Interdependencies among SDGs: evidence-based insights for sustainable development indicators and policy

Achieving the Sustainable Development Goals (SDGs) requires a comprehensive understanding of their complex interdependencies to enhance synergies, reduce trade-offs, and inform effective sustainability indicators. However, most existing studies analyze SDGs in isolation or rely on simple correlations, which are insufficient for capturing the multidimensional relationships necessary for guiding sustainability policy and environmental management. This study develops a novel, three-stage empirical framework to explore and validate interlinkages among SDGs using cross-sectional data from the 2024 SDG dataset, covering 167 economies. First, expectation maximization (EM) clustering identifies performance patterns across countries, revealing latent structures in global SDG achievements. Second, Bayesian network with tree-augmented naïve Bayes (BN-TAN) classification identifies key probabilistic interdependencies among the 17 SDGs. Finally, structural equation modeling (SEM) estimates the strength and statistical significance of these directional associations, offering a robust basis for developing evidence-informed sustainability indicators. Key findings reveal that reducing inequalities (SDG 10) significantly drives poverty reduction (SDG 1), while decent work and economic growth (SDG 8) enhance food security (SDG 2). Additionally, strong institutions (SDG 16) promote marine ecosystem health (SDG 14). However, poverty alleviation (SDG 1) imposes pressures on marine ecosystems (SDG 14), illustrating a critical environmental trade-off. These results underscore the importance of integrated, cross-sectoral policy indicators that align socio-economic goals with ecological sustainability. By combining machine learning with statistical modeling of interdependencies, this study contributes a scalable methodology for monitoring SDG interdependencies and developing actionable indicators for sustainable environmental governance. • Reveals SDG interdependencies using a novel three-stage analytical framework. • Combines clustering, Bayesian networks, and SEM for structured SDG dependency analysis. • Identifies key SDG synergies and trade-offs to support sustainability policymaking. • Provides data-driven insights for integrated SDG implementation strategies. • Advances SDG assessment with machine learning and probabilistic modeling tools.