A Hybrid MCDM and Clustering Framework for Evaluating Sustainable Competitiveness in OECD Countries

Abstract

Sustainable competitiveness has increasingly become an important policy objective for OECD countries, as economic performance is expected to be balanced with environmental protection, social well-being, and effective governance structures. The aim of this study is to evaluate and compare the sustainable competitiveness performance of OECD countries from a holistic perspective. In the analysis, six criteria reflecting the main dimensions of global sustainable competitiveness were considered. Criterion weights were calculated using the CRITIC (Criteria Importance Through Intercriteria Correlation) method, an objective weighting technique that does not rely on subjective judgments. These weights were then integrated with the CoCoSo (Combined Compromise Solution) method to rank the sustainable competitiveness performance of countries. In the final stage, a clustering analysis was applied to group OECD countries exhibiting similar sustainability characteristics. The findings indicate that natural capital emerges as the most influential dimension within the evaluation framework. According to the ranking results, Finland, Sweden, Lithuania, Denmark, and Estonia are positioned among the countries with the highest sustainable competitiveness performance. The results reveal noticeable differences across OECD countries, demonstrating that environmental, social, economic, and governance-related dimensions affect country performance in distinct ways. A direct comparison with the 2025 Global Sustainable Competitiveness Index shows a strong but not perfect association between the two rankings (Spearman’s ρ = 0.977), indicating structural consistency alongside meaningful mid-ranking shifts. Furthermore, the clustering results enable the identification of country groups sharing relatively similar sustainability profiles. Overall, the study contributes methodologically to the sustainable competitiveness literature by integrating objective weighting, multi-criteria decision-making, and clustering analysis within a unified analytical framework, while also offering insights for comparative policy evaluation.

1. Introduction

In recent decades, the concept of competitiveness has undergone a substantial transformation, shifting from purely economic performance toward a multidimensional framework that integrates environmental sustainability, social cohesion, and governance quality [1]. Traditional growth-based indicators are increasingly viewed as insufficient to explain long-term development outcomes, as countries’ resilience and prosperity depend not only on economic efficiency but also on institutional effectiveness, natural resource management, and human capital formation [2,3,4]. Within this evolving perspective, sustainable competitiveness has emerged as a critical policy paradigm, particularly for advanced economies seeking to balance economic growth with environmental and social sustainability.

OECD countries represent a unique empirical context for evaluating sustainable competitiveness due to their diverse economic structures, institutional frameworks, and policy priorities. Previous studies have highlighted that performance differences among OECD members cannot be explained solely by economic indicators but are strongly associated with governance effectiveness, social capital, and innovation capacity [5]. Consequently, composite sustainability assessments increasingly rely on multidimensional indicator systems and integrated analytical approaches to capture structural differences across countries.

The growing application of multi-criteria decision-making (MCDM) methods in sustainability research reflects the need to evaluate complex policy environments involving multiple and often conflicting criteria [6,7,8,9]. Objective weighting techniques such as CRITIC have been widely used to reduce subjectivity in determining criterion importance [10], while compromise-based ranking methods such as CoCoSo provide robust alternatives for evaluating relative performance under multiple dimensions [11,12]. Recent reviews emphasize that hybrid MCDM frameworks enhance decision reliability by combining weighting, ranking, and comparative evaluation mechanisms [13,14,15]. However, existing studies predominantly focus either on ranking procedures or on composite index construction, with limited attention given to integrating ranking outcomes with structural classification techniques such as clustering analysis.

Addressing this gap, the present study proposes an integrated analytical framework that combines CRITIC-based objective weighting, CoCoSo compromise ranking, and K-means clustering to evaluate sustainable competitiveness across OECD countries. By jointly examining performance rankings and structural similarity patterns, the study provides a more comprehensive understanding of how sustainability dimensions shape country-level competitiveness profiles. Unlike prior research that primarily emphasizes composite indices or single-method evaluations, this approach allows both relative performance assessment and group-level structural interpretation.

Beyond methodological integration, this study contributes analytically by re-evaluating the six structural dimensions of sustainable competitiveness under a fully data-driven weighting structure. Unlike the predefined composite weighting scheme embedded in the original Global Sustainable Competitiveness Index (GSCI), the CRITIC-based approach allows the relative informational contribution of each dimension to emerge endogenously from dispersion and inter-criteria correlation patterns. This enables the identification of structural dominance among sustainability dimensions and reveals whether the normative balance implied in composite index construction is empirically supported. In this sense, the study does not replicate the GSCI ranking; rather, it critically reinterprets its dimensional structure under an alternative analytical logic.

In addition to its methodological contribution, this study is closely aligned with the United Nations Sustainable Development Goals (SDGs) [16], particularly SDG 9 (Industry, Innovation and Infrastructure), SDG 12 (Responsible Consumption and Production), SDG 13 (Climate Action), and SDG 16 (Peace, Justice and Strong Institutions). By integrating key sustainability dimensions—including governance quality, human and intellectual capital, natural capital, and efficiency-related indicators—the proposed model provides a multidimensional perspective on sustainable competitiveness. From this perspective, the findings contribute to the broader sustainability transition debate by highlighting how institutional effectiveness and innovation capacity shape long-term sustainable development outcomes across OECD countries.

The remainder of the paper is structured as follows. Section 2 reviews the relevant literature on sustainable competitiveness and MCDM-based evaluation frameworks. Section 3 describes the dataset, methodological design, and analytical procedures. Section 4 presents the empirical findings, followed by a discussion of policy implications in Section 5. Section 6 concludes the study with key contributions and future research directions.

Based on the multidimensional nature of sustainable competitiveness and prior OECD performance research, the study formulates three hypotheses. First, governance- and knowledge-related dimensions are expected to exert stronger effects on sustainable competitiveness compared with resource-oriented indicators (H₁). Second, countries characterized by higher institutional effectiveness and social capital are anticipated to achieve higher competitiveness rankings (H₂). Third, OECD countries are expected to form structurally distinct sustainability clusters reflecting differences in governance, innovation capacity, and socio-economic structures (H₃). These hypotheses aim to provide a theoretically grounded framework for interpreting the empirical findings derived from the integrated MCDM and clustering approach.

2. Literature Review

Sustainable competitiveness refers to a country’s ability to achieve long-term economic performance while maintaining environmental sustainability, social cohesion, and institutional quality. Unlike traditional competitiveness frameworks focused primarily on productivity and macroeconomic indicators, sustainable competitiveness emphasizes multidimensional structural capacity, including natural capital management, governance effectiveness, and knowledge-based development [17,18]. Sustainable competitiveness therefore represents a structural synthesis of environmental, institutional, and socio-economic capacities rather than a purely economic performance measure. Sustainable competitiveness has been discussed in the academic literature as a multidimensional construct linking environmental performance, institutional quality, and long-term economic resilience. Previous empirical studies have emphasized the importance of governance effectiveness, innovation capacity, and resource efficiency in explaining cross-country sustainability differences. In this context, composite sustainability indices have been increasingly used as analytical tools to capture structural development patterns across countries. Recent empirical studies have further examined the relative importance of governance quality, intellectual capital, and environmental dimensions in shaping sustainable competitiveness structures across countries [19,20].

In recent years, composite sustainability indicators have been widely used to evaluate country performance across complex and interrelated dimensions. The Global Sustainable Competitiveness Index (GSCI), developed by SolAbility, measures national sustainability performance across six core dimensions: natural capital, resource intensity, social capital, intellectual capital, economic sustainability, and governance. These dimensions provide a multidimensional framework for assessing long-term structural resilience [17].

Building on this conceptual foundation, the present study applies an integrated CRITIC–CoCoSo and clustering framework to evaluate sustainable competitiveness across OECD countries using the six underlying sustainability dimensions reported in the GSCI framework [18]. Selected studies on the subject are shown in

Table 1. Selected Studies on Sustainable Competitiveness and Composite Indicator Approaches.

Author(s)/Year	Objective(s)	Method(s)	Findings/Results
Preston (1975) [2]	To examine the relationship between income level and life expectancy across countries.	Comparative analysis across countries. Income level and life expectancy data.	Income positively correlates with life expectancy; however, non-economic factors such as health policy and technology explain a substantial share of improvements.
Diakoulaki et al. (1995) [10]	To develop the CRITIC method for objectively determining MCDM criterion weights. Specifically, to reduce the subjective weighting problem by using information on inter-criterion correlation and variance.	CRITIC Objective weight generation based on standard deviation + inter-criteria correlation matrix Application through a multi-criteria firm performance example.	The CRITIC method has produced more stable results compared to alternative objective weighting methods. Its similarity in rankings to PCA supports the reliability of the method.
Opricovic & Tzeng (2004) [21]	Identifying the closest alternative to the ideal solution under conflicting criteria.	VIKOR + Compromise ranking	It provides a balancing solution in multi-criteria rankings. It is applicable in policy analysis.
Jain (2010) [22]	This comprehensive review examines the historical development of data clustering methods, key algorithms (especially K-means), and current trends.	Review article + Classification of clustering algorithms + Theoretical and methodological comparison	There is no single “best” clustering algorithm; method selection depends on the data structure. K-means is still a widely used basic method. Clustering results are strongly influenced by the objective and data representation.
Joumard et al. (2010) [5]	Comparing performance and policy effectiveness in OECD countries.	Cross-country indicator analysis Composite performance indicators	A country’s performance depends on its multifaceted structure.
Mardani et al. (2015) [6]	To analyze how MCDM techniques were used in sustainability and performance evaluations in different fields during the period 2000–2014.	Systematic literature review + Classification of MCDM methods + Comparative analysis according to application areas	It was determined that MCDM techniques are widely used in sustainability and performance evaluations, it was emphasized that hybrid and objective weighting approaches increase the reliability of decision outcomes, and it was concluded that the choice of method depends on the problem structure.
Lorenz et al. (2017) [23]	Analyzing the weighting structures that enable countries to achieve the best rankings in the OECD Better Life Index.	Composite index analysis + Rank-optimal weighting + OECD country comparison	The ranking of countries is strongly influenced by weighting. Weighting plays a critical role in multidimensional welfare indicators.
Popescu et al. (2017) [19]	To measure sustainable competitiveness across European economies and analyze structural sustainability determinants.	Composite index construction + Multidimensional sustainability indicators + Cross-country comparative analysis.	Sustainable competitiveness is shaped by governance quality, innovation capacity, and environmental performance. Institutional strength significantly influences sustainability rankings.
Yazdani et al. (2019) [11]	To develop a new MCDM algorithm, CoCoSo, and to offer an alternative sorting approach to existing MCDM methods.	Development of a new MCDM-based algorithm + hybrid computing derived from WASPAS and gray relational approach and comparison with other methods	The CoCoSo method showed high correlation with different MCDM techniques. Sensitivity analyses revealed that the results were stable.
Kerras et al. (2020) [24]	To examine the impact of the gender digital divide on sustainable development goals and to conduct a comparative analysis between the EU and Maghreb countries.	Comparative country analysis + Multiple linear regression analysis + Empirical analysis via indices and indicators.	The gender digital divide negatively impacts sustainable development. ICT infrastructure and access to technology improve SDG performance. EU countries perform better in terms of technology and equality.
Stević et al. (2020) [13]	Evaluating the performance of countries using MCDM methods.	MCDM hybrid models + CRITIC/Entropy vb.	Objective weighting methods improve ranking stability.
Chakraborty et al. (2023) [14]	The use of MCDM techniques in sustainability assessments.	Hybrid MCDM + Comparative ranking	Hybrid models produce more reliable rankings than single methods.
Qazi & Al-Mhdawi (2024) [20]	To examine the relative importance of sustainable competitiveness pillars and their contribution to national performance.	Empirical analysis of sustainability dimensions + Structural pillar comparison.	Governance and intellectual capital dimensions play a critical role in shaping sustainable competitiveness structures. Multidimensional assessment improves policy relevance.

.

When the studies summarized in Table 1 are examined, it is seen that the literature has largely developed along two main axes. The first group of studies offers methodological contributions to decision-making processes by focusing on the development of MCDM methods and weighting approaches [6,10,11,21]. The second group of studies examines performance comparisons between countries using indicators such as sustainability, well-being, and digital inequality [2,5,13,14,23,24]. A third stream of research directly addresses the structural determinants of sustainable competitiveness, emphasizing governance quality, innovation capacity, and environmental dimensions in shaping national performance [17,20].

Beyond methodological developments, recent global sustainability assessments highlight the increasing importance of governance quality and multidimensional indicators in evaluating national sustainability performance [25,26].

However, much of the current research focuses either solely on ranking methods or sustainable performance indicators, failing to address objective weighting, consensus ranking, and clustering analysis within an integrated framework.

From a theoretical perspective, sustainable competitiveness can be understood as a structural synthesis of ecological capacity, institutional quality, and knowledge-based development. Institutional economics emphasizes the role of governance effectiveness in shaping long-term economic resilience, while ecological modernization theory underlines the integration of environmental sustainability within growth dynamics. Similarly, endogenous growth theory highlights the contribution of intellectual capital and innovation capacity to sustainable development trajectories. By embedding these theoretical perspectives within a multidimensional evaluation framework, this study positions sustainable competitiveness as a structural equilibrium problem rather than a simple aggregate performance index.

This study, unlike existing approaches in the literature [6,10,11,21] presents an integrated evaluation model by combining CRITIC-based objective weighting, the CoCoSo ranking method, and clustering analysis under the same methodological framework to assess the sustainable competitiveness performance of OECD countries.

Accordingly, in order to evaluate sustainable competitiveness performance within a multidimensional framework, the CRITIC method is used as the objective weighting approach, the CoCoSo method as the consensual ranking approach, and clustering analysis is used together to reveal the similarity structure between countries. The following section describes the dataset, criterion structure, and applied methodological steps in detail.

3. Materials and Methods

3.1. Data Set

This study considered a total of 6 indicators for which complete data was available from the 2025 Global Sustainable Competitiveness Index (GSCI) report [18]. The dimensions of sustainable competitiveness and their corresponding indicators, determined based on literature and data availability, are presented in

Table 2. Dimensions, Codes, Directions and Indicators Used in the Analysis.

Dimensions	Codes	Directions	Indicators
Natural Capital	C1	Max	Biodiversity and ecosystem preservation; Forest and land use efficiency; Water resource availability and quality; Air quality and environmental carrying capacity
Resource Intensity	C2	Max	Energy intensity of the economy; Material consumption efficiency; Carbon emissions intensity; Waste generation and recycling performance
Social Capital	C3	Max	Health outcomes and life expectancy; Education attainment and quality; Income distribution and social inclusion; Labor market participation and social cohesion
Intellectual Capital	C4	Max	Research and development (R&D) expenditure; Innovation capacity and patent activity; Human capital and skills development; Digital infrastructure and knowledge diffusion
Economic Sustainability	C5	Max	Macroeconomic stability; Productivity and competitiveness; Investment and savings capacity; Long-term economic resilience
Governance	C6	Max	Institutional quality and regulatory effectiveness; Rule of law and corruption control; Government efficiency and policy implementation; Political stability and accountability

.

Table 2 presents the dimensions, coding structure, optimization directions, and indicator definitions used in constructing the sustainable competitiveness framework. The analysis is based exclusively on the six fundamental dimensions of the Global Sustainable Competitiveness framework as reported by SolAbility (2025) [18]. These dimensions constitute the evaluation criteria used in the CRITIC–CoCoSo analysis.

3.2. Methodology

The study included 38 countries with data from the Global Sustainable Competitiveness Index [18]. The weights of the six indicators were calculated using the CRITIC method. The analysis process followed in the study was structured within a multi-stage framework consisting of indicator selection, data collection, criterion weighting, ranking of alternatives, and cluster analysis. The relationship between these stages and the analysis flow is visually presented in Figure 1.

As shown in Figure 1, the research methodology begins with a literature review and indicator identification phase. This is followed by data collection, the determination of criterion weights using the CRITIC method, and the ranking of countries’ sustainable competitiveness performance using the CoCoSo method. Finally, K-means clustering analysis is applied to identify groups of countries with similar performance profiles. This integrated approach allows for the simultaneous evaluation of both performance rankings and structural similarities. The statistical analyses were performed using IBM SPSS Statistics (Version 23, IBM Corp., Armonk, NY, USA) and Microsoft Excel (Microsoft Corp., Redmond, WA, USA).

3.2.1. CRITIC Method

The CRITIC (Criteria Importance Through Intercriteria Correlation) method is a technique developed to determine the objective weights of criteria in multi-criteria decision-making problems. The method considers the standard deviations of the criteria and the correlation between them when determining the importance levels of the criteria [10]. The steps of the CRITIC method are as follows [10]:

Step 1:

Creating the Decision Matrix

The decision matrix contains the criterion values corresponding to different alternatives. The decision matrix X is shown in Equation (1).

$$X={\left[x_{ij}\right]}_{mxn}=\left[\begin{array}{ccc}{x}_{11}& \dots & x_{1n}\\ ⋮& \ddots & ⋮\\ x_{m1}& \dots & x_{mn}\end{array}\right] \mathrm{i}=\mathrm{1,2},\dots ,\mathrm{m}\mathrm{j}=\mathrm{1,2},\dots ,\mathrm{n}$$

(1)

Step 2:

Creating the Normalized Decision Matrix

The decision matrix is normalized depending on whether the criteria are focused on maximization or minimization.

$$r_{ij}={\displaystyle \frac{x_{ij}-x_j^{min}}{x_j^{max}-x_j^{min }}} \mathrm{max-focused}$$

(2)

$$r_{ij}={\displaystyle \frac{x_j^{max}-x_{ij}}{x_j^{max}-x_j^{min }}} \mathrm{min-focused}$$

(3)

Step 3:

Creating the Correlation Matrix

A correlation matrix is created using the data obtained from the normalized decision matrix (Equation (4)).

$${\rho }_{jk}={\displaystyle \frac{{\sum }_{i=1}^m\left(r_{ij}-{\overline{r}}_j\right)\left(r_{ik}-{\overline{r}}_k\right)}{\sqrt{{\sum }_{i=1}^m{\left(r_{ij}-{\overline{r}}_j\right)}^2{\sum }_{i=1}^m{\left(r_{ik}-{\overline{r}}_k\right)}^2}}} j,k=\mathrm{1,2},\dots ,n$$

(4)

Step 4:

Calculating the $C_j$ value

$C_{ j}$ is calculated using the standard deviations of the columns in the normalized decision matrix. Equation (5) is used for the ${\sigma }_j$ value and Equation (6) is used for the ${ C}_j$ value.

$${\sigma }_j=\sqrt{{\displaystyle \frac{{\sum }_{i=1}^m{\left(r_{ij}-{\overline{r}}_j\right)}^2}{m-1}}}$$

(5)

$$C_j={\sigma }_j{\sum }_{k=1}^n(1-{\rho }_{jk}) j=\mathrm{1,2},\dots n$$

(6)

Step 5:

Calculating Criterion Weights

The objective weights of the criteria are calculated (Equation (7)).

$$W_j={\displaystyle \frac{C_j}{{\sum }_{k=1}^n{C}_k}} j,k=\mathrm{1,2},\dots ,n$$

(7)

3.2.2. CoCoSo Method

The CoCoSo method was developed by Yazdani et al. [11]. The CoCoSo method combines weighted sum and weighted product models in multi-criteria decision-making problems, providing a more balanced evaluation of the performance of alternatives. This integrated structure produces more stable and consistent results compared to methods using only a single computational model [11,12]. The steps for the CoCoSo method are as follows [11,12]:

Step 1:

Creating the decision matrix.

$$X={\left[x_{ij}\right]}_{mxn}=\left[\begin{array}{ccc}{x}_{11}& \dots & x_{1n}\\ ⋮& \ddots & ⋮\\ x_{m1}& \dots & x_{mn}\end{array}\right] \mathrm{i}=\mathrm{1,2},\dots ,\mathrm{m} \mathrm{j}=\mathrm{1,2},\dots ,\mathrm{n}$$

(8)

Step 2:

Creating the Normalized Decision Matrix.

The decision matrix is normalized depending on whether the criteria are focused on maximization or minimization.

$$r_{ij}={\displaystyle \frac{x_{ij}-{min}_i{x}_{ij}}{{{max}_{i}x}_{ij}-{min}_i{x}_{ij}}} \mathrm{max-focused}$$

(9)

$$r_{ij}={\displaystyle \frac{{max}_i{x}_{ij}-x_{ij}}{{{max}_{i}x}_{ij}-{min}_i{x}_{ij}}} \mathrm{min-focused}$$

(10)

Step 3:

Creating the weighted normalized matrix.

$$v_{ij}=w_j\times r_{ij}$$

(11)

Step 4:

Calculation of the weighted sum score $S_i$ and the weighted product score ${ P}_i.$

$$S_i={\sum }_{j=1}^n{v}_{ij}$$

(12)

$$P_i={\sum }_{j=1}^n{\left(r_{ij}\right)}^{w_j}$$

(13)

Step 5:

Calculation of three different composite scores.

$$k_{ia}={\displaystyle \frac{P_i+S_i}{{\sum }_{i=1}^m\left(P_i+S_i\right)}}$$

(14)

$$k_{ib}={\displaystyle \frac{S_i}{{min}_i{S}_i}}+{\displaystyle \frac{P_i}{{min}_i{P}_i}}$$

(15)

$$k_{ic}={\displaystyle \frac{\lambda S_i+\left(1-\lambda \right)P_i}{\lambda {max}_i{S}_i+\left(1-\lambda \right){max}_i{P}_i}} , 0\le \lambda \le 1$$

(16)

Decision-makers generally use $\lambda$ = 0.5.

Step 6:

Calculating the final score.

$$k_i={\left(k_{ia}{k}_{ib}{k}_{ic}\right)}^{{\displaystyle \frac{1}{3}}}+{\displaystyle \frac{1}{3}} \left(k_{ia}+k_{ib}+k_{ic}\right)$$

(17)

3.2.3. Cluster Analysis

K-means clustering is one of the most widely used partition-based clustering algorithms, aiming to divide observations into a predetermined number of clusters. The main goal of the method is to maximize similarity between observations within the same cluster while maximizing the difference between clusters [22,27]. Prior to the clustering procedure, all variables were standardized using z-scores in order to eliminate potential scale effects and ensure comparability among criteria.

The mathematical operation of the K-means algorithm consists of the following steps:

Step 1:

Determining the number of clusters.

First, the number of clusters, k, is determined.

Step 2:

Assigning the initial cluster centers.

$$\mathrm{\mu }1,\mathrm{\mu }2,\dots ,\mathrm{\mu }k$$

Step 3:

Assigning observations to the nearest cluster center.

Each observation $x_i$ is assigned to the nearest cluster center using Euclidean distance.

Step 4:

Updating cluster centers.

For each cluster, a new center is calculated by averaging the observations belonging to that cluster.

Step 5:

Minimizing the objective function.

The k-means algorithm aims to minimize the within-cluster sum of squares. In this process, the objective function is based on minimizing the within-cluster sum of squares and can be expressed with the following general form:

$$su{m}_{k=1}^{K}su{m}_{x_{i}in C_k}{{\displaystyle |}\left|x_i- m{u}_k\right|{\displaystyle |}}^2$$

Step 6:

Checking the convergence criterion.

Step 3 and Step 4 are repeated until the cluster centers remain unchanged or the improvement in the objective function falls below a certain threshold value.

4. Findings

In the first stage of the analysis, a decision matrix was created using sustainable competitiveness indicators of OECD countries, and criterion-based performance values were presented in

Table 3. Decision Matrix of OECD Countries Based on Six Criteria (C1–C6).

Country	Criteria
	C1	C2	C3	C4	C5	C6
Finland	61.35	54.17	53.94	65.05	57.77	70.22
Sweden	55.03	60.41	52.51	67.42	56.76	69.43
Denmark	43.61	60.06	53.42	67.15	60.35	73.53
Estonia	52.02	50.25	54.59	67.05	61.59	70.77
Switzerland	42.70	61.75	53.40	68.30	58.79	69.94
Austria	48.26	56.66	52.94	63.92	62.22	69.85
Lithuania	56.48	55.37	51.07	56.57	62.54	70.74
Iceland	52.24	52.66	56.06	63.16	59.83	68.17
Norway	47.74	56.72	58.13	59.42	54.42	73.84
Latvia	54.28	55.17	51.47	57.42	62.00	68.17
Slovenia	44.70	50.39	58.01	61.36	63.57	70.39
Luxembourg	43.86	57.79	51.10	62.49	59.55	72.01
Ireland	45.97	54.51	55.08	56.21	63.81	70.29
Germany	43.69	57.33	53.40	65.13	55.70	69.79
Portugal	49.15	57.47	53.27	57.13	57.62	69.78
Poland	50.98	50.77	55.11	60.76	58.73	67.02
United Kingdom	43.98	61.90	47.50	67.58	56.59	61.61
Netherlands	38.37	56.49	56.03	59.55	54.54	73.59
Japan	50.65	51.10	55.90	66.32	50.77	63.10
Slovak Republic	54.21	48.74	54.48	56.20	59.01	61.94
Spain	48.85	55.07	52.11	57.36	52.85	67.03
Korea, Rep.	47.98	49.17	51.78	68.32	53.05	62.82
France	42.37	58.64	51.34	61.55	51.43	66.89
New Zealand	50.44	52.43	49.20	56.56	59.74	63.78
Czech Republic	41.77	51.16	53.63	59.32	62.02	64.18
Australia	52.57	53.66	49.65	58.43	51.10	66.52
Italy	43.82	53.46	54.27	59.60	55.21	64.54
Belgium	39.93	51.68	54.02	60.70	54.97	69.28
Canada	59.49	54.79	44.38	55.12	49.61	62.84
Israel	36.61	51.87	50.24	67.42	55.16	58.64
Greece	46.18	48.83	49.73	54.02	53.09	65.01
Hungary	46.75	51.34	47.09	53.03	59.11	59.39
United States	49.81	49.19	37.03	63.27	53.01	61.91
Costa Rica	47.72	52.43	44.19	47.08	64.47	57.79
Chile	44.93	47.77	40.11	50.61	50.72	63.20
Colombia	55.43	51.51	36.60	43.99	54.54	52.54
Türkiye	46.93	44.59	38.27	56.21	51.79	55.16
Mexico	40.85	46.80	41.87	44.99	50.08	52.80

Source: Authors’ compilation based on the Global Sustainable Competitiveness Index 2025 (Solability, 2025) [18].

. This matrix forms the basic data structure for calculating criterion weights using the CRITIC method and performing ranking analysis using the CoCoSo method.

Table 3 presents the decision matrix constructed from the sustainable competitiveness indicators of OECD countries. The matrix includes six criteria (C1–C6) and serves as the input for the CRITIC weighting procedure and the subsequent CoCoSo ranking analysis. The objective weights obtained from the CRITIC method are reported in

Table 4. Weights of criteria.

C1	C2	C3	C4	C5	C6
0.215178	0.148065	0.142628	0.162869	0.201553	0.129707

.

As shown in Table 4, the criterion weights were obtained using the CRITIC method. These weights were then used as the input for the CoCoSo ranking analysis. The ranking results for the standard scenario (λ = 0.5) are presented in

Table 5. Ranking Results (λ = 0.5).

Countries	Si	Pi	kia	kib	kic	ki	Ranking
Finland	0.771206	5.724661	0.029703	9.507301	1	4.168411	1
Sweden	0.757527	5.71012	0.029574	9.371915	0.995656	4.116769	2
Denmark	0.733253	5.637908	0.029133	9.113933	0.980802	4.013216	4
Estonia	0.729432	5.65406	0.029189	9.08384	0.9827	4.003956	5
Switzerland	0.704184	5.584897	0.028758	8.817783	0.968166	3.897733	7
Austria	0.722575	5.668646	0.029225	9.02434	0.98389	3.983648	6
Lithuania	0.731318	5.681436	0.029323	9.112211	0.987205	4.017576	3
Iceland	0.696005	5.634649	0.028948	8.759169	0.974566	3.881751	9
Norway	0.641403	5.515902	0.028155	8.195563	0.94788	3.659708	13
Latvia	0.695748	5.640751	0.028975	8.759069	0.975466	3.882411	8
Slovenia	0.676056	5.547273	0.028457	8.536447	0.958045	3.789424	10
Luxembourg	0.649248	5.540665	0.028304	8.279486	0.9529	3.693588	12
Ireland	0.671208	5.572013	0.028548	8.499886	0.961107	3.778717	11
Germany	0.610959	5.470542	0.027809	7.88929	0.936211	3.541119	16
Portugal	0.631195	5.545604	0.028244	8.110035	0.950881	3.631393	14
Poland	0.624507	5.525402	0.028121	8.038834	0.946742	3.60273	15
United Kingdom	0.592147	5.426876	0.027523	7.694057	0.926593	3.463818	17
Netherlands	0.545044	5.226974	0.026393	7.170606	0.888568	3.247157	21
Japan	0.535158	5.2323	0.026373	7.078818	0.887866	3.213665	22
Slovak Republic	0.573491	5.417576	0.027395	7.513444	0.922289	3.395767	18
Spain	0.520621	5.336412	0.026782	6.980639	0.901655	3.188767	23
Korea, Rep.	0.51085	5.27332	0.026449	6.863786	0.890438	3.138302	25
France	0.497586	5.201126	0.026058	6.710311	0.877283	3.073198	27
New Zealand	0.560707	5.442096	0.027449	7.401483	0.924096	3.356944	19
Czech Republic	0.555704	5.363749	0.027068	7.324056	0.911265	3.319433	20
Australia	0.504897	5.244367	0.026289	6.796228	0.885065	3.109958	26
Italy	0.509101	5.324838	0.026677	6.866874	0.8981	3.145163	24
Belgium	0.491341	5.21944	0.026113	6.658034	0.879141	3.05577	28
Canada	0.474981	4.563287	0.023038	6.252002	0.775611	2.831837	31
Israel	0.421988	4.480854	0.022419	5.717537	0.754763	2.623976	33
Greece	0.396826	5.104124	0.025154	5.716929	0.846838	2.691974	32
Hungary	0.446417	5.225619	0.025936	6.23403	0.873176	2.89842	30
United States	0.389225	4.873248	0.024063	5.556552	0.810126	2.60694	34
Costa Rica	0.468148	5.141179	0.025649	6.408005	0.863522	2.954017	29
Chile	0.247083	4.657409	0.022427	4.125009	0.755017	2.045978	35
Colombia	0.28978	2.616581	0.01329	3.75028	0.447417	1.685132	37
Türkiye	0.228219	3.858509	0.018687	3.640652	0.629128	1.77929	36
Mexico	0.105364	3.89669	0.0183	2.48923	0.616092	1.3451	38

. To examine the parameter sensitivity, the analysis process was repeated for λ = 0.25 and λ = 0.75 values, and the results are presented in

Table 6. Ranking Results (λ = 0.25).

Countries	Si	Pi	kia	kib	kic	ki	Ranking
Finland	0.771206	5.724661	0.029703	9.507301	1	4.168411	1
Sweden	0.757527	5.71012	0.029574	9.371915	0.996807	4.117404	2
Denmark	0.733253	5.637908	0.029133	9.113933	0.983382	4.014636	4
Estonia	0.729432	5.65406	0.029189	9.08384	0.985869	4.005698	5
Switzerland	0.704184	5.584897	0.028758	8.817783	0.9729	3.90033	7
Austria	0.722575	5.668646	0.029225	9.02434	0.987926	3.985864	6
Lithuania	0.731318	5.681436	0.029323	9.112211	0.990551	4.019415	3
Iceland	0.696005	5.634649	0.028948	8.759169	0.980762	3.885143	9
Norway	0.641403	5.515902	0.028155	8.195563	0.957867	3.665146	13
Latvia	0.695748	5.640751	0.028975	8.759069	0.981767	3.88586	8
Slovenia	0.676056	5.547273	0.028457	8.536447	0.965043	3.793251	10
Luxembourg	0.649248	5.540665	0.028304	8.279486	0.962444	3.698788	12
Ireland	0.671208	5.572013	0.028548	8.499886	0.968909	3.782978	11
Germany	0.610959	5.470542	0.027809	7.88929	0.948588	3.547833	16
Portugal	0.631195	5.545604	0.028244	8.110035	0.962264	3.637578	14
Poland	0.624507	5.525402	0.028121	8.038834	0.958514	3.609123	15
United Kingdom	0.592147	5.426876	0.027523	7.694057	0.94024	3.471205	17
Netherlands	0.545044	5.226974	0.026393	7.170606	0.904196	3.255584	21
Japan	0.535158	5.2323	0.026373	7.078818	0.904535	3.222638	22
Slovak Republic	0.573491	5.417576	0.027395	7.513444	0.937645	3.404058	18
Spain	0.520621	5.336412	0.026782	6.980639	0.92113	3.199208	23
Korea, Rep.	0.51085	5.27332	0.026449	6.863786	0.910038	3.148803	25
France	0.497586	5.201126	0.026058	6.710311	0.89723	3.083874	27
New Zealand	0.560707	5.442096	0.027449	7.401483	0.941032	3.366066	19
Czech Republic	0.555704	5.363749	0.027068	7.324056	0.927655	3.328266	20
Australia	0.504897	5.244367	0.026289	6.796228	0.904866	3.120562	26
Italy	0.509101	5.324838	0.026677	6.866874	0.918553	3.156109	24
Belgium	0.491341	5.21944	0.026113	6.658034	0.899944	3.066889	28
Canada	0.474981	4.563287	0.023038	6.252002	0.789339	2.839238	31
Israel	0.421988	4.480854	0.022419	5.717537	0.772605	2.633513	33
Greece	0.396826	5.104124	0.025154	5.716929	0.875399	2.707005	32
Hungary	0.446417	5.225619	0.025936	6.23403	0.898473	2.911833	30
United States	0.389225	4.873248	0.024063	5.556552	0.836378	2.620785	34
Costa Rica	0.468148	5.141179	0.025649	6.408005	0.885568	2.965768	29
Chile	0.247083	4.657409	0.022427	4.125009	0.792374	2.065113	35
Colombia	0.28978	2.616581	0.01329	3.75028	0.453577	1.688471	37
Türkiye	0.228219	3.858509	0.018687	3.640652	0.657767	1.794065	36
Mexico	0.105364	3.89669	0.0183	2.48923	0.657303	1.365468	38

and

Table 7. Ranking Results (λ = 0.75).

Countries	Si	Pi	kia	kib	kic	ki	Ranking
Finland	0.771206	5.724661	0.029703	9.507301	1	4.168411	1
Sweden	0.757527	5.71012	0.029574	9.371915	0.993086	4.115352	2
Denmark	0.733253	5.637908	0.029133	9.113933	0.975043	4.010044	4
Estonia	0.729432	5.65406	0.029189	9.08384	0.975626	4.000062	5
Switzerland	0.704184	5.584897	0.028758	8.817783	0.957599	3.891924	7
Austria	0.722575	5.668646	0.029225	9.02434	0.974882	3.978692	6
Lithuania	0.731318	5.681436	0.029323	9.112211	0.979736	4.013465	3
Iceland	0.696005	5.634649	0.028948	8.759169	0.960736	3.874159	9
Norway	0.641403	5.515902	0.028155	8.195563	0.925585	3.647515	13
Latvia	0.695748	5.640751	0.028975	8.759069	0.961399	3.874689	8
Slovenia	0.676056	5.547273	0.028457	8.536447	0.942421	3.780854	10
Luxembourg	0.649248	5.540665	0.028304	8.279486	0.931593	3.68193	12
Ireland	0.671208	5.572013	0.028548	8.499886	0.943689	3.76917	11
Germany	0.610959	5.470542	0.027809	7.88929	0.90858	3.526046	16
Portugal	0.631195	5.545604	0.028244	8.110035	0.92547	3.617514	14
Poland	0.624507	5.525402	0.028121	8.038834	0.920461	3.588383	15
United Kingdom	0.592147	5.426876	0.027523	7.694057	0.896127	3.447223	17
Netherlands	0.545044	5.226974	0.026393	7.170606	0.853678	3.228206	21
Japan	0.535158	5.2323	0.026373	7.078818	0.850652	3.193476	22
Slovak Republic	0.573491	5.417576	0.027395	7.513444	0.888007	3.377128	18
Spain	0.520621	5.336412	0.026782	6.980639	0.858178	3.165249	23
Korea. Rep.	0.51085	5.27332	0.026449	6.863786	0.846682	3.114643	25
France	0.497586	5.201126	0.026058	6.710311	0.832751	3.049139	27
New Zealand	0.560707	5.442096	0.027449	7.401483	0.886286	3.336423	19
Czech Republic	0.555704	5.363749	0.027068	7.324056	0.874672	3.299566	20
Australia	0.504897	5.244367	0.026289	6.796228	0.840859	3.086065	26
Italy	0.509101	5.324838	0.026677	6.866874	0.852439	3.120494	24
Belgium	0.491341	5.21944	0.026113	6.658034	0.832698	3.030704	28
Canada	0.474981	4.563287	0.023038	6.252002	0.744964	2.815192	31
Israel	0.421988	4.480854	0.022419	5.717537	0.714931	2.602477	33
Greece	0.396826	5.104124	0.025154	5.716929	0.783078	2.657954	32
Hungary	0.446417	5.225619	0.025936	6.23403	0.816701	2.868118	30
United States	0.389225	4.873248	0.024063	5.556552	0.751519	2.57562	34
Costa Rica	0.468148	5.141179	0.025649	6.408005	0.814306	2.927508	29
Chile	0.247083	4.657409	0.022427	4.125009	0.671619	2.002419	35
Colombia	0.28978	2.616581	0.01329	3.75028	0.433665	1.677634	37
Türkiye	0.228219	3.858509	0.018687	3.640652	0.565192	1.745702	36
Mexico	0.105364	3.89669	0.0183	2.48923	0.52409	1.298483	38

.

Table 5 presents the ranking results obtained using the CoCoSo method under the standard parameter setting ($\lambda =0.5$). Finland and Sweden occupy the top two positions, followed by Lithuania, Denmark, and Estonia. These results suggest that countries performing strongly in natural capital (C1), intellectual capital (C4), and economic sustainability (C5) tend to achieve higher overall sustainable competitiveness scores. By contrast, countries located at the lower end of the ranking generally display weaker performance in governance (C6) and social capital (C3). Overall, the results indicate that the CoCoSo method effectively differentiates performance levels across OECD countries. To examine parameter sensitivity, the analysis was repeated for $\lambda =0.25$ and $\lambda =0.75$, and the corresponding results are reported in Table 6 and Table 7.

Table 6 shows the CoCoSo ranking results under the alternative parameter setting ($\lambda =0.25$). The overall ranking pattern remains largely unchanged, with Finland and Sweden preserving the top positions and only minor shifts appearing among middle- and lower-ranked countries. This suggests that the CoCoSo method provides relatively robust results under moderate variation in the parameter value.

Table 7 reports the ranking results obtained under $\lambda =0.75$. The similarity of rankings across the three parameter settings ($\lambda =0.25$, $0.50$, and $0.75$) indicates that the CoCoSo method yields stable country rankings. Finland and Sweden consistently remain at the top, while Lithuania, Denmark, and Estonia also maintain strong positions.

In the final stage of the analysis, countries were grouped according to their similarities using the k-means clustering algorithm. The analysis was conducted in IBM SPSS Statistics 23. The elbow method and Silhouette coefficient were used to determine the optimal number of clusters. The Silhouette coefficient indicated a moderate clustering structure and supported the two-cluster solution. The number of countries in each cluster is presented in

Table 8. Cluster Membership Counts.

Cluster	Number of Country	% of Total
1	8	21.05
2	30	78.95

.

The k-means clustering results based on standardized variables indicate that the 38 OECD countries are grouped into two clusters. Cluster 1 consists of 8 countries (21.05%), while Cluster 2 includes 30 countries (78.95%).

The final cluster centers show that the most pronounced differences between clusters emerge in social capital (C3), intellectual capital (C4), and governance (C6). These variables significantly contribute to the separation of countries, as confirmed by the ANOVA results.

In contrast, natural capital (C1) does not exhibit a statistically significant difference between clusters, suggesting that natural resource endowment alone does not sufficiently explain variations in sustainable competitiveness among OECD countries.

Overall, Cluster 2 represents countries with stronger institutional, social, and knowledge-based structures, whereas Cluster 1 includes countries with comparatively weaker multidimensional sustainability profiles. It should be noted that although clustering was performed using standardized (z-score) variables, the cluster centers are reported in the original scale to facilitate substantive interpretation.

Table 9. Final Cluster Centers Reported in Original Scale.

Code	Cluster 1	Cluster 2
C1	48.99	47.66
C2	49.80	54.52
C3	41.19	53.11
C4	51.79	61.72
C5	54.17	57.49
C6	58.20	67.76

presents the final cluster centers for the six sustainability indicators. The results indicate that Cluster 2 exhibits higher average values in most criteria, particularly in social capital (C3), intellectual capital (C4), and governance (C6). These differences highlight the importance of institutional quality, knowledge-based capacity, and social development in shaping sustainable competitiveness across OECD countries.

By contrast, the difference between the clusters in natural capital (C1) is relatively limited, suggesting that natural resource endowment alone does not sufficiently explain variations in sustainable competitiveness performance among OECD countries.

Overall, Cluster 2 can be interpreted as representing countries with relatively stronger multidimensional sustainability performance, whereas Cluster 1 includes countries with comparatively weaker structural profiles.

When examining the cluster memberships and center distance values presented in

Table 10. Cluster Membership and Distance to Cluster Centers.

Country	Cluster	Distance	Country	Cluster	Distance
Finland	2	14.570	Spain	2	7.099
Sweden	2	11.346	Korea Rep.	2	11.064
Denmark	2	10.855	France	2	9.514
Estonia	2	9.743	New Zealand	2	9.093
Switzerland	2	11.135	Czech Republic	2	9.553
Austria	2	5.999	Australia	2	9.913
Lithuania	2	12.088	Italy	2	6.676
Iceland	2	6.425	Belgium	2	9.252
Norway	2	9.066	Canada	1	14.284
Latvia	2	9.334	Israel	2	16.613
Slovenia	2	9.661	Greece	2	10.468
Luxembourg	2	7.092	Hungary	1	7.821
Ireland	2	9.162	United States	1	12.874
Germany	2	6.368	Costa Rica	1	12.207
Portugal	2	6.076	Chile	1	7.403
Poland	2	5.761	Colombia	1	13.752
United Kingdom	2	13.047	Türkiye	1	9.243
Netherlands	2	12.133	Mexico	1	13.967
Japan	2	10.879	Slovak Republic	2	12.325

, the K-means analysis reveals that the 38 OECD countries are divided into two main clusters. The results indicate that the majority of countries (30 countries) are grouped into Cluster 2, while 8 countries are included in Cluster 1.

Cluster 2 includes Finland, Sweden, Denmark, Estonia, Switzerland, Austria, Lithuania, Iceland, Norway, Latvia, Slovenia, Luxembourg, Ireland, Germany, Portugal, Poland, the United Kingdom, the Netherlands, Japan, the Slovak Republic, Spain, Korea, France, New Zealand, the Czech Republic, Australia, Italy, and Belgium. Examination of the cluster centers shows that this group generally exhibits higher average values across most sustainability criteria. Accordingly, Cluster 2 can be interpreted as representing countries with relatively strong sustainable competitiveness performance. The concentration of Scandinavian and Northern European countries in this cluster reinforces the argument that governance quality, social capital, and intellectual capacity constitute key structural drivers of sustainable competitiveness.

Cluster 1 consists of Canada, Hungary, the United States, Costa Rica, Chile, Colombia, Türkiye, and Mexico. This group includes countries with comparatively lower average values in several sustainable competitiveness indicators. The presence of several Latin American OECD members and structurally heterogeneous economies suggests that institutional capacity, resource efficiency, and social development remain critical areas for improvement. Türkiye’s inclusion in Cluster 1 may reflect structural challenges related to multidimensional sustainability performance.

When the distance values from the cluster centers are examined, certain countries appear to represent the structural characteristics of their respective clusters more distinctly. For example, Austria (5.999), Poland (5.761), and Germany (6.368) are positioned closer to the center within Cluster 2 and therefore reflect the typical characteristics of this cluster more strongly. Similarly, Chile (7.403) and Hungary (7.821) emerge as more representative members of Cluster 1. In contrast, Israel displays a relatively high distance value (16.613) within Cluster 2, suggesting a structural profile that diverges somewhat from the cluster average.

Overall, the clustering results are largely consistent with the ranking results obtained using the CoCoSo method. Most of the top-ranked countries are grouped in Cluster 2, whereas countries with lower rankings are mainly concentrated in Cluster 1. Furthermore, the silhouette coefficient value of 0.372 indicates a moderate yet acceptable level of cluster separation, suggesting that the two-cluster solution captures meaningful structural differences among OECD countries. In general, OECD countries appear to form two broad groups in terms of sustainable competitiveness: countries with stronger multidimensional sustainability performance and countries that still have room for structural improvement.

5. Robustness Analysis

To evaluate the stability of the empirical findings, several robustness checks were conducted. First, parameter sensitivity analysis was performed by varying the CoCoSo compromise coefficient (λ = 0.25, 0.50, and 0.75). The ranking structure remained largely unchanged across scenarios, indicating strong parametric stability. In addition, Spearman’s rank correlation analysis revealed high consistency between alternative ranking results, suggesting that the performance ordering of OECD countries is not sensitive to parameter selection.

Second, an equal-weighting scenario was examined to assess whether the results depend heavily on the CRITIC weighting structure. The overall ranking patterns and top-performing countries remained largely consistent, confirming the robustness of the integrated evaluation framework.

Finally, clustering robustness was examined by comparing alternative cluster solutions and evaluating cluster compactness and separation patterns. The two-cluster structure provided the most interpretable and stable grouping configuration. Figure 2 further presents a graphical illustration of ranking stability under alternative λ values.

As illustrated in Figure 2, the ranking positions of the OECD countries remain largely stable across alternative λ values. The fully horizontal lines indicate complete positional stability, providing additional evidence for the robustness of the CRITIC–CoCoSo-based evaluation framework.

Overall, the robustness checks collectively indicate that the empirical findings are stable across alternative parameter configurations and methodological specifications, strengthening the reliability of the proposed integrated evaluation framework.

To further evaluate whether the CRITIC–CoCoSo framework materially alters the country ordering relative to the original Global Sustainable Competitiveness Index (2025) [18], a direct empirical comparison was conducted. The results of this comparison are presented in

Table 11. Direct Comparison of OECD Country Rankings: GSCI 2025 vs. CRITIC–CoCoSo Results.

Country	GSCI 2025 Rank	CRITIC Rank	Rank Difference (Δ)
Finland	1	1	0
Sweden	2	2	0
Denmark	3	4	1
Estonia	4	5	1
Switzerland	5	7	2
Austria	6	6	0
Lithuania	7	3	−4
Iceland	8	9	1
Norway	9	13	4
Latvia	10	8	−2
Slovenia	11	10	−1
Luxembourg	12	12	0
Ireland	13	11	−2
Germany	14	16	2
Portugal	15	14	−1
Poland	16	15	−1
United Kingdom	17	17	0
Netherlands	18	21	3
Japan	19	22	3
Slovak Republic	20	18	−2
Spain	21	23	2
Korea, Rep.	22	25	3
France	23	27	4
New Zealand	24	19	−5
Czech Republic	25	20	−5
Australia	26	26	0
Italy	27	24	−3
Belgium	28	28	0
Canada	29	31	2
Israel	30	33	3
Greece	31	32	1
Hungary	32	30	−2
United States	33	34	1
Costa Rica	34	29	−5
Chile	35	35	0
Colombia	36	37	1
Turkiye	37	36	−1
Mexico	38	38	0

Note: GSCI ranks are recalculated within the OECD subsample (n = 38) for methodological consistency [18]. The Spearman rank correlation between the two rankings is 0.977.

.

As shown in Table 11, although the overall ranking structure remains broadly consistent (Spearman = 0.977), several middle-ranked countries exhibit noticeable positional shifts under endogenous weighting.

6. Discussion

The empirical findings provide differentiated support for the proposed hypotheses. H1 is partially supported, as intellectual capital and governance contribute to distinguishing sustainability profiles among OECD countries. Notably, natural capital received the highest CRITIC weight, indicating that environmental variability constitutes the most structurally differentiating dimension of sustainable competitiveness. Since CRITIC weights are derived endogenously from dispersion and inter-criteria correlation patterns, this result suggests that sustainability disparities across OECD countries are more sensitive to environmental heterogeneity than to institutional similarity. In this respect, sustainable competitiveness does not appear to be a perfectly symmetric construct, but rather as a multidimensional configuration shaped by unequal informational influence across dimensions.

H₂ is strongly supported, as Nordic and institutionally strong countries consistently occupy the top positions in the ranking. This pattern reinforces the importance of governance quality, social capital, and institutional capacity in shaping national sustainability performance. H3 is also supported by the clustering analysis, which reveals two structurally distinct groups of OECD countries. These clusters indicate that sustainable competitiveness reflects not only performance levels but also broader institutional and socio-economic structures.

To directly evaluate the extent to which the proposed framework alters the original hierarchy, the CRITIC–CoCoSo ranking was compared with the 2025 Global Sustainable Competitiveness Index (GSCI). The Spearman rank correlation coefficient between the two rankings is 0.977, indicating a strong but not perfect association. While the overall structure—particularly among top-performing countries—remains broadly consistent, several middle and lower-ranked countries exhibit substantial positional shifts.

These positional variations primarily occur among middle-ranked countries, suggesting that endogenous weighting alters relative positions without fundamentally reshaping the overall hierarchy. This pattern confirms that the proposed CRITIC–CoCoSo framework does not overturn the general structure of sustainable competitiveness across OECD countries; rather, it refines the relative ordering by emphasizing dispersion-based informational differences across dimensions. In this sense, the framework provides a complementary perspective to the predefined composite index structure rather than a contradictory one.

7. Conclusions

This study offers an integrated evaluation of sustainable competitiveness across OECD countries using a hybrid CRITIC–CoCoSo and clustering framework. The criterion weights were determined using the CRITIC method as an objective weighting approach. Countries were then ranked using the CoCoSo method, and the resulting performance structures were classified through K-means clustering analysis. By combining ranking and classification techniques, the study provides a multidimensional perspective on sustainable competitiveness.

The composite sustainable competitiveness index was not directly incorporated into the empirical decision matrix, as the analysis focuses on the six underlying dimensions (C1–C6) that structurally represent the multidimensional nature of sustainable competitiveness.

The findings reveal noticeable differences among OECD countries in terms of sustainable competitiveness. In particular, the concentration of Northern European countries in the top positions indicates that balanced performance in areas such as governance quality, intellectual capital, and natural capital management plays an important role. At the same time, the results suggest that strong performance in only one dimension does not necessarily lead to high overall competitiveness. Sustainable competitiveness appears to be a multidimensional phenomenon that cannot easily be reduced to a single policy area.

The clustering analysis identified two main country groups. The first cluster includes countries with relatively lower sustainable competitiveness performance, whereas the second cluster consists of countries with more balanced and stronger profiles across the considered indicators. Although the CoCoSo ranking results and clustering outcomes are largely consistent, some differences in ranking positions and cluster memberships can be observed. These differences mainly arise from the analytical perspectives of the methods. While the ranking method focuses on weighted aggregate performance, clustering groups countries according to similarities in their indicator structures. This shows that sustainable competitiveness should be evaluated not only in terms of performance levels but also in terms of structural characteristics.

Contributions of the Study and Limitations

This study makes three distinct contributions. First, methodologically, it integrates objective CRITIC weighting, CoCoSo compromise ranking, and clustering analysis within a unified analytical framework. Second, analytically, it demonstrates that sustainability dimensions exhibit asymmetric informational influence when evaluated through endogenous weighting structures, thereby challenging the implicit normative symmetry of predefined composite indices. This finding implies that composite sustainability indices embed implicit weighting philosophies that may not fully reflect empirical structural dispersion patterns across countries. Third, from a policy perspective, the study highlights that governance quality and intellectual capital function as structural differentiators in clustering patterns, while natural capital variability drives ranking dispersion across OECD countries.

From a policy perspective, the results indicate that sustainable competitiveness cannot be explained solely by economic performance indicators. Governance quality, social capital, and intellectual capacity seem to play particularly important roles. For countries in the lower-performing cluster, policies aimed at improving institutional quality, increasing environmental efficiency, and strengthening knowledge-based production capacity may be especially relevant. At the same time, the generally consistent results obtained under different weighting scenarios and clustering structures support the robustness of the proposed analytical framework. This stability across alternative parameter configurations strengthens confidence in the empirical findings.

Finally, future research may extend this analysis by applying alternative MCDM techniques or panel data approaches in order to examine the time dimension of sustainable competitiveness. Nevertheless, the use of cross-sectional data and the reliance on a specific sustainability indicator framework may impose certain limitations on the generalizability of the findings.