Scientometric Analysis of Circular Innovation: A Novel Approach for Sustainability

The concept of CE has become prominent globally among countries, policymakers, organizations, scientific institutions, and firms in the 21st century (M​e​r​l​i​ ​e​t​ ​a​l​.​,​ ​2​0​1​8). It is increasingly regarded as a solution to environmental, social, and economic challenges, stemming from increasing waste generation, pollution, consumption of non-renewable resources, and resource scarcity (L​i​e​d​e​r​ ​&​ ​R​a​s​h​i​d​,​ ​2​0​1​6). The linear economy that has prevailed until today adopts the “take-make-consume-and-throw away” approach, including over-exploitation of natural resources and accumulation of waste products polluting the environment, which is now about to be ineffective, because it is insufficient in combating climate change and ensuring sustainability. However, the CE, a newly developing phenomenon, endeavors to minimize raw materials per unit of output and recycle wastes as much as possible to reuse them as production inputs. The CE approach recycles wastes or used materials by creating added value, and reuse them in ways that benefit the economy. Therefore, companies need to make radical changes in their manufacturing processes. Transforming the production processes into a circular manner necessitates research and development (R&D) and implementation of novel activities, which requires companies to innovate by adopting circular principles. Inherently, it is more inclusive to use the concept of CI, compared with sustainable, eco, or green innovation in the literature.

Innovation is key to accelerating CE because it runs across the entire CE process and value chain, ranging from business production and delivery of products or services to the end user. After completing its economic life, a product is thrown away and turned into garbage in linear economic model. On the contrary, the end user focuses on maximizing the life of the product by repairing it, delivering it to collection centers for recycling or different creative uses in CE. Therefore, CI refers to a model, in which the end user adopts the open innovation model and all value chain members have sustainability awareness and responsibility, including redesign and production processes of an enterprise.

A joint report prepared by the W​o​r​l​d​ ​B​u​s​i​n​e​s​s​ ​C​o​u​n​c​i​l​ ​f​o​r​ ​S​u​s​t​a​i​n​a​b​l​e​ ​D​e​v​e​l​o​p​m​e​n​t​ ​&​ ​T​h​e​ ​B​o​s​t​o​n​ ​C​o​n​s​u​l​t​i​n​g​ ​G​r​o​u​p​ ​(​2​0​1​8​) asserted that three types of CI with varying degrees of complexity could be considered, namely, process, product, and business model CI. First, circular process innovation focused on developing and implementing new or significantly improved production, logistics, or recycling methods, involving production as well as collecting and recycling waste created during the manufacturing process, which was the easiest to implement and achieve, because it dealt with operational areas under direct control of the organization. These innovations also had lower risks because changes in manufacturing processes were unlikely to affect customers directly. However, strategic planning and change management were required to ensure employees seamlessly adjusted to the new processes. Second, circular product innovation concentrated on developing and implementing new or significantly improved products or services and covered more value chain areas, including procurement, product development, use of products by customers, and ways of “collecting and recycling” products at the end of their life cycle by the firm. This required thorough research on technologies, skills, market trends, and consumer behavior. Success depended on several factors, such as commitment to recruit and train employees, coordination and cross-functional cooperation among employees with different skills across departments, and inclusion of external partners with different skills and/or resources to produce circular products. Circular product innovation may seem more complex and difficult to initiate and achieve as it required the buy-in and active involvement of various divisions in the organization. Third, circular business innovation was the most complicated, because it involved significant changes in the entire value chain, including marketing and sales of products, which required considerable time and resources to educate customers, influence and change consumption habits, disrupt conventional market practices, pricing strategies, etc. This was the most challenging to implement, because it demanded new logic of value generation of a firm, new ways of thinking by designers, new research insights, and collaboration between multiple parties across the organization and from the marketplace.

CI is highly attractive for firms, mainly because it creates and captures lost and hidden value for a firm, the environment, and society (K​o​n​i​e​t​z​k​o​ ​e​t​ ​a​l​.​,​ ​2​0​2​0). In addition, implications of CI often are beyond the individual firm and affect the value chain. Therefore, examining CI, a significant and newly developing concept in the literature, is crucial. This study aimed to answer these following research questions (RQ):

RQ 1: What are the topics discussed extensively in the literature?

RQ 2: What concepts are associated with CI?

RQ 3: What research gaps exist in the field?

This study attempted to assess the CI literature using the scientometric analysis (SA) methodology. However, due to lack of SA in CI, there was a lacuna in objectively and comprehensively understanding the literature, which was solved by critical assessment and analysis using co-occurrence and bibliographic coupling (BC) document analysis methods. This paper was the first to provide an up-to-date and inclusive big picture of CI literature and added value to the innovation and sustainability literature. CE, circular business models, sustainability, and sustainable development were found to be the most prominent issues related to CI.

To answer the above questions, the research framework of this paper included RQs and reasons in Chapter 1, source of data and research method in Chapter 2, and co-occurrence and BC document analysis of CI in Chapter 3. Finally, Chapter 4 and 5 interpreted and summarized the above results and identified the limitation of this paper. Furthermore, the paper presented future research directions of CI concerning business management.

Scientometrics is defined as an informational process where a scientific area is quantitatively studied, using bibliometric links, co-citation, and cognitive structure mapping of scientific knowledge, and is considered as one of the essential disciplines of science. Being first used by P​r​i​t​c​h​a​r​d​ ​(​1​9​6​9​) as early as 1969, bibliometrics has been accepted as a mathematical and statistical method for publications and is also considered a division of scientometrics (I​a​n​c​u​ ​e​t​ ​a​l​.​,​ ​2​0​2​2). Unlike conventional literature review, whose research framework is based on the subjective experience of authors, bibliometrics analyzes objective scientific data (P​a​n​ ​e​t​ ​a​l​.​,​ ​2​0​2​2).

Bibliometric analysis provides valuable, chronological, and practical statistical information, which can further be used by those assessing such systematic progress in the research area (D​u​q​u​e​ ​O​l​i​v​a​ ​e​t​ ​a​l​.​,​ ​2​0​0​6; K​a​t​o​c​h​,​ ​2​0​2​2). Furthermore, quantitatively, scientometric research using bibliographic data (B​r​o​a​d​u​s​,​ ​1​9​8​7) presents a common overview of a scientific area according to various indicators. By adopting a scientometric methodology, this study aimed to determine the most relevant literature and different streams in the field and identify the leading journals, primary authors, and countries at the forefront of CI, which is helpful to get a quick overview of state of the art in a particular area. Moreover, the key results can be precious for newcomers in the field and for identifying research gaps for future CI studies.

The first step in scientometric research involves choosing the databases of documents (“articles” in this study). This study selected the WoS database, because it is widely acknowledged for collecting integrated and reliable data. Furthermore, it provides multidisciplinary research from multiple sources based on impact assessments, meeting the highest quality standards (P​o​d​s​a​k​o​f​f​ ​e​t​ ​a​l​.​,​ ​2​0​0​8). The next step is to select meaningful keywords that clearly return articles on this topic. Two primary keywords “circular” and “inno*” (the symbol * allows for the inclusion of the plural) were selected, because innovation can be written in various forms, such as innovation, innovative, innovativeness, etc. This paper obtained 917 net results between 1975-2022 after searching the database and applying some filters to the search results. To focus on the most revealing parts of the research, merely “journal articles” were selected. In addition, these articles should be written in English and restricted to Social Sciences Citation Index (SSCI). The results showed that the first article in this field was written in 1992. The study was conducted in October 2022, and excluded publications in 2022, because all related articles were categorized in a full year period.

Developed by V​a​n​ ​E​c​k​ ​&​ ​W​a​l​t​m​a​n​ ​(​2​0​1​0​), the VOSviewer is a popular visualization tool that utilizes the VOS (visualization of similarities) for cluster analysis. It was used for bibliometric mapping and visualization of publications, because it allowed algorithmic and automatic analysis of the publications, co-authorship, keyword cluster exploration, and the most productive authors in literature.

The research data was processed by VOSviewer software, which visualized and illustrated authors contributed the most, journals, countries, and most cited articles in tables and maps. Figure 1 demonstrates that the first article found was published in 1992. However, the number of publications and citations about CI increased rapidly since 2010, particularly 2018.

The first ten categories with the largest number of articles in the total 917 WoS articles are presented in Table 1. Accordingly, it was determined that the studies related to CI were mostly in the categories of “Environmental Sciences”, “Green Sustainable Science Technology”, and “Environmental Studies”.

Figure 2 shows the countries with the most publications in the WoS database. In addition, Turkey was also included in the graph to determine its status in the world ranking. According to the figure, Italy ranks first with 127 CI publications (13.8%); England is second with 118 publications (12.9%); the People’s Republic of China ranks third with 105 publications (11.4%); Turkey is in the 25th place with 13 publications (1.4%).

Among the publications in the WoS database, the top ten most cited articles are presented in Table 2. Accordingly, the most cited article was written by K​i​r​c​h​h​e​r​r​ ​e​t​ ​a​l​.​ ​(​2​0​1​8​), which examined four types of CE barriers in the European Union (EU), namely, cultural, market, regulatory, and technological barriers. The second most cited study was written by D​’​A​m​a​t​o​ ​e​t​ ​a​l​.​ ​(​2​0​1​7​), which comprehensively analyzed the diversity within and between the CE, green economy, and bio-economy.

Keywords are the words mentioned after the abstract section in an article and some other index terms associated with the document, which are the principal terms used throughout the article and directly involved with the article’s perspective. Co-occurrence analysis is a preferred method in bibliometric analysis and is made based on the keywords specified by authors in the article in order to understand popular and potential future topics.

Keyword association was created according to the number of times that keywords were repeated in the documents (V​a​n​ ​E​c​k​ ​&​ ​W​a​l​t​m​a​n​,​ ​2​0​1​3). In the co-occurrence map, the circle size indicated the repetition times of keywords. That is, the larger the circle, the more frequent the keywords in that field were repeated. This paper emphasized the relative distance of one concept from another, and the proximity of keywords between points meant that the two concepts were more relevant. Furthermore, the shorter and thicker the lines, the stronger the bond between keywords.

In keyword co-occurrence map in Figure 3, each cluster of the CI thematic area is identified with a different color. The concept repeated the most in each cluster was shown in the larger dimension representing the central concept. Thematic clusters were determined and named by examining the keywords in each cluster.

This paper evaluated the results produced by the VOSviewer program and named the clusters. Accordingly, the themes covered in the articles were classified into thirteen clusters, which were created by the keyword co-occurrence, namely, CE & environment, business models, open innovation, circular urbanization, innovation orientation, institutional & managerial & strategical issues, circular production management, system design, supply chain management, marketing issues, paradigms, entrepreneurship, decision making & methodologies. Clusters and featured keywords are presented in Table 3.

According to the findings, it is noteworthy that the concept of CE is the most studied subject in the field and has the largest circle size. The second most discussed topics in the articles were “circular business models, sustainability, and sustainable development”, shown in purple and yellow.

As the analysis unit, the BC document analysis method offers opportunities to interpret and decipher the intellectual structure of the research field comprehensively and objectively (P​a​t​i​l​ ​&​ ​R​a​h​m​a​n​,​ ​2​0​2​2; Z​u​p​i​c​ ​&​ ​Č​a​t​e​r​,​ ​2​0​1​5). BC links papers referencing an identical set of papers and evaluates the similarity among citing documents, which is forward-looking and suitable for capturing current research trends and studying emergent literature fields within a research area (B​o​y​a​c​k​ ​&​ ​K​l​a​v​a​n​s​,​ ​2​0​1​0; V​o​g​e​l​ ​&​ ​G​ü​t​t​e​l​,​ ​2​0​1​3). In addition, BC associates documents with similarities in their reference lists, indicating the probability of a shared topic (M​a​s​e​d​a​ ​e​t​ ​a​l​.​,​ ​2​0​2​2).

BC uses citations to explain the similarities between two authors, documents, countries, or institutions, under the assumption that two documents referencing the third paper are highly related and should concentrate in a cluster solution of the visualization map (M​a​s​-​T​u​r​ ​e​t​ ​a​l​.​,​ ​2​0​2​1). On the one hand, BC document analysis generates groups of recent documents based on the shared references they carry, representing the current knowledge. On the other hand, the document co-citation analysis clusters old documents cited in recent research, representing basic knowledge (M​u​k​h​e​r​j​e​e​ ​e​t​ ​a​l​.​,​ ​2​0​2​2; P​a​t​i​l​ ​&​ ​R​a​h​m​a​n​,​ ​2​0​2​2).

BC occurs when two documents cite the same paper (M​u​l​e​t​-​F​o​r​t​e​z​a​ ​e​t​ ​a​l​.​,​ ​2​0​1​8), which illustrates the strength of specific article in relation to a set of other articles. The BC documents analysis possibly shows what authors and documents are correlated through multiple citations (C​a​v​a​l​c​a​n​t​e​ ​e​t​ ​a​l​.​,​ ​2​0​2​1). This paper made BC analysis with at least two citations using VOSviewer’s full counting method for documents, which were the unit of analysis. Figure 4 presents the BC of documents and allows us to measure and observe their connection intensity.

The map in Figure 4 reveals three well-defined clusters and concomitant citations between the documents drawn in lines. Topics of the highest-cited authors in the green cluster mostly focus on circular business models, circular supply chains, and circular start-ups. In the second cluster in red, the topics are mainly about CE barriers, EU policies towards CE, and circular supply chains. Finally, the third cluster in blue has the least citation density, whose topics are about circular-oriented innovation.

This study aimed to investigate how CI was handled in the WoS articles because CE needed more systemic innovation approaches. In addition, business and innovation ecosystems rarely shaped around CE or sustainability issues. Therefore, scientometric analysis was made to examine this issue.

The research results showed that CI studies were less cited than other subjects, because CI is an emerging field, and related publications have only begun in the last few years. In other words, very few research has studied CI. WoS-based articles discussed different types of innovation and the concept of CE. Prominent innovation types in CE were business model innovation, collaborative innovation, product and process design, and eco-innovation.

CE is necessary to separate economic growth from resource consumption and environmental impacts. However, its effective implementation requires a systemic change in supply chains, including both technological and non-technological innovations (M​e​n​d​o​z​a​ ​e​t​ ​a​l​.​,​ ​2​0​1​7). Thus, there is a need for innovations beyond enterprises. Apart from eco-innovation, CI requires adopting an open CI approach, with all stakeholders participating in the value chain. Circular-oriented innovation research is nascent (B​r​o​w​n​ ​e​t​ ​a​l​.​,​ ​2​0​2​1). Instead of examining CI in depth, E​i​s​e​n​r​e​i​c​h​ ​e​t​ ​a​l​.​ ​(​2​0​2​1​) only mentioned the necessity of collaborating with external partners, such as research institutions, suppliers, customers, or technical experts. Therefore, research gaps exist in open CI, which needs to be explored.

In addition, N​y​s​t​r​ö​m​ ​e​t​ ​a​l​.​ ​(​2​0​2​1​) pointed out that little effort concentrated on combining adaptable design with CE research. Innovation instruments in most existing circular business models are qualitative and focus on creating initial concepts (B​o​c​k​e​n​ ​e​t​ ​a​l​.​,​ ​2​0​1​9). However, adaptable design techniques are inverse: detailed-oriented and quantitative (F​l​e​t​c​h​e​r​ ​e​t​ ​a​l​.​,​ ​2​0​0​9; L​i​ ​e​t​ ​a​l​.​,​ ​2​0​0​8). Thus, the quantitative and technical approaches required in adaptable design may be prohibitively challenging for use in circular business model development (N​y​s​t​r​ö​m​ ​e​t​ ​a​l​.​,​ ​2​0​2​1).

CI, or circular-oriented innovation mentioned in the literature, is an emerging field, and more research is needed. The so-called collaborative, networked, or ecosystem innovations essentially fall under the umbrella of open innovation. It is essential to study open CI, because CE requires co-creation and collaboration of different stakeholders in the value chain. In addition, it is important to specifically study both circular product innovation to carry out R&D activities of redesign and remanufacturing, which are elements of CE, and transformation of the used materials and remanufacturing of these transformed materials.

This study made the following contributions by applying the scientometrics method to the CI:

(1) Helping newcomers to rapidly know the situations of present research;

(2) Helping experts to understand the research trends in a data-driven way;

(3) Helping scientists to make reasonable plans for prospective studies.

This paper made the keyword co-occurrence analysis in order to respond to RQ 1. The themes of CI articles in the WoS database were classified into thirteen clusters and named, such as CE & environment, business models, innovation orientation, open innovation, circular urbanization, circular production management, system design, institutional & managerial & strategical issues, supply chain management, marketing issues, paradigms, decision-making & methodologies, and entrepreneurship.

Then this paper made BC document analysis using VOSviewer software in order to investigate RQ 2. The most cited articles were reviewed and highlighted in three clusters: circular business models, circular supply chains, and circular start-ups in the first cluster; CE barriers, EU policies towards CE, and circular supply chains in the second cluster; circular-oriented innovation in the third cluster.

As a result of applied analysis, this paper examined the less covered and cited topics of articles. A research gap was observed in several areas in response to RQ 3, such as circular strategies, circular product and process innovation, open CI (also including crowdsourcing, circular ecosystem innovation), digital technology and CE, circular design, case study methodology in CE, product-service systems, circular entrepreneurship, circular supply chain management, circular business models, manufacturing companies, ethical purchase intention, customers’ attitude, nature-based solutions, circular cities/buildings, food supply chains, and sustainable business decisions. In addition, future studies can explore how to apply innovation in CE in various industry sectors using case study methodology and different decision-making techniques, as well as what type of CI is needed in specific sector and product groups.

The research samples of this study are limited to articles listed in the WoS Core Collection database, excluding publications in databases like Scopus and Pubmed, sources not available online, and other publication types. It is suggested that future studies can overcome this limitation by analyzing other publications instead of articles, including publications in different international databases.