Synergies, Trade-offs, and Structural Pathways: A Directed Network Approach to SDG Prioritisation

Abstract

To successfully implement the Sustainable Development Goals (SDGs), it is necessary to understand the process by which the achievement of one goal has a spillover effect in a development system. While existing research studies synergies and trade-offs among the SDGs, most empirical approaches operate at the goal level, treat interactions as undirected, or prioritise indicators without accounting for structural redundancy. In this paper, we propose a direction-sensitive and indicator-level network approach to detect high-impact and diversified entry points for policy intervention. By using statistically significant lagged correlations, we build a directed weighted network of SDG indicators and assign them into groups based on the balance of their positive and negative spillovers. Systemic effects are measured by weighted out-degree Opsahl centrality, and flow-based clustering is used to detect frequent paths of high positive spillovers. Applying the framework to the Indian context, it is found that the interlinkages in the SDGs are highly asymmetric and structured in specific structural subsystems. Although synergies slightly outweigh the total, trade-offs are still embedded in the sectors. Notably, the most influential indicators are focused on a single pathway of propagation, suggesting that influence-based prioritisation itself could result in redundant system-wide impacts. A cluster-based prioritisation approach leads to a more diversified set of interventions, triggering multiple structurally independent channels of beneficial spillovers. The proposed framework combines directionality, trade-off embedding, and structural propagation analysis in a single framework, providing a scalable solution for country-level SDG prioritisation under resource constraints.

Figures (5)

• Figure 1: Overview of the methodological workflow for constructing directed SDG interlinkage networks, identifying synergy-dominated indicators, and prioritising high-impact indicators.
• Figure 2: Schematic illustration of flow-based clustering. Nodes of the same colour indicate clusters identified based on recurrent influence flows, highlighting coherent pathways of propagation rather than simple connection density.
• Figure 3: Heat map of statistically significant dominant lagged correlations among SDG indicators. Rows denote source indicators and columns denote target indicators. Indicators are ordered by SDG, with diagonal blocks corresponding to within-goal interlinkages (SDG 1 at top-left to SDG 17 at bottom-right). Off-diagonal blocks capture cross-goal indicator influences.
• Figure 4: Polar representation of synergy- and trade-off-dominated indicators across the SDGs. For each SDG, outward (inward) radial bars represent the number of synergy- (trade-off-) dominated indicators. The radial scale is discretised such that each unit of bar length corresponds to two indicators; thus, bar length directly encodes indicator counts.
• Figure 5: Flow-based multi-indicator clusters of the strong positive influence network. Green nodes represent synergy-dominated indicators and red nodes represent trade-off-dominated indicators. Among the green nodes, node size is proportional to Opsahl out-centrality and is scaled separately within each cluster; sizes are not comparable across clusters.