Technology decoupling sustainability in emergent markets ML model

Globalization is increasingly defined by technological decoupling (TD), as a factor of national sovereignty, where trade restrictions, information and communication technologies (ICT) barriers, and export controls reshape flows of knowledge and innovation. Yet the implications for sustainability, particularly in emerging markets, remain poorly understood. This paper integrates machine learning analysis with a systematic review of multinational corporations’ (MNCs) nonmarket reporting to investigate how decoupling affects environmental, social, and governance (ESG) outcomes. A country year panel (2014–2023) of over 200 indicators was build, separating sustainability targets from non-sustainability drivers. A Random Forest regression model, trained on 2014–2021 and tested on 2022–2023, achieves an out-of-sample R² of 0.527, MAE of 0.0297, and RMSE of 0.0446. Feature importance identifies high-tech manufacturing, knowledge-intensive employment, R&D personnel, venture capital, macroeconomic stability, and digital services restrictions as the dominant predictors of ESG trajectories. Qualitative analysis of GRI, SASB, CDP, and CPA Zicklin reports (2020–2025) showing MNCs predominantly engage in incremental nonmarket strategies: Modest CSR initiatives, selective CPA campaigns and partial transparency. Although often symbolic, these small steps compound into measurable ESG improvements over time, reflecting an ‘aggregation of marginal gains’ logic. Insights and findings re-frame nonmarket strategy as a persistent process of incremental adaptation, demonstrating that sustainability under decoupling is cumulative, nonlinear, and contested.