European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 20 Examining the dimensionality of Circular Economy metrics using Hierarchical Clustering on Principal Components Collin L. Yobe1* 1 University of South Africa, Graduate School of Business Leadership, Cnr Janadel and, Alexandra Ave, Midrand, 1686 (South Africa) collinyobe@gmail.com *contact author ORCID 0000-0001-5270-2192 Received: 03/01/2024 Accepted for publication: 26/04/2024 Published: 30/04/2024 Abstract The Circular Economy (CE) indices have become a valuable tool for supporting the development of policies that provide information that reduces environmental pressures and impacts. However, highly dimensional data identifying many CE indicators is impractical in application. This paper aims to create a composite index of the CE indicators using the Hierarchical Clustering on Principal Components (HCPC) to extract the meaning of the CE indicators, as reducing dimensionality can improve understanding of indicators and metrics. The advantage of the HCPC methodology over principal components analysis (PCA) alone involves applying objective clustering techniques to the PCA results, which results in a better cluster solution. This study analysed a dataset of 61 indicators obtained from De Pascale, Arbolino, Szopik- Depczyńska, Limosani, and Ioppolo (2021). The composite indices revealed the dimensions of industrial symbiosis (IS), CE strategies, and spatial applications of the CE and IS concepts. The bottom-up and top-down approaches for CE and IS strategies have been the main implementation approaches in different governments and regions. Keywords: Circular Economy, Circular Economy indicators, industrial symbiosis, Hierarchical Clustering on Principal Components European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 21 1. Introduction The circular economy (CE) is being touted as a promising solution to mitigate human activities' growing environmental and resource pressures (Bocken, De Pauw, Bakker, & Van Der Grinten, 2016). Several countries have adopted CE principles to achieve zero waste, a crucial goal for most economies as it reduces greenhouse emissions and environmental impacts from current linear waste production systems, making CE a crucial part of future strategies. A CE promotes system innovations to improve waste management resource efficiency and balance the economy, environment, and society (Kristensen & Mosgaard, 2020). The CE paradigm promotes autonomous production processes that reuse materials, generating increased interest in research and developing metrics for a circular shift. The interest in circularity indicators has led to extensive literature on CE. Currently, there are three levels of indicators for measuring the CE, i.e., micro (companies, product), meso (industrial symbiosis (IS), eco-industrial parks (EIPs)), and macro (governments, global, national, regional, city) (de Oliveira, Dantas, & Soares, 2021; De Pascale et al., 2021; Kristensen & Mosgaard, 2020; Moraga et al., 2019). The rise of EIPs is a significant trend in the new industrial reality, aiming to promote environmental benefits and economic development through collaboration between companies (Felicio, Amaral, Esposto, & Durany, 2016). IS promotes a collective approach to competitive advantage across industries by integrating the physical exchange of materials, energy, water, and by- products into their business processes (Felicio et al., 2016; Geng, Fu, Sarkis, & Xue, 2012; Geng, Zhang, Ulgiati, & Sarkis, 2010; Jacobsen, 2006). Moreover, as such, EIPs must be able to promote IS. The IS and CE are intricate subjects that require meticulous coordination due to the potential differences in preferences among various stakeholders. Therefore, policymakers must take a direct lead in pushing and driving IS and the CE. Through legislation, governments have managed to work towards putting CE plans into action. The Chinese government has played a significant role in the implementation of IS. They can define policy, and everyone within the country follows. CE, adopted in China to promote economic growth despite material and energy limitations, has garnered significant global interest from governments, international bodies, industrial associations, and corporations (Franklin-Johnson, Figge, & Canning, 2016). China leads the ranking of countries that have contributed most to the increasing growth rate in research, implementation, and extensive development of CE concepts in academia and politics (De Pascale et al., 2021; Geng et al., 2012). In 2008, China was the first country to adopt legislation to deploy CE strategies by enacting a specific law, confirming China's prominence (De Pascale et al., 2021; Geng et al., 2012; Moraga et al., 2019). Unlike China, the European Commission and its member countries use self-regulatory, bottom-up strategies for implementing CE strategies, which primarily rely on external factors (Cayzer, Griffiths, & Beghetto, 2017; Linder, Sarasini, & van Loon, 2017). The European Commission and member countries employ self-regulatory, bottom-up approaches to implement CE strategies, unlike China, which primarily relies on external factors. CE is extensively researched, yet its practical application in economic initiatives remains a significant challenge. The extensive literature on CE necessitates rigorous analysis to ensure its relevance for functional purposes due to its disconnection. Several studies have come to the fore and have identified CE indicators (Argüelles, Benavides, & Fernández, 2014; de Oliveira et al., 2021; De Pascale et al., 2021; Parchomenko, Nelen, Gillabel, & Rechberger, 2019; Stanković, Janković-Milić, Marjanović, & European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 22 Janjić, 2021). The current set of circularity metrics is criticised for failing to fully capture the multidisciplinary and systemic nature of the CE (Stanković et al., 2021). These studies critically examine the dimensionality of CE indicators used in studies, highlighting their high dimensionality and potential limitations for practitioners. It suggests that reducing dimensionality can improve understanding of indicators and metrics. The research highlights the need for further knowledge of current research work. Therefore, this paper aims to create a composite index of the CE indicators using the Hierarchical Clustering on Principal Components (HCPC) to extract the meaning of the CE indicators. A dataset of 61 indicators developed in a previous paper identified by De Pascale et al. (2021) will be analysed using the HCPC to achieve the stated aim. This approach reduces data dimensionality and enables further understanding of the already-identified CE issues—IS, spatial applications of the CE concept, and CE strategies. Based on this dataset, the exercise attempts to cluster and classify indicators and papers to provide a structure for the problem. The rest of the document is structured as follows: Section 2 offers an extensive review of Circular Economy (CE) indicators found in previous studies, providing an understanding of the many metrics and approaches used to evaluate CE performance. In Section 3, the data sources and methodology of this study are explored. The analytical frameworks used for CE indicator analysis and classification, such as principal component analysis (PCA) and hierarchical clustering, are explained. Part 4 presents the PCA analysis and hierarchical clustering findings, thoroughly studying the principal components and clusters found, thereby illuminating the underlying patterns and trends in the CE environment. Ultimately, Section 5 summarises the results derived from the research, emphasising the critical takeaways and implications for furthering CE initiatives. 2. Review of CE indicators in research The growing resource demand and environmental issues drive a shift towards sustainable production and consumption (Gallego-Schmid, Chen, Sharmina, & Mendoza, 2020). Contemporary sustainability literature focuses on the potential of the CE to disrupt the unsustainable production and consumption linear economy (Kristensen & Mosgaard, 2020). However, considerable effort is required to transition toward a more CE (Parchomenko et al., 2019). In transitioning to this CE economy, indices are fast becoming a valuable tool to support the development of policies in providing information and reducing environmental pressures and impacts (De Pascale et al., 2021). There three main levels of indicators for measuring CE in the literature so far are macro (global, national, regional, city, governments), meso (IS, EIPs), and micro (single firm, product) (de Oliveira et al., 2021; De Pascale et al., 2021; Geng et al., 2012; Kayal et al., 2019; Kristensen & Mosgaard, 2020; Linder et al., 2017; Mazur-Wierzbicka, 2021; McCarthy, Kapetanaki, & Wang, 2019; Moraga et al., 2019). de Oliveira et al. (2021) present a fourth dimension, nano (products). The lack of detailed measurement and documentation of the CE's progress can hinder understanding of the subject matter, creating barriers for actors at the specific CE level. Governments, policymakers, and business practitioners need more information on CE typologies to promote their business environment towards IS. Transitioning to a CE requires significant effort, but a widely accepted framework for monitoring progress is lacking due to the vast and diverse areas covered by different assessment methodologies (Parchomenko et al., 2019). Despite the concept's lack European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 23 of clarity, CE focuses on defining action plans supported by specific indicators (Moraga et al., 2019). Transitioning to circular systems may not always lead to favourable alternatives, as potential environmental, economic, or social trade-offs may arise (de Oliveira et al., 2021). Past CE research methodologies include reviews and empirical papers, with CE indicators classification, categorising and assessing existing work, with some articles developing a measurement method (de Oliveira et al., 2021; Kristensen & Mosgaard, 2020). Some of the approaches used in developing the metric for measuring CE indicators have included the (a) Principal Component Analysis and PROMETHEE (de Oliveira et al., 2021),(b) The Multiple Correspondence Analysis (MCA) (Parchomenko et al., 2019), and (c) indicators grouped by using a double classification: first according to the three spatial dimensions of sustainability –macro, micro and meso – then based on the 3R Core CE principles (De Pascale et al., 2021). 3. Data and methods 3.1 Data collection The data for this research was obtained from De Pascale et al. (2021) through a systematic literature review, and a final list of indicators was extracted and grouped into three clusters: micro, meso, and macro. The review process involved selecting academic databases, search terms, and screening for practical purposes. Definitions and key concepts were determined, and the gathered material was summarised for future insights. The first step involved a systematic search for published CE indicator implementation studies. The study used a methodological approach to define and select a sampling framework to review Circular Economy (CE) indicators. The framework was based on a time scale of 2000-2019 and a preliminary search of existing literature in Scopus and Web of Science databases. The results were refined using advanced search terms and keywords to identify the level of implementation of CE indicators at micro, meso, and macro spatial levels. A combination of selected keywords was chosen and compared with other CE reviews to ensure high-level significance. The keyword "Circular Economy" was selected to ensure coherence with the main topic. The academic databases were searched for spatial levels (micro, meso, and macro) and identified keywords for each level. The classification system includes micro-level (company), meso-level (circular economy), and macro-level (city, country, region) terms, combining them to create a comprehensive understanding of the circular economy. The categories of indicator, index, assessment, evaluation, and measuring were incorporated into the literature as they are frequently used in CE studies. The study enumerated 61 articles that were used for the analysis. 3.2 Description of the Explanatory Variables Table 1 shows the CE indicators identified by De Pascale et al. (2021). The study quantified these as categorical variables by creating lists of indicators to capture the various attributes of CE indicators. The multivariate approach implemented the Hierarchical Clustering on Principal Components (HCPC). This multivariate approach uses Principal Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) on these attributes. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 24 Table 1: Description of the dimensions of circular economy indicators Attributes Description Approach: It includes five subdivisions: Quantitative, qualitative, VBI, analytical tool, and theoretical results. Application-level: Refers to Product, Materials, (embodied) Energy, Components, Resource Waste (/waste streams), Product families, IS districts and networks, National, Regional, Provincial, the European Union (EU), and World Regions. Core CE principles: This category concerns the following Core CE principles: Reduce, Reuse, Recycle, Recover, Remanufacture, and Redesign Sustainable Development Dimensions: The dimensions are Environmental, Economic, and Social. Level of CE indicators: The spatial level applications of the circularity vary between macro - as a city, region, or nation, meso – as EIPs (eco-EIPs) and IS, and micro levels – as a single company or products using different methods and techniques. Regions: The study considered the following regions: the United States of America (USA), Europe, EU, Belgium, Netherlands, Italy, England/ United Kingdom (UK), India, Spain, Denmark, China, South Korea, Australia, Jordan, Japan, Switzerland, and World Regions. Source: De Pascale et al. (2021) 3.3 Analytical techniques Please use Times New Roman, size 10, for the text of your manuscript, including the main body, references, and footnotes. 3.3.1 Multivariate Approach for Classification - Principal Components Analysis The study utilised Principal Component Analysis (PCA) to create typologies of CE indicators, a statistical method that reduces variables into smaller dimensions with minimal information loss (Jolliffe, 2002). PCA is a feature extraction technique that transforms an initial dataset of variables into a new uncorrelated dataset of orthogonal linear combinations (PCs). It aims to reduce the dimensionality of the data by obtaining the largest variance original variables, accounting for as much variation as possible (Jolliffe, 2002; Manly, 2005); this means that the first PC is the linear combination with the largest variance, while the second one is the linear combination with the second-largest variance, orthogonal to the first PC, etc. According to Costello and Osborne (2005), Varimax rotation simplifies the data's factor structure and makes its interpretation more straightforward and reliable. The authors argue that Varimax rotation, which produces uncorrelated factors, is superior to other rotation orthogonal methods like quartimax and equamax, as it cannot improve fundamental aspects of analysis like variance extraction. De Pascale et al. (2021) grouped the indicators using a double classification: first, according to the three spatial dimensions of sustainability – macro, micro, and meso– and then based on the 3R Core CE principles. The study mapped out dummy variables for indicators' attributes, such as level, country, analytical approach, application level, core CE principles, and sustainable development dimensions, but this approach failed to provide a clear group structure and dimensionality of the CE indicators in European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 25 the research. Jolliffe (2002) posits that the approach to using cluster analysis is vital in cases where such a group structure is lacking. Following Vyas and Kumaranayake (2006) and Achia, Wangombe, and Khadioli (2010), PCA was applied to the dummy variables to reduce the dimensionality of the data and categorise the CE indicators into distinct dimensions (Jolliffe, 2002). The PCA scores of CE indicators initially extracted and retained were followed by Varimax rotation for the above reasons. The study used the Kaiser-Maier-Olkin (KMO) and Bartlett's sphericity tests to assess the suitability of the variables for PCA. Hair, Black, Babin, Anderson, and Tatham (2006) suggest that researchers should consider variables suitable if their KMO values exceed 0.5, and Bartlett's sphericity test yields statistical significance at p<0.05. PCs with Eigenvalues equal to or greater than 0.7 were retained, following Jolliffe (2002). - Hierarchical cluster analysis The study used Hierarchical Clustering (HCA) to classify CE indicators based on their similarity or dissimilarity. The technique uses Euclidean distance as the dissimilarity metric. The research used PCA on CE indicators' characteristics data, a hierarchical clustering method called Hierarchical Clustering on Principal Components (HCPC), to analyse the classification limits of CE indicators. Hierarchical cluster analysis estimates the number of clusters in a dataset using Principal Component Analysis (PCA). The study uses Ward's criterion to perform hierarchical clustering on all principal components derived from PCA on the initial dataset of variables. Ward's method estimates hierarchical clusters that create equal and evenly sized clusters, with solutions estimated based on cases and variables. Several studies, including Argüelles et al. (2014), have used HCPC and K-means clustering as candidate multivariate approaches. HCPC offers two main advantages: it applies objective clustering techniques to PCA results, resulting in a better cluster solution than factorial analysis alone, and Ward's classification enhances the robustness of the final clustering results. Garson (2009), cited in Yobe, Mudhara, and Mafongoya (2019), suggests that hierarchical clustering is the most suitable technique for data sets with less than 250 observations, indicating that datasets below this threshold are unsuitable. According to Kaur and Kaur (2013), the K-means algorithm outperforms the hierarchical algorithm on data sets with over 250 observations. However, preliminary analyses with a smaller sample size confirmed this, as the K-means clustering technique failed to adequately classify cases, leading to using the HCPC technique for multivariate classification. 4. Results 4.1 Descriptive Statistics of the Circular Economy Indicators The dummy variables were created to measure the variation of the CE indicators accurately, enabling a quantitative analysis of their performance. Table 2 shows the descriptive statistics that measure the mean, standard deviation, minimum, and maximum. On average, the Quantitative Approach was extensively used, with a mean of 0.7458. The dominant application level is for product, and the mean value is 0.4068. The dummy variables were created to accurately measure the variation of the CE indicators, enabling a quantitative analysis of their performance. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 26 Table 2: Descriptive statistics of the circular economy indicators Variable Mean Std. Dev. Min Max Approach: Quantitative 0.7458 0.4392 0 1 Qualitative 0.0847 0.2809 0 1 VBI 0.0678 0.2536 0 1 Analytical tool 0.0170 0.1302 0 1 Theoretical results 0.0169 0.1302 0 1 Application-level: Product 0.4068 0.4954 0 1 Materials 0.1695 0.3784 0 1 (embodied) Energy 0.0678 0.2536 0 1 Components 0.0678 0.2536 0 1 Resource Waste (/ waste streams) 0.0508 0.2216 0 1 Product families 0.0169 0.1302 0 1 Industrial symbiosis districts and networks 0.1017 0.3048 0 1 National 0.0508 0.2216 0 1 Regional 0.0847 0.2809 0 1 Provincial 0.0339 0.1825 0 1 EU 0.0678 0.2536 0 1 World Regions 0.0170 0.1302 0 1 Core CE principles: Reduce 0.5593 0.5007 0 1 Reuse 0.6610 0.4774 0 1 Recycle 0.8136 0.3928 0 1 Recover 0.1356 0.3456 0 1 Remanufacture 0.1864 0.3928 0 1 Redesign 0.1017 0.3048 0 1 Sustainable Development Dimensions: Environmental 0.8983 0.3048 0 1 Economic 0.9153 0.2803 0 1 Social 0.4746 0.5036 0 1 Level of CE indicators: Micro-level 0.5085 0.5042 0 1 Macro-level 0.2373 0.4291 0 1 Regions: USA 0.0508 0.2216 0 1 Europe 0.0169 0.1302 0 1 EU 0.1017 0.3048 0 1 Belgium 0.0508 0.2216 0 1 Netherlands 0.0339 0.1825 0 1 Italy 0.0169 0.1302 0 1 England/UK 0.0339 0.1825 0 1 European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 27 Variable Mean Std. Dev. Min Max Approach: India 0.0169 0.1302 0 1 Spain 0.0169 0.1302 0 1 Denmark 0.0169 0.1302 0 1 China 0.0169 0.1302 0 1 South Korea 0.0169 0.1302 0 1 Australia 0.0169 0.1302 0 1 Jordan 0.0169 0.1302 0 1 Japan 0.0169 0.1302 0 1 Switzerland 0.0169 0.1302 0 1 World Regions 0.0169 0.1302 0 1 Source: De Pascale et al. (2021) Observations = 58. VBI is a Value-based indicator/Evaluation indicator system. 4.2 Multivariate analysis results The results of the multivariate analysis, which employed the PCA, are presented below. Table 3 shows the level of the scree plot of the estimated PCs of CE indicators. 28 Table 3: Estimated PCs of circular economy indicators Component 1 2 3 4 5 6 7 8 9 10 11 12 13 15 16 Initial eigenvalues: % of Variance 13.014 7.314 5.636 5.438 5.388 4.722 4.252 3.760 3.540 3.471 3.267 3.034 2.859 2.643 2.582 Cumulative % 13.014 20.328 25.964 31.402 36.790 41.512 45.764 49.524 53.064 56.535 59.801 62.835 65.694 68.337 70.980 Approach: Quantitative -0.464 -0.426 Qualitative 0.887 Value-based indicator/Evaluation indicator system 0.824 Analytical tool Theoretical results 0.910 Core CE principles: Reduce 0.694 0.316 Reuse 0.480 Recycle 0.706 Recover 0.731 Remanufacture -0.350 0.530 -0.499 Redesign 0.604 -0.320 0.568 Sustain. Develop. Dimensions: Environmental 0.448 0.356 0.358 Economic 0.791 Social 0.632 Level of CE indicators: Micro-level -0.857 Macro-level 0.637 Application-level: Product -0.811 Materials -0.445 0.351 0.446 Components 0.538 -0.601 Resource Waste (/ waste streams) 0.836 Product families 0.843 Industrial symbiosis districts and networks -0.524 0.453 European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 29 Component 1 2 3 4 5 6 7 8 9 10 11 12 13 15 16 National 0.547 0.734 Regional 0.549 -0.375 Provincial 0.671 (embodied) Energy 0.868 Regions: USA 0.513 0.743 Europe 0.769 EU 0.342 0.441 0.636 Belgium -0.377 -0.309 Netherlands 0.826 Italy 0.847 England/UK 0.889 India 0.888 Spain -0.808 Denmark China 0.694 0.315 South Korea Australia -0.846 Jordan 0.858 Japan 0.897 World Regions 0.911 Source: De Pascale et al. (2021) Extraction Method: Principal Component Analysis; Rotation Method: Varimax with Kaiser Normalization. The coefficients with a value of 0.3 and less were suppressed and not displayed in the estimated PC scores results. Kaiser-Meyer-Olkin Measure of Sampling Adequacy = 0.291; Bartlett’s Test of Sphericity Approx. Chi-Square = 1301.698; df = 903; Sig. = 0.000. PC 14, which explained 2.77% of the variance, was dropped from the analysis because the variation explained by the factor loadings did not make economic sense. 30 The application of PCA on the attributes of the CE indicators produced initial Eigenvalues 1.015, 19 PCs that explained a cumulative 80.941% of the variance in the dummy variables but only retained 15 PCs for the analysis Table 4. This study dropped PC 14 – which accounts for 2.77% – from the analysis because the variation it explained did not make economic sense. Therefore, the total variance explained by the PCs used in this analysis was 70.92%. The study dropped two variables – the meso-level of CE indicators and EIPs – because the rotation failed to converge in estimating the PCA's rotated component matrix. The study utilised KMO and Bartlett's sphericity measures to assess the feasibility of incorporating 58 cases of indicators into a dataset. The KMO measure showed a 0.291 value, indicating a significant relationship between variables. Bartlett's test showed a p<0.001 correlation, indicating variables could be factored in. Varimax with the Kaiser normalisation rotation method improved PC interpretation. The scree plot below shows the estimated PCs (Figure 1). The study plots Eigenvalue against PCs, with PC1 having the highest score loadings and a high Eigenvalue. PC-2 explains less variation and has a lesser Eigenvalue. The study estimates diminishing declines of Eigenvalue with other PCs, using a cut-off of PCs with an Eigenvalue of 1 and considering PCs that met this criterion. Figure 1: Scree plot of the estimated PCs of CE indicators Source: Author's elaboration The first principal component (PC–1) explains 13.01% of the variance in the indicators of CE, with the estimated component loadings expressed in the following equation: 𝑃𝐶–1 = 0.694 𝑅𝑒𝑑𝑢𝑐𝑒 𝐶𝐸 𝑝𝑟𝑖𝑛𝑐𝑖𝑝𝑙𝑒𝑠 − 0.350 𝑅𝑒𝑚𝑎𝑛𝑢𝑓𝑎𝑐𝑡𝑢𝑟𝑒 𝐶𝐸 𝑝𝑟𝑖𝑛𝑐𝑖𝑝𝑙𝑒𝑠 + 0.448 𝐸𝑛𝑣𝑖𝑟𝑜𝑛𝑚𝑒𝑛𝑡𝑎𝑙 𝑆𝑢𝑠𝑡𝑎𝑖𝑛𝑎𝑏𝑙𝑒 𝐷𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛𝑠 + 0.632 𝑆𝑜𝑐𝑖𝑎𝑙 𝑆𝑢𝑠𝑡𝑎𝑖𝑛𝑎𝑏𝑙𝑒 𝐷𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛𝑠 − 0.857 𝑀𝑖𝑐𝑟𝑜 − 𝑙𝑒𝑣𝑒𝑙 𝑜𝑓 𝐶𝐸 𝑖𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟𝑠 + 0.637 𝑀𝑎𝑐𝑟𝑜 − 𝑙𝑒𝑣𝑒𝑙 𝑜𝑓 𝐶𝐸 𝑖𝑛𝑑𝑖𝑐𝑎𝑡𝑜𝑟𝑠 − 0.811 𝑃𝑟𝑜𝑑𝑢𝑐𝑡 𝐴𝑝𝑝𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛 − 𝑙𝑒𝑣𝑒𝑙 − 0.445 𝑀𝑎𝑡𝑒𝑟𝑖𝑎𝑙𝑠 𝐴𝑝𝑝𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛 − 𝑙𝑒𝑣𝑒𝑙 + 0.549 𝑅𝑒𝑔𝑖𝑜𝑛𝑎𝑙 𝐴𝑝𝑝𝑙𝑖𝑐𝑎𝑡𝑖𝑜𝑛 − 𝑙𝑒𝑣𝑒𝑙 − 0.377 𝐵𝑒𝑙𝑔𝑖𝑢𝑚 + 0.694 𝐶ℎ𝑖𝑛𝑎. 0 1 2 3 4 5 6 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 Ei ge nv al ue Component Number European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 31 In PC–1, Reduce CE principles, Social Sustainable Development Dimensions, and Macro-level CE indicators were the most dominant loadings variables. Therefore, this PC was named Reduce CE principles. PC–2 explains 7.314% of the variance, and the most dominant loadings were on the Materials Application-level and in the USA, EU, England/UK and India regions; this PC was thus called Materials Application-level. PC–3 accounts for 5.64% of the variance with dominant loadings on Recover CE principles, Energy (embodied) Application-level and the EU region and was intuitively named Energy (embodied) Application level. PC–4 was called National Application-level because the most significant loading for this component was the National Application Level; the other loading for this coefficient was the USA and EU regions. The fifth component, PC–5, which accounted for 5.34% of the variation in the PC estimation, was called Diverse Application- level because it had loading for Application-Level variables, namely Components, Resource Waste and Product families. The sixth component, PC–6, Recover and Redesign CE principles Qualitative Approach, accounted for 4.72% of the variation and had significant loadings on qualitative – and negatively so on quantitative – approach, as well as Recover and Redesign CE principles. The seventh component, PC–7, was named Sustainable Development Economic Dimension because this PC had dominant positive loadings on the Sustainable Development Economic. The study revealed that Redesign CE principles, specifically PC-8, accounted for 3.76% of the study's variation. PC-8 had positive coefficients for Recycle and Remanufacture CE principles, positive coefficients for Materials and Components Application-level, and negative coefficients for the research approach. PC-9 had significant coefficient loadings on regional and provincial variables and China region. Hence PC–9 was called Approach Value-based indicator/ Evaluation indicator system and accounted for 3.54% of the variance. PC–10, named Application Level National and World Regions, had dominant coefficient loadings of 0.734 Application Level National and 0.911 for the World regions, accounting for 3.47% of the variation. The next component, PC–11, accounted for 3.26% of the variation. The variables Redesign CE principles and Environmental Sustainable Development Dimensions had positive coefficient loadings of 0.568 and 0.356, respectively; this PC was called Redesign CE principles Environmental Sustainable Development Dimensions. PC–12 is accounted for by 3.034% of the variation and represents the dimension of the CE indicators with significant coefficient loadings on the variables for the Theoretical results approach and the EU region; this PC was thus identified as the Theoretical results approach. PC-13, accounting for 2.86%, showed positive coefficient loadings on Reuse CE principles, while negative coefficient loadings were observed on IS districts, networks Application-level, and Australia. The next component, Reduce CE principle and Environmental Sustainability Development Dimensions, accounted for 2.64% of the variation. The component loadings of the coefficients on the variables are as follows: 𝑃𝐶–14 = 0.316 𝑅𝑒𝑑𝑢𝑐𝑒 𝐶𝐸 𝑝𝑟𝑖𝑛𝑐𝑖𝑝𝑙𝑒𝑠 + 0.358 𝐸𝑛𝑣𝑖𝑟𝑜𝑛𝑚𝑒𝑛𝑡𝑎𝑙 𝑆𝑢𝑠𝑡𝑎𝑖𝑛𝑎𝑏𝑖𝑙𝑖𝑡𝑦 𝐷𝑒𝑣𝑒𝑙𝑜𝑝𝑚𝑒𝑛𝑡 𝐷𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛𝑠 – 0.309 𝐵𝑒𝑙𝑔𝑖𝑢𝑚 + 0.826 𝑁𝑒𝑡ℎ𝑒𝑟𝑙𝑎𝑛𝑑𝑠. The PC-16, the Application-level PC, represents coefficient loadings on IS districts and networks in the Jordan region, exhibiting minimal variation in CE indicator scores. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 32 4.2.1 Hierarchical clustering The HCPC analysis uses the HCA to analyse the PCA outcome. Figure 2 displays a dendrogram representing the HCA result, with the vertical axis representing observations and clusters and the horizontal axis representing the distance between them. Similar observations are grouped based on their dissimilarity, allowing for better understanding and classification. Figure 2 shows a hierarchical dendrogram, where PCs are extracted and arranged into new clusters along the y-axis. Each branch represents connections, and the closer they are to each other, the more related they are. For instance, PC-6 and 15 are members of one cluster from a branch that splits into two, similar to the next group comprising PCs 13, 16, 9, and 8. However, the group composed of PC-2 and PC-4 is far apart in the chart. An imaginary cut-off point in a dendrogram creates a cluster, which becomes larger and more heterogeneous as the branching diagram moves up. This results in more variation within the cluster. For example, two variables next to each other, like PC-6 and PC-15, will be similar. Expanding the group, like one large group with PCs 6, 15, 13, 16, 9, 8, 5, 7, and 1, will result in more similar observations than grouped observations in another branch. However, there is still much variation in each group, necessitating a trade-off decision on where to make the cut-off. The study identifies two clusters in a dendrogram, with each observation in the clusters being similar. Large clusters, with more observations, are more heterogeneous. The choice of meaningful clusters depends on the degree of change in grouping due to slight deviations. The study suggests a two-cluster solution on the dendrogram, while Figure 3 has a four-cluster solution. The ideal variation is minimally affecting solutions. The study's findings align with this rationale. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 33 Clustering solutions by variables Figure 2: Hierarchical clustering and initial partition solutions by variables Source: Author's elaboration on Linkage method—Ward's method. Euclidean distance of all elements European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 34 Table 4 shows references with their title and country for each cluster that result from the dendrogram in Figure 2. The PC-16, also known as the Application-level PC, represents coefficient loadings on IS districts and networks in the Jordan region, exhibiting minimal variation in CE indicator score (Guo-gang, 2011; Li & Su, 2012; Xiong, Dang, & Qian, 2011). Other studies in this set of results focus on industrial parks (Felicio et al., 2016; Geng, Zhang, Côté, & Fujita, 2009; Geng et al., 2010; Tiejun, 2010; Wenbo, 2011; Zhao, Guo, & Zhao, 2018), while the case studies included (Adibi, Lafhaj, Yehya, & Payet, 2017; Bovea & Pérez-Belis, 2018; Huysman, De Schaepmeester, Ragaert, Dewulf, & De Meester, 2017; Huysman et al., 2015; Sałabun, Palczewski, & Wątróbski, 2019). The countries that dominated the studies included China, EU members, Belgium, and the USA. Table 4: List of clusters by references, article and countries Cluster Reference Country 1 Das, Yedlarajiah, and Narendra (2000) Lee, Lu, and Song (2014) MacArthur (2015) Europe Huysman et al. (2017) Belgium; England Mohamed Sultan, Lou, and Mativenga (2017) UK; EU; USA; India Azevedo, Godina, and Matias (2017) Mesa, Esparragoza, and Maury (2018) Vanegas et al. (2018) Belgium 2 Nelen et al. (2014) J. Y. Park and Chertow (2014) USA EU Huysman et al. (2015) Belgium Scheepens, Vogtländer, and Brezet (2016) Netherlands Cayzer et al. (2017) Figge, Thorpe, Givry, Canning, and Franklin-Johnson (2018); Franklin-Johnson et al. (2016) van Schaik and Reuter (2016) Linder et al. (2017) Di Maio, Rem, Baldé, and Polder (2017) Netherlands Favi, Germani, Luzi, Mandolini, and Marconi (2017) Adibi et al. (2017) Figge et al. (2018) Marconi, Germani, Mandolini, and Favi (2019) Mandolini, Favi, Germani, and Marconi (2018) European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 35 Cluster Reference Country Cong, Zhao, and Sutherland (2019) Zwolinski, Lopez-Ontiveros, and Brissaud (2006) Jacobsen (2006) Denmark Karlsson and Wolf (2008) Geng et al. (2010) China Sałabun et al. (2019) Wen and Li (2010) China Geng et al. (2009) China Su, Heshmati, Geng, and Yu (2013) China Geng et al. (2012) China Wenbo (2011) China Li and Su (2012) Beijing - China Wen and Li (2010) China H.-S. Park and Behera (2014) South Korea Pagotto and Halog (2016) Australia Felicio et al. (2016) Zhao et al. (2018) China Tiejun (2010) China Geng, Liu, Liu, Zhao, and Xue (2011) China Guo-gang (2011) China Faizi, Rashid, Sałabun, Zafar, and Wątróbski (2018) China Chun-rong and Jun (2011) Qing, Qiongqiong, and Mingyue (2011) China Xiong et al. (2011) China Wu, Shi, Xia, and Zhu (2014) China Haas, Krausmann, Wiedenhofer, and Heinz (2015) EU Members Haupt, Vadenbo, and Hellweg (2017) Switzerland Tisserant et al. (2017) World Regions Moraga et al. (2019) Mayer et al. (2019) EU Members Fregonara, Giordano, Ferrando, and Pattono (2017) Italy Moriguchi (2007) EU; USA; Japan 3 Bovea and Pérez-Belis (2018) Spain Kayal et al. (2019) Jordan 4 Smol, Kulczycka, and Avdiushchenko (2017) EU Members Source: Author's elaboration European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 36 5. Discussion Firms, governments, and regional authorities have started recognising environmental benefits and economic growth through collaboration between companies that flow directly from IS—generating better collective benefits than could have been achieved from the sum of all individual benefits combined (Bocken et al., 2016; Felicio et al., 2016; Jacobsen, 2006; Karlsson & Wolf, 2008). IS offers several benefits, including reducing waste disposal costs by converting waste into by-products for other industries, enabling innovation and process development, leading to increased profitability, and benefiting from the geographical proximity of businesses (Bocken et al., 2016; Karlsson & Wolf, 2008). Creating a composite index can help better understand the CE indicators and measure the development of the CE guides. Applying PCA based on the above data reduced the dimensionality of the 58 – previously 61 – CE indicators in De Pascale et al. (2021) to 15 PCs, which could be classified. Table 4 shows the list of the HCPC composite indices based on the title of the articles and the regions where these studies took place. This technique provides a structural approach to navigating the multifaceted realm of circular economy indicators. The HCPC reduced estimated PCs, resulting in terminal elements of the dendrogram from hierarchical classification. The dendrogram yielded four clusters of CE indicators, determining the group's CE performance based on specific characteristics, like countries with similar features (Stanković et al., 2021). The understanding of regional and global trends is improved by this hierarchical classification, which provides a systematic framework for identifying distinctive groupings within the CE landscape. The areas in this first cluster include European countries (Huysman et al., 2017; McCarthy et al., 2019; Mohamed Sultan et al., 2017; Vanegas et al., 2018), the USA (Mohamed Sultan et al., 2017) and India (Mohamed Sultan et al., 2017). This first cluster displays a variation of approaches within different national contexts, spotlighting geographical concentrations of CE initiatives. Due to the need to respond to the impacts of unsustainable linear production methods, CE approaches will likely provide a compelling driver across many different governments, countries and regions. Das et al. (2000), Lee et al. (2014), and Mesa et al. (2018) in the first cluster did not focus on a particular region in their research. The first two of these studies focused on estimating the end-of-life product disassembly effort and cost and assessing product End-Of-Life performance. The last of the two studies focused on developing a sustainable circular index. The diverse research focus within this cluster illustrates the breadth of CE initiatives, spanning multiple geographical contexts and ranging from index development to product lifecycle analysis. The following aggregation obtained after applying the HCPC, i.e. Cluster 2, shows studies in Europe (Di Maio et al., 2017; Haas et al., 2015; Huysman et al., 2015; Jacobsen, 2006; Mayer et al., 2019; Scheepens et al., 2016), the USA (J. Y. Park & Chertow, 2014), South Korea (H.-S. Park & Behera, 2014), China (Faizi et al., 2018; Geng et al., 2012; Geng et al., 2011; Geng et al., 2009; Geng et al., 2010; Guo-gang, 2011; Li & Su, 2012; Su et al., 2013; Tiejun, 2010; Wen & Li, 2010; Wenbo, 2011; Wu et al., 2014; Xiong et al., 2011; Zhao et al., 2018), Australia (Pagotto & Halog, 2016) and the World Regions (Tisserant et al., 2017). This second cluster has a broad geographical scope and implicates the variable approaches adopted by different countries and regions and the global reach of CE initiatives. Most of these primarily focused on industrial parks (Felicio et al., European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 37 2016; Geng et al., 2009; Geng et al., 2010; Tiejun, 2010; Wenbo, 2011; Zhao et al., 2018). Industrial parks in China are for environmental and economic growth benefits, and collaboration between companies is critical in realising this achievement; thus, it is crucial to promoting IS. According to Felicio et al. (2016), the IS in EIPs requires the implementation of intense broker involvement. Impediments to the level of symbiosis during operations usually are market changes and technological advancement. Wenbo (2011) further posits that developing CE is the only way to realise the new industrialisation, and building EIPs is an essential means of promoting CE performance. The next area of focus in this cluster that received attention from researchers is that of IS, which includes research done in China (Geng et al., 2009), Denmark (Jacobsen, 2006), South Korea (H.-S. Park & Behera, 2014) and Australia (Pagotto & Halog, 2016). The importance of collaborative efforts in promoting circular economy performance—particularly in industrialised nations—is highlighted by the focus on industrial parks as hubs for sustainability initiatives. The cluster solution suggests a lack of coverage of CE approaches, leading to a lack of studies. These approaches may have been developed to address a specific CE problem but were never expanded to other areas. They may also be challenging to set up in regions with similar CE issues, becoming unsupported and ineffective; this could be the case with two clusters. Nonetheless, the research gaps identified in CE indicate potential areas for further research and strategic intervention to advance more comprehensive and successful circular economy initiatives. Cluster 3 comprises two studies (Fregonara et al., 2017; Moriguchi, 2007) of the HCPC aggregates. Studies conducted in Italy, the EU, the USA, and Japan examined resource productivity indicators and environmental impact. The need to address environmental issues, such as waste reduction and pollution, is a common response to widely accepted action plans. This third cluster explores resource productivity and environmental impacts, highlighting the ongoing need to address pressing environmental issues across geographic boundaries through sustained research and policy interventions. Cluster 4 classifies the studies focusing on circular design guidelines in Spain (Bovea & Pérez-Belis, 2018), measuring a firm's circularity in the water industry in Jordan (Kayal et al., 2019), and CE indicators concerning eco-innovation in the EU member countries (Smol et al., 2017). This final cluster focuses on particular CE initiatives, such as industrial circularity, design guidelines, and eco-innovation, and illustrates the various strategic approaches used in multiple industries and regions.Our journal adheres to the APA 7th edition style for citations and references. Please ensure that all citations in the text and the reference list comply with this format. 6. Conclusions The study used HCPC to identify clusters in a dataset with 61 CE indicators, retaining 58 after data cleaning, utilising complementarities between clustering and principal components methods. The first stage estimation of PCA only allowed for 15 of the 16 PCs to be meaningfully interpreted. The HCPC clustering technique from four clusters provided a better cluster solution than the principal components method, identifying research dimensions such as IS, CE strategies, and spatial arrangement. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 38 Policy recommendations for IS and CE subjects require coordination among stakeholders with diverse agendas. Policymakers must lead the implementation of policies related to these subjects. Despite different approaches, governments and regions are adopting actions that drive the agenda forward. China has successfully used a top-down approach to policy implementation, allowing the government to prescribe policies and have everyone adopt them. This top-down approach could be a crucial lead for other countries seeking similar practices to follow. China's top-down approach could serve as a model for other countries. The bottom-up approach to policy implementation has been successful in European countries, with governments, regions, and businesses realising the benefits of adopting such policies. These systems receive support and spread quickly, often guided by an overarching goal. The central role of government is critical in implementing regulations that fit the purpose, ensuring that well-established and beneficial systems receive support and spread quickly. The specific impacts of top-down and bottom-up policy approaches on IS and CE projects in many geographic and cultural situations may be the subject of future research. Furthermore, examining how well various stakeholder coordination mechanisms work to implement policies will shed light on how to improve cooperation and alignment in IS and CE to achieve shared objectives. Furthermore, looking into possible synergies between top-down and bottom-up techniques may provide insightful information about optimising policy frameworks for scalability and maximum efficacy. Understanding how to successfully negotiate the complicated terrain of corporate environmentalism can be improved by looking into these areas further, which will ultimately lead to better-informed decision-making and sustainable worldwide practices. 7. Conflict of Interest The author has no conflict of interest to declare. 6. References Achia, T. N., Wangombe, A., & Khadioli, N. (2010). A logistic regression model to identify key determinants of poverty using demographic and health survey data. European Journal of Social Sciences, 13(1), 38-45. Adibi, N., Lafhaj, Z., Yehya, M., & Payet, J. (2017). Global Resource Indicator for life cycle impact assessment: Applied in wind turbine case study. Journal of Cleaner Production, 165, 1517-1528. Argüelles, M., Benavides, C., & Fernández, I. (2014). A new approach to the identification of regional clusters: hierarchical clustering on principal components. Applied Economics, 46(21), 2511-2519. Azevedo, S. G., Godina, R., & Matias, J. C. d. O. (2017). Proposal of a sustainable circular index for manufacturing companies. Resources, 6(4), 63. Bocken, N. M., De Pauw, I., Bakker, C., & Van Der Grinten, B. (2016). Product design and business model strategies for a circular economy. Journal of industrial and production engineering, 33(5), 308-320. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 39 Bovea, M. D., & Pérez-Belis, V. (2018). Identifying design guidelines to meet the circular economy principles: A case study on electric and electronic equipment. Journal of Environmental Management, 228, 483-494. doi:https://doi.org/10.1016/j.jenvman.2018.08.014 Cayzer, S., Griffiths, P., & Beghetto, V. (2017). Design of indicators for measuring product performance in the circular economy. International Journal of Sustainable Engineering, 10(4-5), 289-298. Chun-rong, J., & Jun, Z. (2011). Evaluation of regional circular economy based on matter element analysis. Procedia Environmental Sciences, 11, 637-642. Cong, L., Zhao, F., & Sutherland, J. W. (2019). A design method to improve end-of-use product value recovery for circular economy. Journal of Mechanical Design, 141(4). Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical assessment, research & evaluation, 10(7), 1-9. Das, S. K., Yedlarajiah, P., & Narendra, R. (2000). An approach for estimating the end-of-life product disassembly effort and cost. International Journal of Production Research, 38(3), 657-673. doi:10.1080/002075400189356 de Oliveira, C. T., Dantas, T. E. T., & Soares, S. R. (2021). Nano and micro level circular economy indicators: Assisting decision-makers in circularity assessments. Sustainable Production and Consumption, 26, 455-468. doi:https://doi.org/10.1016/j.spc.2020.11.024 De Pascale, A., Arbolino, R., Szopik-Depczyńska, K., Limosani, M., & Ioppolo, G. (2021). A systematic review for measuring circular economy: The 61 indicators. Journal of Cleaner Production, 281, 124942. doi:https://doi.org/10.1016/j.jclepro.2020.124942 Di Maio, F., Rem, P. C., Baldé, K., & Polder, M. (2017). Measuring resource efficiency and circular economy: A market value approach. Resources, Conservation and Recycling, 122, 163-171. Faizi, S., Rashid, T., Sałabun, W., Zafar, S., & Wątróbski, J. (2018). Decision making with uncertainty using hesitant fuzzy sets. International Journal of Fuzzy Systems, 20(1), 93-103. Favi, C., Germani, M., Luzi, A., Mandolini, M., & Marconi, M. (2017). A design for EoL approach and metrics to favour closed-loop scenarios for products. International Journal of Sustainable Engineering, 10(3), 136-146. Felicio, M., Amaral, D., Esposto, K., & Durany, X. G. (2016). Industrial symbiosis indicators to manage eco-industrial parks as dynamic systems. Journal of Cleaner Production, 118, 54-64. Figge, F., Thorpe, A. S., Givry, P., Canning, L., & Franklin-Johnson, E. (2018). Longevity and circularity as indicators of eco-efficient resource use in the circular economy. Ecological Economics, 150, 297-306. Franklin-Johnson, E., Figge, F., & Canning, L. (2016). Resource duration as a managerial indicator for Circular Economy performance. Journal of Cleaner Production, 133, 589-598. Fregonara, E., Giordano, R., Ferrando, D. G., & Pattono, S. (2017). Economic-environmental indicators to support investment decisions: A focus on the buildings’ end-of-life stage. Buildings, 7(3), 65. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 40 Gallego-Schmid, A., Chen, H.-M., Sharmina, M., & Mendoza, J. M. F. (2020). Links between circular economy and climate change mitigation in the built environment. Journal of Cleaner Production, 260, 121115. doi:https://doi.org/10.1016/j.jclepro.2020.121115 Garson, D. (2009). Reliability analysis. [Online]. Available at: http://faculty. chass. ncsu. edu/garson/PA765/reliab. htm. In: Accessed 15 Sep 2017. Geng, Y., Fu, J., Sarkis, J., & Xue, B. (2012). Towards a national circular economy indicator system in China: an evaluation and critical analysis. Journal of Cleaner Production, 23(1), 216-224. doi:https://doi.org/10.1016/j.jclepro.2011.07.005 Geng, Y., Liu, Y., Liu, D., Zhao, H., & Xue, B. (2011). Regional societal and ecosystem metabolism analysis in China: A multi-scale integrated analysis of societal metabolism (MSIASM) approach. Energy, 36(8), 4799-4808. Geng, Y., Zhang, P., Côté, R. P., & Fujita, T. (2009). Assessment of the national eco‐industrial park standard for promoting industrial symbiosis in China. Journal of Industrial Ecology, 13(1), 15-26. Geng, Y., Zhang, P., Ulgiati, S., & Sarkis, J. (2010). Emergy analysis of an industrial park: The case of Dalian, China. Science of The Total Environment, 408(22), 5273-5283. doi:https://doi.org/10.1016/j.scitotenv.2010.07.081 Guo-gang, J. (2011). Empirical analysis of regional circular economy development--study based on Jiangsu, Heilongjiang, Qinghai Province. Energy Procedia, 5, 125-129. Haas, W., Krausmann, F., Wiedenhofer, D., & Heinz, M. (2015). How circular is the global economy?: An assessment of material flows, waste production, and recycling in the European Union and the world in 2005. Journal of Industrial Ecology, 19(5), 765-777. Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006). Multivariate data analysis (Vol. 6). Pearson Prentice Hall Upper Saddle River. Haupt, M., Vadenbo, C., & Hellweg, S. (2017). Do we have the right performance indicators for the circular economy?: insight into the Swiss waste management system. Journal of Industrial Ecology, 21(3), 615-627. Huysman, S., De Schaepmeester, J., Ragaert, K., Dewulf, J., & De Meester, S. (2017). Performance indicators for a circular economy: A case study on post-industrial plastic waste. Resources, Conservation and Recycling, 120, 46-54. doi:https://doi.org/10.1016/j.resconrec.2017.01.013 Huysman, S., Debaveye, S., Schaubroeck, T., De Meester, S., Ardente, F., Mathieux, F., & Dewulf, J. (2015). The recyclability benefit rate of closed-loop and open-loop systems: A case study on plastic recycling in Flanders. Resources, Conservation and Recycling, 101, 53-60. Jacobsen, N. B. (2006). Industrial symbiosis in Kalundborg, Denmark: a quantitative assessment of economic and environmental aspects. Journal of Industrial Ecology, 10(1‐2), 239-255. Jolliffe, I. T. (2002). Principal component analysis and factor analysis. In Principal component analysis (pp. 150-166). Karlsson, M., & Wolf, A. (2008). Using an optimisation model to evaluate the economic benefits of industrial symbiosis in the forest industry. Journal of Cleaner Production, 16(14), 1536-1544. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 41 Kaur, M., & Kaur, U. (2013). Comparison Between K-means and Hierarchical Algorithm Using Query Redirection. International Journal of Advanced Research in Computer Science and Software Engineering, 3(7), 1454-1459. Kayal, B., Abu-Ghunmi, D., Abu-Ghunmi, L., Archenti, A., Nicolescu, M., Larkin, C., & Corbet, S. (2019). An economic index for measuring firm’s circularity: The case of water industry. Journal of Behavioral and Experimental Finance, 21, 123- 129. doi:https://doi.org/10.1016/j.jbef.2018.11.007 Kristensen, H. S., & Mosgaard, M. A. (2020). A review of micro level indicators for a circular economy–moving away from the three dimensions of sustainability? Journal of Cleaner Production, 243, 118531. Lee, H. M., Lu, W. F., & Song, B. (2014). A framework for assessing product End-Of-Life performance: reviewing the state of the art and proposing an innovative approach using an End-of-Life Index. Journal of Cleaner Production, 66, 355-371. Li, R., & Su, C. (2012). Evaluation of the circular economy development level of Chinese chemical enterprises. Procedia Environmental Sciences, 13, 1595-1601. Linder, M., Sarasini, S., & van Loon, P. (2017). A metric for quantifying product‐level circularity. Journal of Industrial Ecology, 21(3), 545-558. MacArthur, E. (2015). Circularity indicators: An approach to measuring circularity. Methodology, 23. Mandolini, M., Favi, C., Germani, M., & Marconi, M. (2018). Time-based disassembly method: how to assess the best disassembly sequence and time of target components in complex products. The International Journal of Advanced Manufacturing Technology, 95(1), 409-430. Manly, B. F. (2005). Tests of significance with multivariate data. In Multivariate statistical methods: a primer, (3rd ed.) (pp. 105-124). Boca Raton, FL: Chapman & Hall. Marconi, M., Germani, M., Mandolini, M., & Favi, C. (2019). Applying data mining technique to disassembly sequence planning: a method to assess effective disassembly time of industrial products. International Journal of Production Research, 57(2), 599-623. Mayer, A., Haas, W., Wiedenhofer, D., Krausmann, F., Nuss, P., & Blengini, G. A. (2019). Measuring progress towards a circular economy: a monitoring framework for economy‐wide material loop closing in the EU28. Journal of Industrial Ecology, 23(1), 62-76. Mazur-Wierzbicka, E. (2021). Circular economy: advancement of European Union countries. Environmental Sciences Europe, 33(1), 1-15. McCarthy, B., Kapetanaki, A. B., & Wang, P. (2019). Circular agri-food approaches: will consumers buy novel products made from vegetable waste? Rural Society, 28(2), 91-107. Mesa, J., Esparragoza, I., & Maury, H. (2018). Developing a set of sustainability indicators for product families based on the circular economy model. Journal of Cleaner Production, 196, 1429-1442. Mohamed Sultan, A. A., Lou, E., & Mativenga, P. T. (2017). What should be recycled: An integrated model for product recycling desirability. Journal of Cleaner Production, 154, 51-60. doi:https://doi.org/10.1016/j.jclepro.2017.03.201 European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 42 Moraga, G., Huysveld, S., Mathieux, F., Blengini, G. A., Alaerts, L., Van Acker, K., Dewulf, J. (2019). Circular economy indicators: What do they measure? Resources, Conservation and Recycling, 146, 452-461. doi:https://doi.org/10.1016/j.resconrec.2019.03.045 Moriguchi, Y. (2007). Material flow indicators to measure progress toward a sound material-cycle society. Journal of Material Cycles and Waste Management, 9(2), 112-120. doi:10.1007/s10163-007-0182-0 Nelen, D., Manshoven, S., Peeters, J. R., Vanegas, P., D'Haese, N., & Vrancken, K. (2014). A multidimensional indicator set to assess the benefits of WEEE material recycling. Journal of Cleaner Production, 83, 305-316. Pagotto, M., & Halog, A. (2016). Towards a circular economy in Australian agri‐food industry: an application of input‐ output oriented approaches for analysing resource efficiency and competitiveness potential. Journal of Industrial Ecology, 20(5), 1176-1186. Parchomenko, A., Nelen, D., Gillabel, J., & Rechberger, H. (2019). Measuring the circular economy - A Multiple Correspondence Analysis of 63 metrics. Journal of Cleaner Production, 210, 200-216. doi:https://doi.org/10.1016/j.jclepro.2018.10.357 Park, H.-S., & Behera, S. K. (2014). Methodological aspects of applying eco-efficiency indicators to industrial symbiosis networks. Journal of Cleaner Production, 64, 478-485. Park, J. Y., & Chertow, M. R. (2014). Establishing and testing the “reuse potential” indicator for managing wastes as resources. Journal of Environmental Management, 137, 45-53. Qing, Y., Qiongqiong, G., & Mingyue, C. (2011). Study and integrative evaluation on the development of circular economy of Shaanxi province. Energy Procedia, 5, 1568-1578. Sałabun, W., Palczewski, K., & Wątróbski, J. (2019). Multicriteria approach to sustainable transport evaluation under incomplete knowledge: Electric bikes case study. Sustainability, 11(12), 3314. Scheepens, A. E., Vogtländer, J. G., & Brezet, J. C. (2016). Two life cycle assessment (LCA) based methods to analyse and design complex (regional) circular economy systems. Case: making water tourism more sustainable. Journal of Cleaner Production, 114, 257-268. doi:https://doi.org/10.1016/j.jclepro.2015.05.075 Smol, M., Kulczycka, J., & Avdiushchenko, A. (2017). Circular economy indicators in relation to eco-innovation in European regions. Clean Technologies and Environmental Policy, 19(3), 669-678. doi:10.1007/s10098-016-1323-8 Stanković, J. J., Janković-Milić, V., Marjanović, I., & Janjić, J. (2021). An integrated approach of PCA and PROMETHEE in spatial assessment of circular economy indicators. Waste Management, 128, 154-166. doi:https://doi.org/10.1016/j.wasman.2021.04.057 Su, B., Heshmati, A., Geng, Y., & Yu, X. (2013). A review of the circular economy in China: moving from rhetoric to implementation. Journal of Cleaner Production, 42, 215-227. doi:https://doi.org/10.1016/j.jclepro.2012.11.020 Tiejun, D. (2010). Two quantitative indices for the planning and evaluation of eco-industrial parks. Resources, Conservation and Recycling, 54(7), 442-448. European Journal of Social Impact and Circular Economy - ISSN: 2704-9906 DOI: 10.13135/2704-9906/7691 Published by University of Turin http://www.ojs.unito.it/index.php/ejsice/index EJSICE content is licensed under a Creative Commons Attribution 4.0 International License 43 Tisserant, A., Pauliuk, S., Merciai, S., Schmidt, J., Fry, J., Wood, R., & Tukker, A. (2017). Solid waste and the circular economy: a global analysis of waste treatment and waste footprints. Journal of Industrial Ecology, 21(3), 628-640. van Schaik, A., & Reuter, M. A. (2016). Recycling indices visualising the performance of the circular economy. World Met Erzmetall, 69, 5-20. Vanegas, P., Peeters, J. R., Cattrysse, D., Tecchio, P., Ardente, F., Mathieux, F., . . . Duflou, J. R. (2018). Ease of disassembly of products to support circular economy strategies. Resources, Conservation and Recycling, 135, 323-334. Vyas, S., & Kumaranayake, L. (2006). Constructing socio-economic status indices: how to use principal components analysis. Health policy and planning, 21(6), 459-468. Wen, Z., & Li, R. (2010). Materials metabolism analysis of China's highway traffic system (HTS) for promoting circular economy. Journal of Industrial Ecology, 14(4), 641-649. Wenbo, L. (2011). Comprehensive evaluation research on circular economic performance of eco-industrial parks. Energy Procedia, 5, 1682-1688. Wu, H.-q., Shi, Y., Xia, Q., & Zhu, W.-d. (2014). Effectiveness of the policy of circular economy in China: A DEA-based analysis for the period of 11th five-year-plan. Resources, Conservation and Recycling, 83, 163-175. Xiong, P., Dang, Y., & Qian, W. (2011). The empirical analysis of circular economy development efficiency in Jiangsu Province. Energy Procedia, 5, 1732-1736. Yobe, C. L., Mudhara, M., & Mafongoya, P. (2019). Livelihood strategies and their determinants among smallholder farming households in KwaZulu-Natal province, South Africa. Agrekon, 1-14. Zhao, H., Guo, S., & Zhao, H. (2018). Comprehensive benefit evaluation of eco-industrial parks by employing the best- worst method based on circular economy and sustainability. Environment, development and sustainability, 20(3), 1229-1253. Zwolinski, P., Lopez-Ontiveros, M.-A., & Brissaud, D. (2006). Integrated design of remanufacturable products based on product profiles. Journal of Cleaner Production, 14(15-16), 1333-1345.