Int. J. Production Economics: Simonov Kusi-Sarpong, Joseph Sarkis, Xuping Wang

Int. J.
Production Economics 181 (2016) 325–341
Contents lists available at ScienceDirect
Int. J. Production Economics

journal homepage: www.elsevier.com/locate/ijpe
Assessing green supply chain practices in the Ghanaian mining

industry: A framework and evaluation
Simonov Kusi-Sarpong a, Joseph Sarkis b,n, Xuping Wang a,c
a
School of Management Science and Engineering, Dalian University of Technology, No. 2 Linggong Road, Ganjingzi District, Dalian Liaoning Province 116023,
PR China
b
Foisie School of Business, Worcester Polytechnic Institute, 100 Institute Road, Worcester, MA 01609-2280, USA
c
School of Business, Dalian University of Technology, Panjin Liaoning Province, 124221, PR China
art ic l e i nf o a b s t r a c t
Article history: Production and consumption in our industrial systems typically begin in the extractive, mining, in-
Received 6 April 2015 dustries. Typically these activities begin in emerging economies, such as Ghana. It is also clear that
Received in revised form supply chain activities in mining operations may have severe environmental and social problems with
16 December 2015
serious economic consequences. Greening the supply chain of mining operations are an important
Accepted 2 April 2016
Available online 6 April 2016
avenue that can provide beneficial consequences. Developing, evaluating, assessing, and selecting es-
sential green supply chain management (GSCM) practices are a goal for successful GSCM implementa-
Keywords: tion. These practices may have interrelated and complex relationships. Understanding them and their
Green supply chain management (GSCM) relative importance is an initial step for achieving the assessment goals for successful GSCM im-
practices
plementation in the mining industry. This study adopts a proposed comprehensive and integrative GSCM
Fuzzy theory
major practices and sub-practices (framework); determines the relative relationships and influences
Decision-making trial and evaluation la-
boratory (DEMATEL) within this GSCM framework, and identifies the perceived impact of the GSCM framework on organi-
Analytic network process (ANP) zational sustainable performance (economic, environmental, and social – triple bottom-line) pertinent to
Mining industry the mining industry, in the emerging economy nation of Ghana. An integrated methodology identifying
Ghana and limiting interdependencies within GSCM factors will be utilized. The methodology uses fuzzy-DE-
MATEL and analytical network process (ANP) for the evaluation. Multiple field studies within Ghana's
mining industry are used to illustrate the applicability of the proposed methodology. The results can
provide valuable clues and guidelines to decision-makers and analysts inside and outside the mining
industry, for improving corporate sustainable production and consumption. Future research and practical
implications are also introduced in the paper.
& 2016 Elsevier B.V. All rights reserved.
1. Introduction include materials purchase and used, the nature of the production
processes and activities (i.e. exploration, mining, mineral proces-
Mining operations supply chain activities are mired with ser- sing and extractive metallurgy), and how waste generated are
ious environmental and social dangers, with economic implica- utilized, whether it is closed-loop, industrial ecosystem,1 or dis-
tions underlying all these activities (Vintró et al., 2014; Söderholm, posal focused.
et al., 2015). Environmental concerns and discourse in mining and Managing these supply chain issues will help address environ-
other industries has grown to include extended producer re- mental problems associated with mining operations. For example,
sponsibility for environmental impacts (Niza et al., 2014) as well as mining activities such as exploration result in the removal of ve-
sustainable production and consumption concerns. These philo- getation causing soil erosion and habitat destruction. Drilling may
sophies support life cycle logic and assessment enabling a holistic lead to serious soil and water contamination through oil spills.
approach to mitigate environmental issues from mining opera- Mining activities (digging, loading and hauling/transportation of the
tions. Thus, supply chain activities represent both production and
consumption dimensions facing organizations. These activities
1
Closed-loop practices within supply chain management bring the product
back into the supply chain system or can be used in another system as a useful
n
Corresponding author. input (industrial ecosystem). This would require the “Re” practices such as re-
E-mail addresses: [email protected] (S. Kusi-Sarpong), cycling, reclamation, reuse or remanufacturing. Reverse logistics or reverse supply
[email protected] (J. Sarkis), [email protected] (X. Wang). chain activities would also be needed to be integrated into these systems.
http://dx.doi.org/10.1016/j.ijpe.2016.04.002
0925-5273/& 2016 Elsevier B.V. All rights reserved.
326 S. Kusi-Sarpong et al. / Int. J. Production Economics 181 (2016) 325–341
ore) cause diversion of natural body water flows, increased sedi- The motivation and objectives for this paper are twofold. First,
ment load in rivers, and waste rock and overburden disposal re- the literature provides some previous studies on the use of DE-
sulting in airborne dust, acid drainage, and erosion. These activities MATEL and ANP methods, but the literature has not explicitly
require careful attention and mitigation as internal mining opera- combined these methods together to form a more efficient deci-
tions. In addition to internal operational mining activities, the sion network for ANP. These two synergistic tools (DEMATEL and
purchasing function, as part of supply chain management, plays a ANP) can improve computational efficiencies and practical deci-
greater role in the reduction of mining operations environmental sion making for managers seeking to evaluate complex initiatives
burdens. The purchasing function needs to consider environmental and programs. These efficiency gains occur by helping to limit the
friendly inputs or equipment, such as inputs with lower environ- number of interrelationships that ANP will have to evaluate based
mental impact and low-energy consumption (Azevedo et al., 2012; on the upfront DEMATEL network formation process.
Mangla et al., 2014). A second motivation is that studies have been completed on
The mining industry is known to have a poor environmental green supply chain management initiatives and programs im-
reputation (Muduli et al., 2013). This reputation and their opera- plementation in the mining industry from both developed and
tions and supply chain activities have forced many mining com- developing nations. China (Haibin and Zhenling, 2010; Si et al.,
panies to face competitive, regulatory, and community/social 2010) and Australia (Van Berkel, 2007; Giurco and Cooper, 2012),
pressures causing increased consideration of greening their supply have seen investigations, but smaller emerging economy nations
chains. Thus far, mining companies’ green solutions have primarily such as Ghana have not been investigated. Studies on environ-
focused on internal supply chain activities of the focal company. mental (green) sustainability in Ghana”s mining industry produce
These localized and reactive green environmental management an unclear picture about the industry”s environmental impacts,
practices do not systemically reduce pollution emissions and focus especially from supply chain-based environmental initiatives (e.g.
on expensive investments in waste management, cleanup or Fei-Baffoe et al., 2013). Thus GSCM and its relationship to sus-
remediation.2 To minimize or eliminate the adverse ecological tainable development in Ghana”s mining industry will help to add
influence of mining company supply chains, there is the need to clarity to this concern. This investigation will set the foundation
holistically address this issue. for future GSCM research in the mining industry and provide
The green supply chain concept can aid in evaluating global and managers and researchers with a better understanding of the
systemic environmental footprint reduction (Muduli et al., 2013; different sustainable operational factors and management inter-
Kumar et al., 2014). Green supply chain management (GSCM) is a ventions that can enhance sustainability performance in Ghana's
systematic and integrated approach that can help companies to mining industry and in general mining.
develop ‘win-win’ strategies resulting in profit and market share The following research questions are addressed in this paper:
objectives achievement and environmental efficiency (Tseng,
(1) How effective a methodology is a joint DEMATEL analysis and
2011; Tseng and Chiu, 2013; Wong et al., 2015). Most studies on
ANP for evaluation purposes?
the practices of GSCM implementation in the mining industry
(2) What are the relative relationships and influences of GSCM
have only considered environmental management practices, fail-
factors and sub-factors in an emerging country mining in-
ing or overlooking the implementation of GSCM practices from a
dustry context, especially Ghana?
holistic perspective. Understanding the importance and relation-
(3) What are the relative relationships of GSCM factors and sub-
ships, as viewed by mining industry managers, especially in
factors on organizational sustainability performance measures
emerging economies can help clarify and aid the implementation
in an emerging country mining industry context?
and management of GSCM practices. Ghana is prime example of an
emerging economy country whose environmental burdens and
The contributions of this study are manifold. First, the issue of
economic development is closely tied to its mining industry. Lit-
GSCM and its relationships to organizational performance has only
erature that has considered GSCM context in Ghana”s mining in-
seen limited discussion in the literature. This paper contributes to
dustry is non-existent. To help address this gap in literature and
this discussion. Second, a focused investigation of GSCM in the
practice the GSCM practices, their relationships, and level of im-
Ghanaian mining industry context is non-existent; this work is the
portance set the stage for their management and implementation
first to investigate this issue. Third, the focus on Ghana represents
in the mining industry in an emerging nation, Ghana, are pre-
an emerging economy country focus on GSCM, an area that has not
sented in this paper. This study uses a multiple criteria decision- seen significant research in general, or specifically to the mining
making (MCDM) tool characterized by multiple conflicting criteria industry. Fourth, for the first time, this paper proposes a hybrid
(Hwang and Yoon, 1981). multi-criteria decision-making (MCDM) methodology based on
The integrated MCDM methodology includes fuzzy-DEMATEL and fuzzy-DEMATEL and ANP with a focus on computational effi-
the Analytical Network Process (ANP) and is composed of two main ciencies that contributes to decision making application.
phases. First, the fuzzy-DEMATEL aspect of the methodology is used The study provides researchers and policy makers with an
to develop the level of relative relationships and influence on each of understanding of how GSCM can be used to reduce mining in-
these GSCM practices and sub-practices. ANP uses the network industry environmental impact. Researchers and policy makers can
terrelationships identified in the fuzzy-DEMATEL stage to determine also use this study to help determine how impact can be lessened
the relative impact of the major GSCM practices and sub-practices to through improved designs, efficient operations and supply chain
organizational sustainable performance. Field study data within synergies (McLellan et al., 2009). At the firm level the results are
Ghana's mining industry are used to illustrate the applicability of the useful managing implementation of GSCM initiatives.
proposed methodology. An assessment of the level of influence of the The rest of the paper is organized in the following manner. The
practices and sub-practices will enable decision-makers to determine paper first reviews GSCM literatures, introduces and adopts a
implementation priority and resources to be allocated to each of previously developed GSCM practices framework for the mining
these practices and sub-practices. industry and discusses GSCM factors implementation benefits and
organizational sustainable performance in Section 2. Section 3
2
U.S. Environmental Protection Agency, Pollution Prevention (P2), Law and
presents the technical background of the various methodologies
Policy, February 16, 2012 〈http://www.epa.gov/p2/pubs/laws.htm〉 (24 October within the proposed novel MCDM methodology. The proposed
2013). novel hybrid MCDM methodology with case application using the
S. Kusi-Sarpong et al. / Int. J. Production Economics 181 (2016) 325–341 327
GSCM practices framework in the Ghanaian Mining Industry is practices (Molla et al., 2014; Koo and Chung, 2014; Bai and
presented in Section 4. Section 5 discusses the results from the Sarkis, 2013; Sarkis et al., 2013). However, there has been a
evaluation and Section 6 concludes by summarizing the findings neglect of the IT function in environmental evaluation pro-
and managerial implications are identified. grams over the years (Savita et al., 2014). In the mining
industry, equipment and employees use ITS. ITS use result in
significant environmental footprints (Faucheux and Nicolaï,
2. Literature review 2011; Uddin and Rahman, 2012). ‘Green’ ITS can help to
mitigate these environmental footprints (Bhadauria et al.,
Industrialization has caused damage to natural environmental 2014) and optimize overall energy consumption of mines
and human systems. These issues have resulted in growing inter- (Bilal et al., 2014). The use of energy efficient hardware and
est on GSCM (Dam and Petkova, 2014; Fabbe-Costes et al., 2014; data center, consolidating servers using virtualization soft-
Tseng et al., 2015). The mining and extractive industries, re- ware, reducing waste associated with obsolete equipment,
presenting the source of most virgin materials, are at the core of collaborative group software and telepresence systems and
these many concerns. eco-labeling of IT products are all part of Green IT initiatives.
GSCM is gaining interest amongst researchers and practitioners (2) Strategic Supplier Partnership (SSP)
(Beske and Seuring, 2014; Brandenburg et al., 2014; Tseng and Chiu, Strategic supplier partnership (SSP) is a long-term and exclusive
2013). GSCM practices include upstream, internal processes (focal alliance between focal organizations and suppliers (Ramanathan
firms) and downstream activities (Tseng, 2011; Tseng and Chiu, 2013; and Gunasekaran, 2014). Mining companies can use strategic
Wong et al., 2015). Literature has shown that many mining compa- supplier partnerships to involve their strategic suppliers in green
nies are gradually adopting some sustainable (green) practices such supply chain planning to communicate sustainability goals and as
as environmental management systems (EMS) and cleaner produc- a baseline to monitor these suppliers’ environmental compliance
tion (CP) (Vintró et al., 2014). Various forces have caused mining status and operational practices. Mining companies can also use
organizations to adopt these practices including regulations (Dupuy, SSP to jointly develop environmental management solutions and
2014; Luthra et al., 2015), community activism (Lin et al., 2015, Moffat programs to reduce or eliminate material use, share environ-
and Zhang, 2014; Falck and Spangenberg, 2014), investors (Dash- mental management techniques and knowledge, and collabora-
wood, 2014) and increasing organizational efficiencies (Mangla et al., tively manage reverse flows of materials and packages (Wong
2014). The rest of this literature review section overviews GSCM et al., 2015; De Giovanni and Vinzi, 2014; Govindan et al., 2015;
practices in the mining industry, multiple criteria decision making Blome et al., 2014; Vachon and Klassen, 2006b).
and evaluation techniques for GSCM, and background on green (3) Operations and Logistics Integration (OLI)
practices in Ghana and mining. These topics set the foundation for Effective operations and logistics integration in the mining
the remainder of the paper. operations will provide time, equipment and capacity utilities
(Wiengarten et al., 2014) with improved economic and re-
2.1. GSCM practices in the mining industry duced environmental impact. Internal and external integration
promote real-time information flow supporting lean produc-
Many environmentally oriented efforts can be applied to help tion, green logistics, green purchasing/electronic-ordering and
reduce the mining industry”s environmental burden (Edraki et al., tracking system replacing paper-based ordering system and
2014). Technology, regulatory and industrial policies have each help minimize environmental impact associated with the flow
facilitated improvement in mining industry environmental per- of materials (Drohomeretski et al., 2014; Wong et al., 2015;
formance (Mathiyazhagan et al., 2014; Govindan et al., 2014a; Govindan et al., 2014b).
Söderholm, et al., 2015). Various barriers have also prevented the (4) Internal Environmental Management (IEM)
implementation of cleaner technologies and cleaner production Environmental concerns in the mining industry require sys-
(CP) practices, as part of GSCM, in the mining industries including tematic and holistic approaches with internal environmental
legislative, technological, and economic barriers (Corder et al., management to help address these problems (Vintró et al., 2014;
2014; Pooe and Mhelembe, 2014). Overcoming these barriers re- Mangla et al., 2014). IEM systematic implementation requires
quires various managerial and educational initiatives, especially in monitoring and auditing for environmental compliance of
the mining industry (Govindan et al., 2014b; Muduli et al., 2013). mining operations. Introducing reward and incentive systems
Although there are many studies on mining industry practices and for environmental suggestions, top management support and
environmental issues (e.g. Vintró et al., 2014), none have focused incorporation of total quality environmental management
on addressing the mining industry socio-environmental problems (TQEM) into IEM systems can help reduce suppliers' environ-
holistically and at the GSCM level. mental degradation (Maslen and Hopkins, 2014; Lee et al., 2014).
(5) Eco-Innovation Practices (ECO)
2.1.1. A green supply chain management evaluation framework for Eco-innovation may be novel systems to an organization and re-
the mining industry sult in environmental risk and resource use burden reduction
A proposed comprehensive and integrative green supply chain throughout the operational life-cycle (Bocken et al., 2014). By-
practices and sub-practices framework for the mining industry has products from mining operations can be transformed into useable
been developed (Kusi-Sarpong et al., 2015). The framework fo- materials and feedback into operations through eco-innovative
cused on six practices and thirty sub-practices. These practices approaches (Lutandula and Maloba, 2013). Substituting chemicals
include Green Information Technology and Systems, Strategic for gold recovery can reduce negative environmental conse-
Supplier Partnership, Operations and Logistics Integration, Internal quences and risk (Azevedo et al., 2012). Mining companies can
Environmental Management, Eco-innovative Practices and End-of- modify their processing plant by shifting from “dirty” to cleaner
life Practices. This framework is evaluated in this study. technology to improve efficiency of mineral recovery and bypro-
Summarizing: duct values and use of resources and fewer inputs, and represent
eco-innovations (Azevedo et al., 2012; Voigt et al., 2014).
(1) Green Information Technology and Systems (GITS) (6) End-of-Life Practices (EOL)
Information Technology and Systems (ITS) are an important End-of-life initiatives can help reduce life cycle environmental
avenue to drive environmental footprints and sustainable burdens of materials (Cucchiella et al., 2014; Wang and
Table 1
GSCM practices (factors) and their sub-factors in the mining industry.
Source: Kusi-Sarpong et al. (2015).
No. GSCM Factors and Sub-factors Literature
1 Green
Information Technology and Systems (GITS)
Sub-factors GITS1Use of energy efficient hardware and data centers Watson et al. (2008), Jenkin et al. (2011), Chou and Chou (2012), Setterstrom (2008),
GITS2Consolidating servers using virtualization software Sarkis and Zhu (2008), Wagner et al. (2009) and Uddin and Rahman (2012)
GITS3Reducing waste associated with obsolete equipment
GITS4Collaborative group software and telepresence
systems
GITS5 Eco-labeling of IT products
2 Strategic Suppliers Partnership (SSP)
Sub-factors SSP1 Jointly develop environmental management solutions Vachon et al. (2001), Rao (2002), Geffen and Rothenberg (2000), Simpson and Power
SSP2 Jointly build programs to reduce or eliminate ma- (2005) and Simpson et al. (2007)
terials use
SSP3 Share environmental management techniques and
knowledge
SSP4 Collaborate with suppliers to manage reverse flows
of materials and packaging
SSP5 Communicate goals of sustainability to suppliers
SSP6 Monitor environmental compliance status and
practices of supplier”s operations
3 Operations and Logistics Integration (OLI)
Sub-Factors OLI1 Lean and green operations Kleindorfer et al. (2005), Hajmohammed et al. (2013), Vachon (2007), Wee and Quazi
OLI2 Process redesign to reduce use of scarce or toxic (2005), Min and Galle (2001), Carter and Easton (2011) and Zsidisin and Hendrick (1998)
resources and energy consumption
OLI3 Community/environmental, employee health and
safety concerns
OLI4 Internal process integration and production
automation
4 Internal Environmental Management (IEM)
Sub-factors IEM1 Total quality environment management Vachon and Klassen (2008), Min and Gall (2001), Azevedo et al. (2012), Simpson et al.
IEM2 Environmental compliance monitoring and auditing (2007), Vachon and Klassen (2006a) and Baram and Partan (1990)
IEM3 Pollution prevention plans
IEM4 Environmental manager and training for employees
IEM5 Environmental standards/ISO14001 certification by
suppliers
IEM6 Employee incentive programs for environmental
suggestions
5 Eco-Innovation practices(ECO)
Sub-Factors ECO1 Substituting toxic inputs with environmentally Carter and Easton (2011), Vachon (2013), Azevedo et al. (2012), Paulraj (2009) and Rao
friendly ones and Holt (2005)
ECO2 Use of fewer inputs to minimize the environmental
risks and impacts
ECO3 Switching from “dirty” to cleaner technologies
ECO4 Internal recycling of inputs, materials and wastes
6 End-of-Life practices (EOL)
Sub-factors EOL1 Resale of used parts or components Stock (2001), Sarkis (2003), Rogers and Tibben-Lembke (2001) and Bell et al. (2013)
EOL2 Recondition and refurbishing of used parts or
components
EOL3 Old/obsolete items being replaced
EOL4 Cyanide and arsenic solution recovery and carbon
regeneration
EOL5 Mining of Tailings
Gaustad, 2012). Mining machinery maintenance produces GSCM implementation offers many important economic ben-
significant wastes (worn-out parts/components) which can efits (Lee et al., 2014; Govindan et al., 2014b, 2015; Dubey et al.,
be put through component exchange programs (reverse logis- 2015) which presents a “win-win” situation for the company and
tics) by returning them to suppliers in exchange of new parts/ the environment (Beckmann et al., 2014). GSCM practices reduce
components with little or no additional cost (Bell et al., 2013). both direct environmental cost and financial costs of the mines
These worn-out parts and components can be recaptured for including reduced environmental fines, reduced energy con-
value or proper disposal to avoid environmental impact sumption cost, reduced cost of material purchasing and improved
(Pishvaee et al., 2014; Li and Wu, 2014; Govindan and Po- tailings residual enabling reduced tailings facility cost. These in-
piuc, 2014). Managing carbon and tailings wastes at the end of
itiatives enhance resource efficiency which relates directly to
life lessens environmental burdens (Edraki et al., 2014).
economic performance (Zhang et al., 2012).
Mining companies can use GSCM practices to evaluate and
The major six factors and thirty sub-factors are summarized in
hence mitigate the impact of their operations on the environment.
Table 1.
The raison d’etre for GSCM is to help mines and industry achieve
better environmental performance. Eco-innovative initiatives can
2.1.2. GSCM implementation benefits and organizational sustainable help reduce solid/liquid waste and emissions mitigating environ-
performance outcomes mental risks and impacts associated with mines' operations.
Performance with a GSCM context may include economic, en- Social performance in the mining industry has been poor re-
vironmental and social performance. sulting in local mining communities and general public opposing
mining developments, questioning, and challenging mining com- and Cooper, 2012); India (Barve and Muduli, 2013; Luthra et al.,
panies to justify their existence and legitimacy (Muduli et al., 2015); Brazil (Gomes et al., 2014) and Spain (Vintró et al., 2014). No
2013; Ranängen et al., 2014). These stakeholder group concerns on studies have been completed in the Ghanaian mining industry
mining operations are linked to environmental and social health context. More specifically, studies into environmental (green)
issues resulting from unhealthy and unsafe practices (Mzembe and sustainability in Ghanaian mining industry produces an unclear
Meaton, 2014; Muduli et al., 2013), impacts on local land-use picture about the industry”s environmental impacts, as environ-
(mainly agricultural land) and lack of community engagement mental initiatives involving the supply chains are rare (e.g. Fei-
(Lawson and Bentil, 2014; Dare et al., 2014). GSCM practices can be Baffoe et al., 2013).
used to help address these socio-environmental issues and im- Fei-Baffoe et al. (2013) recently conducted a study involving
proving social performance. two large mining companies from Ghana to investigate the impact
of ISO 14001 environmental management systems (EMS) on key
2.2. MCDM methods for GSCM evaluation environmental performance indicators. According to the study,
implementation of ISO 14001 EMS by two gold mining companies
Green supply chain evaluation is a multi-criteria task involving result in environmental performance improvement, particularly in
conflicting choices requiring the support of MCDM tools. Many waste management, reported environmental incidents, and energy
researchers have utilized a variety of MCDM tools such as ANP and consumption. Segregation of waste was adopted in both compa-
fuzzy-ANP (Theißen and Spinler, 2014; Büyüközkan and Çifçi, nies to ensure appropriate disposal mechanisms to mitigate pol-
2012a), fuzzy-DEMATEL (Lin, 2013), AHP and fuzzy-AHP (Govindan lution. These are examples where Ghanaian mining company en-
et al., 2014b; Wang et al. 2012; Rostamy et al. 2013), fuzzy-DEA vironmental performance improvement are internally focused and
(Mirhedayatian et al., 2014), fuzzy-AHP-TOPSIS (Wang and Chan, cleanup initiatives. While these attempts are notable efforts, the
2013), fuzzy-Delphi-ANP (Tseng et al., 2015) and ANP-QFD-ZOGP capabilities developed may not fully address the broader ecologi-
(Jayakrishna et al., 2013) for investigating different kinds of green cal influence of supply chains in the mining industry. These loca-
initiatives. We build on this work by integrating fuzzy-DEMATEL lized and reactive environmental management initiatives focus on
and ANP for GSCM multiple criteria evaluation. expensive and remedial investments (Hilson and Nayee, 2002).
Mining in Ghana and specific Ghanaian company field study
2.2.1. Integrating DEMATEL and ANP discussion are further presented in Section 4.1.
The integration of DEMATEL and ANP has seen limited in-
vestigation, although they have natural linkages. In ANP, the cri-
teria within the system are compared through pair-wise compar-
3. Technical background
isons to determine their weights based on predetermined inter-
actions amongst criteria. Too many interactions amongst the cri-
This section presents some background information of the
teria can cause an inordinate amount of comparisons within ANP,
methodologies incorporated into the proposed hybrid multiple
making it difficult for the decision maker to handle, rendering ANP
criteria evaluation tool, fuzzy-DEMATEL and ANP.
less practical and causing decision maker fatigue due to the in-
teractive nature of ANP information elicitation. The focus of this
3.1. Overview of the various aspects of the proposed novel hybrid
integration with DEMATEL is on how these interactions and re-
methodology
lationships can be designed to reduce the volume of pair-wise
comparisons required by decision makers within ANP.
3.1.1. The Fuzzy-DEMATEL method
In the literature, there have been two main attempts to link
The DEMATEL method is a structured analytical tool used for
DEMATEL and ANP. However, these attempts and approaches have
causal mapping (Fontela and Gabus, 1976; Gabus and Fontela,
not really utilized DEMATEL networking results for ANP network
1973). DEMATEL uses graph theory to categorize attributes into
design. One approach transfers the normalized total-relation matrix
cause and effect groups (Senvar et al., 2014). The resulting di-
from the DEMATEL into the ANP inner dependencies part to form
the weighted super-matrix (e.g. Büyüközkan and Çifçi, 2012b). This graphs (directed graphs) from DEMATEL represent a conceptual
approach does not involve the decision-makers involvement in ANP, relationship among the elements in the system, with the strength
nor aids in improving ANP performance. The other approach mul- of influence identified (Miao et al., 2014; Liu et al., 2014; Patil and
tiplies the normalized total-relation matrix from the DEMATEL with Kant, 2014a). This methodology has been successfully applied in
the acquired ANP inner dependencies matrix to form the weighted various fields including emergency management (Li et al., 2014),
super-matrix (e.g. Huang et al., 2014). This approach also does not blogging (Hsu and Lee, 2014), knowledge management (Patil and
use DEMATEL to help achieve efficiencies in ANP. Kant, 2014a, 2014b), and green supply chain management (Wu
In summary, with this integrative approach the criteria re- and Chang, 2015).
lationships and interrelationships established from the DEMATEL The DEMATEL approach will utilize triangular fuzzy numbers
technique provide a structure for determining the factor inter- (Zadeh, 1965). Fuzzy numbers are convex fuzzy sets characterized
dependencies in an ANP network. Fewer interdependencies mean by a given interval of real numbers, with their grade of member-
exponential reduction in the number of relationships to be in- ship between 0 and 1 (Patil and Kant, 2014a, 2014b). This study
vestigated in the ANP stage. Thus, this procedure greatly reduces the adopts triangular fuzzy numbers (TFN) to obtain solutions from
number of pair-wise comparison to be evaluated in the ANP stage. the experts. TFN use three-values: the lowest possible value l , the
most promising value m and the upper possible value u. The TFN,
2.3. Environmental and green supply practices in the ghanaian membership function (μÃ (x)) can be defined by Eq. (1):
mining industry context ⎧ 0, x<l
⎪
Many studies have been completed on green supply chain ⎪ x − l , l < x≤m
⎪
management initiatives and programs implementation in the μÃ (x)=⎨ m − l
mining industry in both developed and developing nations. For ⎪ u − x , m < x≤u
⎪u−m
example, investigations have occurred in China (Haibin and ⎪
Zhenling, 2010; Si et al., 2010); Australia (Van Berkel, 2007; Giurco ⎩ 0, x>u (1)
where l, m and u are real numbers and l ≤ m ≤ u , and l, m and u important partners within the supply chain (Mbendi, 1995–2013).
are the lower, the mean and upper bounds of fuzzy number Ã , The Ghanaian mining industry has been perceived as a socio-
respectively. Let Ã1=(l1, m1, u1) and Ã2=(l2, m2 , u2 ) be two triangular environmentally disruptive industry (Peck and Sinding, 2003).
fuzzy numbers. The triangular fuzzy numbers mathematical op- Negative environmental impacts from the mining industry's sup-
erations are defined by expressions Eqs. ((2)–(6)) (Yu and Hu, ply chain operations are numerous and include toxic reagent re-
2010): leases, acid drainage, air quality reduction, habitat modification or
displacement and pollution (Wasylycia-Leis et al., 2014). Mineral
Ã1 ⊕ Ã2=( l1+l2,m1+m2,u1+u2 ) (2) extraction has resulted in severe and irreversible socio-environ-
mental damages and challenges such as resource degradation,
Ã1 ⊗ Ã2=( l1l2,m1m2,u1u2 ) (3) uprooting and displacement of communities. Environmental im-
pacts such as drinking water contamination and air pollution also
Ã1ΘÃ2=( l1−l2,m1−m2,u1−u2 ) (4) result, translating into serious health problems for residents
(Shandro et al., 2011).
These issues have challenged the mining companies’ license to
Ã1 ⎛ l1 m1 u1 ⎞
=⎜ , , ⎟ operate and legitimacy due to protestations of various socio-en-
Ã2 ⎝ l2 m2 u2 ⎠ (5)
vironmental advocacy groups at both local and international levels
(Bice, 2014; Moffat and Zhang, 2014; Owen and Kemp, 2013).
Ã1 ⊗ λ=( l1xλ , m1xλ , u1xλ ), λ≥0,λ ∈ R (6) These pressures have caused mining organizations to carefully
evaluate their direct and indirect environmental burdens and
Defuzzification to a crisp number is needed. Defuzzification
unsustainable consumption and production practices. Mine op-
takes into account the spread, height and shape of the triangular
erators have to work against a backdrop of a legacy of extensive
fuzzy numbers as imperative characteristics of the fuzzy number
and severe socio-environmental pollution issues associated with
(Cheng and Lin, 2002; Chang et al., 2011). The modified-CFCS
metal mining that have resulted from poor practices and non-
(Converting Fuzzy data into Crisp Scores) defuzzification will be
existent or non-enforced regulatory policy (Johnson, 2013). Fur-
used to identify a crisp value (Opricovic and Tzeng, 2003; Patil and
thermore, tailings generated from mineral processing streams (e.g.
Kant, 2014b). Details on the fuzzy DEMATEL approach will be
gold mines) can also contain large quantities of toxic substances,
presented in Section 4 in the field study application.
such as cyanides reagents and heavy metals, which can pose sig-
nificant human health and ecological risks (Adams, 2013; Kuyucak
3.1.2. The ANP method
and Akcil, 2013).
The ANP method is an extension of the analytical hierarchy
It has become imperative to strike a ‘win–win’ balance between
process (AHP) (Saaty, 1996). ANP does not just use a strict hier-
mineral and mining economic development and environmental
archical network like AHP (Aragonés-Beltrán et al., 2014). ANP is
protection, in this emerging economy. Investigating GSCM and its
capable of modeling interrelationships among the decision eche-
relationship in Ghana”s mining industry is important for both direct
lons and elements, which AHP does not do (Zaim et al., 2014;
and indirect environmental consumption and production sustain-
Meade and Sarkis, 1998; Wong et al., 2014). The major steps of ANP
ability improvement. This study on environmental sustainability
and its relationship with DEMATEL, as introduced in this paper, are
concerns in the mining sector and Ghanaian mining industry con-
described in the next section as part of the field study process.
text is meant to initially address serious negative environmental
consequences from supply chain and organizational operations of
the mining industry, especially in developing economies.
4. Field study of GSCM practices importance and performance
4.2. The proposed hybrid multi-criteria evaluation methodology
The two-stage fuzzy-DEMATEL and ANP methodology is de-
computational steps
tailed in this section with an application to a multiple case field
study in the Ghanaian mining industry. The proposed methodol-
To illustrate the applicability of the hybrid methodology using
ogy is used to assist practitioners in mining and other industries to
the GSCM practices framework, we employ real world multiple case
make strategic, in this case GSCM, decisions. Initially an overview
(field) studies using selected multi-national mining companies from
of the field study environment, Ghana”s mining industry, is pro-
Ghana. The application is in two phases. The first phase applies the
vided as background.
fuzzy-DEMATEL aspect of the model to obtain interrelationships and
influences within the GSCM practices and sub-practices. In the
4.1. The Ghanaian mining industry second phase ANP, using the interrelationships identified in the
DEMATEL step, is applied to determine the relative impact of the
Ghana was selected as the case country for this study because GSCM practices and sub-practices to organizational sustainable
of its unique mining industry positioning (Bloch and Owusu, 2012; performance (environmental, economic and social) based on input
Boon and Ababio, 2009). It consistently ranks as one of Africa”s top from the field study companies and managers.
producers of precious metals and minerals such as gold and dia- Phase 1: The fuzzy-DEMATEL methodology to identify sig-
monds. Mining and minerals contribute 5% of Ghana's GDP and nificant interrelationships. This phase has five major stages with a
37% of their total exports (Boon and Ababio, 2009). Gold, the main number of sub-steps. Each stage is now detailed.
focus of Ghana”s mining and mineral development industry con-
tributes over 90% of the total minerals exports (Aryee, 2001). Stage 1. Determine the decision goal and select participants. In
Ghana's mining industry has attracted nearly US$2billion of for- this step, the decision goal is set and the decision-making team is
eign direct investment (FDI) in both mineral exploration and mine formed. This stage used six selected3 managers, one each from the
development representing over 56% of the total FDI inflows six selected mining companies. Characteristics of the managers in
(Awudi, 2002). The country currently has twenty-three large-scale
mining companies producing gold, diamonds, bauxite and man- 3
Selection of the six mining companies was based on the high interest ex-
ganese, and, there are also over three hundred registered small hibited by these mining companies in greening their mining supply chains. Thus,
scale mining groups and ninety mine support service companies, purposive sampling was used in selecting the final six managers/companies.
this study are summarized in Table 2. The initial direct-relation matrix Z , using Table 5 data, with
crisp numbers for manager 1 and major factors is given in Table 6.
Stage 2. Develop the evaluation model and design fuzzy evalua- Step 3.3.5: Aggregate direct-relation crisp matrices and normalize. All
tion scale. The framework previously introduced in Sections 2.1.1 decision makers’ direct-relation crisp matrices are then aggregated into
and 2.1.2, and depicted in Fig. 1 is adopted. A five-point mea- a single (average) overall crisp direct-relation matrix using Eq. (14):
surement scale ranging from (N) ‘no influence’ to (VH) ‘very high
influence’ is developed for managerial perceptions on influences 1 1 2
amongst factors using pair-wise comparisons. Triangular fuzzy
Zij=
L
(Zij +Zij +… + ZijL ) (14)
number assignments for the linguistic values are shown in Table 3. Then the generalized direct-relation matrix P can be obtained
by normalizing the aggregated direct-relation matrix using Eqs.
Stage 3. Determine causal relationship using fuzzy-based DEMATEL
(15) and (16):
Step 3.1: Linguistic-based DEMATEL survey questionnaire design
1
and pilot testing Let v = n where v > 0
At this stage, a survey questionnaire involving various pair-wise max1 ≤ i ≤ n ∑ j fij , (15)
comparisons is developed and further pilot tested with feedback
incorporated into the questionnaire. and, P = v. Z (16)
Step 3.2: Construct pair-wise comparison matrix for DEMATEL
Given L experts and n factors a pair-wise comparison matrix is The generalized direct-relation matrix P , using Table 6 data for
developed. The influence of factor i compared to factor j is ob- all decision makers’ and major factors, is given in Table 7.
tained from decision maker perceptions. Pair-wise comparisons Step 3.4: Compute the total relation matrix. The total-relation
between any two factors are denoted by fij . The result is an n x n matrix T is determined using Eq. (17), where I represents an
identity matrix
non-negative direct relation matrix Ur ¼ [f ijr ] nxn, with 1 ≤ r ≤ L.
Thus, U1, U2,…UL are response matrices for each of the L experts (
T = P1+P 2+…P m =P (I − P )−1 ) (17)
and each element of U r is a linguistic value denoted by f ijr . The
Step 3.5: Compute Cause/Effect and Prominence of Factors
diagonal elements of each response matrix U r are set to zero.
Using the total-relation matrix T , get to the total row ( Ri ) and
The six selected managers were emailed the questionnaire. The
column ( Cj ) sum using Eqs (18) and (19):
direct-relation matrices are populated with linguistic variables, see
Table 4 for manager 1 linguistic responses to the major factors.
Then the linguistic numbers are replaced by triangular fuzzy
Ri = ( ∑ t ) for each row i
n
j=1
ij
(18)
numbers for the linguistic variables, see Table 5 (matrix A ).
n
Step 3.3: Defuzzify direct-relation matrix into crisp numbers
Let f ijr =(lijr , mijr , uijr ), be an equivalent triangular fuzzy number
Cj = (∑ i=1
tij ) for each column j (19)
for the level of influence of factor i on factor j for expert r rating To determine prominence Pi and net cause/effect Ei of the
with 1 ≤ r ≤ L . Then, the modified-CFCS defuzzification method, factors, use Eqs. (20) and (21);
using (Eqs. (7)–13), is applied to get crisp numbers.
Step 3.3.1: Normalize upper (xu), mean (xm) and lower (xl) fuzzy {
Pi= Cj + Ri /i=j } (20)
numbers:
(
xuijr = uijr −minlijr /∆max
min ) (7)
{
Ei= Cj − Ri /i=j } (21)
A graph can then be plotted by mapping the (Cj+Ri , Cj−Ri ) da-

taset, where prominence Pi represents the X-axis and the net ef-
(
xmijr = mijr −minlijr /∆max
min ) (8) fect Ei represents the Y-axis.
The total-relation matrix T , the row Ri and the column Cj , and
the prominence Pi and the net cause/effect Ei for all major factors,
using Table 7 data for major factors, are shown in Table 8.
(
xlijr = lijr −minlijr /∆max
min) (9)
Step 3.6: Set a threshold value and obtain the network relation-
where ∆max r r
min =maxuij − minlij ,
ship map (NRM). A threshold value β is set to filter and select the
r
maxuij ¼the maximum upper value amongst the upper bound relationships in matrix T with values above the threshold value. A
of fuzzy number values for expert r , and minlijr ¼ the minimum network relationship map (NRM) is determined for those re-
lower value amongst the lower bound of fuzzy number values for lationships that meet or exceed the threshold value. A threshold
expert r . value of the mean4 of all values in the T matrix is set and agreed
Step 3.3.2: Compute upper (xus) and lower (xls) normalized va- upon by the decision makers.
lues: Table 9 (matrix Tβ ) shows the relationships whose values are
greater than calculated threshold of 0.903. Fig. 2 shows the net-
(
xusijr =xuijr / 1 + xuijr −xmijr ) (10) work relationship map, using data from Table 9, for the major
practices. A Tβ is developed for major factors and sub-factors. It is
these factor/sub-factor relationships and interrelationships that
(
xlsijr =xmijr / 1 + xmijr −xlijr ) (11)
will be evaluated by ANP in the next phase of the methodology.
Step 3.3.3: Compute total normalized crisp values (x): Phase 2: Determine the relative perceived impact of factors on
xijr =⎡⎣ xlsijr ( 1 − xlsijr )+xusijr ⁎xusijr ⎤⎦/⎡⎣ 1 − xlsijr +xusijr ⎤⎦ (12) 4
The threshold values can be raised or lowered depending on how many in-
terrelationships are viewed as useful for ANP analysis by analysts. For example,
Step 3.3.4: Compute crisp values (Z):
setting a threshold of one or two standard deviations above the mean can also be
utilized, but that would mean fewer interrelationships from the DEMATEL analysis
(
Zijr =min lijr + xijr *∆max
min ) (13) for ANP analysis.
Table 2
Characteristics of the sixteen mining industry managers and their companies involved in the study.
The Six (6) mining industry managers involved in the DEMATEL method
Manager 1 and Company 1 Manager 4 and Company 4
Position: Supply Manager Position: Assistant Supply Chain Manager
Role: Management of sourcing/procurement, contract & warehouse Role: Management of sourcing/procurement, contract & warehouse
Number of Mining Working Years: 19years Number of Mining Working Years: 10years
Position: Local Supplier & Contractor Development Reg. Manager Position: Commercial Business Optimization Assistant Manager
Role: Develops & monitors local suppliers and contractors capacity Role: Commercial (supply, account & admin) business improvement
Number of Mining Working Years: 15years Number of Mining Working Years:11years
Position: Environmental Manager Position: Senior Procurement Manager
Role: Env’tal program implementations and compliance monitoring Role: Procurement & contract program implementation & training
The ten (10) mining industry managers involved in the ANP method
Position: Supply Manager Position: Assistant Supply Chain Manager
Role: Management of sourcing/procurement, contract & warehouse Role: Management of sourcing/procurement, contract & warehouse
Number of Mining Working Years: 19 Years Number of Mining Working Years: 10years
Position: Finance Manager Position: Commercial Business Optimization Assistant Manager
Role: Management of the company”s financial account and budgetary Role: Commercial (supply, account & admin) business improvement
Number of Mining Working Years: 10 Years Number of Mining Working Years:11years
Position: West Africa Regional Contract Manager Position: Head of Information Communications & Technology-ICT
Role: General management of contracts across the West Africa region Role: ICT program implementation, monitoring & improvement
Position: Parts and Warehouse Manager Position: Senior Procurement Manager
Role: Management of sourcing/procurement, contract & warehouse Role: Procurement & contract program implementation & training
Position: Senior Maintenance Planning Engineer Position: Assistant Environmental Manager
Role: Planning of maintenance and materials for maintenance activities Role: Env'tal program implementations, monitoring and improvement
The six (6) purposively sampled mining companies interested in greening their operations
Company 1 Company 4
Size: 2.1million ounces per year with workforce size of 246 Size: 2.2million ounces per year with workforce size of 700
Age: 4yearsþ Age: 4years
Type of Minerals: Gold Type of Minerals: Gold
Stock listings: TSX(EDV), ASX(EVR) & OTCQX(EDVMF) Stock listings: ASX/TSX (PRU)
Company 2 Company 5
Size: 13.3 million tonnes per year with workforce size of 3500 Size: 3.5 million tonnes per year with workforce size of 1670
Age: 21years Age: 11years
Stock listings: JSE Ltd, NYSE, NASDAO DUBAI, NYX & SWX Stock listings: TSE/NYSE
Company 3 Company 6
Size: 7.5 million tonnes ounces yearly with workforce size of 8539 Size: 2.7 million tonnes per year with workforce size of 700
Age: 9years Age: 15years
Stock listings: NYSE (NEM) Stock listings: TSX (GSC), NYSE (GSS), & GSE (GSR)
Organizational Sustainable Performance
Economic Environmental Social Overall

Goal
Performance Performance Performance
GITS SSP OLI IEM ECO EOL Major

Practices
GITS1
GITS2
GITS3
GITS4
GITS5
ECO1
ECO2
ECO3
ECO4
EOL2
EOL1
EOL3
EOL4
EOL5
EOL6
IEM1
IEM2
IEM3
IEM4
IEM5
IEM6
SSP1
SSP2
SSP3
SSP4
SSP5
SSP6
OLI1
OLI3
OLI4
OLI6
Sub Practices
Fig. 1. Decision structure for GSCM implementation.
organizational sustainability performance using ANP. importance weights on the overall organizational sustainability
goals.
Stage 6.. ANP is applied in four steps to determine factor
Step 6.1: Goal formulation decision structuring (interactions)
Table 3 Table 6
The General Linguistic Scales used in the Fuzzy-DEMATEL analysis. Initial direct-relation matrix Z of major factors after defuzzification for Manager1.
Linguistic terms Triangular fuzzy numbers Major factors GITS SSP OLI IEM ECO EOL
No influence (N) (0, 0, 0.25) GITS 0.000 0.500 0.033 0.267 0.267 0.033
Very low influence (VL) (0, 0.25, 0.50) SSP 0.733 0.000 0.733 0.500 0.733 0.733
Low influence (L) (0.25, 0.50, 0.75) OLI 0.267 0.500 0.000 0.267 0.500 0.500
High influence (H) (0.50, 0.75, 1.00) IEM 0.267 0.500 0.033 0.000 0.500 0.033
Very high influence (VH) (0.75, 1.00, 1.00) ECO 0.033 0.500 0.500 0.733 0.000 0.733
EOL 0.500 0.733 0.500 0.733 0.733 0.000
Table 4
Initial direct-relation linguistic matrix U of major factors from Manager1. Table 7
The generalized direct-relation matrix P for major factors across All Managers.
Major factors GITS SSP OLI IEM ECO EOL
GITS 0 L N VL VL N
SSP H 0 H L H H GITS 0.000 0.191 0.152 0.163 0.152 0.102
OLI VL L 0 VL L L SSP 0.211 0.000 0.202 0.182 0.202 0.202
IEM VL L N 0 L N OLI 0.172 0.192 0.000 0.132 0.142 0.182
ECO N L L H 0 H IEM 0.162 0.142 0.101 0.000 0.182 0.162
EOL L H L H H 0 ECO 0.152 0.132 0.142 0.222 0.000 0.192
EOL 0.112 0.192 0.152 0.231 0.212 0.000
Table 5
Initial direct-relation matrix A of major factors with triangular fuzzy numbers.
GITS 0 (0.25, 0.50, 0.75) (0,0,0.25) (0,0.25,0.50) (0,0.25,0.50) (0,0,0.25)

SSP (0.50, 0.75, 1.00) 0 (0.50,0.75,1.00) (0.25, 0.50, 0.75) (0.50,0.75,1.00) (0.50,0.75,1.00)
OLI (0, 0.25, 0.50) (0.25, 0.50, 0.75) 0 (0,0.25,0.50) (0.25, 0.50, 0.75) (0.25, 0.50, 0.75)
IEM (0, 0.25, 0.50) (0.25, 0.50, 0.75) (0,0,0.25) 0 (0.25, 0.50, 0.75) (0,0,0.25)
ECO (0, 0, 0.25) (0.25, 0.50, 0.75) (0.25, 0.50, 0.75) (0.50,0.75,1.00) 0 (0.50,0.75,1.00)
EOL (0.25, 0.50, 0.75) (0.50,0.75,1.00) (0.25, 0.50, 0.75) (0.50,0.75,1.00) (0.50,0.75,1.00) 0
using the total-relation matrices Tβ from DEMATEL. This step requires managers with their characteristics and companies given in Ta-
a clearly defined goal and the formation of a decision structure. The ble 2. These pair-wise comparisons were rated using the re-
goal for this paper is identifying the perceived impact of major GSCM commended 9-point (1–9) measurement scale as shown in Ta-
practices and sub-practices on organizational sustainability perfor- ble 10 (Saaty, 1996).
mance (environmental, economic and social). The criteria within the Step 6.3: Compute local priority vectors of factors and, form un-
system are compared through pair-wise comparisons to determine weighted and weighted (limiting) super-matrix. Given that A is a
weights through ANP. It is important to most effectively design the pair-wise comparisons matrix, the priority vectors/relative im-
decision interactions within ANP. Too many interactions can cause an portance weights wi can be computed using Eq. (22):
inordinate amount of comparisons within ANP, causing decision
Awi=λmax wi (22)
maker fatigue making ANP less practical, too few and a full interac-
tion effect may be lost. The DEMATEL steps in the previous section where λmax is the largest eigen-value of A . Many authors have
help to reduce the number of factor interdependencies for evalua- proposed several algorithms to approximate the wi value (Saaty and
tion. Using DEMATEL in this way aids ANP to be more practical and Takizawa, 1986; Saaty and Hu, 1998; Meade and Sarkis, 1998; Saaty,
feasible to apply. Although DEMATEL and ANP have been used to- 2004). Several online-based multi-criteria decision-support soft-
gether previously (Büyüközkan and Çifçi, 2012b; Huang et al., 2014),
wares have also been designed to help compute these relative im-
none have explicitly used the two together to form a more efficient
portance weights including Web-HIPRE3þ(http://hipre.aalto.fi/)
decision network for ANP.
(Mustajoki and Hamalainen, 2000) and Super-Decisions (http://
Table 9 is representative of the DEMATEL output used for factor
www.superdecisions.com/).
interactions within ANP. The identified interrelationships are used
The relative importance weights determined from the various
to develop an ANP pair-wise comparison sub-matrix. Sub-matrices
pair-wise comparisons matrices are used to construct the un-
for each of the sub-factor interrelationships are determined in the
weighted super-matrix. The un-weighted super-matrix is formed
same way, but are not shown here.
as a partitioned matrix involving various sub-matrices modeling
Step 6.2: Design survey questionnaire and conduct pair-wise
the factor interrelationships. The un-weighted super-matrix needs
comparison. Once the interdependencies within the major and
to be made column stochastic. The super-matrix is then raised to
sub-factors clusters have effectively been determined, the data
the power of 2k , where k is a large number to converge and arrive
gathering can commence. A survey questionnaire using the ma-
at a long-term stable set of weights.
trices was developed and administered.
Super-Decisions software can generate the un-weighted and
The survey questionnaire was emailed to ten selected5
limiting super matrices after inputting the relative importance
weights. Table 11 shows Manager 1”s final converged super-matrix
5
Purposive sampling was used in selecting these 10 managers. Individually
emails were sent to two managers from each of the six selected mining companies
requesting their participation in the ANP survey and providing them with full in- (footnote continued)
formation on the objective of the research. Two of the managers from two different mining companies declined.
Table 8
The total-relation matrix T and prominence and net cause/effect for major factors across all managers.
Major factors GITS SSP OLI IEM ECO EOL R Prominence Net cause/effect
C þR CR
GITS 0.692 0.876 0.776 0.919 0.884 0.809 5.214 10.170 0.258
SSP 1.044 0.903 0.980 1.139 1.117 1.069 5.411 11.663 0.842
OLI 0.887 0.931 0.692 0.955 0.933 0.920 4.864 10.182 0.453
IEM 0.817 0.828 0.727 0.772 0.897 0.842 5.909 10.792 1.026
ECO 0.875 0.889 0.817 1.029 0.815 0.933 5.692 11.050 0.335
EOL 0.899 0.984 0.873 1.094 1.047 0.827 5.399 11.123 0.324
C 4.956 6.252 5.318 4.883 5.358 5.724
Table 9 Table 10
Matrix Tβ with established relations above threshold value from the total-relation General Linguistic Scale use for the ANP analysis.
matrix T.
Linguistic terms Numerical-rating
Extremely More Important (EM) 9
GITS 0.000 0.000 0.000 0.919a 0.000 0.000 Very Much More Important (VM) 7
SSP 1.044a 0.000 0.980a 1.139a 1.117a 1.069a More Important (M) 5
OLI 0.000 0.931a 0.000 0.955a 0.933a 0.920a Moderately More Important (MM) 3
IEM 0.000 0.000 0.000 0.000 0.000 0.000 Same important (S) 1
ECO 0.000 0.000 0.000 1.029a 0.000 0.933a Moderately Less Important (ML) 1/3
EOL 0.000 0.984a 0.000 1.094a 1.047a 0.000 Less Important (L) 1/5
Very Much Less Important (VL) 1/7
Threshold value¼ 0.903. Extremely Less Important (EL) 1/9
a
Values above the threshold value.
for the major GSCM factor. Table 12 averages the weights of all Table 11
managers for major GSCM factors. Limiting super-matrix with stable weights for manager 1.
Step 6.4: Identification and Selection of best factor influence on the Major ECO EOL GITS IEM OLI SSP
overall organizational sustainability performance. A desirability in-
dex table is used to determine the final aggregation of factor and ECO 0.176972 0.176972 0.176972 0.176972 0.176972 0.176972
EOL 0.257419 0.257419 0.257419 0.257419 0.257419 0.257419
sub-factor weights (local weights when separate) into a single
GITS 0 0 0 0 0 0
numeric score (global weights when aggregated). The greater the IEM 0 0 0 0 0 0
index value, the more important the factor. OLI 0.243619 0.243619 0.243619 0.243619 0.243619 0.243619
Due to the preponderance of zero weighted values from the SSP 0.32199 0.32199 0.32199 0.32199 0.32199 0.32199
previous super-matrix stages, three aggregation models are used:
multiplicative, additive and exponential (multiplicative) powers.
The purpose was to compare and analyze the sensitivity of the global weight is zero. For this aggregation the penalty is very
ANP desirability matrix methodology whether zero values can severe because of a lack of an interrelationship and is reflected
cause a very different ranking. in the overall desirability index value (Natoli and Zuhair,
2011). Although the multiplicative approach is the most
(1) Multiplicative model: In this aggregation, the factor importance popular approach, two other techniques, the additive and
is evaluated by simply multiplying the weights associated exponential (power) multiplication, can provide a more ba-
with each factor/sub-factor. In this situation if any of the factor lanced results from aggregation.
or sub-factor importance weights is equal to zero, then the (2) Additive model: In this aggregation, the factor importance is
1
SSP
0.5
OLI
EOL
0
ECO
C-R
GITIS
-0.5
-1
IEM
-1.5
10 10.2 10.4 10.6 10.8 11 11.2 11.4 11.6 11.8 12
C+R
Fig. 2. Network Relationship Map (NRM)/diagraphs of the GSCM major practices.

Table 12
Group aggregated (mean) weights for major factors cluster interdependencies.
Factors Mgr1 Wght Mgr2 Wght Mgr3 Wght Mgr4 Wght Mgr5 Wght Mgr6 Wght Mgr7 Wght Mgr8 Wght Mgr9 Wght Mgr10 Wght Mean
GITS 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
SSP 0.322 0.404 0.389 0.389 0.413 0.372 0.418 0.403 0.278 0.396 0.378
OLI 0.244 0.378 0.375 0.375 0.500 0.403 0.496 0.508 0.280 0.495 0.405
IEM 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
ECO 0.177 0.038 0.057 0.057 0.014 0.077 0.007 0.012 0.164 0.032 0.063
EOL 0.257 0.180 0.179 0.179 0.074 0.148 0.080 0.077 0.278 0.078 0.153
simply the sum of the weights associated with each factor/ i is the index for the global relative importance desirability indices.
sub-factor. The additive aggregation model allows for a more e is the exponential base value.
balanced inclusion of poor or lower value individual evaluat- An example calculation for each aggregation technique is now
ing criterion (Munda and Nardo, 2005; Nardo et al., 2008). shown for the SSP/SSP1 sub-factor (factor, j ¼2, sub-factor, k ¼6
Less important weights of any of the evaluating factor/sub- and organizational sustainability dimensions, l = 1,2,3) in Table 13
factor will result in relatively less sensitivity in the overall (the italicized and bolded values).
importance of an alternative. Thus, for the additive aggrega-
tion model, there exists better substitution between the (1) Multiplicative model:
individual evaluating criteria influencing the overall desirabil- M
i1,6 = U1D P2,1
D I
P2,1 D
A6,2 I
A6,2 = 0.7144 × 0.2462 × 0.3784 × 0.3492 ×
ity index value for an alternative (Natoli and Zuhair, 2011).
(3) Multiplicative Exponential-weighting model: In this aggregation, 0.3175 = 0.00738
the importance weight is raised to a power, in this case with
M
an exponential base. For this aggregation a complete nullifi- i2,6 = U2D P2,2
D I
P2,2 D
A6,2 I
A6,2 = 0.2027 × 0.1854 × 0.3784 × 0.3492 ×
cation of a factor and associated sub-factors does not occur 0.3175 = 0.00158
due to lack of network interdependencies after application of
M
the DEMATEL interrelationship evaluation. i3,6 = U3D P2,3
D I
P2,3 D
A6,2 I
A6,2 = 0.0829 × 0.2003 × 0.3784 × 0.3492 ×
0.3175 = 0.00070
The global relative importance desirability indices i of the orga-
nizational sustainability sub-factor k for aggregation method z is Sustainability i 6M =(i1,6
M M
+i2,6 M
+i3,6 ) /3 ¼ (0.00738þ 0.00158 þ
denoted as ikz and are computed in two stages. A multiplicative ilkM , 0.00070)/3 ¼0.00322 (column 13)
(2) Additive model:
additive ilkA and exponential multiplicative power ilkE are determined
A
for each factor using (Eqs. (23)–25). Then, the relative factor im- i1,6 = U1D + P2,1
D I
+ P2,1 D
+ A6,2 I
+ A6,2
portance for each of the individual organizational sustainability di-
= 0.7144 + 0.2462 + 0.3784 + 0.3492 + 0.3175 = 2.00569
mensions results under each of the three aggregation models are
further averaged to determine the global relative importance desir- A
i2,6 = U2D + P2,2
D I
+ P2,2 D
+ A6,2 I
+ A6,2
ability indices i of the organizational sustainability sub-factor k for
aggregation method z , ikz . The second stage can be ignored should = 0.2027 + 0.1854 + 0.3784 + 0.3492 + 0.3175 = 1.43318
there be only one organizational sustainability dimension. A
i3,6 = U3D + P2,3
D I
+ P2,3 D
+ A6,2 I
+ A6,2
D D
ilkM =UlD Pjl P Ijl A AkjI (23) = 0.0829 + 0.2003 + 0.3784 + 0.3492 + 0.3175 = 1.32827
kj
Sustainability i 6A=(i1,6
A A
+i2,6 A
+i3,6 ) /3 ¼ (2.00569 þ 1.43318 þ
ilkA =UlD +
D
Pjl +P jlI +
D
A +AkjI 1.32827)/3 ¼ 1.58905 (column 15)
kj (24)
(3) Multiplicative Exponential-weighting model:
= e( U1 + P 2,1+ P 2,1+ A6,2 + A6,2 ) = e (0.7144 + 0.2462 + 0.3784 + 0.3492 + 0.3175)

E D D I D I
ilkE=e (Ul
D+ P D+ P I + A D + A I
jl jl kj kj ) (25)
i1,6
= 7.43122
UlD represents the relative importance weight of organizational
sustainability dimension l as part of the hierarchical (D) relationship.
= e( U2 + P 2,2+ P 2,2+ A6,2 + A6,2 ) = e (0.2027 + 0.1854 + 0.3784 + 0.3492 + 0.3175)
E D D I D I
i2,6
P Djl represents the relative importance weight for major factor j
of organizational sustainability dimension l for the hierarchical (D) = 4.19199
relationship.
P Ijl represents the stable relative importance weight for the = e( U3 + P 2,3+ P 2,3+ A6,2 + A6,2 ) = e (0.0829 + 0.2003 + 0.3784 + 0.3492 + 0.3175)
E D D I D I
i3,6
major factor j of organizational sustainability dimension l for the
= 3.77452
interdependent (I) relationship.
AkjD represents the relative importance weight for sub-factor k Sustainability
of major factor j for the hierarchical (D) relationship. i 6E =(i1,6
E E
+i2,6 E
+i3,6 ) /3 ¼(7.43122þ4.19199 þ3.77452)/3 ¼5.13258
AkjI represents the stable relative importance weight of sub- (column 17)
factor k of major factor j for interdependency (I) relationship. Table 13 summarizes all the results and rankings.
L is the index set for organizational sustainability dimensions
where l = 1,2,3, respectively representing economic, environ-
mental, and social sustainability dimensions. 5. Discussion of results
J is the index set for the major factors where j = 1…6.
K is the index set for the sub-factors where k=1…30. In this section the results of the fuzzy-DEMATEL are first dis-
z is the index set for the aggregated methods where cussed, followed by discussion of the integrated DEMATEL-ANP
z = multiplicative, M , additive A and multiplicative exponential E methodology then an overview of managerial inputs.
336
Table 13
General aggregation desirability index table.
Goal Local Local Local Dimensions Local Local Local local Sub- Local Local Final global Rank Final global Rank Final global Rank
weights- weights- weights- weights weights – weights - stable practices weights - stable weights - over- weights - weights -
econ env sust. social – major major major weights - sub-fac- weights all sustain- overall sus- overall sus-
sust. dim dim on sust. dim factors factors factors major tors on - sub- ability multi- tainability ad- tainability ex-
on goal goal U2D on goal on econ on en- on social practices major factors plicative ikM ditive ikA ponential ikE
U1D U3D dim P D viron. dim P D P Ijl factors I
Akj
j1 j3
S. Kusi-Sarpong et al. / Int. J. Production Economics 181 (2016) 325–341

dim P Dj2
D
Akj
Sustainability 0.7144 0.2027 0.0829 GITS 0.3004 0.2995 0.2017 0.0000 GITS1 0.4118 0.1117 0.00000 1.12398 14 3.22961 13
0.7144 0.2027 0.0829 0.3004 0.2995 0.2017 0.0000 GITS2 0.1600 0.3785 0.00000 1.13906 11 3.27867 10
0.7144 0.2027 0.0829 0.3004 0.2995 0.2017 0.0000 GITS3 0.1579 0.1999 0.00000 0.95833 21 2.73658 21
0.7144 0.2027 0.0829 0.3004 0.2995 0.2017 0.0000 GITS4 0.1406 0.3098 0.00000 1.05093 17 3.00209 16
0.7144 0.2027 0.0829 0.3004 0.2995 0.2017 0.0000 GITS5 0.1298 0.0000 0.00000 0.73031 26 2.17862 26
0.7144 0.2027 0.0829 SSP 0.2462 0.1854 0.2003 0.3784 SSP1 0.3492 0.3175 0.00322 1 1.58905 2 5.13258 2
0.7144 0.2027 0.0829 0.2462 0.1854 0.2003 0.3784 SSP2 0.1817 0.1807 0.00095 4 1.28468 7 3.78574 7
0.7144 0.2027 0.0829 0.2462 0.1854 0.2003 0.3784 SSP3 0.1467 0.1025 0.00044 8 1.17153 9 3.38071 9
0.7144 0.2027 0.0829 0.2462 0.1854 0.2003 0.3784 SSP4 0.1068 0.2795 0.00087 5 1.30865 6 3.87759 5
0.7144 0.2027 0.0829 0.2462 0.1854 0.2003 0.3784 SSP5 0.0967 0.1197 0.00034 10 1.13877 12 3.27176 11
0.7144 0.2027 0.0829 0.2462 0.1854 0.2003 0.3784 SSP6 0.1190 0.0000 0.00000 1.04128 19 2.96785 18
0.7144 0.2027 0.0829 OLI 0.1300 0.1496 0.1292 0.4051 OLI1 0.3820 0.3740 0.00258 2 1.63076 1 5.30697 1
0.7144 0.2027 0.0829 0.1300 0.1496 0.1292 0.4051 OLI2 0.1769 0.3730 0.00119 3 1.42460 3 4.31832 3
0.7144 0.2027 0.0829 0.1300 0.1496 0.1292 0.4051 OLI3 0.2524 0.0000 0.00000 1.12717 13 3.20732 14
0.7144 0.2027 0.0829 0.1300 0.1496 0.1292 0.4051 OLI4 0.1887 0.2530 0.00086 6 1.31637 5 3.87534 6
0.7144 0.2027 0.0829 IEM 0.1084 0.1393 0.1370 0.0000 IEM1 0.3950 0.2878 0.00000 1.14435 10 3.25318 12
0.7144 0.2027 0.0829 0.1084 0.1393 0.1370 0.0000 IEM2 0.2135 0.2547 0.00000 0.92980 22 2.62498 22
0.7144 0.2027 0.0829 0.1084 0.1393 0.1370 0.0000 IEM3 0.1420 0.1845 0.00000 0.78807 24 2.27811 24
0.7144 0.2027 0.0829 0.1084 0.1393 0.1370 0.0000 IEM4 0.0942 0.1504 0.00000 0.70618 27 2.09900 27
0.7144 0.2027 0.0829 0.1084 0.1393 0.1370 0.0000 IEM5 0.0900 0.1224 0.00000 0.67398 29 2.03249 29
0.7144 0.2027 0.0829 0.1084 0.1393 0.1370 0.0000 IEM6 0.0653 0.0000 0.00000 0.52683 30 1.75437 30
0.7144 0.2027 0.0829 ECO 0.0937 0.1164 0.1823 0.0634 ECO1 0.5247 0.3507 0.00041 9 1.40288 4 4.19588 4
0.7144 0.2027 0.0829 0.0937 0.1164 0.1823 0.0634 ECO2 0.1647 0.0000 0.00000 0.69221 28 2.06150 28
0.7144 0.2027 0.0829 0.0937 0.1164 0.1823 0.0634 ECO3 0.1629 0.4193 0.00015 14 1.10973 15 3.12977 15
0.7144 0.2027 0.0829 0.0937 0.1164 0.1823 0.0634 ECO4 0.1478 0.2300 0.00008 15 0.90538 23 2.55130 23
0.7144 0.2027 0.0829 EOL 0.1213 0.1098 0.1496 0.1531 EOL1 0.3164 0.2723 0.00053 7 1.20212 8 3.45434 8
0.7144 0.2027 0.0829 0.1213 0.1098 0.1496 0.1531 EOL2 0.1584 0.2800 0.00027 12 1.05173 16 2.97201 17
0.7144 0.2027 0.0829 0.1213 0.1098 0.1496 0.1531 EOL3 0.1633 0.2723 0.00028 11 1.04898 18 2.96385 19
0.7144 0.2027 0.0829 0.1213 0.1098 0.1496 0.1531 EOL4 0.1660 0.0000 0.00000 0.77939 25 2.26345 25
0.7144 0.2027 0.0829 0.1213 0.1098 0.1496 0.1531 EOL5 0.1958 0.1754 0.00021 13 0.98448 20 2.77871 20
Table 14
Respective GSCM sub-factors rankings from the three desirability and ranking models.
Technique Ranks
1 2 3 4 5 6 7 8 9 10
MW SSP1 OLI1 OLI2 SSP2 SSP4 OLI4 EOL1 SSP3 ECO1 SSP5
AW OLI1 SSP1 OLI2 ECO1 OLI4 SSP4 SSP2 EOL1 SSP3 IEM1
MEW OLI1 SSP1 OLI2 ECO1 SSP4 OLI4 SSP2 EOL1 SSP3 GITS2
MW: Multiplicative Weighting Model; AW: Additive Weighting Model; MEW: Multiplicative Exponential Weighting Model.
5.1. The fuzzy-DEMATEL results – causal relationship (relative re- 5.2. ANP results – weight measurements for ranking the best criteria
lationship and influences) on the overall goal
The results of the fuzzy-DEMATEL (Fig. 2 and Table 8) provide Table 13 columns ilkM , ilkA and ilkE , depict the final results and the
insights for making GSCM managerial decisions. Fig. 2 shows rankings for the aggregated models of perceived GSCM practices
strategic supplier partnership (SSP) as the factor with the highest influence on organizational sustainable performance. Table 14 shows
connectivity degree, making it the most critical GSCM factor. the rank order of the top 10 GSCM most influential sub-practices that
Strategic supplier relationships set the foundation for short- and contribute to sustainability of Ghana”s mining industry.
long-term successful inter-organizational programs, GSCM is not OLI1 “lean and green operations” is the GSCM practice that is
an exception. Prominence rankings also show the following re- perceived to contribute greatly to sustainability in the mining in-
lationships SSP4 EOL4ECO4 IEM 4OLI4GITS. It is not clear if dustry. The multiplicative weighting model (MW) transposes the
this is an implementation path, but does provide a hierarchy of top two sub-factors, but is consistent with the other techniques.
what might be most important to these organizations. This in- Overall, each of the models shows a relatively consistent grouping
dustry and country may be less reliant on information technology of practices in the top 10. SSP1 “Jointly develop environmental
and may explain the lower prominence of GITS. management solutions” are reinforced as one of the other top sub-
The net cause/effect shows that SSP, OLI and EOL factors are net practices. But, even with a general consistency in the rankings,
causes for other factors. IEM, ECO and GITS factors are effected there are some significant shifts, for example SSP2 falls from a 4 to
(resultant causes) by the other factors. Thus, SSP, OLI and EOL are a 7 ranking. Many more such shifts do occur in the broader sets.
the cause factors influencing the other factors more so than being Thus, care must be taken on how values would be aggregated if
effected by them. SSP is the most influential cause factor whilst this tool was to be used as a decision support tool for decided on
IEM is considered the most influenced factor amongst the effect specific programmatic alternatives.
group. For these mining companies implementing GSCM programs The DEMATEL-ANP approach resulted in some factors being left
out of the analysis in the multiplicative aggregation model when
will require an initial focus on SSP and have that foundation built.
using the desirability indices aggregation approach. Practically,
The relative prominence and causation does not necessarily mean
these less linked factors should not be zero valued when calcu-
the most important or least important overall. The reason for this
lating desirability index values and why some non-zero valuation
relative importance may be that foundational activities such as SSP
using additive or exponential aggregation approaches are needed.
may have been developed already, with a need to focus on other
It is more accurate to not completely eliminate factors that lack
less mature GSCM practices that have yet to be implemented. The
interdependencies in ANP. ANP weighting aggregation with de-
linkage of this relationship to ANP relative importance results, in
sirability aggregation approaches; need to be adjusted to not fully
the next section, will help identify which GSCM practices should
penalize the lack of linkages for a factor which occur in the ANP-
be the focus of this Ghanaian mining industry.
DEMATEL approach.
The findings from the empirical study tell us that stronger
Overall, it is not surprising that the results favored the general
strategic supplier partnership is critical to fostering successful
SSP, OLI and EOL factors. These practice factors were found to be
GSCM implementation in the mining industry. The finding sup-
the most influential causes, where SSP was the most important
ports the consensus and the importance placed on inter-organi- and connected factor. This result implies that the relationship
zational partnership during cross-organizational implementation between supplier and buyer may lead to joint greening cap-
programs (Palinkas et al., 2014; Chen et al., 2014). This result does abilities and competencies development which may result in
complement recent empirical studies. For example, environmental collaborative competitive advantages (Gottschalk and Solli-Sæ-
partnerships between organizations and their primary strategic ther, 2006). Thus, building relational capabilities are important
suppliers are positively linked with improved environmental and for addressing the environmental impact of these mining com-
operational performance (Bowen et al., 2001; Vachon and Klassen, panies supply chains. The mining companies can develop these
2008). This result implies that environmental performance of relational greening capabilities and competencies by engaging
mining companies can be improved by extending internal activ- their strategic suppliers in early joint environmental discussions
ities beyond a mining organization”s boundaries (external) to and learning activities. This engagement can result in shared
partner with strategic suppliers. This systemic perspective con- interpretation of environmental concerns to formulate collective
siders both internal and external factors and helps to strengthen environmental solutions to achieve environmental goals (Geffen
greening capabilities and competencies. Engaging strategic sup- and Rothenberg, 2000; Tseng, 2011; Vachon and Klassen, 2008;
pliers’ in early GSCM program planning stages will most likely Wong, 2013). This concurrent result also shows that the mining
result in greater acceptance and performance of other practices. It industry in Ghana is still early in GSCM practices implementation
may also strengthen the program”s capabilities and competencies because the focus and greatest opportunities to improve corpo-
through shared environmental management techniques and rate sustainability are from the more foundational (causation)
knowledge. sub-practices. Further study, after implementation of practices
needs to be completed to determine if the outcomes were as practices and sub-practices. SSP was found to be the most pro-
predicted or expected. minent and networked, to other GSCM practices. This result was
confirmed in the next stage when ANP was utilized with the DE-
5.3. Managerial feedback MATEL input to arrive at the perceived most influential sub-
practices to organizational sustainability.
As a post-hoc analysis of the results, we sent a small survey From a methodological perspective, this study is the first to
asking managers about the approach and results. Three managers fully integrate DEMATEL with ANP. DEMATEL allows for clearly
replied. The major concern of the managers with this technique identifying which interdependencies are most influential, and thus
was the many factors and sub-factors involved in the analysis. reducing, exponentially, the number of pairwise comparisons
Trying to understand the definitions while completing the data needed for ANP. Unfortunately, it was found that if inter-
acquisition survey was one of the major limitations mentioned. dependency did not exist, that the unconnected practice would
This may cause the results to be a bit biased, maybe toward those receive a zero value when using the multiplicative form of desir-
activities and concerns that the managers understood more ability indices. Thus, aggregation techniques that did not com-
completely. Thus, a user friendly or face-to-face detailed im- pletely penalize a set of relationships because of the lack of explicit
plementation with a facilitator who can explain dimensions and interdependent connections were introduced. Both additive and
factors clearly, will be needed for effective implementation of this exponential aggregation approaches were studied. It was found
approach. This consideration further supports the need for filtra- that differences and sensitivities do exist among the three ag-
tion of relationships and factors that are used in ANP. It is assumed gregated techniques, though not as large as would be assumed, at
that the amount of process frustration with ANP was lessened, least not on the highest ranking sub-practices.
although the DEMATEL portion may still have been cumbersome. The practical implications from the results are that managers
Managers were also provided with some mathematical back- believe and should probably focus on early foundational practices
ground associated with the technique. The reason for this is to such as strategic supplier collaborations and operations and lean
allow for transparency and exemplify the robustness of the initiative practices to get the greatest potential sustainability re-
methodology, giving managers a more secure feeling that the turns for their organization. From a methodological perspective,
technique is based on scientific and mathematical principles and researchers and decision analysts should be careful when seeking
logic. Unfortunately, the respondent managers (and informally to integrate DEMATEL as a simplifying agent for ANP network
other managers) felt that the mathematical descriptions of the analysis. Too much simplification in the network connections may
technique was not enlightening and even a hindrance to under- cause significant changes to what practices will arrive at a final
standing. Thus, presentation of these complex techniques may solution, especially when using a multiplicative aggregation for
have been best presented (as one manager stated) at the highest the desirability index. Alternative aggregation measures should be
level of analysis, maybe as a general flow chart. used to not overly penalize those sub-practices or factors that are
The final major issue we tried to address in getting manage- missing interdependencies.
ment replies was the validity and confidence in the final results as The results of this study provide valuable clues and guidelines
summarized in a simple table. Even after some disillusionment to decision-makers and analysts inside and outside the mining
with the process, the managers felt that the final results were industry for making strategic sustainability decisions such as
what was expected and what they wished to convey. Thus, al- GSCM implementation decisions. The methodology introduced in
though the means to arrive at some solution may have been mired this paper has generalizability to many ANP and multi-attribute
in the complexity of the process and definitions, the final results applications. Yet, there are limitations to this study and additional
could be viewed as managerially valid and reliable. investigation is required, which provides fertile ground for further
These are some final feedback results, different settings, managers, studies. Some general limitations are now identified.
preparation and backgrounds may have arrived at different results.
6.2. Limitations of the study
6. Conclusion One of the principle limitations of this study is its reliance on a

small group of managers in one industry in Ghana. Generalizations
6.1. Summary of findings to other countries and other industries cannot be made. This in-
vestigation is exploratory, and more investigation is required with
Mining industry environmental impact is extensive and per- broader empirical studies. Also, the study is just a snap-shot in
sistent; it is especially pernicious in emerging economy countries time. Longitudinal investigation to determine if and how GSCM
such as Ghana. Greening the supply chain is one important and practice requirements and importance change over time is needed.
strategic sustainable production and consumption option for ad- Methodologically we investigated the use of ANP using desir-
dressing these serious environmental impacts. Given the novelty ability indices. There are various other approaches to arrive at ANP
of GSCM in mining it is important to better understand how it can solutions such as complete super-matrix and algebraic matrix
be managed. This study is one of the first to investigate this topic. operations. The sensitivities of the techniques in these other sce-
The focus of this investigation is the first contribution; the second narios need to be investigated. In addition, the mental mapping of
major contribution is introducing a multi-stage DEMATEL and DEMATEL is only one approach to develop the network relation-
ANP-DEMATEL approach to quantitatively investigate influence ships and interdependencies, interpretative structural modeling
and importance perceptions of GSCM practices and their role in and other mental modeling causal analysis tools may be in-
supporting organizational sustainability. vestigated to determine whether the interdependent relationships
This paper adopted a previously developed comprehensive and would change.
integrative GSCM practices framework partially developed and As can be seen there is still significant work that can be com-
practically validated using mining industry managers from Ghana, pleted in integrating ANP with other tools, and further investigation
an important African emerging economy nation. Multiple-field into the mining industry in emerging economy and developed na-
studies were used to gather data and evaluate the methodology. tions. The work presented here helps to set the foundation for ad-
The fuzzy-DEMATEL aspect of the methodology was first applied ditional and important methodological and sustainable supply
to develop the interrelations/interdependencies amongst GSCM chain (organizational) investigations.
6.3. Managerial implications Bocken, N.M.P., Farracho, M., Bosworth, R., Kemp, R., 2014. The front-end of eco-
innovation for eco-innovative small and medium sized companies. J. Eng.
Technol. Manag. 31, 43–57.
From a practical perspective, if there existed greater similarities Boon, E. K., Ababio, F., 2009. Corporate Social Responsibility in Ghana: Lessons from
among the three desirability index rank orders, this would have the mining sector. In: IAIIA09 Conference Proceedings, Impact Assessment and
more clearly provided managers with a consensus set of factors Human Well-Being, May.
Bowen, F.E., Cousins, P.D., Lamming, R.C., Farukt, A.C., 2001. The role of supply
that should be pursued. Given that some sub-factors were left out management capabilities in green supply. Prod. Oper. Manag. 10 (2), 174–189.
(zero valued) in the multiplicative (weighting) aggregation model Brandenburg, M., Govindan, K., Sarkis, J., Seuring, S., 2014. Quantitative models for
analysis when using the desirability aggregation approach, man- sustainable supply chain management: Developments and directions. Eur. J.
Oper. Res. 233 (2), 299–312.
agers and decision makers should be more comfortable with the
Carter, C.R., Easton, P.L., 2011. Sustainable supply chain management: evolution and
alternative ANP AW and MEW weighting schemes. The results future directions. Int. J. Phys. Distrib. Logist. Manag. 41 (1), 46–62.
from the two aggregation weighted approaches (AW and MEW), Chang, B., Chang, C.W., Wu, C.H., 2011. Fuzzy DEMATEL method for developing
supplier selection criteria. Expert. Syst. Appl. 38 (3), 1850–1858.
do provide general consistency in the rankings, and may be an
Chen, Y.H., Lin, T.P., Yen, D.C., 2014. How to facilitate inter-organizational knowledge
option to consider as an initial step for green operational in- sharing: the impact of trust. Inf. Manag. 51 (5), 568–578.
itiatives implementation. In such situations, managers and deci- Cheng, C.H., Lin, Y., 2002. Evaluating the best main battle tank using fuzzy decision
sion-makers in the mining industry may focus on those sub-fac- theory with linguistic criteria evaluation. Eur. J. Oper. Res. 142 (1), 174–186.
Chou, D.C., Chou, A.Y., 2012. Awareness of Green IT and its value model. Comput.
tors that are highly ranked within the top ten sub-factors across Stand. Interfaces 34 (5), 447–451.
the ANP AW and MEW weighting schemes. Corder, G.D., Golev, A., Fyfe, J., King, S., 2014. The status of industrial ecology in
This paper provided some managerial and methodological in- Australia: Barriers and enablers. Resources 3 (2), 340–361.
Cucchiella, F., D’Adamo, I., Gastaldi, M., Koh, S.L., 2014. Implementation of a real
sight into GSCM in an emerging economy (Ghana) mining in- option in a sustainable supply chain: an empirical study of alkaline battery
dustry. The effectiveness of a DEMATEL and ANP linkage was recycling. Int. J. Syst. Sci. 45 (6), 1268–1282.
presented. Clearly, more work across emerging economies with Dam, L., Petkova, B.N., 2014. The impact of environmental supply chain sustain-
ability programs on shareholder wealth. Int. J. Oper. Prod. Manag. 34 (5),
respect to the role of GSCM in sustainable consumption and pro- 586–609.
duction is needed. We believe that this work sets the foundation Dare, M., Schirmer, J., Vanclay, F., 2014. Community engagement and social licence
for additional work on this important sustainable development to operate. Impact Assess. Proj. Apprais. 32 (3), 188–197.
Dashwood, H.S., 2014. Sustainable development and industry self-regulation de-
topic.
velopments in the global mining sector. Bus. Soc. 53 (4), 551–582.
De Giovanni, P., Vinzi, V.E., 2014. The benefits of a monitoring strategy for firms
subject to the Emissions Trading System. Transp. Res. Part D: Transp. Environ.
33, 220–233.
References
Drohomeretski, E., Gouvea da Costa, S.E., Pinheiro de Lima, E., Garbuio, P.A.D.R.,
2014. Lean, Six Sigma and Lean Six Sigma: an analysis based on operations
Adams, M.D., 2013. Impact of recycling cyanide and its reaction products on up- strategy. Int. J. Prod. Res. 52 (3), 804–824.
stream unit operations. Miner. Eng. 53, 241–255. Dubey, R., Gunasekaran, A., Ali, S.S., 2015. Exploring the relationship between
Aragonés-Beltrán, P., Chaparro-González, F., Pastor-Ferrando, J.P., Pla-Rubio, A., leadership, operational practices, institutional pressures and environmental
2014. An AHP (Analytic Hierarchy Process)/ANP (Analytic Network Process)- performance: a framework for green supply chain. Int. J. Prod. Econ. 160,
based multi-criteria decision approach for the selection of solar-thermal power 120–132.
plant investment projects. Energy 66, 222–238. Dupuy, K.E., 2014. Community development requirements in mining laws. Extr. Ind.
Aryee, B.N.A., 2001. Ghana's mining sector: its contribution to the national econ- Soc. 1 (2), 200–215.
omy. Resour. Policy 27 (2), 61–75. Edraki, M., Baumgartl, T., Manlapig, E., Bradshaw, D., Franks, D.M., Moran, C.J., 2014.
Awudi, G.B.K., 2002. The Role of Foreign Direct Investment (FDI) in the Mining Designing mine tailings for better environmental, social and economic out-
Sector of Ghana and the Environment. OECD, Paris. comes: a review of alternative approaches. J. Clean. Prod. 84, 411–420.
Azevedo, S., Cudney, E.A. Grilo, A., Carvalho, H., Cruz-Machado, V., 2012. The in- Fabbe-Costes, N., Roussat, C., Taylor, M., Taylor, A., 2014. Sustainable supply chains:
fluence of eco-innovation supply chain practices on business eco-efficiency. a framework for environmental scanning practices. Int. J. Oper. Prod. Manag. 34
University of Beira Interior, Missouri University of Science and Technology, (5), 664–694.
Universidade Nova de Lisboa, MPRA Paper No. 42704, posted 18. November Falck, W.E., Spangenberg, J.H., 2014. Selection of social demand-based indicators:
2012 (online). Available from: 〈http://mpra.ub.uni-muenchen.de/42704/〉 (as- EO-based indicators for mining. J. Clean. Prod. 84, 193–203.
sessed 10.09.13). Fei-Baffoe, B., Botwe-Koomson, G.K., Mensa-Bonsu, I., F., Agyapong, E.A., 2013. Im-
Bai, C., Sarkis, J, 2013. Green information technology strategic justification and pact of ISO 14001 environmental management system on key environmental
evaluation. Inf. Syst. Front. 15 (5), 831–847. performance indicators of selected gold mining companies in ghana. J. Waste
Büyüközkan, G., Çifçi, G., 2012a. Evaluation of the green supply chain management Manag., 1–6.
practices: a fuzzy ANP approach. Prod. Plan. Control. 23 (6), 405–418. Faucheux, S., Nicolaï, I., 2011. IT for green and green IT: A proposed typology of eco-
Büyüközkan, G., Çifçi, G., 2012b. A novel hybrid MCDM approach based on fuzzy innovation. Ecol. Econ. 70 (11), 2020–2027.
DEMATEL, fuzzy ANP and fuzzy TOPSIS to evaluate green suppliers. Expert. Syst. Fontela, E., Gabus, A., 1976. The DEMATEL Observer, DEMATEL 1976 Report. Battelle
Appl. 39 (3), 3000–3011. Geneva Research Centre, Geneva, Switzerland.
Baram, M.S., Partan, D.G., 1990. Corporate Disclosure of Environmental Risks: US Gabus, A., Fontela, E., 1973. Perceptions of the World Problematique: Commu-
and European Law. Lexis Pub, United States. nication Procedure, Communicating with Those Bearing Collective Responsi-
Barve, A., Muduli, K., 2013. Modelling the challenges of green supply chain man- bility. Battelle Geneva Research Centre, Geneva, Switzerland.
agement practices in Indian mining industries. J. Manuf. Technol. Manag. 24 (8), Geffen, C.A., Rothenberg, S., 2000. Suppliers and environmental innovation: the
1102–1122. automotive paint process. Int. J. Oper. Prod. Manag. 20 (2), 166–186.
Beckmann, M., Hielscher, S., Pies, I., 2014. Commitment strategies for sustainability: Giurco, D., Cooper, C., 2012. Mining and sustainability: asking the right questions.
how business firms can transform trade‐offs into win–win outcomes. Bus. Miner. Eng. 29, 3–12.
Strategy Environ. 23 (1), 18–37. Gomes, C.M., Kneipp, J.M., Kruglianskas, I., Aparecida, L., da Rosa, B., Bichueti, R.S.,
Beske, P., Seuring, S., 2014. Putting sustainability into supply chain management. 2014. Management for sustainability in companies of the mining sector: an
Supply Chain Manag.: Int. J. 19 (3), 322–331. analysis of the main factors related with the business performance. J. Clean.
Bell, J., Mollenkopf, D.A., Stolze, H., 2013. Natural resource scarcity and the closed- Prod. 84, 84–93.
loop supply chain: a resource-advantage view. Int. J. Phys. Distrib. Manag. 43 Gottschalk, P., Solli-Sæther, H., 2006. Maturity model for IT outsourcing relation-
(5), 2–12. ships. Ind. Manag. Data Syst. 106 (2), 200–212.
Bhadauria, V.S., Toms, L., Green, K.W., Meacham, J., 2014. Do green information Govindan, K., Popiuc, M.N., 2014. Reverse supply chain coordination by revenue
systems impact performance? Int. J. Product. Qual. Manag. 13 (4), 377–394. sharing contract: a case for the personal computers industry. Eur. J. Oper. Res.
Bice, S., 2014. What gives you a social licence? An exploration of the social licence 233 (2), 326–336.
to operate in the Australian mining industry. Resour. Policy 3 (1), 62–80. Govindan, K., Kannan, D., Shankar, M., 2014a. Evaluation of green manufacturing
Bilal, K., Malik, S.U.R., Khalid, O., Hameed, A., Alvarez, E., Wijaysekara, V., Khan, S.U., practices using a hybrid MCDM model combining DANP with PROMETHEE. Int.
2014. A taxonomy and survey on Green Data Center Networks. Future Gener. J. Prod. Res., 1–28 (ahead-of-print).
Comput. Syst. 36, 189–208. Govindan, K., Soleimani, H., Kannan, D., 2015. Reverse logistics and closed-loop
Bloch, R., Owusu, G., 2012. Linkages in Ghana's gold mining industry: challenging supply chain: a comprehensive review to explore the future. Eur. J. Oper. Res.
the enclave thesis. Resour. Policy 37 (4), 434–442. 240 (3), 603–626.
Blome, C., Hollos, D., Paulraj, A., 2014. Green procurement and green supplier de- Govindan, K., Kaliyan, M., Kannan, D., Haq, A.N., 2014b. Barriers analysis for green
velopment: antecedents and effects on supplier performance. Int. J. Prod. Res. supply chain management implementation in Indian industries using analytic
52 (1), 32–49. hierarchy process. Int. J. Prod. Econ. 147, 555–568.
Hajmohammad, S., Vachon, S., Klassen, R.D., Gavronski, I., 2013. Reprint of Lean model explaining community acceptance of mining. Resour. Policy 39, 61–70.
management and supply management: their role in green practices and per- Molla, A., Abareshi, A., Cooper, V., 2014. Green IT beliefs and pro-environmental IT
formance. J. Cleaner Prod. 56, 86–93. practices among IT professionals. Inf. Technol. People 27 (2), 129–154.
Haibin, L., Zhenling, L., 2010. Recycling utilization patterns of coal mining waste in Muduli, K., Govindan, K., Barve, A., Geng, Y., 2013. Barriers to green supply chain
China. Resour., Conserv. Recycl. 54 (12), 1331–1340. management in Indian mining industries: a graph theoretic approach. J. Clean.
Hilson, G., Nayee, V., 2002. Environmental management system implementation in Prod. 47, 335–344.
the mining industry: a key to achieving cleaner production. Int. J. Mineral. Munda, G., Nardo, M., 2005. Constructing consistent composite indicators: the issue
Process. 64 (1), 19–41. of weights (EUR 21834 EN). Institute for the Protection and Security of the
Hsu, C.C., Lee, Y.S., 2014. Exploring the Critical Factors Influencing the Quality of Citizen, Ispra. Available at: 〈http://dx.doi.org/http://crell.jrc.ec.europa.eu/Well-
Blog Interfaces Using the Decision-Making Trial and Evaluation Laboratory being/papers/Munda%20Nardo%20euroreport1.pdf〉 (accessed 03.10.14).
(DEMATEL) Method. Behav. Inf. Technol. 33 (2), 184–194. Mustajoki, J., Hamalainen, R.P., 2000. Web-HIPRE: global decision support by value
Huang, C.N., Liou, J.J., Chuang, Y.C., 2014. A method for exploring the inter- tree and AHP analysis. INFOR J. 38 (3), 208–220.
dependencies and importance of critical infrastructures. Knowl.-Based Syst. 55, Mzembe, A.N., Meaton, J., 2014. Driving corporate social responsibility in the Ma-
66–74. lawian mining industry: a stakeholder perspective. Corp. Soc. Responsib. En-
Hwang, C.L., Yoon, K.S., 1981. Multiple Attribute Decision Making: Method and viron. Manag. 21 (4), 189–201.
Applications. Springer, New York. Nardo, M., M., Saisana, A., Saltelli, S., Tarantola, 2008. Handbook on Constructing
Jayakrishna, K., Vinodh, S., Kumar, D.S., 2013. Hybrid ANP: QFD—ZOGP Approach Composite Indicators: Methodology and User Guide. OECD Publishing, Paris,
for Styling Aspects Determination of an Automotive Component. Springer, France.
India. Natoli, R., Zuhair, S., 2011. Rediscovering the condorcet approach as an aggregation
Jenkin, T.A., Webster, J., McShane, L., 2011. An agenda for ‘Green’ information technique for progress measures. Econ. Pap.: J. Appl. Econ. Policy 30 (3),
technology and systems research. Inf. Organ. 21 (1), 17–40. 368–376.
Johnson, D.B., 2013. Development and application of biotechnologies in the metal Niza, S., Santos, E., Costa, I., Ribeiro, P., Ferrão, P., 2014. Extended producer re-
mining industry. Environ. Sci. Pollut. Res. 20 (11), 7768–7776. sponsibility policy in Portugal: a strategy towards improving waste manage-
Kleindorfer, P.R., Singhal, K., Wassenhove, L.N., 2005. Sustainable operations man- ment performance. J. Clean. Prod. 64, 277–287.
agement. Prod. Oper. Manag. 14 (4), 482–492. Opricovic, S., Tzeng, G.H., 2003. Defuzzification within a multicriteria decision
Koo, C., Chung, N., 2014. Examining the eco-technological knowledge of Smart model. J. Uncertain., Fuzziness Knowl.-Based Syst. 11 (5), 635–652.
Green IT adoption behavior: a self-determination perspective. Technol. Fore- Owen, J.R., Kemp, D., 2013. Social licence and mining: a critical perspective. Resour.
cast. Social. Chang. 88, 140–155. Policy 38 (1), 29–35.
Kumar, V., Holt, D., Ghobadian, A., Garza-Reyes, J.A., 2014. Developing green supply Palinkas, L.A., Fuentes, D., Finno, M., Garcia, A.R., Holloway, I.W., Chamberlain, P.,
chain management taxonomy-based decision support system. Int. J. Prod. Res., 2014. Inter-organizational collaboration in the implementation of evidence-
1–18 (ahead-of-print). based practices among public agencies serving abused and neglected youth.
Kusi-Sarpong, S., Bai, C., Sarkis, J., Wang, X., 2015. Green supply chain practices Adm. Policy Ment. Health Ment. Health Serv. Res. 41 (1), 74–85.
evaluation in the mining industry using a joint rough sets and fuzzy TOPSIS Patil, K.S., Kant, R., 2014a. Knowledge management adoption in supply chain:
methodology. Resour. Policy 46 (1), 86–100. Identifying critical success factors using fuzzy DEMATEL approach. J. Model.
Kuyucak, N., Akcil, A., 2013. Cyanide and removal options from effluents in gold Manag. 9 (2), 160–178.
mining and metallurgical processes. Miner. Eng. 50, 13–29. Patil, S.K., Kant, R., 2014b. A hybrid approach based on fuzzy DEMATEL and FMCDM
Lawson, E.T., Bentil, G., 2014. Shifting sands: changes in community perceptions of to predict success of knowledge management adoption in supply chain. Appl.
mining in Ghana. Environ., Dev. Sustain. 16 (1), 217–238. Soft Comput. 18, 126–135.
Lee, V.H., Ooi, K.B., Chong, A.Y.L., Seow, C., 2014. Creating technological innovation Paulraj, A., 2009. Environmental motivations: a classification scheme and its impact
via green supply chain management: An empirical analysis. Expert. Syst. Appl. on environmental strategies and practices. Bus. Strategy Environ. 18 (7),
41 (16), 6983–6994. 453–468.
Li, J., Wu, Z., 2014. Remanufacturing Processes, Planning and Control. In: New Peck, P., Sinding, K., 2003. Environmental and social disclosure and data richness in
Frontiers of Multidisciplinary Research in STEAM-H (Science, Technology, En- the mining industry. Bus. Strategy Environ. 12 (3), 131–146.
gineering, Agriculture, Mathematics, and Health). Springer International Pub- Pishvaee, M.S., Razmi, J., Torabi, S.A., 2014. An accelerated Benders decomposition
lishing, pp. 329–356. algorithm for sustainable supply chain network design under uncertainty: a
Li, Y., Hu, Y., Zhang, X., Deng, Y., Mahadevan, S., 2014. An evidential DEMATEL case study of medical needle and syringe supply chain. Transp. Res. Part E:
method to identify critical success factors in emergency management. Appl. Logist. Transp. Rev. 67, 14–38.
Soft Comput. 22, 504–510. Pooe, R.D., Mhelembe, K., 2014. Exploring the challenges associated with the
Lin, P.T., Li, B., Bu, D., 2015. The relationship between corporate governance and greening of supply chains in the South African manganese and phosphate
community engagement: evidence from the Australian mining companies. mining industry: original research. J. Transp. Supply Chain Manag. 8 (1), 1–9.
Resour. Policy 43, 28–39. Ramanathan, U., Gunasekaran, A., 2014. Supply chain collaboration: impact of
Lin, R.J., 2013. Using fuzzy DEMATEL to evaluate the green supply chain manage- success in long-term partnerships. Int. J. Prod. Econ. 147, 252–259.
ment practices. J. Clean. Prod. 40, 32–39. Ranängen, H., Zobel, T., Bergström, A., 2014. The merits of ISO 26000 for CSR de-
Liu, H.C., You, J.X., Lin, Q.L., Li, H., 2014. Risk assessment in system FMEA combining velopment in the mining industry: a case study in the Zambian Copperbelt. Soc.
fuzzy weighted average with fuzzy decision-making trial and evaluation la- Responsib. J. 10 (3), 500–515.
boratory. Int. J. Comput. Integr. Manuf., 1–14 (ahead-of-print). Rao, P., 2002. Greening the supply chain: a new initiative in South East Asia. Int. J.
Luthra, S., Garg, D., Haleem, A., 2015. An analysis of interactions among critical Oper. Prod. Manag. 22 (6), 632–655.
success factors to implement green supply chain management towards sus- Rao, P., Holt, D., 2005. Do green supply chains lead to competitiveness and eco-
tainability: an Indian perspective. Resour. Policy 46 (1), 37–50. nomic performance? Int. J. Oper. Prod. Manag. 25 (9), 898–916.
Lutandula, M.S., Maloba, B., 2013. Recovery of cobalt and copper through re- Rogers, D.S., Lembke, R., 2001. An examination of reverse logistics practices. J. Bus.
processing of tailings from flotation of oxidised ores. J. Environ. Chem. Eng. 1 Logist. 22 (2), 129–148.
(4), 1085–1090. Rostamy, A.A.A., Shaverdi, M., Ramezani, I., 2013. Green supply chain management
Mangla, S.K., Kumar, P., Barua, M.K., 2014. A flexible decision framework for evaluation in publishing industry based on fuzzy AHP approach. J. Logist.
building risk mitigation strategies in green supply chain using SAP–LAP and IRP Manag. 2 (1), 9–14.
approaches. Glob. J. Flex. Syst. Manag. 15 (3), 203–218. Söderholm, K., Söderholm, P., Helenius, H., Pettersson, M., Viklund, R., Masloboev,
Maslen, S., Hopkins, A., 2014. Do incentives work? A qualitative study of managers' V., Mingaleva, T., Petrov, V., 2015. Environmental regulation and competitive-
motivations in hazardous industries. Saf. Sci. 70, 419–428. ness in the mining industry: permitting processes with special focus on Fin-
Mathiyazhagan, K., Govindan, K., Noorul Haq, A., 2014. Pressure analysis for green land, Sweden and Russia. Resour. Policy 43, 130–142.
supply chain management implementation in Indian industries using analytic Saaty, T.L., 2004. Decision making—the analytic hierarchy and network processes
hierarchy process. Int. J. Prod. Res. 52 (1), 188–202. (AHP/ANP). J. Syst. Sci. Syst. Eng. 13 (1), 1–35.
Mbendi, 1995–2013. Mining in Ghana-Overview, (online). Available from: 〈http:// Saaty, T.L., Takizawa, M., 1986. Dependence and independence: from linear hier-
www.mbendi.com/indy/ming/af/gh/p0005.htm〉 (assessed: 10.04.13). archies to nonlinear networks. Eur. J. Oper. Res. 26 (2), 229–237.
McLellan, B.C., Corder, G.D., Giurco, D., Green, S., 2009. Incorporating sustainable Saaty, T.L., Hu, G., 1998. Ranking by eigenvector versus other methods in the ana-
development in the design of mineral processing operations – review and lytic hierarchy process. Appl. Math. Lett. 11 (4), 121–125.
analysis of current approaches. J. Clean. Prod. 17 (16), 1414–1425. Saaty, T.L., 1996. Decision Making with Dependence and Feedback: the Analytic
Meade, L., Sarkis, J., 1998. Strategic analysis of logistics and supply chain manage- Network Process.
ment systems using the analytical network process. Transp. Res. Part E: Logist. Sarkis, J., 2003. A strategic decision framework for green supply chain management.
Transp. Rev. 34 (3), 201–215. J. Clean. Prod. 11 (4), 397–409.
Miao, R., Xu, F., Zhang, K., Jiang, Z., 2014. Development of a multi-scale model for Sarkis, J., Zhu, H., 2008. Information technology and systems in China's cir-
customer perceived value of electric vehicles. Int. J. Prod. Res. 52 (16), 4820–4834. cular economy: implications for sustainability. J. Syst. Inf. Technol. 10 (3),
Min, H., Galle, W.P., 2001. Green purchasing practices of US firms. Int. J. Oper. Prod. 202–217.
Manag. 21 (9), 1222–1238. Sarkis, J., Koo, C., Watson, R.T., 2013. Green information systems & technologies–this
Mirhedayatian, S.M., Azadi, M., Farzipoor Saen, R., 2014. A novel network data generation and beyond: Introduction to the special issue. Inf. Syst. Front. 15 (5),
envelopment analysis model for evaluating green supply chain management. 695–704.
Int. J. Prod. Econ. 147, 544–554. Savita, K.S., Dominic, P.D.D., Ramayah, T., 2014. The adoption of green information
Moffat, K., Zhang, A., 2014. The paths to social licence to operate: an integrative technologies and systems as a driver within green SCM. In: Computer and
Information Sciences (ICCOINS), 2014 International Conference on, IEEE, pp. 1– Wagner, S., Hespenheide, E., Pavlovsky, K., 2009. The responsible and sustainable
6), June. board. Deloitte Rev. 4, 59–71.
Senvar, O., Tuzkaya, U.R., Kahraman, C., 2014. Supply chain performance measure- Wang, X., Gaustad, G., 2012. Prioritizing material recovery for end-of-life printed
ment: an integrated DEMATEL and fuzzy-ANP approach. In: Supply Chain circuit boards. Waste Manag. 32 (10), 1903–1913.
Management Under Fuzziness, Springer Berlin Heidelberg, pp. 143–165. Wang, X., Chan, H.K., 2013. A hierarchical fuzzy TOPSIS approach to assess im-
Setterstrom, A., 2008. The natural resource-base view of a firm: Strategic oppor- provement areas when implementing green supply chain initiatives. Int. J. Prod.
tunities in IT. Paper Presented at the Annual Academy of Management Meet- Res. 51 (10), 3117–3130.
ings, Anaheim, CA. Wang, X., Chan, H.K., Yee, R.W., Diaz-Rainey, I., 2012. A two-stage fuzzy-AHP model
Shandro, J.A., Veiga, M.M., Shoveller, J., Scoble, M., Koehoorn, M., 2011. Perspectives for risk assessment of implementing green initiatives in the fashion supply
on community health issues and the mining boom–bust cycle. Resour. Policy 36 chain. Int. J. Prod. Econ. 135 (2), 595–606.
(2), 178–186. Wasylycia-Leis, J., Fitzpatrick, P., Fonseca, A., 2014. Mining communities from a
Si, H., Bi, H., Li, X., Wang, C., 2010. Environmental evaluation for sustainable de- resilience perspective: managing disturbance and vulnerability in itabira, Bra-
velopment of coal mining in Qijiang, Western China. Int. J. Coal Geol. 81 (3), zil. Environ. Manag. 53 (3), 481–495.
163–168. Watson, R.T., Boudreau, M.C., Chen, A., Huber, M.H., 2008. Green IS: Building sus-
Simpson, D.F., Power, D.J., 2005. Use the supply relationship to develop lean and tainable business practices. In: Watson, R.T. (Ed.), Information Systems. Global
green suppliers. Supply Chain Manag.: Int. J. 10 (1), 60–68. Text Project, Athens, GA, USA.
Simpson, D., Power, D., Samson, D., 2007. Greening the automotive supply chain: a Wee, Y.S., Quazi, H.A., 2005. Development and validation of critical factors of en-
relationship perspective. Int. J. Oper. Prod. Manag. 27 (1), 28–48. vironmental management. Ind. Manag. Data Syst. 105 (1), 96–114.
Stock, J.R., 2001. Reverse logistics in the supply chain. Rev. Transp. Logist., 44. Wiengarten, F., Pagell, M., Ahmed, M.U., Gimenez, C., 2014. Do a country's logistical
Theißen, S., Spinler, S., 2014. Strategic analysis of manufacturer-supplier partner- capabilities moderate the external integration performance relationship? J.
ships: an ANP model for collaborative CO 2 reduction management. Eur. J. Oper. Oper. Manag. 32 (1), 51–63.
Res. 233 (2), 383–397. Wong, C.W., 2013. Leveraging environmental information integration to enable
Tseng, M.L., 2011. Green supply chain management with linguistic preferences and environmental management capability and performance. J. Supply Chain
incomplete information. Appl. Soft Comput. 11 (8), 4894–4903. Manag. 49 (2), 114–136.
Tseng, M.L., Chiu, A.S.F., 2013. Evaluating firm's green supply chain management in Wong, C.Y., Wong, C.W., Boon-itt, S., 2015. Integrating environmental management
linguistic preferences. J. Clean. Prod. 40, 22–31. into supply chains: a systematic literature review and theoretical framework.
Tseng, M.L., Lim, K.M., Wong, W.P., 2015. Sustainable supply chain management: a Int. J. Phys. Distrib. Logist. Manag. 45 (1/2), 43–68.
closed-loop network approach. Ind. Manag. Data Syst. 115 (3), 436–461. Wong, W.P., Ignatius, J., Soh, K.L., 2014. What is the leanness level of your organi-
Uddin, M., Rahman, A.A., 2012. Energy efficiency and low carbon enabler green IT sation in lean transformation implementation? An integrated lean index using
framework for data centers considering green metrics. Renew. Sustain. Energy
ANP approach. Prod. Plan. Control. 25 (4), 273–287.
Rev. 16 (6), 4078–4094.
Wu, H.H., Chang, S.Y., 2015. A case study of using DEMATEL method to identify
Vachon, S., 2007. Green supply chain practices and the selection of environmental
critical factors in green supply chain management. Appl. Math. Comput. 256,
technologies. Int. J. Prod. Res. 45 (18–19), 4357–4379.
394–403.
Vachon, S., Klassen, R.D., 2006a. Extending green practices across the supply chain:
Yu, V.F., Hu, K.J., 2010. An integrated fuzzy multi-criteria approach for the perfor-
the impact of upstream and downstream integration. Int. J. Oper. Prod. Manag.
mance evaluation of multiple manufacturing plants. Comput. Ind. Eng. 58 (2),
26 (7), 795–821.
269–277.
Vachon, S., Klassen, R.D., 2006b. Green project partnership in the supply chain: the
Zadeh, L.A., 1965. Fuzzy sets. Inf. Control. 8 (3), 338–353.
case of the package printing industry. Journal. Clean. Prod. 14 (6–7), 661–671.
Zaim, S., Sevkli, M., Camgöz-Akdağ, H., Demirel, O.F., Yayla, A.Y., Delen, D., 2014. Use
Vachon, S., Klassen, R.D., 2008. Environmental management and manufacturing
of ANP weighted crisp and fuzzy QFD for product development. Expert. Syst.
performance: the role of collaboration in the supply chain. Int. J. Prod. Econ. 111
Appl. 41 (9), 4464–4474.
(2), 299–315.
Zhang, J.J., Joglekar, N.R., Verma, R., 2012. Exploring resource efficiency bench-
Vachon, S., Klassen, R.D., Johnson, P.F., 2001. Customers as green suppliers: mana-
marks for environmental sustainability in hotels. Cornell Hosp. Q. 53 (3),
ging the complexity of the reverse supply chain. In: Sarkis, J. (Ed.), Greening
Manufacturing: From Design to Delivery and Back. Greenleaf Publisher, Shef- 229–241.
field, UK. Zsidisin, G.A., Hendrick, T.E., 1998. Purchasing”s involvement in environmental is-
Vachon, S., 2013. Technological capacity and environmental performance: a re- sues: a multi-country perspective. Ind. Manag. Data Syst. 98 (7), 313–320.
search note using country level data. J. Oper. Supply Chain Manag. 1 (1), 21–28.
Van Berkel, R., 2007. Eco-efficiency in the Australian minerals processing sector. J.
Clean. Prod. 15 (8), 772–781.
Vintró, C., Sanmiquel, L., Freijo, M., 2014. Environmental sustainability in the
mining sector: evidence from Catalan companies. J. Clean. Prod. 84, 155–163.
Voigt, S., De Cian, E., Schymura, M., Verdolini, E., 2014. Energy intensity develop-
ments in 40 major economies: structural change or technology improvement?
Energy Econ. 41, 47–62.

Int. J. Production Economics: Simonov Kusi-Sarpong, Joseph Sarkis, Xuping Wang

Uploaded by

Copyright:

Available Formats

Int. J. Production Economics: Simonov Kusi-Sarpong, Joseph Sarkis, Xuping Wang

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Int. J. Production Economics: Simonov Kusi-Sarpong, Joseph Sarkis, Xuping Wang

Uploaded by

Copyright:

Available Formats

Int. J.

Production Economics 181 (2016) 325–341

Contents lists available at ScienceDirect

Int. J. Production Economics

Assessing green supply chain practices in the Ghanaian mining

No. GSCM Factors and Sub-factors Literature

A graph can then be plotted by mapping the (Cj+Ri , Cj−Ri ) da-

Organizational Sustainable Performance

Economic Environmental Social Overall

GITS SSP OLI IEM ECO EOL Major

Fig. 1. Decision structure for GSCM implementation.

Major factors GITS SSP OLI IEM ECO EOL

GITS 0 (0.25, 0.50, 0.75) (0,0,0.25) (0,0.25,0.50) (0,0.25,0.50) (0,0,0.25)

Fig. 2. Network Relationship Map (NRM)/diagraphs of the GSCM major practices.

= e( U1 + P 2,1+ P 2,1+ A6,2 + A6,2 ) = e (0.7144 + 0.2462 + 0.3784 + 0.3492 + 0.3175)

S. Kusi-Sarpong et al. / Int. J. Production Economics 181 (2016) 325–341

6. Conclusion One of the principle limitations of this study is its reliance on a

You might also like