Weak governance keeps some EU regions stuck in SDG performance traps

Europe's progress toward the Sustainable Development Goals (SDGs) is being distorted by national averages that hide sharp regional divides, according to new research published in Sustainability. The study finds that high-performing and lagging regions across the European Union remain locked into persistent geographic clusters.

The study, titled "Sub-National SDG Progress and Spatial Inequality: A Composite Index Framework for Multi-Level Governance," analyzes 218 NUTS 2 regions across the EU from 2015 to 2022, using 1,744 region-year observations to build a sub-national SDG composite index and test how spatial inequality, governance quality and regional development patterns shape sustainable development outcomes.

National averages hide Europe's regional SDG fault lines

The study points out a major weakness in global SDG monitoring: most progress reports still rely heavily on national data. This approach can show whether a country is moving forward overall, but it often fails to reveal whether development gains are concentrated in only some regions.

The researchers argue that this blind spot is no longer a technical issue; has become a policy risk. If governments rely on national averages, regions trapped in low social, economic, educational or institutional performance can remain invisible. That can lead to broad national strategies that miss the areas most in need of targeted intervention.

To address this gap, the researchers created a composite SDG index designed for sub-national governance assessment. The index brings together four dimensions: social, economic, educational and institutional. The social dimension includes poverty risk and life expectancy. The economic dimension includes GDP per capita, employment and unemployment indicators. The education dimension uses tertiary education attainment. The institutional dimension relies on governance quality indicators.

The study originally considered broader SDG coverage, but data availability shaped the final framework. Some indicators, including youth not in employment, education or training and environmental variables, were excluded because valid NUTS 2-level data were insufficient. The authors replaced the environmental dimension with education, arguing that educational attainment is a foundational enabler of sustainable development because it shapes institutional capacity, economic resilience and green transition readiness.

The index was built through a transparent method using min-max normalization, equal weighting across four dimensions and robustness checks with alternative weighting methods. The results were stable. Rankings under equal weighting were strongly correlated with rankings using principal component analysis and entropy-based weighting. That finding matters because it suggests the regional gaps identified in the study are not artifacts of the index design.

The analysis found strong and persistent spatial clustering in EU SDG performance. Global Moran's I statistics remained high and statistically significant across every year from 2015 to 2022. This means regions with strong SDG performance tend to be near other strong performers, while weaker regions tend to be surrounded by similarly weak regions.

High-performing clusters are concentrated in Northern and Western Europe. Low-performing clusters are concentrated in Eastern and Southern peripheries. The study identifies 38 High-High regions and 48 Low-Low regions in 2022, while only a small number fall into outlier categories where a high-performing region borders low-performing neighbors or vice versa. This limited number of outliers signals strong spatial polarization rather than gradual transition zones.

The finding points to a core-periphery divide in European sustainable development. Some regions appear to benefit from accumulated institutional quality, stronger education systems, labor mobility, better infrastructure and more integrated innovation networks. Others appear caught in development traps shaped by geographical isolation, weaker governance and inherited institutional disadvantages.

Governance quality emerges as strongest regional predictor

The study's regression analysis identifies governance quality as the most consistent predictor of sub-national SDG performance. Across pooled, fixed-effects and spatial models, the governance index remained positive and statistically significant. This result clearly suggests that institutions matter. Regions with stronger governance capacity are better able to convert resources, education, employment and development programs into measurable SDG progress. Regions with weak governance may receive investment or policy support, but still fail to turn those inputs into sustained outcomes.

The spatial error model was the preferred specification after model selection tests. Its significant spatial error parameter indicates that neighboring regions are linked by unobserved common shocks, including shared institutional legacies, geographical conditions, cross-border dynamics and regional spillovers. In practical terms, the results show that SDG performance cannot be understood region by region in isolation. Geography and institutional context shape outcomes across borders.

GDP per capita did not remain a decisive predictor once spatial dependence and governance quality were accounted for. The authors caution that this does not mean economic output is irrelevant. Rather, it suggests that income alone does not explain regional SDG performance when institutional capacity and spatially linked development patterns are considered. Economic resources must pass through capable governance systems to produce broad sustainability gains.

Employment showed a positive association with SDG performance, reinforcing the role of labor-market strength in regional development. But the results again point to a wider story: jobs and income contribute to sustainable development when supported by institutions able to deliver services, coordinate policy and manage public investment effectively.

Education produced a more complex result. In the fixed-effects model, increases in tertiary education did not immediately translate into higher SDG performance. In the spatial model, regions with stronger education stocks performed better. The study interprets this contrast as evidence of time lags and institutional dependence. Education investment may not quickly raise SDG scores unless governance systems can convert human capital into better jobs, innovation, public services and institutional performance.

This finding is particularly important for EU regions attempting to catch up through education alone. The study suggests that human capital gains can stall without institutional reform. Regions may build education capacity but still fail to achieve stronger SDG outcomes if governance quality remains weak.

The study also warns that standard non-spatial models are not enough for SDG analysis. Ignoring spatial dependence risks biased conclusions because regional development outcomes are geographically connected. Policies aimed at one region may be affected by neighboring regions, shared labor markets, infrastructure corridors, institutional spillovers and regional investment networks.

Four regional pathways expose uneven convergence

Apart from identifying spatial clusters, the study uses k-means clustering to classify EU regions into four development archetypes: Disadvantaged, Leaders, Educated and Transitional.

The Disadvantaged group includes 51 regions, or 23.4 percent of the sample. These regions score low across social, economic, educational and institutional dimensions, with particularly weak education and governance scores. Their profile suggests compounded development problems rather than a single-sector weakness. For these regions, the study indicates that narrow interventions are unlikely to work. They need integrated packages combining institutional reform, social protection, education investment and economic support.

The Leaders group includes 17 regions, or 7.8 percent of the sample. These regions score highest on institutional quality and perform strongly across social and educational dimensions. The group represents the EU's most advanced sub-national units and highlights governance as a key differentiator. These regions are not only richer or better educated; they are better positioned institutionally to coordinate and sustain progress.

The Educated group includes 16 regions, or 7.3 percent of the sample. These regions show strong education performance but only moderate institutional quality. This profile points to a development pathway where human capital exists but is not fully converted into broader SDG gains. The implication is that education alone cannot drive sustainable development without stronger governance and institutional capacity.

The Transitional group is by far the largest, with 134 regions, or 61.5 percent of the sample. These regions show moderate social, economic and institutional performance but weaker education scores. The study presents this group as a broad middle tier that may be stuck in uneven convergence. These regions are not deeply disadvantaged, but they have not achieved multidimensional sustainability either.

Nearly two-thirds of EU NUTS 2 regions fall into a mid-level performance profile, suggesting that the EU's SDG challenge is not limited to a small set of lagging regions. It also involves a much larger group of regions that may appear stable but lack the educational and institutional depth needed for long-term sustainable convergence.

EU cohesion policy and structural funding should integrate sub-national SDG performance into eligibility and targeting systems. Low-Low cluster regions should receive more targeted support because their disadvantages are spatially concentrated and self-reinforcing, the research stresses.

The study also recommends making governance capacity a central condition for regional development transfers. Funding alone may not deliver results if local institutions cannot plan, implement and monitor development programs. Technical assistance, administrative capacity building and institutional reform should therefore become core parts of regional development strategy, not secondary add-ons.

The authors also call for stronger annual SDG reporting at the NUTS 2 level. Current Eurostat infrastructure provides a foundation, but gaps remain in several indicators. Without better sub-national data, policymakers will continue to depend on national indicators that conceal inequality within countries.

The study acknowledges several limitations. Spatial models were estimated on the 2022 cross-section rather than the full panel, meaning dynamic spatial processes require further study. Some indicators were excluded due to weak data availability. Governance indicators rely on survey waves that may smooth short-term changes. The authors also note that the findings apply to EU member states and should not be automatically extended to candidate countries or non-EU regions.