Keywords

1 Introduction

The population of the world has been predicted to be about 9 billion by 2050 and 10.1 billion by 2100 [1]. This massive population growth puts tremendous stress on the environment and the resource [2]. So, there is the requirement of more extraction of raw materials and consequently more waste.

The current economic model is linear in nature [3] and aims at producing products from raw material, sold and dumped as waste after serving the purpose of use [4]. This economic model is called the take-make-use-destroy model [5]. This linear model is not sustainable in the long run as this model continuously destroys the limited natural resources [6]. The adoption of the circular economy (CE) can surmount the limitations of the linear model of the economy. A CE as described by the authors of [7] is ‘an economic system that is based on business models which replace the ‘end-of-life’ concept with reducing, alternatively reusing, recycling and recovering materials in production/distribution and consumption processes, thus operating at the micro level (products, companies, consumers), meso level (eco-industrial parks) and macro level (city, region, nation and beyond), with the aim to accomplish sustainable development, which implies creating environmental quality, economic prosperity and social equity, to the benefit of current and future generations’.

The aim of the CE was to limit the extraction of raw materials and the production of waste by recovering and reusing as many of the materials possible. The CE has many significant advantages over the traditional economy as it is evident in literature through various studies and researches [8]. The linear economy flow model is causing several serious environmental impacts such as global warming and natural resource depletion. Unlike linear economy practices, CE-oriented approach focuses on reuse, remanufacturing, refurbishment, repair of products by cascading and upgrading. CE also emphasise on the exploitation of renewable energies such as solar, wind and biomass energy throughout the product value chain.

The switching of a linear economy to a CE is a very challenging task and demands motivational factors. These drivers facilitate in the adoption of the CE which makes it necessary for the identification and evaluation of the drivers before the adoption of the CE. With the use of DEMATEL, this study acknowledges the position of each driver in the CE system along with their degree of effectiveness on the system and on the other drivers. With the help of experts, the results obtained by the DEMATEL provide the major dominant drivers [9]. The focus is put on these influential drivers to make better the overall system. The other benefits of this technique are that the DEMATEL technique is suitable with limited data, [10]. Understanding the drivers and their driving potential will be beneficial in formulating the strategies in order to decrease the effect of barriers. In this paper, the objective of research is to find the drivers of the CE and explore the causal relationships among the drivers.

The remaining paper is structured as follows: Sect. 2 reviews the literature on the CE. Section 3 identifies the significant drivers of the CE. Section 4 presents the research methodology of the paper. The results are illustrated in Sect. 5. Sections 6, 7 and 8 present the discussions, implications and conclusions, respectively.

2 Literature Review

A CE approach is to facilitate growth and, at the same, time minimises resource usage by closing all resource loops and reconnecting them at various nodes, thereby reducing and ultimately eliminating waste [11].

To reveal the current status of adoption of the CE, various research scholars have generally reviewed the concepts of CE. Ghisellini et al. [5] first reviewed the CE and compared CE adoption and practices of China with Europe, Japan and the world. Govindan et al. [8] presented a comprehensive review of barriers, drivers and practices that affect the adoption of the CE through supply chain perspective.

Sun et al. [12] assessed the concepts, practices and assessment tools of CE in china. Geng et al. [13] evaluated the advancement and extent of adoption of the CE in China. Kirchherr et al. [7] examined the various conceptualisations of the CE to clarify the present interpretation of CE concepts by analysing 114 definitions. Lewandowski [14] assessed the current business models of circular to find out gaps that continued to be in the literature and presented a proposal for a new structure for better adoption of the CE. Ellen McArthur Foundation [15] presented a study that provides a toolkit for the policymakers who want to design a policy to step up the changes towards the CE. Liguori and Faraco [16] analysed the various conceptions of biotreatment such as biorefineries and waste management that encourage the adoption of the CE. Ghisellini et al. [5] analysed the changes involved with interlinking the systems of environment and economy.

3 Drivers Towards the Adoption of the Circular Economy

Drivers are the motivational requirement for the effective adoption of CE in a supply chain. For the successful implementation, these drivers are identified and investigated, and Table 1 describes the selected drivers.

Table 1 Drivers towards the adoption of circular economy
Full size table

4 Research Methodology

The paper aims to investigate the drivers of the CE through causal relationship. After finding the drivers through literature review and expert’s input, the drivers are examined using DEMATEL technique [27].

DEMATEL is the MCDM technique which was proposed and developed by the Battle Memorial Institute of Geneva during the period of 1972 and 1976. DEMATEL technique is summarised in the following steps.

  • Step I: Construction of direct influence matrix

A questionnaire is used to collect the response experts about the effect of one factor over another factor. The pairwise comparison values between factor ‘i’ and ‘j’, by kth expert, were expressed using a five-point linear scale ranging between 0 and 4; the numerals are used to indicate the strength of a relationship (please see Table 2).

Table 2 Scale and their interpretation
Full size table

The notation of xij indicates the influence of variable i on variable j. The elements of diagonal (i.e. i = j) of the direct relation matrix are zero. For every expert, a non-negative n × n matrix is acquired as Xk =  \(x_{{ij}}^{k}\) where k represents the kth experts (1 ≤ k ≤ H). The direct relation matrix is acquired from each expert (number of expert is H) in the form of X1, X2, X3 … XH.

  • Step II: Construct an overall direct relation matrix from H experts, the average matrix A = [aij] can be acquired from the Expression [1]:

$$a_{ij} = \frac{{\mathop \sum \nolimits_{k = 1}^{H} x_{ij}^{k} }}{H}$$
(1)

  • Step III: Construct a normalised initial direct relation matrix, B from Expressions (2) and (3).

$$B = A \cdot S$$
(2)
$${\text{where}}\,S = \frac{1}{{\max_{1 \le i \le n} \mathop \sum \nolimits_{j = 1}^{n} a_{ij} }}$$
(3)

The elements of matrix B lie between zero and one.

  • Step IV: Construct the total relation matrix ‘T’ from Expression (4)

$$T = B \cdot (I - B)^{ - 1}$$
(4)

where ‘I’ is the identity matrix.

  • Step V: Evaluate the causal parameters through Expressions (5) and (6):

$$R_{i} = \mathop \sum \limits_{j = 1}^{n} t_{ij} \,\text{for}\,\text{all}\,i$$
(5)
$$C_{j} = \mathop \sum \limits_{i = 1}^{n} t_{ij} \,\text{for}\,\text{all}\,j$$
(6)

where Ri indicates the sum of rows and Cj indicates the sum of columns.

  • Step VI: Construct a cause and effect diagram using the data of (R + C) and (R–C) given in Table 6, computed from Eqs. (7) and (8):

$$P_{i} = R_{i} + C_{j} |i = j$$
(7)
$$E_{i} = R_{i} - C_{j} |i = j$$
(8)

The difference (Ri− Ci) indicates the net influence that variable i adds to the system. Furthermore, if the value of (Ri − Ci) is positive, variable i is a net cause, and if the value of (Ri − Ci) is negative, variable i is a net receiver.

5 Results

The drivers towards the adoption of the CE are obtained through literature review. After discussing, the drivers with a five-member experts’ group for the deeper insights and ten drivers were finalised on the recommendations of experts (Table 1). After finalising the drivers, the decision makers were asked to assess the direct relation among the drivers of the CE on the scale of 0–4 (Table 2). The overall direct relation matrix is calculated using Expression 1 (Table 3).

Table 3 Overall direct relation matrix as obtained through expert inputs
Full size table

Then, the normalised direct relation matrix (B) is computed from Expressions (2) and (3) as shown in Table 4.

Table 4 Normalised direct relation matrix (B)
Full size table

After that, this matrix is changed into total relation matrix (T) using Expression (4) as shown in Table 5.

Table 5 Total relation matrix (T)
Full size table

In matrix T, the summation of rows and columns is denoted by R and C, respectively, using Expression (5).

Where Ri denotes the total influence of driver ‘i’ to the other drivers, and Ci denotes net effect on driver ‘j’ from other drivers. Following the computations of values R and C for every row and column, the prominence (Pi) and net effect (Ei) is computed utilising the Expressions (7) and (8) as shown in Table 6. The net cause/effect of each driver is determined through ‘Ei (i.e. R − C)’. If the value Ei is positive, then the driver produces the net cause, and if negative, then the driver is the net effect. Figure 1 shows the plots of R + C and R − C, which thereby depicts the causal relationship among the drivers of the CE. We have discussed these results with experts for further insights.

Table 6 Prominence (Pi) and net effect (Ei)
Full size table
Fig. 1
figure 1

Cause and effect diagram

Full size image

6 Discussion on Results

In this research, the relationship among the various factors was investigated by using DEMATEL, which crammed the space left by conventional models that have only considered the direct effects of a factor [28].

The results that show the driver effects on the system alongside on each other are shown in Table 5. Based on the ‘R + C’ values, the importance order of the drivers is D_05 > D_01 > D_10 > D_06 > D_09 > D_08 > D_02 > D_03 > D_07 > D_04. Thus, from the ‘R − C’ values, the drivers are classified into cause and effect groups. The drivers ‘government policies for cleaner production (D_01)’, ‘urbanisation (D_05)’, ‘consumer awareness (D_06)’, ‘cost reduction and financial profitability (D_09)’ and ‘scarcity of resources (D_10)’ having the positive values of R − C are categorised under cause group. The drivers ‘economic growth through adoption of CE (D_02)’, ‘environmental protection laws (D_03)’, ‘demand for renewable energy (D_04)’, ‘enhanced material and energy efficiency (D_07) and ‘improving the product quality (D_08)’ having the negative values of R − C are categorised under effect group.

The categorisation of drivers into cause and effect groups will facilitate the experts to recognise the cause group drivers, which are necessary to be controlled [29]. The matter of fact is that the variables in the influential group are hard to transform while the variables in the influenced group can be easily transformed [30]. The drivers who need improvement on priority basis will be recognised by the decision makers with the help of prioritisation of drivers while focusing on that driver will improve the other drivers as well as the whole system [31].

6.1 Influential Drivers

Among the influential drivers, the ‘government policies for cleaner production (D_01)’ has been identified as the most influential driver. This is justified because the government laws are vital for adopting and implementing CE practices in the supply chain. The second most influential driver is ‘urbanisation (D_05)’; the developmental changes necessitated by the shifting of people towards cities put pressure on the environment and hence demand a change. The adoption of the CE addresses this demand. The next influential driver is ‘scarcity of resources (D_10)’, as the natural resources such as water, non-renewable energy and raw material are limited. Growing population stresses the limited availability of resources and leads to non-sustainable development. The CE reduces the need for new raw materials by reusing existing materials. ‘Cost reduction and financial profitability (D_09)’ is another influential driver for the implementation of the CE practices. The least influential driver identified is ‘consumer awareness (D_06)’ which is significant for the adoption of the CE.

6.2 Influenced Drivers

The most influenced driver identified is ‘improving the product quality (D_08)’. It can be influenced by consumer awareness (D_06). The next most influenced driver identified is ‘enhanced material and energy efficiency (D_07)’; it is influenced by ‘scarcity of the resource (D_10)’ and ‘government policies for cleaner production (D_01)’. The next influenced driver is ‘demand for renewable energy (D_04)’; it could be influenced by ‘consumers awareness (D_06)’ and ‘environmental protection laws (D_03)’. The least influenced drivers are ‘economic growth through adoption of the CE (D_02)’ and ‘environmental protection laws (D_03)’.

7 Implications

This research can be useful for the successful implementation of the CE. The decision makers can get help from identified interrelationships between drivers. The categorisation of the drivers into an influential and influenced group may help the policymakers to develop the policies for the adoption of the CE. The organisation needs to primarily focus on the influential group drivers for the successful adoption of the CE. This study can be beneficial for the academia to develop the understanding of the drivers of the CE and interrelationship among them.

8 Conclusion

The drivers of the successful implementation of the CE practices are recognised through the systematic literature review and affirmed through expert’s opinion. After finalising the drivers, the relationship among the identified driver is determined using the DEMATEL approach. These drivers are classified into ‘influenced’ and ‘influential’ groups. The results of this study are discussed with the experts, and helpful cognisance is given in the discussion section. Finally, the research implication of this paper is provided. The result of this study suggested that most influential are ‘government policies for cleaner production’ and ‘urbanisation’. The most influenced drivers are ‘enhanced material and energy efficiency’ and ‘improving product quality’. The influential drivers are the major focus while drafting the policies to implement the CE. This study also has some limitations such as the expert’s opinion may be biased and subjective. To overcome such limitations, DEMATEL can be integrated with fuzzy and grey theory in future studies. Further, these drivers can be evaluated using the other MCDM technique such as fuzzy AHP, TOPSIS and BWM techniques.