An Empirical Investigation of a Green Supply Chain Management Practices Scale
Confirmation of a Measurement Model for Green Supply Chain Management Practices Implementation
Qinghua Zhu Corresponding authorSchool of Management
Dalian University of Technology
Dalian, Liaoning Province (116024)
P.R. of China
Tel: 86-411-8470-7331
Fax: 86-411-8470-8342E-mail: [email protected] SarkisGraduate School of Management
Clark University
Worcester, MA 01610-1477
Phone: 508-793-7659
Fax: 508-793-8822Email: [email protected]
Kee-hung Lai
Department of Logistics
The Hong Kong Polytechnic University
Hung Hom, Kowloon, Hong Kong
Tel: 852-2766-7920
Fax: 852-2330-2704
E-mail: [email protected]: This study aims to empirically investigate the construct of and the scale for evaluating green supply chain management (GSCM) practices implementation among manufacturers. With data collected from 341 Chinese manufacturers, two measurement models of GSCM practices implementation were tested and compared by confirmatory factor analysis. Our empirical findings suggest that both the first-order and the second-order models for GSCM implementation are reliable and valid. Our study contributes to the literature on empirical examination of the construct of GSCM practices implementation and to the practices of managers with a validated measurement scale to evaluate their strengths and weaknesses in different facets of implementing GSCM practices in their organizations.
Key words: Green supply chain management; Construct; Practice measurement; Confirmatory factor analysis; Manufacturers
Confirmation of a Measurement Model for Green Supply Chain Management Practices Implementation1. Introduction
Environmentally sustainable (green) supply chain management (GSCM) has emerged as an important organizational philosophy to achieve corporate profit and market share objectives by reducing environmental risks and impacts while improving ecological efficiency of these organizations and their partners (van Hock and Erasmus, 2000). As a synergistic joining of environmental and supply chain management, the competitive and global dimensions of these two topics cannot go unnoticed by organizations. For example, multinational enterprises have established global networks of suppliers to take advantage of country-industry specific characteristics to build competitive advantage (Dunning, 1993). Simultaneously, due to stricter regulations and increased community and consumer pressures, manufacturers need to effectively integrate environmental concerns into their regular practices and onto their strategic planning agenda. As a result, integrating environmental concerns into supply chain management has become increasingly important for manufacturers to gain and maintain competitive advantage. Thus, the study of this topic is timely and necessary to better aid organizations in the management of GSCM principles.
To advance investigation and practice in GSCM, appropriate measurement scales are needed. In general, identification of appropriate measurement scales for emerging concepts and theories is necessary to complete robust research and to advance the body of knowledge in a field. Devellis (2003) states measurement is a fundamental activity of science and that measurements and broader scientific questions interacting with each other within their boundaries are almost imperceptible. The field of GSCM is arguably in its early development phases, both academically and practically. Academically, to effectively and empirically advance theory within this field, some useful and testable multi-item measurement scales are needed. Thus, greater attention will need to be focused on employing multi-item latent constructs, assessing them for content validity, and purifying them through field-based testing (Malhotra and Grover, 1998). Using literature in supply chain (operations) and environmental management, we introduce a number of scales that may be used to help evaluate practices in this area. Practically, organizations can also benefit from development of reliable and valid scales to measure GSCM practices implementation. Practitioners can use these scales for benchmarking, continuous improvement and project management activities when seeking to implement GSCM practices. One contribution of this study is to help manufacturers understand the different facets of GSCM practices implementation and identify the strengths and weaknesses of the implementation of their GSCM practices.
Given the academic (theoretical) and practical importance of developing a GSCM practices implementation measurement scale, we introduce a study based on an empirical survey of Chinese manufacturing organizations. Chinese manufacturers provide a microcosm of international markets. In some ways they are very advanced in practices and technology due to their relationships with international partners and community. In other ways, many Chinese organizations still lack the knowledge and economic resources of developed country institutions and function within a developing and emerging economy context. Thus, in one study we can include situations where the GSCM practice concepts are relatively novel to some organizations but concurrently more mature and acceptable for more advanced organizations (see Zhu et al., 2005, for examples of the diffusion of these concepts). Thus, measurement scales that have general agreement in this diverse economic environment, and across numerous industries, provide an extra degree of robustness.
The objective of this study is to investigate the GSCM practice implementation construct and its defining measurement items emphasizing Chinese manufacturers with broader implications for application of these scales to other environments (e.g. research in other countries and further development of environmental and supply chain research). In this paper we initially introduce literature reviews of measurement items for GSCM implementation in Section 2. The methodology used to develop and validate the GSCM implementation scale will be presented in Section 3. Section 4 will present results of this study followed by discussions and implications of these results in Section 5. Section 6 concludes our discussion by summarizing our findings, implications, limitations, and potential for future research.
2 Literature review of GSCM practices implementationGSCM has emerged as an effective management tool and philosophy for proactive and leading manufacturing organizations. The scope of GSCM practices implementation ranges from green purchasing to integrated life cycle management supply chains flowing from supplier, through to manufacturer, customer and closing the loop with reverse logistics. A number of definitions of GSCM exist (Zhu and Sarkis, 2004). Similar to the concept of supply chain management, the boundary of GSCM is dependent on researcher goals and the problems at hand, e.g., should it be just the procurement stage or the full logistics channel that is to be investigated (Lai et al., 2004)?Prescriptive models for measures of GSCM practices implementation with a focus on green purchasing and GSCM have been developed. Handfield et al. (2002) developed a decision model to measure environmental practice of suppliers using a multiattribute utility theory approach. Kainuma and Tawara (2006) proposed the multiple attribute utility theory method for assessing a supply chain including re-use and recycling throughout the life cycle of products and services. Using the tool of life-cycle assessment, Faruk et al. (2002) put forward aspects to measure GSCM practices implementation, that is, material acquisition, preproduction, production, use, distribution and disposal. Sarkis (2003) developed a strategic decision framework for GSCM practices implementation to evaluate alternatives adopted by companies that will affect their external relationships with suppliers and customers. Sheu et al. (2005) developed a linear multi-objective programming model which optimized the operations of both forward and reverse logistics in a given green supply chain. These models and frameworks included and defined a variety of characteristics, attributes, and scales for GSCM practices implementation, yet none attempted to rigorously validate these scales.
Empirically, Carter et al. (1998) developed and validated a scale to measure environmental purchasing, and compared the activities between US and German purchasing managers. They put forward six key factors related to green purchasing, including top management support, middle management support, firms mission, department goals, training for personnel to buy environmentally friendly input, and evaluation of purchasing management on green purchasing. Using confirmatory factor analysis, they empirically demonstrated that middle management support and department goals are related to green purchasing, while the relationship between training for personnel and green purchasing was only supported in US firms. Their measurement scale was thus limited to green purchasing and activities surrounding its management.Using case studies of five companies in the furniture industry, Walton et al. (1998) identified several dimensions of change to increase the impact of purchasing on environmental results. Included among these dimensions are materials used in product design for the environment, product design process, supplier process improvement, supplier evaluation and inbound logistics processes. They also identified the top ten environmental suppliers evaluation criteria, including suppliers ISO 14001 certification and second-tier supplier (suppliers supplier) environmental friendly practice evaluation. However, their research was descriptive in nature with little empirical evidence for support to help confirm their identified factors and measurement scales. By investigating purchasing managers in Germany, the UK and the USA, Zsidisin and Hendrick (1998) identified four GSCM factors, namely hazardous materials, investment recovery, product design and supply chain relationships, and determined the existence of these four factors with an exploratory factor analysis (EFA).
Even with the above research, broadly integrative and confirmed measures of GSCM practices implementation have yet to be fully investigated. This research is predicated upon and a direct outcome of our previous works related to GSCM practices implementation and performance in China. Previously, using a convenience sample of 186 respondents, Zhu and Sarkis (2004) investigated relationships between GSCM practices implementation and performance with a focus on the moderating effects of quality and just-in-time (lean) practices. Zhu et al. (2005), using a broader study with a random sample of Chinese organizations with 314 responses, derived groupings of GSCM pressures, practice and performance. Building on our previous studies and earlier exploratory research (Zhu and Sarkis, 2004; Zhu et al., 2005), this study aims to examine the measurement model of GSCM practices implementation focusing on its five underlying dimensions (factors) and a measurement scale for it. These dimensions and the scale are represented in the form of questionnaire items, for measuring the different facets of GSCM practices implementation, enabling organizations to evaluate their strengths and weaknesses in the course of implementing these practices. The five underlying GSCM practices implementation factors to be confirmed in our study include internal environmental management (IEM), green purchasing (GP), cooperation with customers including environmental requirements (CC), eco-design practices (ECO), and investment recovery (IR).
3 Methodology
Following Churchills (1979) paradigm for construct development and measurement, we first conceptualize the construct of GSCM practices implementation and then operationalize the construct by developing a multi-item five-point Likert measurement scale to evaluate the different facets of GSCM practices implemented among Chinese manufacturers. To help support scale generalization it is important to collect data from a broad variety of organizational and contextual characteristics. Even though we focus on one country, the diversity of organizations (ownership, size, industry, and level of development) within China provides a robust contextual environment. China is a typical developing country with increasing manufacturing, but evidence does exist that Chinese manufacturers have initiated various GSCM practices implementation (Zhu and Sarkis, 2004; Zhu et al., 2005).The measurement scale instrument in the form of a survey questionnaire (developed from the various literature sources summarized in Section 2) was initially pilot-tested with respondents from the Environmental Protection Bureau in two industrial zones, then refined with feedback from the pilot test to improve the wording and seminal meanings of some individual measurement items. Subsequently, the refined scale was administered to capture GSCM practices implementation in a cross-sectional survey among Chinese manufacturers. To evaluate the construct of GSCM practices implementation, confirmatory factor analysis (CFA) tests were performed in our study to examine the measurement properties of GSCM practices implementation, followed by a comparison of the test results of two alternative measurement models evaluating the GSCM practices implementation construct. The methodology employed to guide the research process for this study is summarized in Figure 1. This methodology is further embellished in the following sections.
3.1 GSCM practices implementation and the Chinese manufacturing context
GSCM has become an increasingly important management approach for Chinese manufacturers to help achieve cost and service advantages. Based on results of previous exploratory research (Zhu and Sarkis, 2004) we conceptualize GSCM practices implementation as encompassing five different dimensions of practices including IEM, GP, CC, ECO and IR.
IEM has been the most widely adopted set of GSCM practices by Chinese manufacturers (Zhu et al., 2005). GP is one main aspect of GSCM practices implementation. In some cases GP has been considered as the complete scope of GSCM practices implementation (Walton et al. 1998) while in other studies GP is just an element of GSCM practices implementation (Nagel, 2000). Compared to GP, CC has gained less attention. Researchers have identified opportunities for suppliers to cooperate with their customers and even affect the design and development of their environmental practices (BSR, 2001; Lai et al., 2005). External (to the organization) GSCM practices implementation such as GP and CC have been only at the consideration stage rather than on actual implementation by many Chinese manufacturers (Zhu et al., 2005). Yet, the adoption of these external GSCM practices implementation is growing in the Chinese manufacturing context. ECO is a critical factor governing the environmental impacts of a manufactured product since materials and processes are selected at the design stage (Lewis and Gretsakis, 2001). ECO has become an emerging environmental practice in China but has significant internal and external influence on GSCM practices implementation (Zhu and Geng, 2001). Both US and European enterprises have considered IR as a critical aspect for GSCM practices implementation (Zsidisin and Hendrick, 1998). In China, IR has received much less attention than in developed countries such as the U.S. and Germany due to the inadequacy of Chinese waste management policies and lack of recycling systems (TEDA, 2003). This complex environmental situation with variations in the practices and their level of adoption sets the context for this study to examine the measurement of GSCM practices implementation and a scale for evaluating the different facets of the implementation of its practices.
3.2 Measurement development
To develop a measurement scale for the implementation of GSCM practices, we developed a list of 21 measurement items on GSCM practices implementation that are generally deemed important for implementation by manufacturers (see details in Appendix A). The measurement items were developed on the basis of inputs by industrial experts and the literature (Zsidisin and Hendrick, 1998; Walton et al., 1998; Carter et al., 2000), as discussed earlier. In our previous work (Zhu and Sarkis, 2004), we refined these 21 measurement items with comments from academics and practitioners in environmental studies and supply chain management. This procedure ensured content and face validity of the measurement items. All the measurement items are organized into a survey questionnaire administered to manufacturers in the Chinese mainland. The target respondents of our survey were requested to indicate, using a five-point Likert scale (1=not considering it, 2=planning to consider it, 3=considering it currently, 4=initiating implementation, 5= implementing successfully), the extent to which they perceived their companies implementing each of the dimensions of GSCM practices, underpinned by the 21 individual questionnaire items for measurement.3.3 Data collection
To control the contextual differences among our target survey respondents, we focused on two types of manufacturers, that is, traditional heavy polluters and manufacturers exporting products or suppliers of foreign manufacturers in China. Due to the difficulties in obtaining data as well as the likelihood of our survey respondents to misunderstand the survey items, we used convenience sampling as the first step. Through training workshops in the School of Management at Dalian University of Technology, we interviewed all of our target respondents, explaining the purpose of our study. We collected their completed questionnaire several days later. In this step, we received 213 valid and usable questionnaires for data analysis.
To avoid potential bias from convenience sampling we also completed a random mailing to manufacturers with follow-up telephone calls in the Dalian metropolitan area. The targeted companies were identified from the list of Dalian Manufacturers with a focus on the two major groupings of manufacturers mentioned above. Out of a total of 1,000 survey questionnaires mailed, a total of 128 usable responses from manufacturing enterprises were received.
There are advantages of using both convenience and random surveys in data collection. First, the convenience samples are representative of heavy Chinese manufacturing polluters. By collecting data from them, it can help to assure the validity of sample selection in this study. Second, the respondents in our convenience samples are key informants on the environmental management practices that are being planned or implemented in their companies. These groups of respondents are knowledgeable on the topic of GSCM practices implementation under investigation and help to ensure the quality of the data collected in this study. Third, the random survey contributed to triangulate the quality of the responses in our convenient samples and make it more generalizable in the wider Chinese manufacturing context. As the study samples collected in the three inter-related stages are key manufacturing polluters in China, they are operating under similar industrial environments and face similar pressures for embracing such environmental practices as GSCM to improve their operations. It is therefore reasonable to combine the samples collected from the three stages for data analysis.
Further, we completed a Chi-square test to compare organizational characteristics of the two groups of respondent manufacturers, i.e., the convenience samples including respondents from workshops and the mail survey samples. The test results indicate that no difference, at a 5% level of significance, exists between the two groups on ownership and firm size.In sum, a total of 341 valid and usable manufacturer responses were received for data analysis. Our survey respondents are mainly from four industries, that is, automobile (109, 32.0%), power generating (70, 20.5%), chemical/petroleum (50, 14.7%), and electrical & electronic (39, 11.4%). The remaining 73 respondents (21.4%) are from textile, steel, food processing, pharmaceutical, etc. Organizational sizes range from under 500 to over 8,000 employees with the majority of manufacturers falling into the relatively large company classification of between 500 and 8,000 employees. The distribution can be described as 29 manufacturers (8.5%) with over 8,000 employees, 65 manufacturers (19.1%) with employees from 3,000 to 8,000, 104 manufacturers (30.5%) with employees from 1,000 to 3,000, 82 manufacturers (24.0%) with employees from 500 to 1,000, and 61 manufacturers (17.9%) with employees below 500.
4 Results4.1 Validity and reliability testingWe first tested the measurement properties of the GSCM practices implementation construct using reliability and item-to-total correlation analyses, followed by confirmatory factor analysis (CFA) (c.f. Lai et al., 2002). CFA was used to assess how well the observed variables, i.e., measurement items, reflect unobserved or latent variables in the hypothesized structure. A strong a priori basis from our previous research warrants the use of CFA instead of EFA.
The reliability test and item-to-total correlation analysis results provided in Table 1 suggest a reasonable fit of the latent factors to the data collected. Cronbach alpha values for all the five factors of GSCM practices implementation are greater than 0.83 and the item loadings on the factors are all acceptable, i.e., >0.48.
A series of pairwise CFAs were conducted to assess the discriminant validity of the factors using 2 difference tests (Anderson and Gerbing, 1988). This step of the analysis was conducted by forcing measurement items of each pair of factors into a single underlying factor. If there is a significant deterioration of model fit relative to a two-factor model, then this result implies the presence of discriminant validity between the pair of factors (Bagozzi and Phillips, 1982). This test was performed on all possible pairs of the factors. Table 2 reports the results of the 10 pairwise tests of the factors. Discriminant validity is achieved for all cases. The significant results of the 2 difference tests attest to the presence of discriminant validity between any two factors (Anderson and Gerbing, 1988). Before examining the measurement models of the GSCM practices implementation construct, we tested the criterion validity of the five factors of GSCM with three outcome variables, i.e., environmental performance (EP), economic performance (ECP), and operational performance (OP). The items measuring these three perceived performance variables are listed in Appendix B. We evaluated the criterion validity of the GSCM practices implementation construct by computing the correlations of its five underlying factors, i.e., IEM, GP, CC, ECO, and IR with these three performance measures, i.e., EP, ECP, and OP. The correlations of the five GSCM factors with the three performance measures are all significant at p < 0.05 level, with the largest coefficient between IEM and EP (r = 0.624) and the lowest coefficient between CC and OP (r = 0.293). The test results provide evidence for the criterion validity of the GSCM practices implementation construct.4.2 Testing first-order and second-order models
In the previous discussion, IEM, GP, CC, ECO and IR are specified as a priori factors of GSCM practices implementation. In the first-order model, IEM, GP, CC, ECO and IR are correlated measurement factors for GSCM practices implementation. Alternatively, GSCM practices implementation may be operationalized as a second-order model, where the five factors are governed by a higher order factor, i.e., GSCM practices implementation. The results of the model estimation are shown in Figures 2 and 3.
The first-order model for testing the GSCM construct (see Figure 2) implies that IEM, GP, CC, ECO and IR are correlated but not governed by a common latent factor. Although the 2 statistic is significant (2 =616.928; df = 179; p 11, p < 0.001.
Development of concepts on GSCM practices implementation
Review related literature
Understand GSCM practices implementation generally and obtain practical input from Chinese practitioners
Develop two potential factor models
Development of measurement items for evaluating GSCM practices implementation
Select measurement items for evaluating GSCM practices implementation
Design survey questionnaire for the measurement items
Interview academics and practitioners among Chinese manufacturers
Model testing
Validity and reliability tests of factors
CFA tests for measurement models
Comparison of measurement models
Data collection
Pilot test
Convenience sampling
Random surveys
Figure 1 Research Process
Chi Square (179)=616.928 (P
