内部統制の質、サステナビリティ報告の広さ、ESG格付の不一致:中国上場企業からの証拠
Internal Control Quality, Sustainability Reporting Breadth, and ESG Rating Disagreement: Evidence from Chinese Listed Firms (原題)
Shiyi Chen, Yuanyuan Wang
🤖 gxceed AI 要約
日本語
中国A株企業4万件超の2009〜2023年データを用い、内部統制の質およびサステナビリティ報告の広さと、ESG格付プロバイダー間の不一致の関係を検証。内部統制の質が高い企業ほど格付不一致が小さく、報告の広さは逆に不一致を拡大させる。効果は統計的に頑健だが経済的には小幅で、開示情報の標準化・比較可能性の重要性を示唆する。
English
Using 40,310 Chinese A-share firm-year observations (2009-2023), this study links internal control quality and sustainability reporting breadth to cross-provider ESG rating disagreement. Higher internal control quality is associated with modestly lower disagreement, while broader reporting is associated with greater disagreement. The results highlight the role of information discipline and disclosure standardization in rating comparability.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
SSBJ基準・有報でのサステナビリティ開示が進む日本でも、格付機関間の不一致は投資家対応の実務課題。内部統制と開示の質が評価のばらつきを左右するという知見は、開示ガバナンス整備の根拠となる。
In the global GX context
As ISSB/CSRD push mandatory sustainability disclosure, rating disagreement remains a barrier to comparability. This paper adds evidence that internal control quality and reporting breadth shape disagreement, informing debates on disclosure standardization and rating regulation.
👥 読者別の含意
🔬研究者:ESG格付不一致の決定要因として内部統制と開示の広さを定量化した点が、格付研究・開示研究の双方に貢献する。
🏢実務担当者:開示の量を増やすだけでなく、内部統制・情報品質を高めることが格付評価の安定に寄与しうる点を実務に示す。
🏛政策担当者:格付不一致の是正には開示の標準化・比較可能性の確保が鍵であり、制度設計上の示唆を与える。
📄 Abstract(原文)
Persistent disagreement among environmental, social, and governance (ESG) rating providers limits the comparability of corporate sustainability assessments but does not, by itself, indicate rating inaccuracy. This study examines whether firm-level internal control quality and sustainability reporting breadth are associated with cross-provider ESG rating disagreement. The dataset contains 40,310 firm-year observations for Chinese A-share listed firms from 2009 to 2023. The principal specifications hold provider identity constant using ratings from Huazheng and CNRDS and normalize scores within provider-year distributions. Firm and year fixed effects are included, with standard errors clustered at the firm level. Internal control quality is negatively associated with ESG rating disagreement across the principal and alternative specifications. In the main percentile-based model, a one-standard-deviation increase in internal control quality is associated with a 0.0098 decrease in disagreement, equivalent to 3.16% of the dependent-variable mean and 4.43% of its standard deviation. The association is therefore statistically systematic but economically modest. Internal control quality is also positively associated with sustainability reporting breadth, whereas reporting breadth is positively associated with contemporaneous rating disagreement. These relationships are examined as separate conditional associations rather than as evidence of a mediated effect. Coverage diagnostics, fixed-provider samples, alternative rating transformations, alternative internal-control measures, and text-based validation of reporting breadth indicate that the principal ICQ result is not attributable solely to changes in provider composition or measurement construction. Lagged specifications preserve the negative ICQ association in the fixed-two-provider models, while the SRB estimates are more sensitive to normalization and timing. The evidence suggests that stronger internal information discipline is associated with modestly lower ESG rating disagreement, whereas broader reporting may increase the scope for provider-specific interpretation when the disclosed information is insufficiently standardized or comparable.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.3390/su18189450first seen 2026-09-17 04:47:15
- semanticscholar https://doi.org/10.3390/su18189450first seen 2026-09-19 05:28:02 · last seen 2026-09-22 04:58:27
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録(無料・3キーワードまで)。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。