ESG開示におけるグリーンウォッシング検出：パーム油企業を対象としたNLP分析
Deteksi Greenwashing dalam Pengungkapan ESG: Analisis Berbasis NLP pada Perusahaan Perkebunan Kelapa Sawit (原題)
Rinaldi Sinaga, Enny Susilowati Mardjono, Nila Tristiarini
🤖 gxceed AI 要約
日本語
本研究は、8カ国のパーム油プランテーション企業198社を対象に、企業のESG自己開示と外部メディア・NGO報道との乖離をNLPで定量化し、グリーンウォッシング深刻度指数（GSI）を構築した。ESG焦点度、感情分析、TF-IDF、トピックモデリングを統合し、規制成熟度やRSPO・ISPO・MSPO認証、上場の有無が乖離に与える影響を分析している。結果、グリーンウォッシングは中程度だが広範に存在し、RSPO認証企業や上場企業ほど叙述の整合性が高いことが示された。形式的な透明性だけでは信頼性あるESGコミュニケーションを保証できないと結論づける。
English
Using NLP on 198 palm oil plantation firms across eight producing countries, this study builds a Greenwashing Severity Index (GSI) measuring divergence between corporate ESG self-disclosure and external media/NGO narratives. It integrates ESG focus, sentiment analysis, TF-IDF, and topic modelling, linking divergence to regulatory maturity, RSPO/ISPO/MSPO certification, and listing status. Greenwashing is moderate but widespread, with certified and listed firms showing greater narrative alignment. Formal transparency alone does not ensure credible ESG communication.
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
日本企業にとって、ESG開示と外部評価の乖離を定量化する手法は、SSBJ基準や有報でのサステナビリティ開示の信頼性検証に応用可能。特に、認証・上場が開示の整合性を高めるという知見は、日本企業のESG格付対応や統合報告書の質的向上に示唆を与える。
In the global GX context
This work advances global disclosure scholarship by operationalizing greenwashing detection through NLP, offering a replicable method for assessing narrative credibility beyond formal compliance. It speaks directly to ISSB/CSRD assurance debates and the role of certification and market discipline in aligning corporate claims with external evidence, with relevance for transition finance and ESG rating integrity.
👥 読者別の含意
🔬研究者:NLPを用いたグリーンウォッシング定量化の手法と、認証・上場が開示整合性に与える実証的証拠を提供する。
🏢実務担当者:自社のESG開示と外部報道の乖離をモニタリングし、グリーンウォッシング批判を回避するための自己点検ツールとして活用できる。
🏛政策担当者:認証制度や上場規制が開示の信頼性を高める可能性を示し、ESG開示規制の設計や執行に示唆を与える。
📄 Abstract（原文）
The rapid expansion of corporate sustainability reporting in the palm oil sector has intensified demands for transparency, however, while simultaneously raising concerns about greenwashing driven by selective narrative disclosure. This study examines the credibility of Environmental, Social, and Governance (ESG) communication by comparing corporate sustainability reports with external media coverage across 198 palm oil plantation companies operating in eight major producing countries Indonesia, Malaysia, Thailand, Colombia, Nigeria, Papua New Guinea, Guatemala, and Honduras for the year 2025. Employing Natural Language Processing (NLP) techniques, the study constructs a Greenwashing Severity Index (GSI) that measures the divergence between corporate ESG self-representation and independent external narratives drawn from news media coverage and NGO reports. The index integrates ESG-specific focus measures, sentiment analysis, TF-IDF word weighting, and topic modelling to quantify imbalances in ESG communication. Findings reveal moderate yet widespread greenwashing across countries, company types, and plantation sizes, with substantial heterogeneity linked to regulatory maturity, certification adoption (RSPO, ISPO, MSPO), and stakeholder oversight. Greater narrative alignment is observed among RSPO-certified companies and exchange-listed plantation groups. Propensity score matching analysis further indicates that companies exhibiting unbalanced emphasis across ESG dimensions display significantly higher GSI values, consistent with strategic disclosure behaviour. Overall, findings suggest that formal transparency alone is insufficient to ensure credible ESG communication in the global palm oil industry.
🔗 Provenance — このレコードを発見したソース
- semanticscholar https://owner.polgan.ac.id/index.php/owner/article/download/3790/1990first seen 2026-10-06 05:47:35
🔔 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。