Governance Pressures and Performance of Sustainable Supply Chains UK Manufacturing PDF

Journal of Cleaner Production 155 (2017) 66e78 Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www....

Journal of Cleaner Production 155 (2017) 66e78 Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro Governance pressures and performance outcomes of sustainable supply chain management e An empirical analysis of UK manufacturing industry Ali Esfahbodi*, Yufeng Zhang, Glyn Watson, Tao Zhang Department of Procurement and Operations Management, Birmingham Business School, University of Birmingham, United Kingdom a r t i c l e i n f o a b s t r a c t Article history: Although sustainable supply chain management (SSCM) has recently received increasing attention Received 15 December 2015 among UK manufacturing firms, there is a concern as to whether SSCM practices are being implemented Received in revised form because they are profitable or only because of governance coercive pressure. Thus, the aims of this paper 14 July 2016 are twofold: first, determining the role of governance in the adoption of SSCM practices; second, Accepted 18 July 2016 investigating whether SSCM practices can be both environmentally beneficial and commercially viable. Available online 21 July 2016 In light of these issues, this paper develops and empirically assesses an integrated model of governance pressures-SSCM practices-performance. Data was collected from 146 UK manufacturing managers, and Keywords: Sustainable supply chain management analysed using the structural equation modelling method. Exogenous driving forces of governance were (SSCM) found to be precursors to the successful implementation of SSCM practices. The empirical results further Coercive pressure suggest that while the implementation of sustainable supply chain management has a positive effect on Manufacturing firms environmental performance, it does not necessarily lead to improved economic performance, as only Environmental performance sustainable procurement was found to have a positive effect on economic performance. This paper Economic performance contributes to the literature by highlighting the role of governance in SSCM adoption and performance Structural equation modelling (SEM) gains in environmental protection while economic performance is partially compromised. The results UK also provide useful insights for both managers seeking to adopt sustainable practices and policy-makers seeking to further promote sustainable supply chain. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction et al., 2015). While a variety of institutional pressures (gover- nance pressure in particular) are major motivating forces that lead There are increasing stakeholder expectations for firms to be firms to pursue SSCM practices (Zhu et al., 2013), a number of de- fully responsible for their business operations and to clearly bates have recently opened up to question whether the governance demonstrate their environmental and ethical behaviour (Lozano, pressure can actually drive the adoption of SSCM practices further 2008). Such expectations, along with growing institutional pres- €m et al., 2015). (Sarkis et al., 2010; Bostro sures, have caused manufacturing firms to adjust their traditional Environmental concerns and the inclusion of sustainable prac- supply chains to incorporate sustainable inputs in order to provide tices within the context of the supply chain are a subject that has more sustainable products, services and product-service combi- become popular in both academia and industry, with the hope of nations (Vezzoli et al., 2012). In response to these pressures, a mitigating environmental damage while achieving financial per- growing number of manufacturing firms have begun to undertake formance gains (Lozano et al., 2015). As institutional pressures push environmental initiatives across their supply chains and to imple- firms to adopt and maintain environmental initiatives and produce ment sustainable supply chain management (SSCM) practices (Su environmentally friendly products and services, firms must consider the impact of the adoption of such environmental initia- tives on both their environmental performance and their business * Corresponding author. Birmingham Business School, University House, Edg- performance (Gimenez et al., 2012). The recognition of financial baston Park Road, Edgbaston, Birmingham B15 2TT, United Kingdom. benefits gained from environmental initiatives in firms' bottom line E-mail addresses: [email protected], [email protected] (A. Esfahbodi), [email protected] (Y. Zhang), [email protected] (G. Watson), t.zhang.1@ is crucial for the adoption of such environmental initiatives. This bham.ac.uk (T. Zhang). contention is grounded on the fact that the economic performance http://dx.doi.org/10.1016/j.jclepro.2016.07.098 0959-6526/© 2016 Elsevier Ltd. All rights reserved.  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 67 has traditionally been, and continues to be, the key priority for 2.1. Governance pressures and SSCM practices: an institutional firms (Zhu and Sarkis, 2007). perspective While previous studies have attempted to link SSCM practices with organisational performance (Zhu and Sarkis, 2004; Rao and From an institutional perspective, firms' responsiveness to the Holt, 2005), they have neglected to include the driving role of adoption of new organisational practices such as environmental governance as a key antecedent to the adoption of such environ- initiatives may be influenced by three institutional isomorphic mental initiatives. We argue that the existence of and response to pressures: normative, mimetic, and coercive pressures (DiMaggio governance pressures will influence the relationships between and Powell, 1983; Sarkis, 2011). In essence the aforementioned SSCM practices and performance outcomes. In that sense, such institutional pressures can influence a firm's environmental align- governance pressures have a causal impact on SSCM adoption, ment. Several studies show how firms have been driven by insti- which in turn influences the relationship between SSCM practices tutional norms to enhance environmental performance by and organisational performance, particularly from a holistic and adopting proactive environmental programmes such as SSCM integrated perspective (Green et al., 2012a). Therefore, we should practices (Zhu et al., 2007; Sarkis et al., 2010). Among institutional consider the effects of governance pressures when examining the pressures, coercive pressure, which refers to the conformity relationship between the adoption of SSCM practices and perfor- occurring through influence exerted by those in power, is the most mance outcomes. Currently in supply chain management literature important factor that drives environmental initiatives among there is a dearth of studies empirically examining the relationship manufacturing firms (Zailani et al., 2012). From a practical between adopting SSCM practices and performance outcomes perspective, government agencies as powerful groups can influence considering the effects of governance pressures from a holistic and the actions of an organisation by enacting environmental regula- integrated perspective. Moreover, there is a level of ambiguity tions (Rivera, 2004). In this paper, governance pressure is referred within the existing knowledge concerning the merits of under- to as coercive pressure in the form of environmental regulations taking an SSCM agenda- whether it can deliver both environmental that drive the implementation of SSCM practices. improvements and economic benefits (Zhu and Sarkis, 2007; Scarcity of resources and environmental degradation have Seuring and Müller, 2008a). The existing research provides some prompted governmental organisations, at both national and inter- direction but remains inconclusive as they are of contradiction with national level, to exert pressures on manufacturers, the main one another (Zhu et al., 2005; Green et al., 2012a). resource consumers and polluters, through increasing environ- In order to bridge the knowledge gap in the literature and mental regulatory and tax policies (Taylor and Taylor, 2013; Yu and advance the contemporary understanding of SSCM, this research Ramanathan, 2015). For example, the European Union has enacted seeks to investigate an integrated Governance Pressures/SSCM various Environmental Directives such as Waste Electrical and Practices/ Performance Outcomes relationships. This paper serves Electronic Equipment (WEEE), Energy-using Products (EuP), Re- two purposes: (1) determining the role of governance in the striction of Hazardous Substances (RoHS), End of Life Vehicle (ELV), adoption of SSCM practices; (2) investigating the consequences of and so forth, which lead manufacturers to minimise negative governance within sustainable supply chains in terms of perfor- environmental impact (Gerrard and Kandlikar, 2007; Koh, 2012). mance outcomes. We utilise the institutional theory lens to help Such international regulatory policies not only push the EU man- explain the exogenous driving force of governance in the adoption ufacturers to undertake environmental initiatives across their of SSCM practices and their related influence on performance supply chains, but also force the manufacturers in developing outcomes. countries which intend to export and sell to Europe to embark upon The remainder of this paper is organised as follows. A literature the adoption of SSCM practices (Seuring et al., 2008; Sarkis et al., review and theoretical foundation are presented in Section 2, 2010). addressing SSCM drivers (i.e. governance pressures), practices and Therefore, while all institutional pressures can be valid moti- performance. Section 3 presents the research method along with a vators for the adoption of environmental practices, coercive pres- summary of the samples and data. Section 4 presents the analysis of sures exerted from government agencies and national or the data and reports the results of this study, followed by a dis- international regulators are deemed to be the major driving forces cussion of the key findings in Section 5. Section 6 concludes the that influence the adoption of SSCM practices (Lai et al., 2011; Lin paper. and Ho, 2011; Zailani et al., 2012). Thus, we argue that gover- nance pressures are the main driving forces in the adoption of SSCM practices. With this set of arguments, we hypothesise: H1. Governance pressures are directly and positively associated with the adoption of SSCM practices. 2. Theoretical development and hypotheses It is worth mentioning that other possible antecedents to the Firms involved in today's sophisticated supply chains face implementation of SSCM practices exist that may contribute to a increasing pressure from various stakeholders, especially govern- firm's decision to adopt sustainable practices in addition to gover- ments, to accept responsibility for social and environmental mat- nance pressures such as market orientation (Green et al., 2015). ters beyond their immediate organisational boundaries (Bostro €m However, this was not feasible within the scope of this paper, other et al., 2015). These regulatory pressures lead firms to pay close external antecedents to the implementation of SSCM practices attention to their commensurate environmental alignment in an deserves future research. SCM context in particular. Several organisational theories including resource dependence, transaction cost economics and institutional 2.2. SSCM practices theory have been used to understand how firms adopt and un- dertake environmental initiatives across their supply chains (Sarkis, In essence, a traditional supply chain is a set of business actions 2011; Niesten and Jolink, 2014; Lozano et al., 2015). Using the that directly involves the upstream or downstream flows of infor- theoretical anchors of the institutional theory, we argue that firms mation, products and services from a point-of-origin to a point-of- adopt SSCM practices in response to intuitional pressures mainly consumption (Lambert et al., 1998). Unlike the traditional supply exerted by governance. chain, the sustainable supply chain considers the environmental 68 A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 impacts of the production process as goods flow through the supply SSCM practices on performance outcomes (Zhu et al., 2013). Hence, chain (Hsu and Hu, 2009; Seuring and Müller, 2008b). Hence, a the contemporary knowledge of SSCM has been mixed on the re- sustainable supply chain extends: lationships between environmental and economic performance and adoption of SSCM practices, reporting inconclusive findings (Geffen […] the traditional supply chain to include activities that aim at and Rothenberg, 2000; Rao and Holt, 2005; Zhu et al., 2010). These minimizing environmental impacts of a product throughout its uncertainties and ambiguities within the existing knowledge in this entire life cycle, such as green design, resource saving, harmful area necessitate further empirical investigation. material reduction, and product recycle (Beamon, 1999, p.332) 2.3.1. Relationships between SSCM practices and environmental Sustainable supply chain management (SSCM) simply integrates performance environmental concerns into supply chain management (Linton Although early studies found little relationship between envi- et al., 2007). The literature in SSCM has been growing as both ronmental practice and environmental performance (e.g. Levy, practitioners and researchers begin to realise that the management 1995), in most cases the literature demonstrates that SSCM prac- of environmental programmes and operations do not end at the tices can improve environmental performance (Rao, 2005; Vachon boundaries of the organisation (Jolink and Niesten, 2015; Su et al., and Klassen, 2006a; Zhu et al., 2007). Particularly, recent studies 2015). Organisations may undertake a set of SSCM initiatives to tend to agree on a positive relationship between sustainable supply minimise the negative environmental impacts associated with the chain management and environmental performance (Hollos et al., entire lifecycle of its products or services, starting from design 2012; Taylor and Taylor, 2013). Some prominent research has through the acquisition of raw materials to consumption and offered insights into the potential patterns of supply-chain-based product disposal (Zsidisin and Siferd, 2001; de Bakker et al., 2002). relations for improving environmental performance (Zhu et al., The concept of SSCM has evolved to include boundary-spanning 2012; Lee et al., 2012). For example, Lee et al. (2012) suggested activities such as sustainable procurement (Carter and Carter, that inter-organisational linkage and collaboration with suppliers 1998), eco-design (Seuring and Müller, 2008b), sustainable distri- aid the adoption and development of innovative environmental bution (Vachon and Klassen, 2006b), and investment recovery (Zhu technologies and bring environmental improvement. Tsoulfas and and Sarkis, 2006; Zhu et al., 2007). Given the multi-dimensional Pappis (2006) also asserted that the interaction of customer and expansion of SSCM literature, our study focuses on four major supplier staff, partnership agreements and joint research and SSCM practices: sustainable production, sustainable design, sus- development lead to improvements in environmental perfor- tainable distribution and investment recovery. These four areas mance. Hence, we argue that there is consensus within the litera- represent the major implementation of SSCM practices, as they ture that the implementation of SSCM practices results in relatively encompass the main internal and external activities and functions improved environmental performance. within sustainable supply chain management (Zhu et al., 2005; Su In essence, the four major SSCM practices of sustainable pro- et al., 2015). Therefore, governance pressures as the main driving curement, sustainable design, sustainable distribution, and in- forces in the adoption of SSCM practices now can be linked to each vestment recovery are environmental-friendly initiatives in nature. of the four major SSCM practices in the form of coercive pressures These initiatives are designed to minimise a product's environ- encompassing environmental regulations. As previously discussed, mental impact without creating a negative trade-off with other we define environmental regulations as the coercive pressures performance dimensions, such as costs and functionality (Zhu et al., driving the implementation of SSCM practices. In light of this, we 2005; Green et al., 2012a). This is also in line with Porter and Van break down our primary hypothesis into four sub-hypotheses and der Linde's (1995) argument for a balance between a society's propose the following: desire for environmental protection and the economic burden on H1a. Coercive pressures are directly and positively associated industry. The implementation of each individual SSCM practice is with sustainable procurement. deemed to have direct environmental results as measured by re- ductions in air emissions, effluent waste, solid waste, and the H1b. Coercive pressures are directly and positively associated consumption of toxic materials (Zhu and Sarkis, 2004; Seuring and with sustainable distribution. Müller, 2008b). Therefore, the implementation of sustainable pro- H1c. Coercive pressures are directly and positively associated curement, sustainable design, sustainable distribution, and in- with sustainable design. vestment recovery is expected to result in improved environmental performance due to their capability in reduction of material con- H1d. Coercive pressures are directly and positively associated sumptions, wastes, emissions, energy usage, and excessive in- with investment recovery. ventory (Zhu and Sarkis, 2007; Seuring and Müller, 2008a). Accordingly, we propose that: 2.3. SSCM practices and performance outcomes H2a. Sustainable procurement is directly and positively associ- ated with environmental performance. Previous research has explored the relationships between the H2b. Sustainable distribution is directly and positively associated adoption of SSCM practices and performance outcomes, including with environmental performance. environmental, operational and economic performance (Zhu and Sarkis, 2007; Green et al., 2012a; Yu and Ramanathan, 2015). H2c. Sustainable design is directly and positively associated with Porter and Van der Linde (1995) contributed to the environment- environmental performance. competitiveness related relationships by developing a paradigm of H2d. Investment recovery is directly and positively associated balancing a society's desire for environmental protection with the with environmental performance. economic burden on industry. They advanced the idea that envi- ronmental regulations through coercive pressure can generate better economic outcomes in terms of innovation. However, recent emer- 2.3.2. Relationships between SSCM practices and economic gent studies have started to throw doubt on this issue, questioning performance the stance of previous literature concerning the positive impacts of In essence, SSCM practices mainly focus on the elimination of  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 69 wastes associated with material acquisitions, product design, de- antecedents and environmental and economic performance as livery and disposal (Vachon and Klassen, 2006a; Seuring et al., consequences. In addition, governance pressures in the form of 2008). Such waste minimisation should lead to reduced costs, environmental regulations drive the adoption of SSCM practices resulting in improved economic performance. Whether SSCM including sustainable procurement, sustainable distribution, sus- practices are positively or negatively related to economic perfor- tainable design, and investment recovery (Zhu et al., 2013). mance is still an open question (Vachon and Klassen, 2006b; Green Changes made as a result of governance pressures impact the et al., 2012a; Hollos et al., 2012). Some studies have shown that adoption of SSCM practices, which will impact both environmental environmental management programmes such as SSCM practices and economic performance. Fig. 1 outlines the theoretical model have a positive relationship with an organisation's economic per- that guides this research. formance (Rao and Holt, 2005). Such environmental management The relationships between governance pressures, implementa- practices can improve corporate reputation and customer satis- tion of SSCM practices and environmental and economic perfor- faction, which can in turn bring improved economic performance mance are theorised in order to first investigate whether the (Tan et al., 2014). In general, inter-firm linkage provides formal and exogenous driving force of governance exists in the adoption of informal mechanisms that promote trust, reduce risk and in turn SSCM practices and understand if it is necessary and sufficient; and increase cooperation, commitment, and hence profitability (Zhu second to assess the impact of such SSCM implementation, which et al., 2012). Furthermore, Seuring and Müller (2008b) and Sarkis has inwardly included the effects of the governance-based ante- et al. (2010) argued that the success in addressing environmental cedents, on performance outcomes. Definitions of the constructs issues may provide new opportunities for competition and new incorporated in the model are provided in Table 1. ways to add value to core business programmes. However, some other researchers have stated that economic performance is not 3. Method reaped in short-term profitability and sales performance with the implementation of SSCM practices (Bowen et al., 2001). 3.1. Measures In fact, the literature over the past two decades is still not clear whether benefits or costs dominate when adopting SSCM practices The measures of coercive pressures, SSCM practices and per- (Seuring and Müller, 2008a). The reasons for the variation in these formance components are developed on the basis of previous findings may be due to the heterogeneity of the types of environ- studies (e.g. Zhu and Sarkis, 2006; Sarkis et al., 2010). Zhu et al. mental management practices adopted by the firm and industry (2013) developed the measuring components of coercive pres- (Zhu et al., 2005). Given the strength of the overall literature in sures with reference to the isomorphic forces within institutional supporting improved economic performance when SSCM practices theory mainly associated with coercive forces (DiMaggio and are implemented, we postulate the following hypotheses. Powell, 1983). We adopt Zhu et al.'s (2013) measurement items for the coercive pressures construct. Furthermore, Zhu and Sarkis H3a. Sustainable procurement is directly and positively associ- (2006) developed and tested a measurement model for SSCM ated with economic performance. practice implementation and performance outcomes. They found H3b. Sustainable distribution is directly and positively associated four underlying constructs which reflect the dimensions of SSCM with economic performance. practices and performance measures, including economic, opera- tional and environmental performance. In this paper, we select two H3c. Sustainable design is directly and positively associated with dimensions, i.e. environmental performance and economic per- economic performance. formance. We adopt the measurement items of Zhu and Sarkis H3d. Investment recovery is directly and positively associated (2006) for the “sustainable procurement” and “investment recov- with economic performance. ery” constructs. For “sustainable distribution” and “sustainable design”, we utilise additional items found in Esty and Winston (2009) and Green et al. (2012b). 2.3.3. Relationship between environmental and economic The measurement items for evaluating coercive pressures, SSCM performance practices and performance outcomes are summarised in Appendix As the measurement scales of the environmental and economic A. These measurement items have been operationalised in a survey performance are constructed, economic performance reflects sav- questionnaire to evaluate performance outcomes as a result of ings that result from improved environmental performance (Zhu SSCM practice implementation driven by governance pressures. and Sarkis, 2006). The cost saving nature of environmental per- Respondents were asked to evaluate the importance of each coer- formance should lead to improved economic performance, sus- cive pressure on the adoption of SSCM practices, using a five-point taining decreases in the associated costs (Green et al., 2012a). Likert scale ranging from 1 ¼ unimportant to 5 ¼ very important. In Therefore, an additional hypothesis linking environmental perfor- addition, a five-point interval scale for evaluating individual SSCM mance with economic performance is proposed: practice items was provided, ranging from 1 ¼ no implementation H4. Environmental performance is directly and positively associ- to 5 ¼ implemented fully. Respondents were further asked to ated with economic performance. evaluate the significance level of performance improvement due to SSCM practices with a five-point scale ranging from 1 ¼ not at all to 5 ¼ significant. To avoid confusing respondents with three different 2.4. Theoretical model five-point Likert scales, we provided a brief explanation of the three groups of items at the beginning of each survey section. The theoretical model is a path analytical model with six latent variables (see Fig. 1): Coercive pressures, sustainable procurement, 3.2. Sampling sustainable distribution sustainable design, investment recovery, environmental performance and economic performance. Each of Given the fact that resource consumption, waste production and the hypotheses depicted in our research model is theorised as being implementation of environmental management practices are direct and positive. Generally, SSCM practices are the focal con- mostly associated with manufacturing, our survey focused on structs in the theorised model, with governance pressures as manufacturers, particularly those in the chemical, electronic, 70 A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 Fig. 1. Governance pressures-SSCM practices-performance model with hypotheses. Table 1 Construct definitions. Construct Definition Coercive pressures Pressures exerted from governance in the form of environmental regulations and legislations which enhance a firm's environmental alignment (Zhu et al., 2013) Sustainable Sustainable procurement focuses on cooperating with suppliers for the purpose of developing products that are environmentally sustainable procurement (Carter and Carter, 1998; Zhu and Sarkis, 2006) Sustainable Sustainable distribution refers to any means of transportation of products or services from suppliers to manufacturers to final customers with the distribution purpose of having the least possible negative environmental impact (Zhu and Sarkis, 2006; Esty and Winston, 2009; Green et al., 2012b) Sustainable design Sustainable design requires that manufacturers design products that minimise consumption of materials and energy, facilitate the reuse, recycling and recovery of component materials and parts, and avoid or reduce the use of hazardous products within the manufacturing process (Zhu and Sarkis, 2006) Investment recovery Investment recovery refers to the process of recovering the value of unused or end of life assets through effective reuse or surplus sales. It requires the sale of excess inventories, scraps and used materials and excess equipment (Zhu et al., 2007) Environmental Environmental performance relates to the ability of manufacturing plants to reduce air emissions, effluent waste and solid wastes and the ability performance to decrease consumption of hazardous and toxic materials (Zhu and Sarkis, 2006) Economic performance Economic performance relates to the manufacturing plant's ability to reduce costs associated with purchased materials, energy consumption, waste treatment, waste discharge and fines for environmental accidents (Zhu and Sarkis, 2006) automotive and mechanical engineering sectors. The (Bristol Online Surveys) from March 2014 to April 2014 and data manufacturing firms in these sectors are the main polluters and was collected from 146 qualified managers from various resource consumers. These manufacturers have been traditionally manufacturing firms. The data collection process was managed by associated with higher-than-average resource consumption, waste Bristol Online Surveys and was structured so as to ensure unique generation and implementation of environmental management responses from validated managers of UK manufacturers. The practices (Zhu et al., 2013). Considering this manufacturing focus, theorised model was assessed following a Structural Equation data was collected from a sample of manufacturing managers Modelling (SEM) method. Hair et al. (2010) suggested that sample working for the UK manufacturers. The UK was chosen as the sizes from 150 to 400 are generally suitable for SEM analysis, with empirical setting for this study because of the regional importance sample size varying according to the complexity of the model and of British manufacturers in terms of their share of Europe's total the number of parameters to be estimated. Hence, we targeted a manufacturing outputs and resource requirements. Anticipating sample size of 200 so we could safely reach the minimum threshold the difficulties associated with recruiting knowledgeable and of 150 required for SEM analysis given the practical implications of experienced respondents, we employed convenience sampling to data collection. target managers with knowledge of their firm's SSCM practices and Although response rates vary, Nulty (2008) asserted that the organisational performance (Malhotra and Grover, 1998). The average response rate to online surveys in social research is questionnaire survey was administered using convenience sam- generally 33 per cent. With this in mind, we set the threshold of the pling to a subset of the population of manufacturing managers, e.g., participation requests to 600, as this could potentially provide us plant managers, logistics managers, operations managers, pur- with 200 responses (600  (33/100)) which falls within the rec- chasing managers, supply chain managers, sales managers, engi- ommended range. A total of 600 managers were contacted via neering managers and industrial waste managers. email, 36 were screened out as non-managers and 194 managers completed the survey. Of the 194 respondents, 48 selected the “other manager” category. Because of concerns related to a lack of 3.3. Data collection knowledge of sustainable supply chain management practices and organisational performance, data from these 48 was not included in The surveys were conducted via a Web-based survey service  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 71 the dataset analysed. Finally, data from 146 manufacturing man- Harman's one-factor test was also used as a post-hoc statistical agers who have the necessary knowledge to fully complete the test to examine the possibility of the CMB problem (Podsakoff and survey was included in the dataset that is subsequently analysed. Organ, 1986). In this test, a principal component factor analysis was The effective response rate, therefore, is 25 per cent (146/ performed using Varimax rotation method with all research vari- (60036)). ables in the model. The results of the factor analysis revealed that All of the respondents hold management positions in four factors explain 73.16% of the variance of the variables with manufacturing firms. The majority of respondents (43 per cent) are 26.44% by the first extracted factor. According to Podsakoff and operations and supply chain managers. The respondents selected Organ (1986) and Inman et al. (2011), substantial common 11 different industry classifications representing a diverse array of method bias is signalled by the emergence of either a single factor manufacturing firms. They are experienced in their current posi- or one “general” factor that explains a majority of the total variance. tions, with an average of 6.85 years work experience. They work for firms with an average of 594.62 employees. According to the EU 3.5. Non-response bias criteria, firms are classified as large ones if they have 250 em- ployees or more and small/medium ones if they have fewer than To test for non-response bias, the responses of early and late 250 employees (Gimenez et al., 2012). We sought to target large waves of returned surveys were compared, based on the assump- manufacturing firms because they are likely to have undertaken tion that the opinions of late respondents are representative of the some sustainable supply chain initiatives (Zhu et al., 2008). The opinions of the theoretical non-respondents (Rogelberg and sample is diverse as intended and is made up of individuals with Stanton, 2007). Respondents were categorised as responding to knowledge of SSCM initiatives and organisational performance. either the initial or follow-up requests that were sent approxi- Table 2 displays the sample demographics. mately two weeks later. Of the respondents, 61 per cent (89) were categorised as early respondents and 39 per cent (57) were cat- egorised as late respondents. A comparison of the means of the 3.4. Common method bias demographic variables and the summary variables for the two groups was conducted using one-way ANOVA. The comparisons We have considered that common method bias (CMB) might resulted in statistically non-significant differences at the 0.01 level. arise in our survey as a potential threat in survey research. To avoid Since non-respondents have been found to descriptively resemble the ‘item characteristic’ effect as one of the key causes for common late respondents (Lambert and Harrington, 1990; Green et al., method bias (Podsakoff et al., 2003) due to ambiguous items that 2012a), this finding of general equality between early and late re- can result in unreliable answers, we pretested the measurement spondents indicates that non-response bias has not impacted the items in the questionnaire survey by interviewing a number of data set, suggesting that non-response bias was not a problem. plant-level manufacturing managers from the above-mentioned industrial sectors. We sought to determine whether the question- 4. Analysis and results naire items could be fully understood and if more items should be included to ensure that the questionnaire was practically capable of Because of our objective to assess the theorised model as a obtaining answers for the research inquiries. We made minor whole, we opted to assess the model using the Structural Equation modifications based on interviewees' feedback, mainly on how to Modelling (SEM) method. LISREL 8.80 software was used to better present the measurement items in the questionnaire. perform both the confirmatory factor analysis necessary to assess Furthermore, as a result of these minor adjustments, a number of the measurement model and the structural analysis necessary to items that were overlapping were clarified to eliminate potential assess the theorised model, because the software package can show confusion. the important model fit information. Table 2 4.1. Scale assessment process Sample demographics summary. The measurement scales were assessed for reliability and val- Title Number idity. Since all scales were taken directly from prior research (Zhu Plant manager 16 and Sarkis, 2006; Esty and Winston, 2009; Green et al., 2012b; Logistics manager 19 Operations manager 28 Zhu et al., 2013), content validity was assumed. Following Purchasing manager 13 Gerbing and Anderson (1988), each pair of scales under consider- Supply chain manager 36 ation was tested for discriminant validity using a chi-square dif- Sales manager 11 ference test. Chi-square difference tests for pairings of each scale Engineering manager 17 with other scales returned significant differences at the 0.01 level, Industrial waste manager 6 Total 146 indicating sufficient discriminant validity for all scales (Garver and Industry classification (UK SIC e Standard Industrial Classification) Mentzer, 1999; Gerbing and Anderson, 1988). Manufacture of food products 14 The construct validity of the theoretical constructs was also Manufacture of beverages 6 assessed using exploratory factor analysis (EFA) to “determine Manufacture of wood and of products of wood and cork 5 Manufacture of paper and paper products 8 whether the majority of the variance can be accounted for by one Manufacture of chemicals and chemical products 17 general factor” (Podsakoff et al., 2003, p. 890). As a result of Manufacture of basic metals 12 exploratory factor analysis, the “Sustainable Distribution” construct Manufacture of electrical equipment 23 (SDIST6) and “Environmental Performance” construct (ENV6 and Manufacture of machinery and equipment 14 ENV7) measurement items were excluded (see Appendix A for Manufacture of motor vehicles, trailers and semi-trailers 42 Manufacture of other transport equipment 3 details of the measurement scales). Consequently, all of the con- Other manufacturing 2 structs yield an average factor loading of 0.84, exceeding the rec- Total 146 ommended 0.70 level, providing evidence of self-reporting scales Mean years in current position 6.85 (Kline, 1998). This confirmed that all of the remaining measure- Mean number of employees 594.62 ment items in each construct represent one factor, indicating 72 A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 sufficient construct validity (Hair et al., 2010). Table 4 Furthermore, all standardised coefficients for scale items pre- Descriptive statistics. sented in Table 3 exceed the recommended 0.70 minimum and are Minimum Maximum Mean Std. Deviation Skewness Kurtosis significant at the 0.01 level, indicating sufficient convergent validity CP 2.17 4.62 3.584 0.563 0.259 1.020 (Garver and Mentzer, 1999). All of the reliability coefficients SP 1.88 4.42 3.015 0.418 0.183 1.168 (Cronbach alpha values) for all of the measurement scales exceed SD 2.37 5.00 3.369 0.554 0.496 0.873 the recommended 0.70 level, indicating sufficient reliability SDIST 2.10 4.75 3.128 0.425 0.385 1.119 IR 1.47 3.52 2.313 0.539 0.746 0.086 (Garver and Mentzer, 1999). The standardised coefficients for ENV 2.00 4.79 3.367 0.433 0.552 0.479 measuring items and their associated t-values along with the ECP 2.00 4.56 3.086 0.526 0.384 0.787 Cronbach alpha values for all of the scales are displayed in Table 3. 4.2. Measurement model assessment The relative chi-square value is less than the 3.00 maximum recommended by Kline (1998) and the RMSEA value is below the The measurement items were also assessed within the context recommended maximum of 0.08 (Schumacker and Lomax, 2004). of the full measurement model using confirmatory factor analysis Byrne (1998) points out that the Comparative Fit Index (CFI) and (CFA), as recommended by Koufteros (1999). The results of the Incremental Fit Index (IFI) are more appropriate when the sample confirmatory factor analysis (CFA) are in Table 3. The measurement size is small. The CFI (0.981) and IFI (0.988) both exceed the rec- model fits the data well, with a relative chi-square (chi-square/ ommended 0.90 level (Byrne, 1998). The results relating to fit of the degrees of freedom) value of 1.368, a root mean square error of model generally support a claim of good fit. Furthermore, none of approximation (RMSEA) value of 0.052, a comparative fit index the standardised residuals exceeds the 4.00 maximum recom- (CFI) value of 0.981, an incremental fit index (IFI) value of 0.988, and mended by Hair et al. (2010), suggesting that there is no concern a non-normed fit index (NNFI) value of 0.964. regarding a potential unacceptable degree of error. Table 3 displays the measurement model results along with the model fit indices. Table 3 4.3. Structural equation modelling results Measurement model results. Constructs Alpha Standardised coefficient t-values Multicollinearity can be a threat for structural models, as it can Coercive pressures (CP) 0.745 distort the effects of an individual construct, leading to incorrect CP1 0.90 14.33 estimations of regression weights (Hair et al., 2010). As such, we CP2 0.84 12.82 tested for multicollinearity by calculating the variance inflation CP3 0.82 12.26 factors (VIF1) for each regression coefficient before reporting hy- CP4 0.88 14.16 CP5 0.92 14.94 pothesis testing results. A threshold of a VIF value that is less than CP6 0.82 12.18 or equal to 10.0 is a commonly used criterion to determine the Sustainable procurement (SP) 0.816 presence of multicollinearity (Hair et al., 2010). The VIF values SP1 0.83 12.46 range from 1.136 to 1.874, significantly below the recommended SP2 0.85 13.24 SP3 0.85 13.50 threshold of 10.0, hence suggesting that multicollinearity does not SP4 0.79 11.39 pose a problem to our model. SP5 0.88 14.11 Summary values for the research variables were computed by SP6 0.89 14.25 averaging the items in the scales. Descriptive statistics are pre- Sustainable distribution (SDIST) 0.738 sented in Table 4. All variables are sufficiently normally distributed SDIST1 0.82 12.41 SDIST2 0.86 13.66 with skewness and kurtosis coefficients within the 2.00 SDIST3 0.86 13.58 and þ2.00 range (Field, 2009). Furthermore, correlation coefficients SDIST4 0.78 11.28 are positive and significant at the 0.01 level for all variable pairings. SDIST5 0.80 11.62 The correlations are presented in Table 5. Sustainable design (SD) 0.853 SD1 0.90 14.69 The theorised model was assessed with the Structural Equation SD2 0.89 14.22 Modelling (SEM) technique. Fig. 2 illustrates the model with the SD3 0.86 13.77 SEM results specified in the LISREL 8.80 output. SD4 0.92 15.52 The SEM results related to individual hypothesis tests are dis- SD5 0.88 14.13 played in Table 6. All hypotheses are positive and significant with SD6 0.88 14.05 Investment recovery (IR) 0.705 the exception of H3b (SDIST / ECP), H3c (SD / ECP), and H3d (IR IR1 0.77 10.84 / ECP). H1a, H1b, H1c and H1d are positive and significant, indi- IR2 0.78 11.08 cating that coercive pressures are deemed to be powerful pre- IR3 0.78 11.20 cursors that give rise to the implementation of SSCM practices. H2a, Environmental performance (ENV) 0.795 ENV1 0.84 12.71 H2b, H2c and H2d, which predict positive associations between ENV2 0.85 13.15 SSCM practices and environmental performance, are positive and ENV3 0.82 12.32 significant as expected. However, H3a, H3b, H3c and H3d, which ENV4 0.81 12.08 predict positive associations between SSCM practices and eco- ENV5 0.90 14.48 nomic performance, have another version of story: the sustainable Economic performance (ECN) 0.788 ECP1 0.87 13.93 procurement to economic performance link is significant and ECP2 0.82 12.33 positive; Sustainable distribution and investment recovery do not ECP3 0.91 15.06 ECP4 0.88 14.19 ECP5 0.81 11.89 1 The variance inflation factor (VIF) quantifies the severity of multicollinearity in Notes: Chi-Square Ratio ¼ 1.368; RMSEA ¼ 0.052; NFI ¼ 0.95; NNFI ¼ 0.96; regression analyses and indicates the degree to which each predictor variable is CFI ¼ 0.98; IFI ¼ 0.98. explained by other predictor variables (Hair et al., 2010).  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 73 Table 5 Table 6 Correlation matrix. Structural model results. CP SP SDIST SD IR ENV ECP Model link Standardised Hypothesis tests results coefficient CP 1 SP 0.672** 1 Hypothesis tests SDIST 0.588** 0.536** 1 CP SD 0.691** 0.544** 0.610** 1 SP 0.67** H1a: Supported IR 0.345** 0.364** 0.377** 0.442** 1 SDIST 0.62** H1b: Supported ENV 0.536** 0.578** 0.616** 0.638** 0.486** 1 SD 0.71** H1c: Supported ECP 0.484** 0.604** 0.593** 0.615** 0.407** 0.619** 1 IR 0.33* H1d: Partially supported SP Notes: ** indicates significance at the 0.01 level; CP¼Coercive Pressures; ENV 0.42** H2a: Supported SP¼Sustainable Procurement; SDIST¼Sustainable Distribution; SD¼Sustainable ECP 0.36** H3a: Supported Design; IR¼Investment Recovery; ENV ¼ Environmental Performance; SDIST ECP ¼ Economic Performance. ENV 0.30** H2b: Supported ECP 0.26 ns H3b: Not supported SD impact on economic performance; and sustainable design nega- ENV 0.56** H2c: Supported tively impacts on economic performance. Lastly, H4 is positive and ECP e 0.28** H3c: Not supported significant as expected indicating that environmental performance IR ENV 0.23** H2d: Supported positively impacts economic performance. It is worth noting that ECP 0.17 ns H3d: Not supported Zhu and Sarkis (2006) did not originally consider the relationship ENV between environmental and economic performance. This paper ECP 0.59** H4: Supported advanced their work by developing an inter-linkage between these Notes: * significant at the 0.05 level; ** significant at the level 0.01; ns: not signif- performance constructs. icant; Chi-Square ratio ¼ 1.492; RMSEA ¼ 0.059; NFI ¼ 0.919; NNFI ¼ 0.938; Furthermore, in order to further examine the impact of firm size CFI ¼ 0.970; IFI ¼ 0.970; CP ¼ Coercive Pressures; SP ¼ Sustainable Procurement; on relationships between embedded constructs, we tested if the SDIST ¼ Sustainable Distribution; SD ¼ Sustainable Design; IR ¼ Investment Re- covery; ENV ¼ Environmental Performance; ECP ¼ Economic Performance. results of the regression coefficients differ between large and me- dium/small firms. The results show that the coefficient for size is insignificant, hence indicating that the effects of the SSCM practices implementation driven by governance pressures and economic on performance outcomes are similar irrespective of the firm size. performance is less clear-cut. Overall, the results show that struc- tural relationships between governance pressures, SSCM practices, 5. Discussion environmental and economic performances exist, supporting further development of more proactive environmental practices Whilst not all of the individual hypotheses are supported, the within the context of supply chain management. They also support constructs stand together reasonably well, grounded on the good fit the merits of the idea that governance gives rise to the adoption of of the structural model and the statistical support for the majority sustainable supply chain management which is capable of deliv- of the hypotheses. We believe that our model is a good represen- ering environmental and ecological benefits. tation of the relationships among the constructs. The empirical results demonstrate a significant linkage between governance 5.1. Governance pressures and SSCM practices pressures, the implementation of SSCM practices and the envi- ronmental performance of manufacturing firms operating in the This study shows that coercive pressures generally relate to UK UK. However, the results related to the linkage between SSCM manufacturers implementing SSCM practices including sustainable Fig. 2. SEM results on the antecedent and performance outcomes of implementing SSCM. 74 A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 procurement, sustainable distribution, sustainable design and in- of environmental performance. However, the results about the vestment recovery. Coercive pressures directly impact all of the impact of SSCM implementation on economic performance are less SSCM practices and indirectly impact firms' environmental and clear-cut. In particular we found that sustainable design is nega- economic performances through these SSCM practices. Overall, tively related to economic performance. coercive pressures are found to be major driving forces that lead The empirical results show that sustainable procurement posi- manufacturers to pursue SSCM practices, indicating the critical role tively and significantly impacts on both environmental and eco- of governance in the adoption of an SSCM agenda. nomic performance, with standardised coefficients b ¼ 0.42 (sig. at The statistical results indicate that coercive pressures are most the 0.01 level) and b ¼ 0.36 (sig. at the 0.01 level), respectively. significantly associated with sustainable design, with standardised Generally, the practice of sustainable procurement is environ- coefficients b ¼ 0.71 (sig. at the 0.01 level), followed by sustainable mentally friendly in nature and facilitates the development of procurement (b ¼ 0.67, sig. at the 0.01 level) and sustainable dis- environmentally friendly products and services, decreasing the tribution (b ¼ 0.62, sig. at the 0.01 level). This observation is levels of wastes and emissions. From an economic standpoint, the consistent with the findings of previous studies (Zhu et al., 2013), responsibility for sustainable procurement may lie with the sup- and can be explained by the fact that most of the environmental pliers rather than the manufacturers. Thus it is less costly for impact on a product and its processes are ‘locked’ into the product manufacturers to implement than other SSCM practices (Sarkis at the design stage when materials are selected and product per- et al., 2010; Hollos et al., 2012). formance is largely determined, irrespective of where the product On the other hand, the practice of sustainable design positively lies in the product life cycle (Lewis and Harvey, 2001). Therefore, impacts environmental performance (b ¼ 0.56, sig. at the 0.01 environmental regulatory bodies mainly attempt to enact strict level), and negatively and significantly impacts economic perfor- environmental regulations at the product design stage as the most mance (b ¼ 0.28, sig. at the 0.01 level). These results are in line effective measure to mitigate the product's negative environmental with the findings of Zhu and Sarkis (2007), Green et al. (2012a) and impact (Yu and Ramanathan, 2015). Zhu et al. (2013). According to Grote et al. (2007, p. 4100), the aim of Coercive pressures are also significantly and positively associ- sustainable design is to “reduce a product's environmental impacts ated with sustainable procurement and sustainable distribution. without creating a negative trade-off with other design criteria, During the past few years, the UK government, in order to improve such as functionality and costs”. It appears, then, that sustainable sustainable distribution and better design of processes and logis- design has not fully accomplished this intended aim. We therefore tics, has enacted stringent regional and national environmental suggest that the current sustainable design requires further regulations to limit the use of non-renewable resources such as development and improvement. Furthermore, another reason for diesel and petrol (Yu and Ramanathan, 2015). The UK government this negative relationship between sustainable design and eco- has also offered financial incentives such as grants and tax re- nomic performance may lie with the fact that eco-design requires ductions to encourage manufacturers to embark upon using bio- capital investment (Zhu et al., 2013). In addition, the capacity of fuels for their transportation systems (Taylor and Taylor, 2013). sustainable design to reduce environmental pollutants is counter- Furthermore, regional and national resource saving and conserva- balanced by the increases in the associated costs, perhaps related to tion regulations and their associated compliance issues have led the materials purchases (Zhu et al., 2012; Green et al., 2012a). UK manufacturers to pursue sustainable procurement, which al- Sustainable distribution directly impacts environmental per- lows the initiation of the development of environmentally friendly formance but does not significantly impact economic performance. products or services. Coercive pressure is partially associated with This can be explained by the fact that the practice of sustainable investment recovery, with standardised coefficients b ¼ 0.33 (sig. at distribution focuses on decreasing the levels of environmental the 0.05 level), This can be explained by the fact that the practice of pollutants, which can directly enhance the environmental perfor- investment recovery is more external to a firm, making the firm mance. From an economic perspective, the lack of appropriate have less control over the implementation of this practice (Zhu green infrastructures hinders the benefits of sustainable distribu- et al., 2012). Exogenous driving forces of governance may not tion (e.g. profitability and sales performance). It also requires more greatly influence manufacturing firms to embark upon external supporting technologies for green distribution initiatives (Zhu SSCM practices. Such empirical results are generally consistent et al., 2007). This necessitates further infrastructure investment with the findings of previous studies (Zhu et al., 2007, 2013). in order to tackle the potential lack of green capabilities and green Overall these findings suggest that exogenous pressures of distribution characteristics. While Zhu and Sarkis (2007) do not governance cause manufacturing firms to embark upon SSCM find sustainable distribution to be significantly linked to either implementation and should be in place before SSCM practices are economic or environmental performance, a recent study by Green expected to be adopted. In effect, the exogenous driving forces of et al. (2012b) reports that sustainable distribution directly im- governance are found to be necessary for the adoption of SSCM pacts on environmental performance, which is in line with our practices. Therefore, governance pressures are identified as neces- findings. This consolidates our findings, providing evidence of sary precursors to the implementation of SSCM practices: sus- consensus within existing knowledge. tainable procurement, sustainable distribution, sustainable design Investment recovery also positively impacts on environmental and investment recovery. performance but does not significantly impact on economic per- formance. This may be explained by the fact that the practice of 5.2. SSCM and performance outcomes investment recovery has received less attention in the UK (Yu and Ramanathan, 2015). Interestingly, our finding about the relation- As the model depicts (see Fig. 2), all of the SSCM practices are ship between investment recovery and environmental perfor- positively and significantly associated with environmental perfor- mance is opposite to that of Zhu and Sarkis (2007), where the mance. This indicates that SSCM practices lead to improved envi- authors reported that investment recovery is positively associated ronmental performance. These results are in line with the work of with economic performance but not significantly associated with Zhu and Sarkis (2007), Lee et al. (2012) and Green et al. (2012a). environmental performance. The difference may be attributed to Generally, SSCM practices are designed with a focus on minimising the different sample sizes, regions and segments. However, our environmental impacts. Such environmental impact minimisation finding is consistent with the findings of more recent work by Zhu should have direct environmental results and lead to higher levels et al. (2013) and Green et al. (2012a). Overall, the empirical findings  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 75 suggest that the adoption of SSCM practices leads to higher levels of the implementation of SSCM practices on performance outcomes environmental performance, resulting in environmental improve- while taking into account the influential effects of governance ments, but does not necessarily lead to improved economic per- pressures as driving forces, generating more credible findings. formance, as only sustainable procurement has a positive impact on Generally, our analysis shows that structural relationships exist economic performance. This indicates that SSCM implementation among governance pressures, SSCM practices, and performance driven by governance pressures increases environmental and outcomes. The results reveal that coercive pressures have a signif- ecological benefits and potentially has the capacity to enhance icant positive effect on SSCM practices, hence suggesting that economic performance. governance pressures exist in the adoption of SSCM practices and Lastly, the environmental performance construct positively are deemed necessary for the implementation of SSCM practices. impacts economic performance (b ¼ 0.59, sig. at the 0.01 level). The These results indicate that the governance pressures for environ- relationship among the performance constructs seems to make mental protection have driven UK manufacturers to embark upon logical sense, as the economic performance construct reflects sav- environmentally orientated organisational initiatives across the ings that result from improved environmental performance (Zhu supply chain and implement SSCM practices. Therefore, we and Sarkis, 2006). The impact of the implementation of SSCM conclude that governance pressures are necessary precursors to the practices on economic performance can be further explained successful implementation of SSCM practices, and should be in through the linkage among environmental and economic perfor- place before these environmental initiatives are expected to be mance. In view of this, the sustainable distribution, sustainable adopted. The empirical results further suggest that the imple- design, and investment recovery constructs that do not directly and mentation of SSCM driven by governance pressures leads to a positively impact economic performance, indirectly impact eco- higher level of environmental performance, while the economic nomic performance through environmental performance. performance is somewhat compromised. We find that SSCM practices positively affect the environmental performance of UK 6. Implications and conclusions manufacturing firms, resulting in environmental improvements. However, we find evidence that the adoption of sustainable prac- The findings of this study effectively achieve the research aims, tices across the supply chain does not necessarily lead to improved shedding some important light on the merits of governance within economic performance, as only sustainable procurement positively the sustainable supply chain management in terms of ecological affects economic performance. benefits and performance improvements. The results of this study Overall, the results of our research clarify the proposition that suggest a number of interesting insights concerning sustainable the role of governance in the adoption of SSCM practices is indeed supply chain research. First, this paper theorises and assesses a environmentally necessary, but that its merits might not be being comprehensive Governance Pressures/SSCM Practi- reaped in terms of short-term profitability. While the short-term ces/Performance Outcomes model using the Structural Equation benefits of governance within sustainable supply chain may not Modelling (SEM) method to investigate whether exogenous pres- be evident, long-term benefits can be accrued. Governance plays a sures of governance exist in SSCM adoption and thereafter what critical role in the transition to a more sustainable society and is this can deliver in terms of performance outcomes. It is our belief capable of bringing not only ecological and environmental benefits that one of the major contributions of this study lies in the but also economic benefits by supporting financial incentives in the comprehensive nature of the theorised model, as it is capable of form of subsidies or tax reductions to encourage environmental assessing the merits of governance within a sustainable supply management. This promises to allow manufacturing firms to bal- chain context, covering both ecological benefits and performance ance economic benefits with environmental protection and further gains. Second, this research extends the literature on sustainable ensures a ‘win-win’ opportunity for the supply chain partners, supply chain management by applying the insights of governance minimising potential trade-offs between environmental and eco- to bridge sustainable supply chain governance (SSCG). The contri- nomic performance. bution to SCM knowledge is to demonstrate the linkage between The findings of our study are generally consistent with the SSCM and SSCG by exhibiting the theoretical linkages between majority of prior investigations, and where contradictory results governance pressures, SSCM practices and commensurate perfor- exist, our findings stand with more recent studies reporting similar mance outcomes. results. This consolidates our findings and diminishes any potential This paper also makes a significant contribution to on-going contradictory directions. Hence, we conclude that this paper has research that relates sustainable practices along the supply chain reported relatively conclusive results on the topic of governance to performance outcomes, through the inclusion of exogenous within sustainable supply chains and its commensurate perfor- driving forces of governance as a major antecedent to the adoption mance gains. of such environmental initiatives. Thus, this paper bridges the existing gap surrounding the lack of empirical evidence in under- 6.1. Managerial implications standing the relationships between governance pressures, SSCM practices and performance outcomes from a comprehensive This study provides practical implications for both practitioners perspective. Furthermore, adopting Structural Equation Modelling in the manufacturing context and policy-makers. Manufacturers (SEM) technique over the traditional path analysis has enabled this are given insights into how they can gain improved economic study to obtain more credible results by taking into account the performance from implementing SSCM practices. For instance, influential effects of governance exogenous forces in examining the manufacturers need to understand the structural relationships relationships between SSCM practices and performance outcomes. between governance arrangements and internal and external as- The significance of our study can be further extended, as this is one pects of implementing SSCM and ensure the coordination of their of a few empirical studies that bridges the mixed views on SSCM respective activities to arrive at better environmental performance performance implications, by reporting relatively conclusive re- for economic gains to be achieved. Public policy-makers and reg- sults, reaching consensus on the recent findings conducted by ulators can further understand how to motivate manufacturers to various authors. Moreover, the methodological contribution of this embark upon SSCM adoption. In particular, coercive forces are research broadly lies in the capability of the employed data analysis influential antecedents affecting the implementation of environ- technique, i.e. SEM analysis, to rigorously examine the impacts of mental initiatives in manufacturing. It is useful for the government 76 A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 and related bodies to promote SSCM by creating awareness of the forces for sustainability adoption. Furthermore, internal support at benefits and sharing successful experiences. Such promotion can senior management level is also found to be a key precursor to the help to lessen firms' doubts about adopting SSCM and reduce the implementation of environmental management practices (Carter perception of the risks associated with the adoption of environ- et al., 1998). Future studies may carry out a comparative analysis mental innovations. This study informs policy-makers and regula- between exogenous driving forces and endogenous factors that tors that effective governance arrangements in the supply chain drive SSCM adoption, so that a more holistic view can be achieved. along with financial incentives can provide ‘win-win’ opportunities Second, research on SSCM is still at the early stage in developing for both environmental performance and economic benefits in appropriate measures of SSCM related organisational performance, implementing SSCM practices. particularly for economic performance (Tseng et al., 2015). This Furthermore, practitioners are provided with a validated study has only measured economic performance in terms of framework for assessing the synergistic impact of SSCM practices reduced cost and could have further considered other quantifiable on environmental and economic performances. In addition, the measurement criteria of economic performance such as increased SSCM initiatives validated in this work can help manufacturing profits, growth, brand image and so forth. Future research may use firms operating in both developed countries and emerging econo- and validate other economic measures than reduced costs to avoid mies to identify those areas of SSCM that require improvement and limiting the economic performance dimension to cost performance the prioritisation of their green efforts. This work can be useful for through integration of other measures of supply chain performance manufacturing industries that need to convert their traditional (He et al., 2013). Third, we further recommend longitudinal studies supply chains into sustainable supply chains. In the resource- to determine whether long-term economic performance is constrained environment of the EU, our framework points to the enhanced by implementing SSCM practices. Fourth, this study did key environmental initiatives in the supply chain which need to be not consider the aspects of organisational culture interaction be- implemented, i.e. sustainable procurement, sustainable design, tween key customers and suppliers in the sustainable supply chain sustainable distribution and investment recovery. Collectively, the context, which can provide future research opportunities. key SSCM initiatives can serve as an audit tool and later on as a Moreover, future studies may examine other developed markets benchmarking tool for managers to evaluate the perceptions of to eliminate the potential effect of country-level variance such as SSCM in their organisations. Policy-makers need to invest more in market size, economic development and legal systems. To increase appropriate infrastructures that enhance green capabilities and the generalisability of the research, repeating this study for expertise. This will also facilitate the benefits of SSCM practices comparative analysis in different developed countries would be being reaped in terms of long-term profitability and sales another research direction. In addition, future research can performance. examine emerging economies such as China, the world's largest manufacturing country, and compare the findings with developed markets. Lastly, this research has been developed primarily with a 6.2. Limitations and future directions focus on manufacturing firms, without consideration of other sec- tors such as the service sector and servitisation. Therefore, future As with any research, this study has some limitations that pro- studies may examine the applicability of our model to these sectors. vide opportunities and directions for further research. First, we acknowledge that customer and competitor pressures were excluded in developing our research model. Future research may Appendix consider other antecedents to the implementation of SSCM prac- tices such as customer and competitor pressures as motivating Table A1 Measurement scales. Coercive Pressures (Zhu et al., 2013) Please indicate the importance of each of the following on SSCM practices adoption. (Five-point scale: 1 ¼ unimportant, 2 ¼ somewhat unimportant, 3 ¼ neutral, 4 ¼ somewhat important, and 5 ¼ very important) CP1 National environmental regulations (such as waste emissions, cleaner production etc.). CP2 National resource saving and conservation regulations. CP3 Regional environmental regulations (such as waste emission, cleaner production etc.). CP4 Regional resource saving and conservation regulations. CP5 Export countries' environmental regulations. CP6 Products potentially conflict with laws (such as circular economy, EPR, EHS etc.). Sustainable Procurement (Zhu and Sarkis, 2006) Please indicate the extent to which you perceive that your company is implementing each of the following. (five-point scale: 1 ¼ no implementation; 2 ¼ planning to consider implementation; 3 ¼ currently considering implementation; 4 ¼ initiating implementation; 5 ¼ implementing fully) SP1 Eco labelling of products. SP2 Cooperation with suppliers for environmental objectives. SP3 Environmental audit for suppliers' internal management. SP4 Suppliers' ISO 14000 certification. SP5 Second-tier supplier environmentally friendly practice evaluation. SP6 Providing design specification to suppliers that include environmental requirements for purchased item. Sustainable Distribution (Zhu and Sarkis, 2006; Green et al., 2012b) Please indicate the extent to which you perceive that your company is implementing each of the following. (five-point scale: 1 ¼ no implementation; 2 ¼ planning to consider implementation; 3 ¼ currently considering implementation; 4 ¼ initiating implementation; 5 ¼ implementing fully) SDIST1 Cooperation with customers for using less energy during product transportation. SDIST2 Cooperation with customers for green packaging. SDIST3 Use of renewable energy in any mode of products transportation. SDIST4 Use of renewable energy in the process of products packaging.  A. Esfahbodi et al. / Journal of Cleaner Production 155 (2017) 66e78 77 Table A1 (continued ) SDIST5 Upgrade freight logistics and transportation systems (either software or hardware such as minimising empty miles, reducing container weight, improving refrigeration, etc.). SDIST6 Tracking and monitoring emissions caused in products distributions (e.g., carbon footprint). Sustainable Design (Zhu and Sarkis, 2006; Esty and Winston, 2009) Please indicate the extent to which you perceive that your company is implementing each of the following. (five-point scale: 1 ¼ no implementation; 2 ¼ planning to consider implementation; 3 ¼ currently considering implementation; 4 ¼ initiating implementation; 5 ¼ implementing fully) SD1 Design of products for reduced consumption of material. SD2 Design of products for reduced consumption of energy. SD3 Design of products for reuse, recycle, recovery of material, component parts, and by-products. SD4 Design of products to avoid or reduce use of hazardous materials in their manufacturing process. SD5 Cooperation with customers for eco design. SD6 Cooperation with customers for cleaner production. Investment Recovery (Zhu and Sarkis, 2006) Please indicate the extent to which you perceive that your company is implementing each of the following. (five-point scale: 1 ¼ no implementation; 2 ¼ planning to consider implementation; 3 ¼ currently considering implementation; 4 ¼ initiating implementation; 5 ¼ implementing fully) IR1 Sale of excess inventories or materials. IR2 Sale of scrap and used materials or by-products. IR3 Sale of excess capital equipment. Environmental Performance (Zhu and Sarkis, 2006) Please indicate the extent to which you perceive that your company has achieved each of the following during the past year. (five-point scale: 1 ¼ not at all; 2 ¼ a little bit; 3 ¼ to some degree; 4 ¼ relatively significant; 5 ¼ significant) ENV1 Reduction of air emission. ENV2 Reduction of waste emission. ENV3 Reduction of solid wastes. ENV4 Reduction of effluent wastes. ENV5 Decrease of consumption for hazardous/harmful/toxic materials. ENV6 Decrease of frequency for environmental accidents. ENV7 Improvement of an enterprise's environmental situation. Economic Performance (Zhu and Sarkis, 2006) Please indicate the extent to which you perceive that your company has achieved each of the following during the past year. (five-point scale: 1 ¼ not at all; 2 ¼ a little bit; 3 ¼ to some degree; 4 ¼ relatively significant; 5 ¼ significant) ECP1 Decrease of cost for purchased materials. ECP2 Decrease of cost for energy consumption. ECP3 Decrease of fee for waste treatment. ECP4 Decrease of fee for waste discharge. ECP5 Decrease of fine for environmental accidents. References Green, K.W., Toms, L.C., Clark, J., 2015. Impact of market orientation on environ- mental sustainability strategy. Manag. Res. Rev. 38 (2), 217e238. Green, K.W., Zelbst, P.J., Bhadauria, V.S., Meacham, J., 2012b. Do environmental Beamon, B.M., 1999. Designing the green supply chain. Logist. Inf. Manag. 12 (4), collaboration and monitoring enhance organizational performance? Ind. 332e342. €m, M., Jo€nsson, A.M., Lockie, S., Mol, A., Oosterveer, P., 2015. Sustainable and Manag. Data Syst. 112 (2), 186e205. Bostro Green, K.W., Zelbst, P.J., Meacham, J., Bhadauria, V.S., 2012a. Green supply chain responsible supply chain governance: challenges and opportunities. J. Clean. management practices:impact on performance. Supply Chain Manag. Int. J. 17 Prod. 107 (16), 1e7. (3), 290e305. Bowen, F.E., Cousins, P.D., Lamming, R.C., Faruk, A.C., 2001. The role of supply Grote, C., Jones, R., Blount, G., Goodyer, J.J., Shayler, M.M., 2007. An approach to the management capabilities in green supply. Prod. Oper. Manag. 10 (2), 174e189. EuP directive and the application of the economic eco-des