International Journal of Sustainable Development &

World Ecology

ISSN: 1350-4509 (Print) 1745-2627 (Online) Journal homepage: https://www.tandfonline.com/loi/tsdw20

An AHP-ELECTRE framework to evaluate barriers

to green supply chain management in the leather
industry

S. Uddin, S. M. Ali, G. Kabir, S. A. Suhi, R. Enayet & T. Haque

To cite this article: S. Uddin, S. M. Ali, G. Kabir, S. A. Suhi, R. Enayet & T. Haque (2019):
An AHP-ELECTRE framework to evaluate barriers to green supply chain management in the
leather industry, International Journal of Sustainable Development & World Ecology, DOI:
10.1080/13504509.2019.1661044

To link to this article: https://doi.org/10.1080/13504509.2019.1661044

Published online: 05 Sep 2019.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tsdw20
INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY
https://doi.org/10.1080/13504509.2019.1661044

An AHP-ELECTRE framework to evaluate barriers to green supply chain

management in the leather industry
a
S. Uddin , S. M. Alib, G. Kabir c
, S. A. Suhid, R. Enayetd and T. Haqued
a
Institute of Leather Engineering and Technology, University of Dhaka, Dhaka, Bangladesh; bDepartment of Industrial and Production
Engineering, Bangladesh University of Engineering and Technology (BUET), Dhaka, Bangladesh; cIndustrial Systems Engineering, Faculty
of Engineering and Applied Science, University of Regina, Regina, SK, Canada; dDepartment of Industrial & Production Engineering,
Bangladesh University of Engineering & Technology, Dhaka, Bangladesh

ABSTRACT ARTICLE HISTORY

The leather-processing industry (LPI) is constantly polluting the environment in Bangladesh. As Received 27 June 2019
a result, stakeholders are continuously pressurizing managers working in LPI to embrace green Accepted 24 August 2019
leather-processing activities. Thus, the green concept is attracting significant attention from KEYWORDS
managers in the Bangladeshi LPI. However, the industry is struggling with many barriers to Green supply chain
implementing green supply chain management (GSCM). There are many studies regarding management; analytic
barriers to GSCM. However, those studies failed to show the possible pathways to implement hierarchy process; ELECTRE-I;
GSCM. This study addresses the gap by evaluating barriers to GSCM considering effective path- barrier analysis; leather
ways to GSCM. In this study, the Analytical Hierarchy Process (AHP) is integrated with Elimination industry; sensitivity analysis
Et Choix Traduisant La Realite (ELECTRE-I) method to identify and prioritize the barriers and to
rank the possible pathways to implementing GSCM in the leather industry. To accredit the
proposed framework, it is implemented on a leather-processing factory in Bangladesh.
A sensitivity analysis is performed to inspect the strength of the outcome of this method. The
outcome of this study indicates that the high cost of advanced technology is the most important
barrier to implement GSCM while green technology and techniques are the most effective
pathways to GSCM. The findings of this research will support researchers and practitioners by
giving insights on barriers and possible pathways to implementing GSCM.

1. Introduction produced in the tanning process. With a high level of

biological oxygen demand and chemical oxygen
The leather-processing industry (LPI) was one of the ear-
demand, the generated wastewater is discharged into
liest industries in Bangladesh. This sector has potential
the sewage system without any treatment (Paul et al.
due to its remarkable contribution to the creation of
2013). This leads to the extensive breakdown of the
national wealth as well as its worldwide status. Because
ecosystem of the river. It can also be noted that when
of the natural grain surface, noble texture, identical fibre
chromium-III is oxidized to chromium-VI, it turns into
composition, and fine touch, Bangladeshi leather has
a carcinogen (Chowdhury et al. 2013). There are about
achieved a strong position all over the world (Paul et al.
200 tanneries in Bangladesh. Paul et al. (2013) presented
2013). According to the Bangladesh export promotion
that tanneries discharge about 20,000 m3of effluents
bureau (2017), the involvement of the leather and leather
and 232 tons of solid waste per day. This huge amount
product sector in the nation’s economy was worth US
of solid waste (sludge) is dumped into the residential
$434 million in 2018–19 fiscal-years. Furthermore, many
area, which causes serious air pollution. It also causes
industrially advanced countries have shown their interest
a threat to human health. Besides, incineration of
in investing in the Bangladeshi leather sector. This is due
leather trimmings in an open place causes severe envir-
to the low labour cost compared with developed coun-
onmental pollution. The living conditions of this area are
tries. The process of leather tanning consists of various
below a satisfactory level (Azom et al. 2012). It is worth
steps that require a huge quantity of pure water and
pointing out that all tanneries in Hazaribagh, Dhaka
generates bulk amounts of solid and liquid wastes.
have already been transferred to a leather park at
In most of the tanning process, chromium salt is used
Savar. However, the leather industries are still suffering
as a tanning agent because more than 90% of the
due to the lack of advanced technology and inappropri-
leather produced in Bangladesh is chrome tanned
ate infrastructure. This points out a challenge to the
(Ahmed and Chowdhury 2016). Many organic and inor-
upcoming sustainability and cleaner production of the
ganic compounds, such as ammonia, chloride, chro-
leather sector with an increasing number of barriers,
mium, sulfates, dissolved and suspended solids, and
including environmental legislation and eco-criteria
other heavy metals, characterize the wastewater
derived from the exclusive export market.

CONTACT G. Kabir golam.kabir@uregina.ca

© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 S. UDDIN ET AL.
Experts show that if environmental requirements
are followed strongly, exports are expected to be
increased by $5 billion (Luthra et al. (2016); Parmar
(2016)). GSCM enables such a firm to become more
viable in the worldwide marketplace (Rao and Holt
2005). The literature provides masses of studies and
proof that illuminate the benefits of environmental
compliance for business (Hsu and Hu 2008; Jabbour
and Puppim-de-Oliveira 2012; Sarkis 2003; Revell and
Rutherfoord 2003). Common supply chain manage-
ment has turned into GSCM by furnishing ecological
necessities (Mathiyazhagan et al. 2014; Gandhi et al.
2015). GSCM is defined by the inclusion of green
elements, i.e., green supply and procurement, green
plan, green process, green promotion and distribu-
tion, reverse logistics, etc., in every functional stage
of a supply chain (Min and Kim 2012; Mohanty and
Prakash 2013). GSCM has contributed to the financial
escalation and progression of companies while pro-
tecting the environment (Vanpoucke 2014; Toke and
Kalpande 2018).GSCM is participating to support the
continuous sustainable improvement of any organiza-
tion (Somsuk and Laosirihongthong 2017). GSCM is
also playing a significant role in maintaining the total
information analysis report of any organization
engaged in supply chain acts (Chin et al. 2015).
While implementing GSCM in this field, a lot of bar-
riers have come to the management system and the
management found areas to implement alternatives
for improvement (Wang and Chan 2013; Govindan
et al. 2014).
The study aims at analyzing such barriers and find-
ing possible pathways. Therefore, this study helps the
firm to identify the major barriers to implementing
GSCM through the hybrid Multi-Criteria Decision
Analysis (MCDA) method, namely AHP and ELECTRE-
I. The AHP (Saaty 1980) and ELECTRE-I (Roy 1968) are
well-known multi-criteria decision-making methods
and extensively used in different areas. AHP has
become an important technique in MCDA. It works
on the competence of humans to gain applicable
findings of problems. The adhesion of AHP merges
multi-dimensional scales of measurement into a one-
dimensional scale of precedence. AHP is very reliable
because the pairwise comparison makes the proce-
dure impervious to comparison inaccuracy. One of the
most important facilities is that the values are allo-
cated with respect to know-how, perception, and
valid data in the pairwise comparisons (Rodrigues
et al. 2015). In addition, the AHP integrates a vital
method for inspecting the consistency of evaluations
of the stakeholder’s judgments, thus lessening the
bias in decision-making (Samaras et al. 2014). Above
all, the transparent construction process of AHP
makes the technique easily understood by academics
and practitioners. For the above reasons, the AHP
method has been chosen for this study. However,
AHP is often criticized for its inability to process
ambiguous variables. Kabir and Hasin (2011) indicated
that, due to the utilization of a scale with one to nine
discrete values, AHP cannot determine the uncer-
tainty in calculating the weights of various
characteristics.
On the other hand, ELECTRE-I has been selected for
this research because it is applied in multi-criteria
decision-making (MCDM) problems involving alterna-
tives. It is also able to reflect the stakeholder’s choice
and is widely used as an outranking technique in
various sectors. This method also represents the prior-
ity of comparisons with criteria through outranking
comparisons to lessen the bias of decision-makers
(Pang et al. 2011). The ELECTRE-I method can be
applied in problems where the alternatives can be
shown in unusual scales. The most important facility
of ELECTRE is the capability to handle data in
a diversity of measurement scales, both quantitative
and qualitative. The strength of ELECTRE lies both in
its adaptability and its mathematical soundness
(Samaras et al. 2014).
Therefore, the strengths of the AHP and ELECTRE-
I methods can be combined to develop an effective
decision support tool to evaluate barriers of GSCM
by establishing relationships with respective dimen-
sions and to link possible pathways with respective
barriers by decision-making matrix. The developed
research framework will motivate practitioners to
involve a new field of research. It will also assist
the decision-makers and stakeholders to identify
new barriers related to GSCM implementation. The
developed hybrid AHP-ELECTRE-I framework will
also guide the decision-makers to choose effective
pathways for barriers. The outcome of this study
will also motivate the policymaker to patronize the
leather sector to attract foreign direct invest-
ment (FDI).
This research focuses on the following research
questions:
(1) What are the barriers and possible pathways to
the implementation of GSCM in LPI in the con-
text of Bangladesh?
(2) How can industrial managers prioritize impor-
tant barriers and rank the most effective
pathways?
(3) Do the outcomes become helpful for industrial
managers to formulate strategies to implement
GSCM?
To address the above research questions, this
research has the following objectives:
(a) To identify barriers and to select possible path-
ways for the implementation of GSCM in the
LPI of Bangladesh.
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY
(b) To prioritize the barriers and to identify the
most critical and effective pathways using
a hybrid AHP-ELECTRE-I method.
(c) To recommend theoretical and managerial
implications for the implementation of GSCM
in LPI.
The rest of the article is arranged as follows: Section 2
shows a survey of associated writing. Section 3 describes
the development of hybrid methodology of AHP and
ELECTRE-I. Section 4 depicts an explicatory use of the
hybrid methodology in a company. The following sec-
tion discussed the managerial implication and theoreti-
cal implication. Section 6 concludes the article by
highlighting the limitations of the study and future
research direction.
2. Literature review
This segment shows the idea of GSCM and barriers
and possible pathways associated with its implemen-
tation in Bangladeshi leather industry.
2.1. Green supply chain management (GSCM)
Chowdhury et al. (2016) addressed the idea of GSCM
by using the hypothesis of reprocessing, recycling, and
reuse. Srivastava (2007) described the GSCM as a tool
for ensuring the sustainability of an environment by
reducing the amount of waste and carbon emission.
Srivastava (2007) also delineated the concept that
GSCM helps improve the attitude on research in var-
ious industrial sectors. Perron (2005) proposed that the
acknowledgement of availing of the ecological oppor-
tunities and achievement of industries is indispensable
for circulating those opportunities in small and med-
ium enterprises (SMEs) and large enterprises. Shipeng
and Linna (2011) noticed that in all sorts of business
enterprises, GSCM considers the ecological problems in
supply chain management. Wu et al. (2011) found
GSCM to be a helping hand to support the organiza-
tion for gaining turnover and market share through the
reduction of ecological hazards and influences while
improving their productivity.
Chowdhury et al. (2016) outlined that the theory of
GSCM contributes not only in the processing area but
also in technology improvement. Currently, it is
applied in other sectors, e.g., administration of the
state, academic institutions, and many service sectors.
Hoque and Clarke (2013) proposed that the adoption
of GSCM could improve the efficiency and security of
the inherent environment. GSCM has remarkable pos-
sibilities to help in monetary and ecological sustain-
ability as well. Ali and Bentley (2015) suggested that
exercising GSCM can facilitate the improvement of
productivity, employee morale, reputation of trade-
mark, and effectiveness, and thus enhance the
3
efficacy of an organization. Abdel-Baset et al. (2019)
concluded that GSCM practices can assist to reduce
waste, to decline cost and it can be used to recom-
mend economic advantages and to ensure better
utilization of resources. Seman et al. (2019) mentioned
that GSCM has a remarkable effect to accelerate the
green innovation of enterprises and it can affect the
establishments of the manufacturing process, which
eventually improve the environmental performance.
Tumpa et al. (2019) outlined GSCM as the path to
decrease the adverse environmental effects on an
emerging economy of industries worldwide.
2.2. GSCM implementation: barriers
This subsection presents the literature related to bar-
riers in implementing GSCM. After the review of exist-
ing literature, it is obvious that there are many studies
on the proposed analysis specifically on the field of
GSCM. It has also been found that only a few studies
discuss GSCM barriers. Table 1 shows the studies that
facilitate analysis of the barriers to implementing
GSCM.
A brief discussion about the major barriers identi-
fied from literature and experts’ opinions is as follows:
High cost of advanced technology
At the early stage, every successful GSCM faces the
barrier of accepting advanced technology due to high
cost. This issue creates challenges for the adoption of
useful advanced technology instead of older technol-
ogy of the organization (Balaji et al. 2014).
Apprehension of failing
Firms often feel fear of failing in adopting GSCM
because the firm thinks about suffering from mone-
tary losses or product failure as well as lead to loss of
competitive advantage (Perron 2005; Rao and Holt
2005; Govindan et al. 2014).
Design complexity to recycle the product
According to Mathiyazhagan et al. (2013), design
complexity to recycle the product plays an important
role to implement GSCM. This barrier often slows
down the manufacturing process. Use of environmen-
tal technology reduces the complexity associated with
product recycling.
Lack of knowledge and experience
Balaji et al. (2014) addressed the lack of knowledge
and experience for implementing GSCM among sup-
ply chain stakeholders as a barrier to executing the
GSCM. This barrier also leads to a feeling of ‘too
complex’ to implement GSCM among stakeholders.
Lack of recognition about reverse adoption of
logistics
Due to the lack of awareness about reverse logistics
adoption, the implementation of GSCM becomes challen-
ging for the firm (Mathiyazhagan et al. 2013; Govindan
et al. 2014). If organizations aware of reverse logistics it
results in an expenditure reducing the area that increases
4 S. UDDIN ET AL.
Table 1. Summary of barriers to GSCM implementation.
Category of Barriers Barriers to implementing GSCM
Technology i) Reluctance to adopt advanced technology
ii) Low level of technological know-how
iii) Apprehension of failing
iv) Absence of expertise in technology
v) Design complexity to recycle the product
vi) Insufficient internal technological resources expertise
on environment
Financial i) Unavailable bank loans for motivation of green product
ii) High cost for hazardous waste disposal
iii) Cost implication
iv) Financial constraints
v) High cost for using environmental packaging
vi) High investment and fewer return-on-investments
Knowledge i) Lack of knowledge and experience
ii) Disbelief about environmental benefits
iii) Absence of experts revealed to green method
iv) Conception of ‘out-of-responsibility’ zone
v) Lack of recognition about reverse adoption of logistics Woofi & Zailani (2010); Parmar (2016); Balon et al. (2016);
vi) Lack of environmental knowledge
Government i) Lack of government support to adopt GSCM
support and
policies ii) Lack of government regulation and legislation
iii) Inhibitory business policy towards product or process Govindan et al. (2014); Mathiyazhagan et al. (2013); Stokes &
maintenance
iv) Absence of motivational agenda or incentives
v) Lack of rewards and encouragement programs
vi) Absence of security and moral exercises
Outsourcing i) Difficulty in evaluating, supervising, and controlling
green practices among suppliers
ii) Problems in maintaining environmental suppliers
iii) Lack of sustainability certification (ISO 14001)
iv) Insufficient technical institutions to instruct, supervise Mathiyazhagan et al. (2013); Govindan et al. (2014)
and counselor the improvement particular to each
industry
v) Lack of corporate social responsibility
vi) Eco-literacy among supply chain partners
the profit margin of the firm (Subramanian et al. 2014;
Wang et al. 2015; Balon et al. 2016).
Absence of experts revealed to green method
The lack of skilled professionals revealed to the green
system becomes a significant barrier to implement
GSCM. To chase the management of the green system,
quality of professionals must be remarkable
(Mathiyazhagan et al. 2013; Govindan et al. 2014). To
eradicate the barrier, the skill training program may be
organized.
Lack of government regulation and legislation
Regulation and legislation of both local and central
governments are said to be a critical layout to set up
GSCM for firms (Luthra et al. 2016). Strict governmental
law will force the organization to implement GSCM.
Lack of rewards and encouragement programs
To motivate the entrepreneurs for implementing
GSCM, the government should provide a reward or
award for best green practices. Various GSCM related
programs, seminar, and symposium may encourage
the entities to adopt GSCM (Luthra et al. 2016).
References
Parmar (2016); Balon et al. (2016); Balaji et al. (2014)
Parmar (2016)
Govindan et al. (2014); Rao and Holt (2005); Perron (2005)
Stokes & Rutherfoord (2000); Van Hemel and Cramer (2002);
Mathiyazhagan et al. (2013)
Mathiyazhagan et al. (2013)
Zhu and Geng (2013)
Govindan et al. (2014); Parmar (2016)
Parmar (2016); Govindan et al. (2014); Zhu and Geng (2013)
Parmar (2016)
Balon et al. (2016); Govindan et al. (2014); Mathiyazhagan et al.
(2013)
Zhu and Geng (2013); Wang et al. (2015)
Govindan et al. (2014); Mathiyazhagan et al. (2013)
Balaji et al. (2014)
Govindan et al. (2014); Mathiyazhagan et al. (2013); Stokes &
Rutherfoord (2000); Wang et al. (2015)
Mathiyazhagan et al. (2013); Govindan et al. (2014); Wang et al.
(2015)
Govindan et al. (2014); Mathiyazhagan et al. (2013)
Govindan et al. (2014); Wang et al. (2015); Mathiyazhagan
et al. (2013)
Govindan et al. (2014); Mathiyazhagan et al. (2013)
Parmar (2016); Govindan et al. (2014); Wang et al. (2015);
Mathiyazhagan et al. (2013)
Balaji et al. (2014); Luthra et al. (2016); Balon et al. (2016); Singh
and Sarkar (2019)
Rutherfoord (2000)
Luthra et al. (2016); Govindan et al. (2014); Mathiyazhagan et al.
(2013); Sarkis (2011)
Luthra et al. (2016)
Luthra et al. (2016)
Govindan et al. (2014); Wang et al. (2015)
Wang et al. (2015); Govindan et al. (2014)
Parmar (2016)
Parmar (2016); Govindan et al. (2014); Mathiyazhagan et al.
(2013); Balon et al. (2016)
Balon et al. (2016)
Inhibitory business policy towards product or
process maintenance
Restrictions of the company policies add less
importance to product or process stewardship. It
also reduces the management’s interest to implement
GSCM and leads to harmful effects on business per-
formance (Mathiyazhagan et al. 2013; Stokes &
Rutherfoord 2000).
High cost for hazardous waste disposal
There are various threats engaged during the disposal
of hazardous waste. These threats lead to the high cost to
implement GSCM (Zhu and Geng 2013; Parmar 2016). In
addition, carrying out the inventory for hazardous mate-
rials leads to the high possibility of monetary losses.
High cost for using environmental packaging
The high cost of environmental packaging demoti-
vates the entrepreneurs to adopt GSCM in their plant.
The government should provide incentives to buy the
environment-friendly packaging. Therefore, the entre-
preneurs feel the necessity of GSCM adoption (Zhu
and Geng 2013; Wang et al. 2015).
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY 5

Financial constraints should be according to the different chemical process

Financial investment plays a critical role in the (Paul et al. 2013).
implementation of GSCM. In the sense of the indus- Green technology and techniques
trial sector, to take an effective decision for the invest- It is one of the most important possible pathways
ment is very crucial in the commencement and which comprise several attributes like mitigation, low
approval of GSCM practices (Govindan et al. 2014; risk and recyclability materials, energy consumption,
Balon et al. 2016). If the primary investment becomes green practices or procedures, and lifecycle. For suc-
high, then the return on investment is high following cessful implementation of GSCM, the adaptation of
the implementation of GSCM. these attributes is vital due to the strong relationships
among themselves (Lam et al. 2010).
Cleaner technology
2.3. Possible pathways to implement GSCM Cleaner technology refers to production equip-
ment, methods, product designs and supply system
There are some research articles that discuss the path-
which reduce the adverse impacts of the product on
ways to implementing GSCM. Researchers in various
the environment. This technology helps to generate
industrial applications and countries identify many
products with less harmful parts by efficient use of
pathways. Table 2 depicts these pathways.
natural resources. Long-lasting products are produced
A brief discussion about the possible pathways
with the help of this technology which will be
selected from literature and experts’ opinions is as follows:
recycled to the greatest possible extent. Therefore,
Waste management
the innovation and installation of cleaner technology
Waste management is a foremost pre-requisition to
accelerate the ecological performance and GSCM
implement GSCM. Waste management covers the
practices (Hosseini 2007).
management of both solid and liquid waste. Proper
Supplier management
management of waste accelerates business perfor-
Supplier management is defined as the manage-
mance (Udomleartprasert 2004). The wastewater pre-
ment of suppliers’ activities, capabilities, and educa-
treatment process is usually used in the LPI. Careful
tion regarding environmental production and
treatment of solid waste generated from tannery
innovation. The supplier must have knowledge of
plays an important role to keep balance in the eco-
ecology for facing adverse environmental situations.
system (Paul et al. 2013).
The consciousness of the GSCM implementation
Appropriate infrastructure
among suppliers plays a vital role in creating
Appropriate infrastructure setup plays a vital role
a competitive field (Hu et al. 2007).
to accelerate the GSCM process. Waste management
becomes easier due to the appropriate infrastructure
in the firms and thus improves the implementation 2.4. Research gap and research contributions
process of GSCM. In the tannery sector, infrastructure
The extant literature makes it evident that both aca-
demics and practitioners are paying attention toward
Table 2. Summary of possible pathways to GSCM
implementation. barrier analysis of GSCM adoption. GSCM studies have
Possible pathways to shown significant progression in developed countries,
Category implementing GSCM References compared with the developing ones. Despite increas-
Environmental i) Environmental Chien and Shih (2007); ing awareness and concerns for environmental sustain-
regulations Mudgal et al. (2009)
ii) Eco-design Zhu et al. (2010); Eltayeb ability in the world, the Bangladeshi LPI has shown
et al. (2011) very little development regarding GSCM practices.
iii) Reverse logistics Eltayeb et al. (2011)
iv) Pollution prevention Hosseini (2007) Considering GSCM as a desirable tool to improve the
v) Cleaner technology Hosseini (2007) supply chain and overall business operation (Srivastava
Operational i) Guide & benchmarking Lam et al. (2010)
systems
2007; Hoque and Clarke 2013; Chowdhury et al. 2016),
ii) Green technology and Lam et al. (2010) this study aims to provide a perspective for GSCM
techniques practices by reviewing the existing literature, which
iii) Reducing energy Diabat and Govindan
consumption (2011) can utilize the practices in the leather industry of
iv) Reliability and quality Lam et al. (2010) Bangladesh. There are some studies regarding the
specifications
v) Appropriate Paul et al. (2013) leather sector of Bangladesh. Moktadir et al. (2018d)
infrastructure modelled the interrelationships among different bar-
Management i) Waste management Udomleartprasert (2004)
ii) Internal environment Zhu et al. (2010) riers to sustainable supply chain management in the
management Bangladeshi LPI. Moktadir et al. (2018a) tried to identify
iii) Leadership and Lam et al. (2010)
responsibilities
barriers to big data analytics in leather manufacturing
iv) Certification to ISO: Mudgal et al. (2009) supply chain. Moktadir et al. (2018e) examine the dri-
14001 EMS vers to sustainable leather supply chain and circular
v) Supplier management Hu et. al. (2010)
economy. Moktadir et al. (2018b) prioritize the drivers
6 S. UDDIN ET AL.

of corporate social responsibility (CSR) in the context of
the footwear industry. Existing closely related studies
revealed important barriers and drivers regarding sus-
tainable supply chain management in the Bangladeshi
LPI, but were unable to present effective pathways to
eliminate barriers by interlinking barriers to possible
pathways to GSCM. This study addresses the gap in
research using a hybrid method named AHP and
ELECTRE-I.
3. Methodology
This paper presents a standard framework to identify
the most important barriers and select the most effec-
tive pathways to the implementation of GSCM in the
Bangladeshi leather industry. Figure1 illustrates this
structure.
The goal is to select the most important barriers
and most effective pathways to the implementation
of GSCM. It consists of four steps as described below.
Step 1: Identify the suitable dimensions, major
barriers, and pathways to implementing GSCM
In this step, major barriers with suitable dimen-
sions and possible pathways are identified from
the review of literature and opinion of experts. It
can be noted that 30most common barriers with
five dimensions and 15possible pathways were
identified from the existing literature.
Step 2: Determine the most important barriers by
applying AHP
The most important barrier has been determined
by major barriers, which were identified in Step
1. This step has been carried out by applying the
AHP method. The highest weighted barrier has
been considered as the most important barrier,
which can be easily identified from global
weights of specific barriers.
Step 3: Determine the comparative superiorities of
possible pathways concerning each barrier by
applying the AHP method
In this stage, the relative priorities of identified
possible pathways concerning each of the major
barriers have been determined by applying the
AHP technique. A decision-making matrix has
been constructed in this step.
Step 4: Determine the most effective pathways by
applying the ELECTRE-I method
In the final step, the most effective pathway has been
determined from a general matrix G. This step has been
carried out by applying the ELECTRE-I method.

Identify the most influential barriers and select the most effective
Goal pathways to implementing GSCM in Bangladeshi leather industry

Determine the suitable dimensions & major barriers and some

Step-1 possible pathways for analysis

Data processing

AHP

Determine the weights of dimensions as well as barriers for

Step-2 prioritizing barriers by AHP

Determine the most influential barriers

Establish the comparative superiority of possible pathways

Step-3
concerning each barrier by AHP

ELECTRE-I

Employed ELECTRE-I method to rank the possible pathways to

implementing GSCM
Step-4

Select the most effective pathways

Figure 1. Framework of the proposed method.

 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY
3.1. Determination of major barriers with
suitable dimensions and possible pathways
The major barriers to implementing GSCM in the
Bangladeshi leather industry have been identified from
the literature and professionals’ opinions. It is worth
pointing out that a specific firm may have distinctive
observations concerning barriers to implementing
GSCM. Therefore, barriers defined for any firm might
not be the same in others and may have discrete influ-
ences as well. Based on the existing literature and
experts’ input, 12 major barriers and five possible path-
ways to the implementation of GSCM have been deter-
mined (details of finalizing the barriers with suitable
dimensions and possible pathways are discussed in
the Data Collection section). The validated barriers
have been categorized into four dimensions (technolo-
gical, knowledge, government support and policies, and
financial) based on the experts’ opinion. Table 3 shows
the list of major barriers with suitable dimensions. Five
possible pathways considered in this study are waste
management (P1), appropriate infrastructure (P2), green
technology and techniques (P3), cleaner technology
(P4), and supplier management (P5). Figure 2 shows
the decision hierarchy of the analysis.
In addition, it is important to note that, to deter-
mine the most important barriers hindering the
implementation of GSCM and select the most effec-
tive pathways to implementing GSCM in the leather
industry, a standard questionnaire was designed that
consists of common barriers and 15possible pathways
collected from a literature review as per the
Bangladeshi perspective (see Appendix A, Table A1).
A survey was conducted by sending the questionnaire
by mail and personal contacts to different experts.
The survey was conducted from May to October in
2017 and was a Yes or No type question format.
In this stage, it is necessary to depict the applic-
ability of this study in different sectors of different
countries. There are various studies focusing on
Table 3. Major barriers to implement GSCM in Bangladeshi
leather-processing industry.
Dimensions Barriers
Technology i) High cost of advanced technology
ii) Apprehension of failing
iii) Design complexity to recycle the product
Knowledge i) Lack of knowledge and experience
ii) Lack of recognition about reverse adoption
of logistics
iii) Absence of experts revealed to green
method
Government i) Lack of government regulation and
support legislation
and ii) Lack of rewards and encouragement
policies programs
iii) Inhibitory business policy towards product
or process maintenance
Financial i) High cost for hazardous waste disposal
ii) High cost for using environmental
packaging
iii) Financial constraints
7
MCDM analysis on selecting the barriers to implement
GSCM. For instance, Govindan et al. (2014) employed
the AHP method to identify barriers to implementing
GSCM in various industrial sectors of India. Balaji et al.
(2014) applied the interpretive structural modelling
(ISM) technique to analyse barriers associated with
implementing GSCM in the foundry sector of India.
Zhu and Geng (2013) worked with drivers and barriers
of extended supply chain practices for energy saving
and emission reduction among Chinese manufac-
turers. Parmar (2016) used Fuzzy AHP and ISM to
analyze barriers to implementing GSCM in SMEs in
India. Moktadir et al. (2018a) explored barriers to
adopting big data analytics in supply chains using
AHP. Moktadir et al. (2018c) investigated supply
chain risks in the pharmaceutical industry using AHP.
All MCDM methods presented in these studies can
help the decision-makers to select the influential bar-
riers among multiple barriers to implementing GSCM.
However, there are still two issues to be enhanced,
one is the determination of the possible pathways for
GSCM, and the other is about linking and ranking the
possible pathways with respect to the barriers. This
study is initiating a standard framework for complet-
ing these two issues, which is usually an MDM pro-
blem with many barriers. This framework can help the
decision-makers/stakeholders of different sectors in
different countries to select the most important bar-
riers and most effective pathways.
3.2. A hybrid MCDA technique by the combination
of AHP and ELECTRE
Barrier analysis to implement GSCM, in which several
barriers remain under certain dimensions, is a problem
of MCDA. The establishment of the decision-making
matrix is a precondition to accomplish the analysis.
According to Hajkowicz and Higgins (2008), MCDA
has been commonly used to attain a fixed quantity of
alternatives that stand on an evaluation criteria set
that has a subsequent distinctiveness: (a) capability of
decision framework management; (b) ability to
Notation explain complicated criteria with the unit that does
T1 not have a common measure; and (c) aiding in the
T2 process of making decisions (Mendoza and Martins
T3
K1 2006). Furthermore, it helps in calculating various
K2 goals with contradictory standards or instinctive fea-
K3 tures (Teixeira de Almeida 2007). Broad application of
MCDA has occurred in many distinguishable chal-
G1
lenges such as management of water resources, stra-
G2 tegic technology development solutions of fuel cells
in the automotive industry, sustainability assessment
G3
of biogas production, assessment of technologies
F1 relating to concentrated thermal and solar energy,
F2
management of natural resources, selection and eva-
F3 luation of suppliers (An et al. 2017). All problems in
8 S. UDDIN ET AL.

Level 3 Level 4
Level 1 Level 2
Barriers Possible
Goal Dimensions
Pathways

High cost of advanced technology

(T1)

Technology (T) P1
Apprehension of failing (T2)

Identify the most influential barriers and select the most effective pathways
Design complexity to recycle the
product (T3)

Lack of knowledge and

experience(K1) P2
Knowledge (K) Lack of recognition about reverse
adoption of logistics (K2)

Absence of experts revealed to

green method (K3)

P3
Lack of government regulation
and legislation (G1)
Government
support and Lack of rewards and
policies (G) encouragement programs (G2)

Inhibitory business policy towards

product or process maintenance
(G3)
P4

High cost for hazardous waste

disposal (F1)

Financial (F) High cost for using environmental

packaging (F2)

Financial constraints (F3) P5

Figure 2. Decision hierarchy of the analysis.

the process of making decisions are parallel to barrier a precondition to make the right judgments. There
analysis of GSCM in the leather industry. are various ways to find out the weights of the bar-
Several MCDA methodologies have been published riers, i.e. the factual processes and the emotional
in the literature, such as ISM (Parmar 2016), AHP processes. The factual methods, i.e. TOPSIS and
(Luthra et al. 2016), ELECTRE (Pang et al. 2011; An entropy, can help the administrators find out the
et al. 2017), Technique for Order of Preference by weights in relation to the features of the data of
Similarity to Ideal Solution (TOPSIS). TOPSIS (Wang possible pathways concerning the barriers neutrally.
and Chan 2013). Among these, ELECTRE is one of Nevertheless, these equitable techniques are unable
the most commonly used methods for its various to indicate the likings and eagerness of the adminis-
advantages (Pang et al. 2011). It is a widely recog- trator or decision-maker. Consequently, the opinion-
nized multi-standard approach standing on the based weight determining techniques, i.e., AHP and
research of dominating relationships, that utilizes Delphi, which can indicate the views of the stake-
indexes of concordance and discordance to examine holders, are extensively used in establishing the
the outranking associations between the options weights of the factors or barriers in question. AHP is
(Teixeira de Almeida 2007). Family procedures of the most prominent emotional weight-determining
ELECTRE consist of various privileges, considering process that can assist the administrators to find out
the ELECTRE-I system as an instance, it can mirror the weights of the barriers by means of one to nine
the real qualities between each pair of options (An discrete scales and their reciprocals to set up the
et al. 2017). relationship matrix (Saaty 1980).
There are two most important branches in the The AHP method, developed by Saaty (1980), is
matrix of making decisions, i.e. (a) the weights of the used to find out both the weights of the barriers to
barriers to implementing GSCM and (b) the weights of implement GSCM and the comparative precedence
the possible pathways concerning each of the barriers of the possible pathways concerning each of the
to implementing GSCM. To determine the weights of barriers to implement GSCM. It is obviously tricky
the barriers, picking an appropriate method is for decision-makers to find out the precedence of
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY 9

the possible pathways concerning each barrier to 1

zji ¼ ; zij > 0; i; j ¼ 1; 2; ::::n (2)
GSCM implementation. There are two main causes: zij
(1) rigorous ambiguity troubles that exist in deci-
sion-making: information of the possible pathways
3.2.1.2. Calculation of weights. By means of the
concerning the barriers to implementing GSCM can-
relationship matrix, the coefficients of weights con-
not be achieved owing to different causes; (2) many
cerning each barrier are obtained through computing
flexible barriers cannot be illustrated in a straight
the principal eigenvector of the relationship matrix, as
line via quantitative techniques. After determining
exposed in Equation (3)
the decision-making matrix, the ELECTRE-I techni-
que has been applied to find out the precedence 2 3 w1 w1
1 z12 ::: z1n
of possible pathways to implement GSCM in 6 z21 1:::: z2n 7 w1 w1
6 7
Bangladeshi LPI. The following subsection defines 6 .. . . 7 j j ¼ λmax j j (3)
4 .    . . .. 5 .. ..
the AHP and ELECTRE-I method. . .
zn1 zn2 :::: 1
wn wn
3.2.1. AHP method to find out the weights of the where λmax is the maximum eigenvalue of matrix Z.
barriers Equation (4) can determine the maximal eigenvector
In this research, the proposed method, AHP, deter- of the relationship matrix.
mines the weights of barriers used in the ELECTRE
Wmax ; ¼ ½w1 ; w2 ;   ; wn  (4)
method. As declared above, the process is a decision
investigation technique that judges the quality-based By normalizing, the maximum eigenvector of the
as well as quantity-based data and mirrors the likings weights of the barrier can be obtained, as presented
of the decision-makers. The utilization of the AHP in Equation (5)
setting offered by Saaty in 1980 to evaluate the bar- w
1 w2 wn  T Xn Xn Xn
rier weightings in MCDA has recently become familiar W¼ ; ; :::; wi wi wi (5)
in many sections of process engineering (Cengiz and n n n i¼1 i¼1 i¼1
Akbulak 2009; Martin-Utrillas et al. 2015; An et al. where W is the weight coefficient vector, wi repre-
2017). sents the weight of the barriers, and n represents the
total number of the barriers.
3.2.1.1. Establishment of relationship matrix.
Assume there are n numbers of barriers in a number 3.2.1.3. Consistency check. If zik = zijzjk, j, k = 1, 2, · · ·,
of gradations, the pair-wise relationship technique n, then the relationship matrix Z can be documented
suggested in 1980 by Saaty (see Table 4) can be as a consistent matrix. Hypothesis confirms that if the
applied to compute the matrix of comparison (indi- n-dimensional relationship matrix is a consistent
cated by matrix Z), as displayed in Equation (1) matrix, its maximum eigenvalue must be equal to n.
2 3 However, it is hard to set up a relationship matrix,
1 z12 :::::: z1n
6 z21 1 which is consistent with matrix. In real applications,
6 z2n 7 7
Z ¼ 6 .. .. .7 (1) various relationship matrixes that satisfy the consis-
4 . ... . .. 5 tency check denote the consistent matrixes. The con-
zn1 zn2 ::::1 sistency ratio is the frequent way to evaluate whether
where zij indicates the comparative significance of a comparison matrix is consistent or not, as shown in
barrier i comparing with j. Equation (6)
Equation (2) can calculate the comparative signifi- Cl
cance of barrier j comparing to i. CR ¼ (6)
Rl
where CR denotes consistency ratio, CI denotes
Table 4. Relationship scale (Saaty 1980). consistency index; RI indicates average random index
Scales Definition Note with the same dimension with Z.
1 Equal importance i is equally important to j The value of RI can be found from Table 5 and the
3 Moderate importance i is moderately important to j
5 Essential importance i is essentially important to j value of CI can be calculated by Equation (7)
7 Very strong i is very strongly important to j
importance λmax  n
i is very absolutely important to j Cl ¼ (7)
9 Absolute importance n1
2, 4, 6, 8 Intermediate value The relative importance of i to j is
between two adjacent
judgment
Reciprocal Reciprocals of above The value had been assigned to Table 5. The value of the average random consistency index RI.
i when compared to j, then
j has the reciprocal value n 1 2 3 4 5 6 7 8 9
compared to i RI 0 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45
10 S. UDDIN ET AL.

where λmax symbolizes the maximum eigenvalue of from the sets of intervals of the concordance and dis-
the relationship matrix Z, n symbolizes the dimension cordance. The sets of attributes may get past this system.
of this matrix. Let A ={a, b, c . . . } signifies a finite set of alternatives.
When the CR of a relationship matrix is less than There are two different sets of concordance interval set
0.1, then the matrix is about to be satisfactory one as (Cab) and discordance interval set (Dab). The concordance
consistent matrix. Perversely, if CR is greater than or interval shows the dominancy if the requirement is met.
equal to 0.1, the matrix should be customized until  
Cab ¼ jxaj  xbj (10)
a satisfactory one.
On the other hand, we can establish the discor-
dance interval set (Dab) by using Equation (11)
3.2.2. ELECTRE method  
By means of outranking relationships, ELECTRE techni- Dab ¼ jxaj < xbj ¼ J  Cab (11)
que reveals the superiority of relationships among dif-
ferent alternatives. Therefore, it is feasible that, these
outranking relationships can differentiate among the Step 3: Calculation of the concordance interval matrix
alternatives. There are two types of indices used in the Based on the preferences of the decision-makers for alter-
ELECTRE method for pairwise comparison of alterna- natives, the concordance interval index (Cab) between Aa
tives. Concordance and discordance are the two and Ab can be obtained by using Equation (12).
indices (Pang et al. 2011). ELECTRE-I method was X
Cab ¼ wj (12)
applied by analyzing these two indices to select the j2Cab
possible pathways in this paper.
We assume that A1, A2, . . ., Am are m possible path- The concordance index shows the consent of the
ways for GSCM implementation in the leather indus- assertion ‘A outranks B.’
try, C1, C2, . . ., Cn are barriers, which can describe the Equation (13) shows the formulation of the concor-
properties of possible pathways. xij defines the level dance interval matrix.
of pathways Ai concerning barriers Cj. Again, Wn  
 cð1; 2Þ:::::::::::cð1; mÞ 

denotes the weight of the significance of Cn, which  cð2; 1Þ  :::::::::::::cð2; mÞ 

is obtained from the AHP method. Following para- C¼ . .. .. ..  (13)
 .. . . . 
graphs show the stepwise formulation of ELECTRE-I  
 cðm; 1Þ cðm; 2Þ:::  
method.
Step 1: Determination of Normalization matrix and
weighted matrix Step 4: Calculation of the discordance interval matrix
Taking into consideration theory on the interval To begin with, take the discordance index of d (a, b),
numbers of decision matrix, the normalization matrix which can be considered the priority of dissatisfac-
of Rij =[rij] is calculated by Equation (8) tion in the resolution of alternative a instead of
xij alternative b. Equation (14) defines this preference
rij ¼ sffiffiffiffiffiffiffiffiffiffiffi ; i ¼ 1; 2; :::::; n j ¼ 1; 2; :::; m (8) more particularly.
Pm
xij2  
i¼1 Maxvaj  vbj 
j2Dab
dða; bÞ ¼   (14)
Thus, equation (9) shows the weighted matrix, Max vmj  vnj 
j2J; m; n2I
which depends on normalization matrix, assigned to it.
Here m and n help to calculate the weighted nor-
Vij ¼ R  R  W
  malized value of all possible pathways. Therefore, we
 r11 :w1 r12 :w2 rln :wn 
  can establish the discordance interval matrix by using
 r21 :w1 r22 :w2 r2n :wn 
¼  .. .. .. .. 
 (9) discordance interval index sets.
 . . . .  Equation (15) shows this interval matrix.
 
rm1 :w1 rm2 :w2 rmn :wn  
 :::::::::dð1; 2Þ:::::::::::::dðl; mÞ 
where 0 ≤ w1, w2, . . ., wn ≤ 1. These constants express the  
 dð2; 1Þ:::::::::  ::::::::::::dð2; mÞ 
 
weights of the attributes. Besides, the correlation coeffi- D¼ . .. .. ..  (15)
 .. . . . 
cients of normalized interval lie between zero and one.  
 dðm; 1Þ dðm; 2Þ  
Step 2: Ascertainment of concordance and discor-
dance interval sets
Suppose selecting the effective possible pathways is Step 5: Determination of concordance index matrix
a multiple standard decision. By the use of information Equation (16) describes the concordance index for the
of preference, the decision-making system is obtained measurement of satisfaction difficulty.
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY 11

X
m X
m
4.1. Company profile
c ¼ cða; bÞ=mðm  1Þ (16)
a¼1 b The proposed method has been executed through
where c denotes the average significance of con- a practical and sensible question. The selected firm
cordance index. trades with the production of crust and finished
Equation (17) describes the Boolean matrix (E) of leather. The yearly production of this company is
the dominance index. 32million square metres. The profile of the case com-
 pany is shown in Table 6.
eða; bÞ ¼ 1 if cða; bÞ  c With the shifting of its facilities to the newly built
(17)
eða; bÞ ¼ 0 if cða; bÞ < c Government ‘Leather Economic Zone’ with a common
effluent treatment plant (CETP), the company has
created a benchmark in the tannery sector. The new
Step 6: Determination of discordance index matrix
production unit has been built with the most modern
In opposite, the index of discordance measures the
and compliant factors into consideration. The new
level of the dissatisfaction. The discordance index
unit will not only improve efficiency in production
follows Equation (18).
but also increase in-house production capacity by at
P
m P
m
least 20%. From a discussion with the firm’s manage-
dða; bÞ
 a¼1 b ment, it is obvious that they have an outline to imple-
d¼ (18)
mðm  1Þ ment GSCM for improving their business
performance. There are many reasons to show their
According to the basis of the discordance index stated
positive sense to implement GSCM in their plant. One
before, the matrix of discordance index (F) can be
of the most crucial reasons is the buyers’ requirement.
determined. Equation (19) shows the matrix (F).
 Buyers always push them to maintain ecological bal-
f ða; bÞ ¼ if dða; bÞ  d  ance and to reduce environmental pollution. They
 (19)
f ða; bÞ ¼ 0 if dða; bÞ > d also force the management to have a smooth produc-
tion process and enhanced product quality.
Nowadays, CSR is the topmost demand from buyers.
Step 7: Construction of the general matrix
Merchant wants to engage the production house with
The general matrix G is measured by multiplying the peer-
CSR on a regular basis. Besides these reasons, the
to-peer elements of matrix E and matrix F (Hatami-Marbini
government’s monitoring pressurizes them to attain
et al. 2013). The general matrix follows Equation (20).
the criteria of sustainable development goals 2030
G ¼ EF (20) proposed by the United Nations General Assembly.
Above all, if the management agrees to implement
GSCM, they get more orders from buyers and can
contribute to the nation’s economy at large by gain-
4. Case study
ing more profits. Therefore, the objectives of this
The implementation of GSCM is challenging due to the research are in synergy with the vision of the selected
contradictory nature of the objectives. Organizations company. For this, a comprehensive process has been
want extensive profits to be increased and negative executed to determine the comparative significance
environmental impacts to be diminished. For the time of identified barriers and possible pathways asso-
being, they disagree on augmenting the expenditure ciated with GSCM implementation.
to a large value. However, GSCM implementation can
increase trade contribution. Here, a real-life case exam-
ple to a leather-processing factory (named ABC) is 4.2. Data collection
presented to demonstrate how the proposed hybrid
technique that combines AHP and ELECTRE helps/aids For the data collection process, a group of five experts
decision-making. was formed. Two of the experts are from the relevant

Table 6. Profile of the case company.

Years of experience & role Company size
Name of Company Respondent of area Types of products (Fiscal year 2016)
ABC Leather Production (i) 41 years i) Cow Area: 3 acres,
processing manager (ii) Procurement, -full chrome Number of
company Manufacturing, -semi chrome Employees:788,
distribution - vegetable tanned leather Annual sales turnover:
ii) Goat USD $40 million
-full chrome
-semi chrome leather and various types of finished
leathers for shoes and leather goods
12 S. UDDIN ET AL.

Table 7. Profile of experts. meeting. The experts also considered different litera-
Years of ture for the barriers and possible pathways concern-
Experts Position experience Academic qualification
ing each barrier for implementing GSCM. The
X1 Associate professor 10 years Doctor of Philosophy
(Ph.D.), majoring in objectives along with a brief methodology have also
supply chain been presented to the expert panel and they were
management
X2 Executive director 25 years Master of Business requested to fill pairwise matrices. A synchronizer has
Administration been allotted to direct the meeting for shaping every
(MBA)
X3 Production manager 17 years MBA
one of the comparison matrices. It is important to
X4 Leather engineer 12 years B.Sc. in Leather note that frequently there were various arguments
Engineering in the conversation, and the synchronizer had to syn-
X5 Leather engineer 10 years B.Sc. in Leather
Engineering chronize the conversation to arrive at a harmony for
each relationship matrix.

industry, two are from the case company, and one is

from academia. Table 7 shows the profiles of the
experts. The required data were gathered from the
industry professionals and academic expert. Here, this
research recruited only five experts in the data collec-
tion process. Sadly, there is no consensus in the lit-
erature on how many experts are required to model
an MCDM problem. For example, Zubayer et al. (2019)
took the opinion of five experts to model a fuzzy-
TOPSIS problem. Moktadir et al. (2018d) considered
four experts to model a grey-DEMATEL problem. Due
to time constraints and for simplicity, this research
considered five experts in the data collection process.
The process of data collection is performed in two
stages as described below.
Stage-1: Finalize the major barriers with suitable
dimensions and possible pathways to the implemen-
tation of GSCM
At first, 30 common barriers with many dimensions
and 15 possible pathways have been identified
through a survey of existing literature that investi-
gates the relevant sector. After that, 12barriers and
five possible pathways were identified for this study. It
can be noted here that ‘reluctance to adopt advanced
technology’ barrier has been substituted by ‘high cost
of advanced technology’ following experts’ opinions.
Experts opined that managers are reluctant to adopt
advanced technology as the cost is high. Further, the
12barriers have been categorized into four dimen-
sions by taking experts’ opinions.
Stage-2: Evaluation of comparison among dimen-
sions, identified barriers, and possible pathways
This phase has been carried out by taking experts’
input for developing a hybrid methodology named
AHP and ELECTRE-I. To support the analysis and to
determine the decision-making matrix, the experts
have been requested to contribute to a focus group
4.3. Data analysis and results
The collected data were analyzed by inputting the
data into the AHP and ELECTRE-I formulations. All
required matrices have been obtained to determine
the most important barriers and most effective path-
ways. First, the AHP technique has been applied to
determine the most important barriers. According to
the AHP technique, the computation of the weights of
the four dimensions (technology, knowledge, govern-
ment support and policies, and financial dimensions)
is considered. For instance, the specialists apprehend
the view that the comparative significance of technol-
ogy issue evaluated with knowledge issue is of mod-
erate importance (scale 3). Therefore, the number ‘3’
has been placed in the cell (1, 2) of the relationship
matrix to compare the comparative significance of
these four dimensions. Similarly, all the other inputs
in this matrix can also be found, as shown in Table 8.
With the help of the AHP technique, it can be shown
from Table 8 that the maximal eigenvalue of the relation-
ship matrix, (λmax) is 4.186 and the values of the consis-
tency index (CI) and ratio (CR) are 0.062 and 0.069,
respectively. It is noticeable that the computed value of
CR is less than 0.1, which implies that for consistency
checking the relationship matrix is acceptable. In the
same way, the local weights of the barriers in each
dimension can be established. After that, for determining
the global weight of each barrier to implement GSCM,
the local weight of every barrier is multiplied by the
parallel dimension’s weight with which it is associated.
The global weight of each barrier is shown in Table 9.
Table 9 shows that the ranking of the barrier dimension in
the downward array is Technology > Finance >
Knowledge > Government support and policies.
Technology has been accepted as the most important
barrier dimension in which specific barrier ‘reluctance to

Table 8. Relationship matrix to obtain the weights of four dimensions.

Technology Knowledge Government support and policies Financial Weights
Technology 1 3 2 1 0.362
Knowledge 1/3 1 2 1 0.209
Government support and policies 1/2 1/2 1 1/3 0.124
Financial 1 1 3 1 0.305
λmax = 4.186, CI = 0.062, CR = 0.069 < 0.1
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY 13

Table 9. Global weights of the barriers to implementing GSCM.

Dimensions Weights of Dimensions Barriers Weights of Barriers Global Weights Overall Priority/sequence
Technology 0.3620 T1 0.6483 0.2347 1
T2 0.2297 0.0831 6
T3 0.1220 0.0442 8
Knowledge 0.2090 K1 0.6491 0.1356 3
K2 0.2790 0.0583 7
K3 0.0719 0.0150 11
Government support and policies 0.1239 G1 0.7514 0.0931 5
G2 0.1782 0.0221 10
G3 0.0704 0.0087 12
Financial 0.3051 F1 0.5917 0.1806 2
F2 0.3332 0.1017 4
F3 0.0751 0.0229 9

adopt advanced technology’ is prioritized as first. The Table 11. Determine the concordance and discordance index
matrices (Boolean matrices).
final result is summarised as follows: T1 > F1> K1> F2 >
Pathways P1 P2 P3 P4 P5
G1> T2> K2 > T3> F3 > G2 > K3 > G3. P1 0 1 1 1 1
The comparative superiorities of five possible path- P2 1 0 1 1 1
ways concerning each barrier have been computed by P3 1 1 0 0 0
P4 0 0 0 0 0
the same method of regulating the local weight of P5 0 0 0 0 0
each barrier with the help of the AHP technique (see C Bar = 0.194788
Appendix, Tables A2–A13). It can also be noted that P1 0 1 0 1 1
P2 0 0 0 1 1
preferences of experts are needed to run this compar- P3 1 1 0 1 1
ison. It is important to say that all matrices have been P4 0 0 0 0 1
P5 0 0 0 0 0
normalized. Both the global weight of each barrier to
D Bar = 0.616027
implementing GSCM and the comparative superiori-
ties of five possible pathways concerning each barrier
Table 12. General matrix.
are arranged in a matrix. This matrix implies the deci-
Pathways P1 P2 P3 P4 P5
sion-making matrix. Table 10 presents the matrix.
P1 0 1 0 1 1
To obtain the rank of the five possible pathways to P2 0 0 0 1 1
implement GSCM, ELECTRE-I is applied. At first, the P3 1 1 0 0 0
P4 0 0 0 0 0
weighted normalized matrix has been acquired by find- P5 0 0 0 0 0
ing the product of the global weight of each barrier and
the corresponding comparative priorities of five possi-
ble pathways (see Appendix, Table A14). The concor- With the general matrix shown in Table 12, the
dance and discordance index matrices of the five superiority order of the five possible pathways is
possible pathways have been computed by means of obtained. From Table 12 with the formulation of
concordance interval index and discordance interval ELECTRE, it is shown that P1 is preferred; then, P2,
index (see Appendix A, Tables A15 and A16). The con- P4, P5 (i.e. P1> P2, P4, P5) and P2 is preferred toP4, P5
cordance and discordance interval indexes have been (i.e. P2> P4, P5). Again, it is shown that P3 is preferred
obtained from concordance and discordance interval to P1, P2 (i.e. P3> P1, P2). It is clear from the above
sets (see Appendix A, Tables A15 and A17). Table 11 inferences that P3 is the most preferable pathways,
shows the concordance and discordance index matrices followed by P1, P2. P4 and P5 have the same prefer-
(Boolean matrices). ences. Therefore, the final order is P3> P1 > P2 > P4,
To establish the general matrix (E), we then aggre- P5. From this order, it is found that green technology
gated these two Boolean matrices by multiplying and techniques (P3) have been recognized as the
their peer-to-peer elements. The constructed general most effective possible pathway to implement
matrix is shown in Table 12. From the matrix E shown GSCM. The waste management (P1), appropriate infra-
in Table 12, we get the final sequence of the five structure (P2), cleaner technology (P4), and supplier
possible pathways. management (P5) follow next in the ranking.

Table 10. Weights of the possible pathways with respect to the barriers.
Barriers T1 T2 T3 K1 K2 K3 G1 G2 G3 F1 F2 F3
Weights 0.2347 0.0831 0.0442 0.1356 0.0583 0.0150 0.0931 0.0221 0.0087 0.1806 0.1017 0.0229
P1 0.2067 0.3991 0.3231 0.3911 0.3780 0.1856 0.3782 0.2249 0.3539 0.3362 0.3273 0.2197
P2 0.1530 0.3817 0.1636 0.1085 0.2983 0.0794 0.1608 0.4643 0.3081 0.3550 0.1632 0.4000
P3 0.4322 0.0869 0.3549 0.3171 0.1941 0.5609 0.3036 0.1572 0.1875 0.1781 0.3625 0.1535
P4 0.1444 0.0650 0.0829 0.1205 0.0404 0.1345 0.0978 0.0843 0.0752 0.0734 0.0764 0.1278
P5 0.0637 0.0673 0.0754 0.0627 0.0891 0.0396 0.0595 0.0693 0.0752 0.0574 0.0705 0.0989
14 S. UDDIN ET AL.

4.4 Comparison with existing literature
It is very important to evaluate the barriers to implement
GSCM in the leather industry in order to sustain in the
global market. Barriers to GSCM may vary from country to
country. The AHP-ELECTRE method presented the most
important barriers related to Bangladeshi LPI. The high
cost of advanced technology (T1) has been identified as
the most important barrier. Due to the high cost of
advanced technology, owner of tanneries often fails to
incorporate required technology to implement GSCM in
their plant. Balaji et al. (2014) represented this barrier as
one of the most important barriers to implement GSCM
in Indian foundry industry through ISM method. Lack of
advanced technology creates more difficulties during the
disposal of hazardous waste. These difficulties increase
the cost of the disposal process (Mudgal et al. 2010). In
this study, high cost for hazardous waste disposal (F1)
barrier ranks second. Govindan et al. (2014) mentioned
that high cost to dispose of hazardous materials has
a remarkable effect on the improvement of environmen-
tal technology. Due to the absence of knowledge about
environmental technology and the lacking of experience,
it will become difficult to implement GSCM in LPI. Lack of
knowledge and experience (K1) barrier ranks the third
position in this research. Balaji et al. (2014) cited this
barrier as the top-ranked barrier in the context of foundry
domain. The outcome of this research showed that inhi-
bitory business policy towards product or process main-
tenance (G3) barrier occupies the least importance to
implement GSCM in LPI. Product maintenance refers to
the management standard to handle the products safely
throughout their life cycle (Mudgal et al. 2010). Govindan
et al. (2014) asserted this barrier as the least valued one.
Presently most of the Bangladeshi LPI is able to handle
the products safely, but more attention is needed to
ensure 100% progression.
4.5. Sensitivity analysis
A sensitivity analysis is usually done to observe the
strength of the outcomes (Kaya and Kahraman 2011;
Agrebi et al. 2017; An et al. 2017). In fact, the investigation
deals with one query, i.e., ‘How responsive is the final
result to tiny alterations in each weight obtained through
the pairwise evaluation method?’ (Agrebi et al. 2017).This
question can be answered by carrying out a sensitivity
analysis. To validate the proposed method, this investiga-
tion was carried out (Govindan et al. 2014). The barrier
with the highest weight should be identified first. In this
research, this process is done by putting different values
on the barriers. Thirteen events have been studied, which
are shown in Table 13. The definition of an event is given
below:
(1) Identical significance: the same weight of
0.2347 is considered in every barrier (Event-1).
(2) A leading and other equal significance circum-
stances (Events 2–13): this implies that a leading
weight is considered to a barrier, and the same
weight is put to the remaining barriers. For event
k (k= 2, 3, . . ., 13), a leading weight of 1 is consid-
ered for the (k – 1)th barrier (according to the
priority sequence of every barrier shown in
Table 9) and the same weight of 0.2 to the remain-
ing barriers. Considering Event 2 as an instance,
a leading weight of 1 is put to ‘high cost of
advanced technology adoption’ and the weight
of 0.2 to the remaining 11 barriers. Similarly, all
the events have been done. Table 13 shows the
outcomes of total events.
According to Table 13, it is noticeable that the
ranking of pathways is altered when the weights of
the barrier are adjusted. In addition, ELECTRE-I cannot
differentiate among five possible pathways in several
events. Therefore, it may be concluded that the rank-
ing of possible pathways is not identical if the experts
had been altered. There are two reasons for this
deviation: i) as the experts have dissimilar likings
and motivation, the weights of the barriers for GSCM
implementation are altered; ii) ranking of possible
pathways concerning individual barrier to implement
GSCM obtained by individual experts may have dis-
tinctive priorities. Overall, individual experts can

Table 13. Sensitivity analysis.

Pathways
Events P1 P2 P3 P4 P5
Base Event 2 3 1 4 4
Event −1 Not available Not available Not available Not available Not available
Event −2 2 2 1 3 3
Event −3 Not available Not available Not available Not available Not available
Event −4 1 2 1 3 3
Event −5 1 2 1 3 3
Event −6 1 2 1 3 3
Event −7 1 1 2 2 2
Event −8 Not available Not available Not available Not available Not available
Event −9 1 2 1 3 3
Event −10 2 1 3 3 3
Event −11 2 1 3 3 3
Event −12 2 3 1 4 4
Event −13 Not available Not available Not available Not available Not available
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY
decide on a distinctive relationship or comparative
matrices that may direct to characteristic outcomes.
5. Managerial and theoretical implications
5.1. Managerial implications
While implementing GSCM, it is not feasible to eliminate
all barriers identified immediately. That is why managers
have to determine the barrier that occupies the first
priority among barriers to implement GSCM. As technol-
ogy dimension is ranked first as a barrier, they should
focus more on technological issues. In addition, they
have to show a positive sense about advanced technol-
ogy. Managers need to be able to choose the most
suitable advanced technology at minimum cost.
Managers can accelerate the implementation pro-
cess of GSCM by selecting an effective pathway. This
research reveals that green technology and techni-
ques are the most effective pathway. Therefore, man-
agers can execute green technology and techniques
such as green purchasing, green manufacturing, and
green distribution in their system. If they can intro-
duce green strategy in their system, then it will be
fruitful to implement GSCM. It can also help to
improve cleaner production and transport system
and to increase the performance of CETP.
5.2. Theoretical and global implications
The present study identified major barriers and
ranked the effective pathways to GSCM implementa-
tion in the Bangladeshi LPI. The present study
encourages local and global stakeholders to imple-
ment GSCM in the Bangladeshi LPI by advancing the
level of GSCM theory and understanding in an emer-
ging economy context. Global stakeholders put pres-
sure on the Bangladeshi LPI to implement GSCM in
the LPI supply chains to reduce adverse environmen-
tal, economic and social effects. This study may assist
industrial managers in improving GSCM practices in
the LPI supply chains.
The methodology of this research can be applied
to other sectors, including the automotive industry,
the pharmaceutical industry, textiles, chemicals, and
the plastic industry, to evaluate major barriers and
select effective pathways to implement GSCM.
6. Conclusions, limitations and future work
GSCM is now becoming an important toolkit for greening
the world rapidly. In the advancement of products and
processes, GSCM is acting as an important tool for ful-
filling the ecological criteria. It is also facilitating the
improvement of productivity, reputation of trademark,
and thus increasing the performance of an organization.
In greening the industries, GSCM is a must and needs
15
harmonization among different stages in a supply chain.
This study tried to generate responsiveness among
leather practitioners of an emerging economy,
Bangladesh, regarding the importance of GSCM, and to
identify the most important barriers and define some
effective pathways to GSCM implementation in
Bangladeshi LPI. Identifying major barriers to implement
GSCM is a complex task. The present study proposed
a framework to lessen the complexities involved with
the identification of major barriers and pathways to
GSCM. The proposed framework would help practitioners
improve ecological practices in the LPI supply chain.
This research presents a hybrid technique by com-
bining AHP and ELECTRE. The ranking of the barriers
and pathways to implement GSCM determined by
various experts can show biases due to the presence
of distinctive consent of individual expert. Another
limitation is that this research considered only 20
barriers and 15 possible pathways. Besides, AHP can-
not determine the uncertainty in calculating the
weights of various attributes. In this study, all barriers
and pathways were examined without considering
any uncertainties. Therefore, future work can be
focused on barrier analysis to implement GSCM
under uncertainties. This study can be further
explained by using other MCDA methods such as
ANP, Fuzzy-AHP, ISM, and DEMATEL.
Disclosure statement
No potential conflict of interest was reported by the
authors.
ORCID
S. Uddin http://orcid.org/0000-0002-0126-8397
G. Kabir http://orcid.org/0000-0001-9591-0629
References
Abdel-Baset M, Chang V, Gamal A. 2019. Evaluation of the
green supply chain management practices: a novel neu-
trosophic approach. Comput Ind. 108:210–220.
doi:10.1016/j.compind.2019.02.013.
Agrebi M, Abed M, Omri MN. 2017. ELECTRE I based rele-
vance decision-makers feedback to the location selection
of distribution centers. J Adv Transp. 2017:1–10.
Ahmed T, Chowdhury ZU 2016. Environmental burden of tan-
neries in Bangladesh. 36th Annual Conference of the
International Association for Impact Assessment; Nagoya,
Japan.
Ali A, Bentley Y 2015. The impact of green supply chain
management on performance: a study of garments industry
in Bangladesh. International Conference on Sustainable
Development (ICSD); Bangladesh: Southern University.
An D, Xi B, Ren J, Wang Y, Jia X, He C, Li Z. 2017.
Sustainability assessment of groundwater remediation
technologies based on multi-criteria decision making
method. Resour Conserv Recycl. 119:36–46.
16 S. UDDIN ET AL.
Azom MR, Mahmud K, Yahya SM, Sontu A, Himon SB. 2012.
Environmental impact assessment of tanneries: a case
study of Hazaribag in Bangladesh. Int J Environ Sci Dev.
3(2):152–156.
Balaji M, Velmurugan V, Prasath M. 2014. Barriers in green
supply chain management: an indian goundry
perspective. Int J Res Eng Technol. 3(7):423–429.
Balon V, Sharma AK, Barua MK. 2016. Assessment of barriers
in green supply chain management using ISM: a case
study of the automobile industry in India. Global Bus
Rev. 17(1):116–135.
Bangladesh export promotion bureau. 2017. http://www.
epb.gov.bd/site/files/51916ae6-a9a3-462e-a6bd-9ef%
20074d835%20af/%20Statistic-%20Data-2016-2017.
Cengiz T, Akbulak C. 2009. Application of analytical hierar-
chy process and geographic information systems in land-
use suitability evaluation: a case study of Dümrek village
(Çanakkale, Turkey). Int J Sustainable Dev World Ecol. 16
(4):286–294.
Chien MK, Shih LH. 2007. Relationship between manage-
ment practice and organisation performance under
European Union directives such as RoHS: a case study of
the electrical and electronic industry in Taiwan. Afr
J Environ Sci Technol. 1(3):37–48.
Chin TA, Tat HH, Sulaiman Z. 2015. Green supply chain
management, environmental collaboration and sustain-
ability performance. Procedia CIRP. 26:695–699.
Chowdhury M, Mostafa MG, Biswas TK, Saha AK. 2013.
Treatment of leather industrial effluents by filtration and
coagulation processes. Water Resour Ind. 3:11–22.
Chowdhury M, Upadhyay A, Briggs A, Belal M 2016. An empiri-
cal analysis of green supply chain management practices in
Bangladesh construction industry. Annual Conference; Lake
Buena Vista Hilton, Orlando, Florida, U.S.A.
Diabat A, Govindan K. 2011. An analysis of the drivers
affecting the implementation of green supply chain
management. Resour Conserv Recycl. 55(6):659–667.
Eltayeb TK, Zailani S, Ramayah T. 2011. Green supply chain
initiatives among certified companies in Malaysia and
environmental sustainability: investigating the
outcomes. Resour Conserv Recycl. 55(5):495–506.
Gandhi S, Mangla SK, Kumar P, Kumar D. 2015. Evaluating
factors in implementation of successful green supply
chain management using DEMATEL: a case study.
Int Strategic Manage Rev. 3:96–109.
Govindan K, Kaliyan M, Kannan D, Haq AN. 2014. Barriers
analysis for green supply chain management implemen-
tation in Indian industries using analytic hierarchy
process. Int J Prod Econ. 147:555–568.
Hajkowicz S, Higgins A. 2008. A comparison of multiple
criteria analysis techniques for water resource
management. Eur J Oper Res. 184(1):255–265.
Hatami-Marbini A, Tavana M, Moradi M, Kangi F. 2013.
A fuzzy group ELECTRE method for safety and health
assessment in hazardous waste recycling facilities. Saf
Sci. 51(1):414–426.
Hoque A, Clarke A. 2013. Greening of industries in
Bangladesh: pollution prevention practices. J Clean
Prod. 51:47–56.
Hosseini A. 2007. Identification of green management sys-
tem’s factors: a conceptualized model. Int J Manage Sci
Eng Manage. 2(3):221–228.
Hsu CW, Hu AH. 2008. Green supply chain management in the
electronic industry. Int J Environ Sci Technol. 5(2):205–216.
Hu AH, Hsu CW. 2010. Critical factors for implementing
green supply chain management practice: an empirical
study of electrical and electronics industries in taiwan.
Manag Res Rev. 33(6):586–608. doi:10.1108/
01409171011050208.
Jabbour CJC, Puppim-de-Oliveira JA. 2012. Barriers to envir-
onmental management in clusters of small businesses in
Brazil and Japan: from a lack of knowledge to a decline in
traditional knowledge. Int J Sustainable Dev World Ecol.
19(3):247–257.
Kabir G, Hasin MAA. 2011. Comparative analysis of ahp and
fuzzy ahp models for multicriteria inventory classification.
International Journal of Fuzzy Logic Systems. 1(1):1–16.
Kaya T, Kahraman C. 2011. An integrated fuzzy AHP-ELECTRE
methodology for environmental impact assessment.
Expert Syst Appl. 38(7):8553–8562.
Lam PTI, Chan EHW, Poon CS, Chau CK, Chun KP. 2010.
Factors affecting the implementation of green specifica-
tions in construction. J Environ Manage. 91(3):654–661.
Luthra S, Mangla SK, Xu L, Diabat A. 2016. Using AHP to
evaluate barriers in adopting sustainable consumption
and production initiatives in a supply chain. Int J Prod
Econ. 181:342–349.
Martin-Utrillas M, Juan-Garcia F, Canto-Perello J, Curiel-
Esparza J. 2015. Optimal infrastructure selection to
boost regional sustainable economy. Int J Sustainable
Dev World Ecol. 22(1):30–38.
Mathiyazhagan K, Govindan K, Noorul Haq A. 2014. Pressure
analysis for green supply chain management implemen-
tation in Indian industries using analytic hierarchy
process. Int J Prod Res. 52(1):188–202.
Mathiyazhagan K, Govindan K, NoorulHaq A, Geng Y. 2013. An
ISM approach for the barrier analysis in implementing green
supply chain management. J Clean Prod. 47:283–297.
Mendoza GA, Martins H. 2006. Multi-criteria decision analy-
sis in natural resource management: a critical review of
methods and new modelling paradigms. For Ecol
Manage. 230(1–3):1–22.
Min H, Kim I. 2012. Green supply chain research: past, pre-
sent, and future. Logist Res. 4(1–2):39–47.
Mohanty RP, Prakash A. 2013. Green supply chain manage-
ment practices in India: an empirical study. Prod Plann
Control. 25(16):1322–1337.
Moktadir A, Ali SM, Paul SK, Shukla N. 2018a. Barriers to big
data analytics in manufacturing supply chains: a case
study from Bangladesh. Comp Ind Eng. 128:1063–1075.
doi:10.1016/j.cie.2018.04.013.
Moktadir A, Rahman T, Jabbour CJC, Mithun Ali S, Kabir G.
2018b. Prioritization of drivers of corporate social respon-
sibility in the footwear industry in an emerging economy:
a fuzzy AHP approach. J Clean Prod. 201:369–381.
doi:10.1016/j.jclepro.2018.07.326.
Moktadir M, Ali S, Mangla S, Sharmy T, Luthra S, Mishra N,
Garza-Reyes J. 2018c. Decision modeling of risks in phar-
maceutical supply chains. Ind Manage Data Syst. 118
(7):1388–1412. doi:10.1108/IMDS-10-2017-0465.
Moktadir MA, Ali SM, Rajesh R, Paul SK. 2018d. Modeling the
interrelationships among barriers to sustainable supply
chain management in leather industry. J Clean Prod.
181:631–651. doi:10.1016/j.jclepro.2018.01.245.
Moktadir MA, Rahman T, Rahman MH, Ali SM, Paul SK.
2018e. Drivers to sustainable manufacturing practices
and circular economy: a perspective of leather industries
in Bangladesh. J Clean Prod. 174:1366–1380. doi:10.1016/
j.jclepro.2017.11.063.
Mudgal RK, Shankar R, Talib P, Raj T. 2009. Greening the
supply chain practices: an Indian perspective of enablers’
relationships. Int J Adv Oper Manage. 1(2/3):151–176.
Mudgal RK, Shankar R, Talib P, Raj T. 2010. Modelling the
barriers of green supply chain practices: an Indian
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY
perspective. Int J Logist Syst Manage. 7(1):81.
doi:10.1504/ijlsm.2010.033891.
Pang J, Zhang G, Chen G. 2011. ELECTRE I decision model of
reliability design scheme for computer numerical control
machine. J Software. 6(5):894–900.
Parmar NK. 2016. Analysis of barriers for implementing
green supply chain management in small and medium
sized enterprises (SMEs) of India. Int J Hum Manage Sci. 4
(3):219–223.
Paul HL, Antunes APM, Covington AD, Evans P, Phillips PS. 2013.
Bangladeshi leather industry: an overview of recent sustain-
able developments. Soc Leather Technol Chem. 97:25–32.
Perron GM. 2005. Barriers to environmental performance
improvements in Canadian SMEs. Canada: Dalhousie
University.
Rao P, Holt D. 2005. Do green supply chains lead to compe-
titiveness and economic performance? Int J Oper Prod
Manage. 25(9):898–916.
Revell A, Rutherfoord R. 2003. UK environmental policy and
the small firm: broadening the focus. Bus Strategy
Environ. 12(1):26–35.
Rodrigues SP, Costa LA, Produc D, Minho U 2015.
Comparing AHP and ELECTRE I for prioritizing software
requirements. International Conference on Software
Engineering, Artificial Intelligence, Networking and
Parallel/Distributed Computing; Takamatsu, Japan.
Roy B. 1968. Classement et choix en présence de points de
vue multiples. Revue française d’informatique et de
recherche opérationnelle. 2(8):57–75.
Saaty TL. 1980. Multicriteria decision making: the analytic
hierarchy process. New York: McGraw-Hill.
Samaras GD, Gkanas NI, Vitsa KC. 2014. Assessing risk in
Dam projects using AHP and ELECTRE I. Int J Constr
Manage. 14(4):255–266.
Sarkis J. 2003. A strategic decision framework for green
supply chain management. J Clean Prod. 11(4):397–409.
Sarkis J, Zhu Q, Lai KH. 2011. An organizational theoretic review
of green supply chain management literature. Internat J Prod
Econ. 130(1):1–15. doi:10.1016/j.ijpe.2010.11.010.
Seman NAA, Govindan K, Mardani A, Zakuan N, Mat
Saman MZ, Hooker RE, Ozkul S. 2019. The mediating
effect of green innovation on the relationship between
green supply chain management and environmental
performance. J Clean Prod. 229:115–127. doi:10.1016/j.
jclepro.2019.03.211.
Shipeng Q, Linna D 2011. A study on green supply chain
management of enterprises based on self-locking theory.
International Conference on E -Business and E –
Government; Shanghai, China.
Singh PK, Sarkar P. 2019. A framework based on fuzzy
AHP-TOPSIS for prioritizing solutions to overcome the
barriers in the implementation of ecodesign practices in
SMEs. Int J Sustainable Dev World Ecol. 1–16. doi:10.1080/
13504509.2019.1605547.
Somsuk N, Laosirihongthong T. 2017. Prioritization of applic-
able drivers for green supply chain management imple-
mentation toward sustainability in Thailand.
Int J Sustainable Dev World Ecol. 24(2):175–191.
17
Srivastava SK. 2007. Green supply-chain management: a
state-of-the-art literature review. Int J Manage Rev. 9
(1):53–80.
Subramanian N, Gunasekaran A, Abdulrahman MD, Liu C,
Su D. 2014. Reverse logistics in the Chinese auto-parts
firms: implementation framework development through
multiple case studies. Int J Sustainable Dev World Ecol. 21
(3):223–234. doi:10.1080/13504509.2014.907216.
Teixeira de Almeida A. 2007. Multicriteria decision model for
outsourcing contracts selection based on utility function
and ELECTRE method. Comput Oper Res. 34(12):3569–3574.
Toke LK, Kalpande SD. 2018. A framework of enabler’s rela-
tionship for implementation of green manufacturing in
Indian context. Int J Sustainable Dev World Ecol. 25
(4):303–311.
Tumpa TJ, Ali SM, Rahman MH, Paul SK, Chowdhury P,
Rehman Khan SA. 2019. Barriers to green supply chain
management: an emerging economy context. J Clean
Prod. 236:117617. doi:10.1016/j.jclepro.2019.117617.
Udomleartprasert P 2004. Roadmap to green supply chain
electronics: design for manufacturing implementation
and management panyaluck. International IEEE confer-
ence on Asian Green Electronics; Hong Kong, China.
Van Hemel C, Cramer J. 2002. Barriers and stimuli for ecode-
sign in SMEs. J Clean Prod. 10(5):439–453.
Vanpoucke E 2014. Motivation to pursue green supply chain
management. International Conference on Industrial
Engineering and Operations Management; Bali, Indonesia.
Wang XJ, Chan HK. 2013. A hierarchical fuzzy TOPSIS
approach to assess improvement areas when implement-
ing green supply chain initiatives. Int J Prod Res. 51
(10):3117–3130.
Wang Z, Mathiyazhagan K, Xu L, Diabat A. 2015. A decision
making trial and evaluation laboratory approach to ana-
lyze the barriers to green supply chain management
adoption in a food packaging company. J Clean Prod.
117:19–28.
Wooi GC, Zailani S. 2010. Green supply chain initiatives:
investigation on the barriers in the context of smes in
Malaysia. Internat Busin Manag. 4(1):20–27. doi:10.3923/
ibm.2010.20.27.
Wu K-J, Tseng M-L, Vy T. 2011. Evaluation the drivers of
green supply chain management practices in
uncertainty. Procedia Social Behav Sci. 25(2011):384–397.
Zhu Q, Geng Y. 2013. Drivers and barriers of extended supply
chain practices for energy saving and emission reduction
among Chinese manufacturers. J Clean Prod. 40:6–12.
Zhu Q, Geng Y, Fujita T, Hashimoto S. 2010. Green supply
chain management in leading manufacturers. Manage
Res Rev. 33(4):380–392.
Zhu Q, Sarkis J, Lai KH. 2007. Green supply chain manage-
ment: pressures, practices and performance within the
chinese automobile industry. J Clean Prod. 15(11):1041–
1052. doi:10.1016/j.jclepro.2006.05.021.
Zubayer AA, Ali SM, Kabir G. 2019. Analysis of supply chain
risk in the ceramic industry using the TOPSIS method
under a fuzzy environment. J Modell Manage. (Accepted
Manuscript). doi: 10.1108/JM2-06-2018-0081.
18 S. UDDIN ET AL.

Appendix

Table A1. Questionnaire.

No. Questionnaire Yes/No
1 Does the manager show reluctance to adopt advanced technology due to high cost?
2 Are there any organizational difficulties with environmental technology and apprehension of failing to implement GSCM?
3 Are there any lacking of technological knowledge and resources?
4 Are the cost of hazardous waste disposal and environmental packages high?
5 Are there fewer return-on-investments?
6 Does the financial constraint seem to be more critical barrier?
7 Is the awareness about reverse logistics adoption unsatisfactory?
8 Does the absence of professionals hinder the acceleration of green system?
9 Can the reward and encouragement programs accelerate the GSCM Implementation?
10 Do you think government needs to revise regulation and legislation about environment?
11 Are there any restrictive company policies towards product or process stewardship?
12 Can the green technology and techniques accelerate the implementation process?
13 Does the management of waste and appropriate infrastructure make the GSCM implementation process easier?
14 Do you think supplier management and cleaner technology can influence the GSCM implementation process effectively?

Table A2. Relationship matrix of possible pathways with Table A6. Relationship matrix of possible pathways with
respect to T1. respect to K2.
T1 P1 P2 P3 P4 P5 Weights K2 P1 P2 P3 P4 P5 Weights
P1 1 2 1/2 1 3 0.207 P1 1 1 2 7 7 0.378
P2 1/2 1 1/3 2 2 0.153 P2 1 1 2 5 3 0.298
P3 2 3 1 5 4 0.432 P3 1/2 1/2 1 7 2 0.194
P4 1 1/2 1/2 1 2 0.144 P4 1/7 1/5 1/3 1 1/7 0.040
P5 1/3 1/5 1/4 1/2 1 0.064 P5 1/7 1/7 1/2 7 1 0.089
λmax = 5.277, CI = 0.069, CR = 0.062 < 0.1 λmax = 5.295, CI = 0.074, CR = 0.066 < 0.1

Table A3. Relationship matrix of possible pathways with Table A7. Relationship matrix of possible pathways with
respect to T2. respect to K3.
T2 P1 P2 P3 P4 P5 Weights K3 P1 P2 P3 P4 P5 Weights
P1 1 1 5 5 7 0.399 P1 1 2 1/4 2 5 0.186
P2 1 1 7 4 5 0.382 P2 1/2 1 1/7 1 1 0.079
P3 1/5 1/7 1 3 1 0.087 P3 4 7 1 5 9 0.561
P4 1/5 1/4 1/5 1 2 0.065 P4 1/2 1 1 1 2 0.135
P5 1/7 1/3 1 1/2 1 0.067 P5 1/5 1/5 1/9 1/2 1 0.040
λmax = 5.243, CI = 0.061, CR = 0.054 < 0.1 λmax = 5.435, CI = 0.109, CR = 0.097 < 0.1

Table A4. Relationship matrix of possible pathways with Table A8. Relationship matrix of possible pathways with
respect to T3. respect to G1.
T3 P1 P2 P3 P4 P5 Weights G1 P1 P2 P3 P4 P5 Weights
P1 1 2 1 5 3 0.323 P1 1 3 2 4 3 0.378
P2 1/2 1 1/3 3 2 0.164 P2 1/3 1 1/2 2 3 0.161
P3 1 3 1 8 2 0.355 P3 1/2 2 1 6 4 0.304
P4 1/5 1/3 1/2 1 1 0.083 P4 1/4 1/2 1/3 1 2 0.098
P5 1/3 1/8 1/2 1 1 0.075 P5 1/3 1/6 1/4 1/2 1 0.060
λmax = 5.391, CI = 0.098, CR = 0.087 < 0.1 λmax = 5.320, CI = 0.080, CR = 0.071 < 0.1

Table A5. Relationship matrix of possible pathways with Table A9. Relationship matrix of possible pathways with
respect to K1. respect to G2.
K1 P1 P2 P3 P4 P5 Weights G2 P1 P2 P3 P4 P5 Weights
P1 1 3 2 3 5 0.391 P1 1 1/3 2 3 3 0.225
P2 1/3 1 1/3 1/3 4 0.109 P2 3 1 5 3 5 0.464
P3 1/2 3 1 3 7 0.317 P3 1/2 1/5 1 5 2 0.157
P4 1/3 3 1/4 1 1 0.121 P4 1/3 1/3 1/5 1 2 0.084
P5 1/5 1/3 1/7 1 1 0.063 P5 1/3 1/5 1/2 1/2 1 0.069
λmax = 5.362, CI = 0.090, CR = 0.081 < 0.1 λmax = 5.444, CI = 0.111, CR = 0.099 < 0.1
 INTERNATIONAL JOURNAL OF SUSTAINABLE DEVELOPMENT & WORLD ECOLOGY 19

Table A10. Relationship matrix of possible pathways with Table A12. Relationship matrix of possible pathways with
respect to G3. respect to F2.
G3 P1 P2 P3 P4 P5 Weights F2 P1 P2 P3 P4 P5 Weights
P1 1 2 2 4 3 0.354 P1 1 3 1 4 3 0.327
P2 1/2 1 2 3 8 0.308 P2 1/3 1 1/3 2 5 0.163
P3 1/2 1/2 1 2 4 0.187 P3 1 3 1 4 5 0.363
P4 1/4 1/3 1/8 1 2 0.075 P4 1/4 1/2 1/5 1 1 0.076
P5 1/3 1/2 1/4 1/2 1 0.075 P5 1/3 1/4 1/5 1 1 0.070
λmax = 5.401, CI = 0.100, CR = 0.090 < 0.1 λmax = 5.185, CI = 0.046, CR = 0.041 < 0.1

Table A11. Relationship matrix of possible pathways with Table A13. Relationship matrix of possible pathways with
respect to F1. respect to F3.
F1 P1 P2 P3 P4 P5 Weights F3 P1 P2 P3 P4 P5 Weights
P1 1 1 2 8 3 0.336 P1 1 1/2 2 1 3 0.220
P2 1 1 3 3 7 0.355 P2 2 1 3 5 2 0.400
P3 1/2 1/3 1 4 3 0.178 P3 1/2 1/3 1 3 1 0.154
P4 1/8 1/3 1/7 1 4 0.073 P4 1 1/5 1/2 1 2 0.128
P5 1/3 1/4 1/3 1/4 1 0.057 P5 1/3 1/3 1 1/2 1 0.099
λmax = 5.408, CI = 0.102, CR = 0.091 < 0.1 λmax = 5.393, CI = 0.098, CR = 0.088 < 0.1

Table A14. Weighted normalized matrix.

Barriers T1 T2 T3 K1 K2 K3 G1 G2 G3 F1 F2 F3
Weights 0.2347 0.0831 0.0442 0.1356 0.0583 0.0150 0.0931 0.0221 0.0087 0.1806 0.1017 0.0229
P1 0.0485 0.0332 0.0143 0.0531 0.0220 0.0028 0.0352 0.0050 0.0031 0.0607 0.0333 0.0050
P2 0.0359 0.0317 0.0072 0.0147 0.0174 0.0012 0.0150 0.0103 0.0027 0.0641 0.0166 0.0092
P3 0.1014 0.0072 0.0157 0.0430 0.0113 0.0084 0.0283 0.0035 0.0016 0.0321 0.0369 0.0035
P4 0.0339 0.0054 0.0037 0.0163 0.0024 0.0020 0.0091 0.0019 0.0007 0.0133 0.0078 0.0029
P5 0.0149 0.0056 0.0033 0.0085 0.0052 0.0006 0.0055 0.0015 0.0007 0.0104 0.0072 0.0023

Table A15. Concordance interval index and interval set.

P1 P2 P3 P4 P5
P1 0 0.7745 0.6045 0.6045 0.5213
P2 0.2255 0 0.3757 0.3757 0.2926
P3 0.395535 0.6243 0 0 0
P4 0 0.0000 0.0000 0 0.0000
P5 0 0.0583 0.0583 0.0251 0

Table A16. Discordance interval index.

Pathways P1 P2 P3 P4 P5
P1 0 0.1379 1.8537 0.0162 0.0436
P2 7.250708 0 2.051243 0.03195721 0.011108
P3 0.539465 0.4875 0 0.00871834 0.011325
P4 61.79739 31.29184 114.7008 0 0.150038
P5 22.94635 90.02325 88.30298 6.66498026 0
20 S. UDDIN ET AL.

Table A17. Discordance sets.

Barriers T1 T2 T3 K1 K2 K3 G1 G2 G3 F1 F2 F3
Weights 0.2347 0.0831 0.0442 0.1356 0.0583 0.0150 0.0931 0.0221 0.0087 0.1806 0.1017 0.0229
P1 0.0485 0.0332 0.0143 0.0531 0.0220 0.0028 0.0352 0.0050 0.0031 0.0607 0.0333 0.0050
P2 0.0359 0.0317 0.0072 0.0147 0.0174 0.0012 0.0150 0.0103 0.0027 0.0641 0.0166 0.0092
P3 0.1014 0.0072 0.0157 0.0430 0.0113 0.0084 0.0283 0.0035 0.0016 0.0321 0.0369 0.0035
P4 0.0339 0.0054 0.0037 0.0163 0.0024 0.0020 0.0091 0.0019 0.0007 0.0133 0.0078 0.0029
P5 0.0149 0.0056 0.0033 0.0085 0.0052 0.0006 0.0055 0.0015 0.0007 0.0104 0.0072 0.0023
P1-P2 0.0126 0.0014 0.0070 0.0383 0.0046 0.0016 0.0202 −0.0053 0.0004 −0.0034 0.0167 −0.0041
P1-P3 −0.0529 0.0260 −0.0014 0.0100 0.0107 −0.0056 0.0069 0.0015 0.0015 0.0286 −0.0036 0.0015
P1-P4 0.0146 0.0278 0.0106 0.0367 0.0197 0.0008 0.0261 0.0031 0.0024 0.0474 0.0255 0.0021
P1-P5 0.0336 0.0276 0.0109 0.0446 0.0168 0.0022 0.0297 0.0034 0.0024 0.0503 0.0261 0.0028
P2-P1 −0.0126 −0.0014 −0.0070 −0.0383 −0.0046 −0.0016 −0.0202 0.0053 −0.0004 0.0034 −0.0167 0.0041
P2-P3 −0.0655 0.0245 −0.0084 −0.0283 0.0061 −0.0072 −0.0133 0.0068 0.0011 0.0319 −0.0203 0.0056
P2-P4 0.0020 0.0263 0.0036 −0.0016 0.0150 −0.0008 0.0059 0.0084 0.0020 0.0508 0.0088 0.0062
P2-P5 0.0210 0.0261 0.0039 0.0062 0.0122 0.0006 0.0094 0.0087 0.0020 0.0537 0.0094 0.0069
P3-P1 0.0529 −0.0260 0.0014 −0.0100 −0.0107 0.0056 −0.0069 −0.0015 −0.0015 −0.0286 0.0036 −0.0015
P3-P2 0.0655 −0.0245 0.0084 0.0283 −0.0061 0.0072 0.0133 −0.0068 −0.0011 −0.0319 0.0203 −0.0056
P3-P4 0.0675 0.0018 0.0120 0.0267 0.0090 0.0064 0.0192 0.0016 0.0010 0.0189 0.0291 0.0006
P3-P5 0.0865 0.0016 0.0123 0.0345 0.0061 0.0078 0.0227 0.0019 0.0010 0.0218 0.0297 0.0013
P4-P1 −0.0146 −0.0278 −0.0106 −0.0367 −0.0197 −0.0008 −0.0261 −0.0031 −0.0024 −0.0474 −0.0255 −0.0021
P4-P2 −0.0020 −0.0263 −0.0036 0.0016 −0.0150 0.0008 −0.0059 −0.0084 −0.0020 −0.0508 −0.0088 −0.0062
P4-P3 −0.0675 −0.0018 −0.0120 −0.0267 −0.0090 −0.0064 −0.0192 −0.0016 −0.0010 −0.0189 −0.0291 −0.0006
P4-P5 0.0189 −0.0002 0.0003 0.0078 −0.0028 0.0014 0.0036 0.0003 0.0000 0.0029 0.0006 0.0007
P5-P1 −0.0336 −0.0276 −0.0109 −0.0446 −0.0168 −0.0022 −0.0297 −0.0034 −0.0024 −0.0503 −0.0261 −0.0028
P5-P2 −0.0210 −0.0261 −0.0039 −0.0062 −0.0122 −0.0006 −0.0094 −0.0087 −0.0020 −0.0537 −0.0094 −0.0069
P5-P3 −0.0865 −0.0016 −0.0123 −0.0345 −0.0061 −0.0078 −0.0227 −0.0019 −0.0010 −0.0218 −0.0297 −0.0013
P5-P4 −0.0189 0.0002 −0.0003 −0.0078 0.0028 −0.0014 −0.0036 −0.0003 0.0000 −0.0029 −0.0006 −0.0007