ohlson (1980) bankruptcy probability score (financial
                distress risk).
KZ              Kaplan-Zingales financing-constraints index.
GES             Government environmental subsidies compiled from CSMAR
                subsidy line items filtered by environmental keywords and
                scaled by operating revenue.
Size            Firm size: natural logarithm of year-end total assets.
BM              Book-to-market ratio.
Lev             Leverage: total liabilities / total assets.
SOE             = 1 if the ultimate controlling owner is the state.
CashRatio       Cash holdings / total assets.
Top1            Shareholding ratio of the largest shareholder.
Media           Media attention: natural logarithm of one plus the count
                of positive newspaper and online news reports in the year
                (CFND and the CSMAR news sub-database).

4. DATA PROCESSING
--------------------------------------------------------------------------
- The regressions reported in the manuscript use 1%/99% winsorized versions
  of all continuous variables (see winsor2 commands in the Stata code). This
  file stores the pre-winsorization values so that the full processing chain
  is reproducible.
- DID, event-study dummies (Pre3/Pre2/Current/Post1/Post2) and interaction
  terms are constructed from LCC_pilot and LCC_first_year as described in
  Section 3 of the manuscript.
- Stock codes are anonymized to FirmID. Industry codes and years are retained.

5. DATA SOURCE AND RESTRICTIONS
--------------------------------------------------------------------------
ESG scores: Bloomberg. Financial and ownership variables: CSMAR. Media
attention: China Financial News Database (CFND). Low-carbon pilot lists:
official Chinese government documents.

The underlying commercial data are licensed and cannot be redistributed in
raw form. This minimal dataset contains the processed firm-year values
needed to verify the reported results. Owing to these licensing
restrictions, the raw source data cannot be shared publicly; the
variable-construction documentation and the full analysis code are
available from the corresponding author upon reasonable request, as stated
in the manuscript's Data Availability Statement.
