Optimizing Green GDP Accounting: An Entropy-Based Model and G20 Evidence

Abstract

Green GDP accounting has re-emerged as a practical instrument for aligning economic evaluation with environmental sustainability, yet cross-country measures often suffer from indicator arbitrariness and insufficient empirical validation. This study develops an entropy-based Green GDP accounting framework and applies it to the G20 over 2016–2020, using country-level economic and environment-related indicators retrieved from the Trading Economics database. First, an entropy-weighting scheme is implemented to generate a transparent composite Green GDP metric that integrates economic output with ecological constraints in a comparable manner across countries. Second, the robustness of the accounting results is assessed through three complementary validation strategies. Grey relational analysis (GRA) is used to examine consistency between the proposed Green GDP outcomes and benchmark sustainability-relevant indicators, yielding a high relational degree (0.850). Kendall’s coefficient of concordance further confirms strong agreement in country rankings (W = 0.925), indicating stable ordering and reduced sensitivity to single-indicator perturbations. Finally, partial least squares regression (PLSR) is employed as a predictive validation tool to evaluate how key environmental and development factors are associated with Green GDP performance, achieving satisfactory explanatory capacity (R² = 0.666) and identifying influential drivers with VIP > 1. Overall, the findings suggest that entropy-based Green GDP accounting provides a replicable and empirically validated alternative to conventional GDP-centric evaluation, supporting evidence-based policy design for sustainable growth. This research contributes to monitoring and policy implementation of SDG 8, SDG 12, and SDG 13.