The Use of Visuals in Sustainability Reporting
サステナビリティ報告におけるビジュアルの活用 (AI 翻訳)
Amir Amel-Zadeh, Tami Dinh, Andreas Seebeck, Robin Wolter
🤖 gxceed AI 要約
日本語
本論文は、ディープラーニングを用いて3,923の欧州企業のサステナビリティ報告書（2013-2021年）を分析し、図表と写真の機能的役割の分離を実証した。ESGパフォーマンスの高い企業は図表を多く使用するが写真は使用せず、ESGパフォーマンスの低い企業は写真を戦略的に配置する傾向がある。図表の一致度は将来のESGインシデント減少と評価向上に関連する一方、写真は印象操作的な役割を果たす。
English
This paper uses deep learning to analyze visuals and text in 3,923 European sustainability reports (2013-2021), documenting a functional separation between graphics and photographs. Firms with stronger ESG performance use more graphics but not photographs, while poor performers strategically place photographs in sustainability-dense sections. Graphics congruence predicts fewer future ESG incidents and higher ratings, whereas photographs show weaker or opposite associations, highlighting the informative relevance of multimodal reporting.
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
本論文は、欧州データに基づくが、日本企業がサステナビリティ報告（有報・統合報告書）においてビジュアルを戦略的に活用する際の示唆を与える。特にSSBJ基準での開示充実が進む中、図表と写真の使い分けがESG評価やインシデント防止に与える影響は、日本の開示実務にも応用可能。
In the global GX context
This paper advances global sustainability disclosure scholarship by using AI to uncover strategic use of visuals in reporting. It shows that graphics and photographs serve different communicative functions, with graphics linked to genuine sustainability performance and photographs to impression management. This has implications for regulatory scrutiny (e.g., SEC, ISSB) and ESG rating methodologies that increasingly consider qualitative reporting.
👥 読者別の含意
🔬研究者:Empirical evidence on the signaling role of visuals in sustainability reports, using deep learning to analyze large-scale multimodal data.
🏢実務担当者:Insights into how graphics vs. photographs in sustainability reports affect stakeholder perceptions and ESG ratings; useful for designing more credible and effective reporting.
🏛政策担当者:Findings on visual congruence and its link to ESG outcomes can inform guidelines on the use of visuals in mandatory sustainability disclosures.
📄 Abstract（原文）
This paper documents a functional separation between graphics and photographs in sustainability reporting. Using deep-learning-based visual and textual analyses of 3,923 European reports from 2013 to 2021, we show that firms' use, placement, and congruence of visuals vary systematically with sustainability performance and the surrounding narrative. At the report level, firms with stronger sustainability performance are more likely to use graphics, but not photographs. At the micro level, sections with a stronger sustainability focus contain more graphical elements but fewer photographs, and photographs in these sections are less congruent with their surrounding narratives and use more green colors and salient designs. These patterns are amplified among poor ESG performers, suggesting that photographs play a more aesthetic, impression-related role, whereas graphics serve a primarily content-oriented communicative function. Consequences tests show that graphics, particularly when congruent, are associated with fewer future ESG incidents and higher subsequent ESG ratings, whereas photographs exhibit weaker or opposite associations. Across outcomes, particularly photographs are more likely to be strategically positioned in sustainability-dense disclosure passages that contain decision-relevant information, highlighting the informative relevance of multimodal sustainability reporting.
🔗 Provenance — このレコードを発見したソース
- openaire https://doi.org/10.2139/ssrn.6581518first seen 2026-06-11 05:10:32
🔔 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。