Unlocking the institutional foundations of green innovation: A machine learning analysis
Document Type
Research-Article
Journal Name
Technological Forecasting and Social Change
Keywords
Green innovation, Institutional elements, Machine learning, Shap values
Abstract
Institutional elements are crucial drivers of corporate green innovation. Although existing research has acknowledged the important role of institutional elements in corporate green innovation and explored their interrelations, a comprehensive understanding of their relative importance and nonlinear impacts remains limited. To address this gap, this study draws on institutional theory and employs multiple machine learning algorithms, along with SHAP value analysis, using data from Chinese A-share listed companies (2011−2022) to systematically assess the influence of regulative, normative, and cultural-cognitive elements on firms' green innovation. The findings reveal that, among institutional elements, cultural-cognitive elements exert the most significant influence on firm green innovation. Specifically, central inspections, public attention, and industry-level green orientation are the predominant factors within their respective institutional categories. Most institutional elements exhibit significant nonlinear relationships with green innovation. Further analysis indicates that cultural-cognitive elements can, under certain conditions, impede green innovation, whereas regulative and normative elements generally foster it. Moreover, the impact of institutional elements demonstrates considerable heterogeneity across different regions, industries, and firm sizes. This study highlights the importance and interplay of institutional elements in shaping firm green innovation, offering insights for emerging economies to tailor policies and support firms' sustainable transformation. © 2026 Elsevier Inc.