The motivation behind the index is grounded in hard geopolitical arithmetic. In 2023, 95 percent of the EU’s imports of rare earth elements came from just three countries: China accounted for 46 percent, Russia 28 percent, and Malaysia 20 percent. Energy tells a similar story of exposure and rapid, painful adjustment. The share of natural gas imported from Russia fell from 48 percent in early 2021 to 17 percent by 2025, with the United States now supplying nearly half of the bloc’s liquefied natural gas. Against this backdrop, the European Green Deal, the Circular Economy Action Plan and the Raw Materials Initiative have all reframed circularity not merely as an environmental strategy but as a pillar of what EU policy calls open strategic autonomy — a deliberate balance between global openness and reduced dependency.

What distinguishes the new research from earlier attempts to rank countries on circularity is its conceptual target. Existing composite indices, the authors argue, essentially ask how circular a country’s economy has become, relying heavily on environmental indicators such as recycling rates, resource efficiency and emissions control. The SRI instead measures resilience-enabling capacity: the structural, productive and innovation-related conditions that determine whether an economy can absorb shocks, adapt and transform. To build it, the researchers standardized seven candidate indicators — circular material use rate, energy import dependency, economic globalization, consumption-based greenhouse gas footprint, environmental patents per capita, the UNCTAD Productive Capacities Index, and an energy diversification measure — and applied principal component analysis, a statistical technique that compresses correlated variables into a smaller set of uncorrelated components.

The statistics delivered a strikingly clean result. A single dominant component explained 56.4 percent of the total variance, with all remaining components showing eigenvalues below one, and a factor analysis confirmed the identical structure. Internal consistency, measured by Cronbach’s alpha, reached 0.80, comfortably above the conventional 0.70 threshold. Energy indicators, despite their obvious theoretical relevance, refused to cohere with the rest: across extensive robustness testing that included renewable energy share and energy intensity, their loadings on the first component ranged only from 0.06 to 0.25, compared with 0.35 to 0.54 for the core five indicators. Energy performance, the authors conclude, constitutes an empirically distinct dimension — a separate energy resilience construct — rather than part of the structural capacity captured by the SRI.

The weighting structure of the final index is where the study delivers its headline finding. The formula reads SRI = 0.209·NCMU + 0.251·NEIG − 0.263·NGHGFP + 0.273·NIC + 0.322·NPCI, with the coefficients derived from the data itself rather than assigned by judgment. Productive capacity carries the heaviest weight at 0.322, followed by innovation capacity at 0.273 and economic globalization at 0.251. The circular material use rate contributes positively but modestly at 0.209, while the greenhouse gas footprint enters negatively at −0.263. Innovation and productive capacity also showed the strongest pairwise correlation in the raw data, at r = 0.741. The message is that the technological and industrial base of an economy — not its recycling bin — is currently the strongest statistical companion of sustainable resilience.

Applied across the EU-26, the index reveals a Union split into three tiers along a pronounced north-west to south-east gradient. The high-resilience group comprises the Netherlands, Denmark, Luxembourg, Germany, Belgium, Finland, Ireland and Sweden, with average scores between 62.0 and 87.9 on a rescaled 0–100 range. A broad middle tier, including Estonia, Austria, France, Czechia, Spain, Italy, Poland and Latvia, clusters around the EU-26 average of 45.7. At the bottom sit Portugal, Lithuania, Greece, Slovakia, Croatia, Bulgaria and Romania, with averages from 5.5 to 28.7 — countries where low circularity coexists with weak productive and innovative capacity, a pattern a circularity-only index would misread as a simple recycling gap.

The longitudinal picture is equally revealing. The cross-country standard deviation of SRI scores fell from 26.34 in 2010 to 18.72 in 2022, a decline of nearly 29 percent, signalling partial convergence driven mainly by catch-up among laggards. Croatia posted the largest gain at +18.83 points, followed by Latvia, Romania, Lithuania and Slovakia. Yet the convergence is partly a story of decline at the top: Finland lost 23.88 points and Luxembourg 22.57, with falling innovation capacity — not environmental deterioration — driving the slide in Finland and Denmark. Slovenia’s drop from mid-table to the lower tier was propelled by a sharp erosion of productive capacity, despite improving circularity and emissions performance. The EU-26 average itself barely moved, fluctuating between 43.8 and 47.2 over twelve years.

Perhaps the most compelling validation of the index comes from its response to crises. Using fixed-effects panel regressions with structural break tests, the researchers found statistically significant downward shifts in the SRI only around the two great global shocks of the period: the COVID-19 pandemic, which produced an immediate drop of 4.44 points followed by a recovery trend of 4.72 points per year, and the Russia–Ukraine war, which coincided with a 5.09-point decline. Neither the European debt crisis nor Brexit registered a significant panel-wide break. The index, in other words, appears to measure exactly what it claims: the capacity of economies to withstand systemic, exogenous disruption rather than localized political turbulence.

For policymakers, the implications cut against one-size-fits-all circularity mandates. In Eastern and South-Eastern Europe, the barriers to a circular transition are structural rather than regulatory, and the authors suggest embedding circular economy conditionalities within existing Cohesion Policy instruments such as the European Regional Development Fund and the Just Transition Fund, rather than pursuing separate funding lines. For mid-tier countries with productive capacity but uneven innovation, cross-border initiatives like the Interregional Innovation Investments mechanism could help. For the leaders losing ground, the priority is sustaining innovation intensity through instruments such as Horizon Europe partnerships. The study also points to accelerating the Critical Raw Materials Act and related regulations as the most direct current bridge between circularity and resilience policy.

So is circular the new resilient? The authors’ own answer, delivered with empirical precision, is: not yet, but increasingly so. The circular economy contributes meaningfully to sustainable resilience, and its weight is expected to grow as the EU pushes toward its target of doubling the circular material use rate by 2030 — an ambition whose full effects fall largely beyond the study’s 2022 cutoff. What the evidence demands, the researchers argue, is not a retreat from circular policy but better sequencing: countries lacking the productive and innovation infrastructure to scale circular strategies cannot be expected to harvest resilience gains from recycling targets alone. The Sustainable Resilience Index offers a benchmarking tool for identifying exactly where those enabling conditions are missing — and, as the EU’s resource dependencies tighten, that diagnostic may prove as valuable as the transition itself.

Subject of Research: Measuring the relationship between circular economy performance and sustainable economic resilience across EU member states

Article Title: Is circular the new resilient? Rethinking EU performance through a new sustainable resilience index

Article References: Chovancová, J., & Petruška, I. (2026). Is circular the new resilient? Rethinking EU performance through a new sustainable resilience index. Environmental and Sustainability Indicators, 32, Article 101509. https://doi.org/10.1016/j.indic.2026.101509

Image Credits: AI Generated

DOI: 10.1016/j.indic.2026.101509

Keywords: circular economy, economic resilience, Sustainable Resilience Index, European Union, principal component analysis, innovation capacity, productive capacity, critical raw materials, energy security, composite indicators, EU policy, convergence

The study’s central conceptual move is deliberately cautious. Rather than claiming that digitalization is inherently green, the authors — T.T. Pham, T.D. Nguyen, H.D. Luu, and colleagues — treat digital transformation as an enabling capability. They construct a five-dimensional Digital Transformation Indicator (DTI) covering digital infrastructure, digital adoption, digital innovation, digital governance, and digital industry development. The index deliberately does not measure environmental performance itself; a province can score high on digital capability while still performing poorly environmentally. What the DTI captures is the capacity to make resource use visible, lower coordination costs, and support the adoption of cleaner technologies when institutions and firms are able to use them.

Building the index required careful handling of Vietnam’s fragmented administrative data. Backbone indicators — such as fixed broadband subscriptions per 100 inhabitants, enterprise broadband adoption, active internet use, ICT-registered enterprise density, e-government readiness, and the digital economy’s share of provincial GDP — were standardized and combined using principal component analysis. The first component explains 63.7 percent of total variance, with loadings ranging from 0.38 to 0.47, and the index passes standard reliability diagnostics including a Kaiser-Meyer-Olkin measure of 0.83 and Cronbach’s alpha of 0.87. Alternative constructions using equal weights and entropy-weighted TOPSIS correlate at 0.971 and 0.924 with the baseline, confirming that no single dimension drives the findings.

The spatial picture that emerges is stark. Hanoi, Ho Chi Minh City, Binh Duong, and Da Nang consistently occupy the upper quartile of digital capability throughout the sample period, while provinces in the Northern Uplands and Central Highlands remain in the lowest. Yet relative rankings are remarkably stable — the rank correlation between 2010 and 2022 DTI values is 0.81 — even as absolute digital levels rise nationwide. Energy efficiency displays a parallel hierarchy, with industrially diversified Red River Delta and Southeast provinces outperforming resource-dependent northwestern and highland regions. Global Moran’s I statistics confirm significant positive spatial autocorrelation in both digital capability (0.341) and energy efficiency (0.287) in 2022.

To move beyond description, the researchers employed two-way fixed-effects panel models and Spatial Durbin Models (SDM), which allow both outcomes and explanatory variables to be spatially interdependent. In the preferred specification with full development controls — including log GDP per capita, industrial structure, human capital, FDI intensity, environmental regulation, and population density — a one-standard-deviation increase in DTI corresponds to a 0.049-unit difference in the energy-efficiency measure. Adding the full control vector attenuates the DTI coefficient by 12.3 percent, from 0.381 to 0.334, indicating that observable development gradients explain part, but not all, of the association. An instrumental-variable specification using distance to pre-2009 fiber-optic trunk lines yields a larger estimate of 0.512, but the authors treat this strictly as directional robustness evidence because the instrument’s exclusion restriction is contestable.

The spillover findings are arguably the most consequential. Under queen-contiguity weights, the SDM decomposition attributes a direct association of 0.312 and a spillover association of 0.187 to DTI, meaning 37.5 percent of the total association is spatially indirect. In plain terms, digital capability in one province is associated with better energy efficiency in neighboring provinces — a pattern consistent with infrastructure corridors, labor mobility, supply chains, and policy learning. Positive spillovers persist, though smaller, under inverse-distance and economic-proximity weight matrices, suggesting the phenomenon is not confined to shared borders. A non-spatial model, the authors note, would systematically understate the resource-sustainability relevance of digital investment.

Mechanism analysis points to innovation capacity as a key transmission channel. Provinces with higher DTI scores show stronger patent activity per capita, and when patent activity is included in the model, the DTI coefficient falls from 0.334 to 0.191. The bootstrapped pathway estimate of 0.144 yields a descriptive ratio of 43.1 percent; using R&D intensity instead produces an analogous figure of 34.7 percent. The authors are careful to label this an innovation-capacity pathway rather than a green-innovation mechanism, because the patent data cover all patents, not specifically environmental ones. These ratios describe mechanism-consistent covariance under sequential assumptions, not causal mediation shares.

Perhaps the most intriguing result is nonlinear. Using Hansen’s threshold regression, the study identifies a critical DTI value of 0.487 — near the 60th percentile of the distribution — above which the association with energy efficiency more than doubles, from 0.187 to 0.412. This is consistent with absorptive-capacity theory: digital capability pays off most when complementary skills, innovation assets, and administrative readiness have accumulated. Provinces such as Vinh Phuc, Khanh Hoa, and Can Tho sit near this benchmark, while Hanoi and Ho Chi Minh City operate well above it. The threshold is explicitly a Vietnam-specific benchmarking point, not a universal policy cutoff, and remains in the 55th to 63rd percentile range across alternative index constructions and outcome measures.

The authors are unusually candid about limitations. Provincial energy consumption is reconstructed from sectoral activity data rather than directly metered, with roughly 7 percent of activity cells requiring imputation; eight provinces with imputation rates above 15 percent were excluded in a sensitivity check, and Green Total Factor Productivity — computed via a Malmquist-Luenberger index treating CO2 as an undesirable output — serves as an alternative outcome with qualitatively consistent results. Placebo tests randomly permuting provincial DTI series place the observed coefficient at the 99.8th percentile of the placebo distribution. Still, the design remains observational, and the authors explicitly decline to claim causal policy effects.

The policy implications are differentiated rather than uniform. For provinces well below the readiness threshold, the evidence favors foundational investments — reliable broadband, basic digital skills, interoperable reporting standards — over advanced applications that cannot substitute for these bases. Provinces near the benchmark are candidates for bundled support combining energy-management systems, data interoperability, and technology-extension services. High-DTI provinces are better positioned to pilot real-time energy dashboards, digital permitting, and cross-agency analytics, ideally with transparent evaluation designs. And because spillovers operate through multiple proximity channels, the study supports harmonized data standards along shared corridors, infrastructure sharing among geographically proximate regions, and benchmarking platforms among economically similar provinces — a nuanced architecture that treats digital transformation not as an environmental end in itself, but as the connective tissue that determines whether green-growth commitments become measurable efficiency gains.

Subject of Research: The association between provincial digital transformation and resource efficiency in Vietnam

Article Title: Digital transformation as environmental implementation capacity: resource-efficiency associations, spatial spillovers, and threshold evidence from Vietnam

Article References: Pham, T., Nguyen, T., Luu, H., Pham, D., Tran, A., Le, K., & Nguyen, M. (2026). Digital transformation as environmental implementation capacity: resource-efficiency associations, spatial spillovers, and threshold evidence from Vietnam. Environmental Challenges, 25, Article 101645. https://doi.org/10.1016/j.envc.2026.101645

Image Credits: AI Generated

DOI: 10.1016/j.envc.2026.101645

Keywords: digital transformation, resource efficiency, energy efficiency, Vietnam, spatial spillovers, threshold effects, environmental implementation capacity, green growth, spatial econometrics, innovation capacity, absorptive capacity, provincial panel data