Measuring Greenhushing: A Firm-Level Index Based on Environmental Performance and Digital Communication

Abstract

As the global climate crisis intensifies, corporations face mounting pressure to demonstrate credible decarbonisation. While exaggerated environmental claims, known as greenwashing, are widely recognised, a countering phenomenon called greenhushing is emerging. Greenhushing occurs when environmentally progressive firms deliberately under-communicate or conceal substantive environmental initiatives to avoid scrutiny or stakeholder backlash. Despite its growing prevalence, this form of strategic silence remains empirically and methodologically underexplored. This dissertation addresses this gap by developing and validating a data-driven framework to quantify the performance-disclosure misalingment. The methodology employs a longitudinal dual-dataset design tracking 70 firms across 10 sectors from 2019 to 2024. Corporate environmental performance is assessed using emissions data from the World Benchmarking Alliance to calculate changes in carbon intensity. Conversely, corporate communication is evaluated by applying ClimateBERT, a fine-tuned natural language processing model, to contextually classify the environmental relevance of over 90,000 Instagram posts. These dimensions are integrated to construct the Greenhushing Index (GHI), a geometric index that isolates directional misalignments between climate action and public visibility relative to a baseline. Additionally, Dynamic Time Warping is applied to cluster firm-level strategic trajectories. The primary contribution is the GHI, a scalable diagnostic tool that distinguishes greenhushing and greenwashing from neutral alignment in a dynamic, high-frequency digital setting. The findings reveal that while neutral alignment is the dominant corporate posture, greenhushing cases consistently outnumber greenwashing ones and have intensified significantly by 2024. Greenhushing is highly context-dependent, appearing most prominently in high-impact sectors such as Utilities, Energy, and Industrials. In addition, the clustering results indicate that greenhushing can operate as a persistent, multi-year strategic trajectory rather than a static or episodic anomaly. These outcomes offer significant implications for both research and practice. By equipping researchers, investors, and regulators with an advanced analytical tool to detect systemic silence, the study enhances the critical evaluation of corporate transparency, ensures more accurate sustainability benchmarking, and reduces the risks that selective under-disclosure poses to collective climate action.